uri.c 217 KB
Newer Older
1 2
/*
 * Copyright 2010 Jacek Caban for CodeWeavers
3
 * Copyright 2010 Thomas Mullaly
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
 *
 * This library is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * This library is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with this library; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA
 */

#include "urlmon_main.h"
#include "wine/debug.h"

23 24 25
#define NO_SHLWAPI_REG
#include "shlwapi.h"

26 27
#include "strsafe.h"

28
#define UINT_MAX 0xffffffff
29
#define USHORT_MAX 0xffff
30

31 32 33
#define URI_DISPLAY_NO_ABSOLUTE_URI         0x1
#define URI_DISPLAY_NO_DEFAULT_PORT_AUTH    0x2

34 35 36 37 38 39
#define ALLOW_NULL_TERM_SCHEME          0x01
#define ALLOW_NULL_TERM_USER_NAME       0x02
#define ALLOW_NULL_TERM_PASSWORD        0x04
#define ALLOW_BRACKETLESS_IP_LITERAL    0x08
#define SKIP_IP_FUTURE_CHECK            0x10
#define IGNORE_PORT_DELIMITER           0x20
40

41 42
#define RAW_URI_FORCE_PORT_DISP     0x1
#define RAW_URI_CONVERT_TO_DOS_PATH 0x2
43

44 45
#define COMBINE_URI_FORCE_FLAG_USE  0x1

46 47
WINE_DEFAULT_DEBUG_CHANNEL(urlmon);

48 49
static const IID IID_IUriObj = {0x4b364760,0x9f51,0x11df,{0x98,0x1c,0x08,0x00,0x20,0x0c,0x9a,0x66}};

50
typedef struct {
51 52 53
    IUri                IUri_iface;
    IUriBuilderFactory  IUriBuilderFactory_iface;

54
    LONG ref;
55 56

    BSTR            raw_uri;
57

58
    /* Information about the canonicalized URI's buffer. */
59 60 61
    WCHAR           *canon_uri;
    DWORD           canon_size;
    DWORD           canon_len;
62
    BOOL            display_modifiers;
63
    DWORD           create_flags;
64 65 66 67

    INT             scheme_start;
    DWORD           scheme_len;
    URL_SCHEME      scheme_type;
68

69 70 71
    INT             userinfo_start;
    DWORD           userinfo_len;
    INT             userinfo_split;
72

73 74 75
    INT             host_start;
    DWORD           host_len;
    Uri_HOST_TYPE   host_type;
76

77
    INT             port_offset;
78
    DWORD           port;
79
    BOOL            has_port;
80 81 82

    INT             authority_start;
    DWORD           authority_len;
83 84

    INT             domain_offset;
85 86 87

    INT             path_start;
    DWORD           path_len;
88
    INT             extension_offset;
89 90 91

    INT             query_start;
    DWORD           query_len;
92 93 94

    INT             fragment_start;
    DWORD           fragment_len;
95 96
} Uri;

97
typedef struct {
98
    IUriBuilder IUriBuilder_iface;
99
    LONG ref;
100

101 102 103 104 105
    Uri *uri;
    DWORD modified_props;

    WCHAR   *fragment;
    DWORD   fragment_len;
106 107 108

    WCHAR   *host;
    DWORD   host_len;
109 110 111

    WCHAR   *password;
    DWORD   password_len;
112 113 114

    WCHAR   *path;
    DWORD   path_len;
115 116 117

    BOOL    has_port;
    DWORD   port;
118 119 120

    WCHAR   *query;
    DWORD   query_len;
121 122 123

    WCHAR   *scheme;
    DWORD   scheme_len;
124 125 126

    WCHAR   *username;
    DWORD   username_len;
127 128
} UriBuilder;

129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149
typedef struct {
    const WCHAR *str;
    DWORD       len;
} h16;

typedef struct {
    /* IPv6 addresses can hold up to 8 h16 components. */
    h16         components[8];
    DWORD       h16_count;

    /* An IPv6 can have 1 elision ("::"). */
    const WCHAR *elision;

    /* An IPv6 can contain 1 IPv4 address as the last 32bits of the address. */
    const WCHAR *ipv4;
    DWORD       ipv4_len;

    INT         components_size;
    INT         elision_size;
} ipv6_address;

150 151 152 153
typedef struct {
    BSTR            uri;

    BOOL            is_relative;
154 155
    BOOL            is_opaque;
    BOOL            has_implicit_scheme;
156 157
    BOOL            has_implicit_ip;
    UINT            implicit_ipv4;
158 159 160

    const WCHAR     *scheme;
    DWORD           scheme_len;
161
    URL_SCHEME      scheme_type;
162

163 164 165 166 167
    const WCHAR     *username;
    DWORD           username_len;

    const WCHAR     *password;
    DWORD           password_len;
168 169 170 171

    const WCHAR     *host;
    DWORD           host_len;
    Uri_HOST_TYPE   host_type;
172 173 174

    BOOL            has_ipv6;
    ipv6_address    ipv6_address;
175

176
    BOOL            has_port;
177 178
    const WCHAR     *port;
    DWORD           port_len;
179
    DWORD           port_value;
180 181 182

    const WCHAR     *path;
    DWORD           path_len;
183 184 185

    const WCHAR     *query;
    DWORD           query_len;
186 187 188

    const WCHAR     *fragment;
    DWORD           fragment_len;
189 190
} parse_data;

191 192
static const CHAR hexDigits[] = "0123456789ABCDEF";

193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221
/* List of scheme types/scheme names that are recognized by the IUri interface as of IE 7. */
static const struct {
    URL_SCHEME  scheme;
    WCHAR       scheme_name[16];
} recognized_schemes[] = {
    {URL_SCHEME_FTP,            {'f','t','p',0}},
    {URL_SCHEME_HTTP,           {'h','t','t','p',0}},
    {URL_SCHEME_GOPHER,         {'g','o','p','h','e','r',0}},
    {URL_SCHEME_MAILTO,         {'m','a','i','l','t','o',0}},
    {URL_SCHEME_NEWS,           {'n','e','w','s',0}},
    {URL_SCHEME_NNTP,           {'n','n','t','p',0}},
    {URL_SCHEME_TELNET,         {'t','e','l','n','e','t',0}},
    {URL_SCHEME_WAIS,           {'w','a','i','s',0}},
    {URL_SCHEME_FILE,           {'f','i','l','e',0}},
    {URL_SCHEME_MK,             {'m','k',0}},
    {URL_SCHEME_HTTPS,          {'h','t','t','p','s',0}},
    {URL_SCHEME_SHELL,          {'s','h','e','l','l',0}},
    {URL_SCHEME_SNEWS,          {'s','n','e','w','s',0}},
    {URL_SCHEME_LOCAL,          {'l','o','c','a','l',0}},
    {URL_SCHEME_JAVASCRIPT,     {'j','a','v','a','s','c','r','i','p','t',0}},
    {URL_SCHEME_VBSCRIPT,       {'v','b','s','c','r','i','p','t',0}},
    {URL_SCHEME_ABOUT,          {'a','b','o','u','t',0}},
    {URL_SCHEME_RES,            {'r','e','s',0}},
    {URL_SCHEME_MSSHELLROOTED,  {'m','s','-','s','h','e','l','l','-','r','o','o','t','e','d',0}},
    {URL_SCHEME_MSSHELLIDLIST,  {'m','s','-','s','h','e','l','l','-','i','d','l','i','s','t',0}},
    {URL_SCHEME_MSHELP,         {'h','c','p',0}},
    {URL_SCHEME_WILDCARD,       {'*',0}}
};

222 223 224 225 226 227 228 229 230 231 232 233 234 235
/* List of default ports Windows recognizes. */
static const struct {
    URL_SCHEME  scheme;
    USHORT      port;
} default_ports[] = {
    {URL_SCHEME_FTP,    21},
    {URL_SCHEME_HTTP,   80},
    {URL_SCHEME_GOPHER, 70},
    {URL_SCHEME_NNTP,   119},
    {URL_SCHEME_TELNET, 23},
    {URL_SCHEME_WAIS,   210},
    {URL_SCHEME_HTTPS,  443},
};

236
/* List of 3-character top level domain names Windows seems to recognize.
237 238 239 240 241 242 243 244 245 246 247 248 249 250
 * There might be more, but, these are the only ones I've found so far.
 */
static const struct {
    WCHAR tld_name[4];
} recognized_tlds[] = {
    {{'c','o','m',0}},
    {{'e','d','u',0}},
    {{'g','o','v',0}},
    {{'i','n','t',0}},
    {{'m','i','l',0}},
    {{'n','e','t',0}},
    {{'o','r','g',0}}
};

251 252 253 254 255 256 257 258 259
static Uri *get_uri_obj(IUri *uri)
{
    Uri *ret;
    HRESULT hres;

    hres = IUri_QueryInterface(uri, &IID_IUriObj, (void**)&ret);
    return SUCCEEDED(hres) ? ret : NULL;
}

260
static inline BOOL is_alpha(WCHAR val) {
261
    return ((val >= 'a' && val <= 'z') || (val >= 'A' && val <= 'Z'));
262 263 264
}

static inline BOOL is_num(WCHAR val) {
265
    return (val >= '0' && val <= '9');
266 267
}

268 269 270 271 272 273 274 275
static inline BOOL is_drive_path(const WCHAR *str) {
    return (is_alpha(str[0]) && (str[1] == ':' || str[1] == '|'));
}

static inline BOOL is_unc_path(const WCHAR *str) {
    return (str[0] == '\\' && str[0] == '\\');
}

276 277 278 279
static inline BOOL is_forbidden_dos_path_char(WCHAR val) {
    return (val == '>' || val == '<' || val == '\"');
}

280
/* A URI is implicitly a file path if it begins with
281
 * a drive letter (e.g. X:) or starts with "\\" (UNC path).
282 283
 */
static inline BOOL is_implicit_file_path(const WCHAR *str) {
284
    return (is_unc_path(str) || (is_alpha(str[0]) && str[1] == ':'));
285 286
}

287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305
/* Checks if the URI is a hierarchical URI. A hierarchical
 * URI is one that has "//" after the scheme.
 */
static BOOL check_hierarchical(const WCHAR **ptr) {
    const WCHAR *start = *ptr;

    if(**ptr != '/')
        return FALSE;

    ++(*ptr);
    if(**ptr != '/') {
        *ptr = start;
        return FALSE;
    }

    ++(*ptr);
    return TRUE;
}

306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330
/* unreserved    = ALPHA / DIGIT / "-" / "." / "_" / "~" */
static inline BOOL is_unreserved(WCHAR val) {
    return (is_alpha(val) || is_num(val) || val == '-' || val == '.' ||
            val == '_' || val == '~');
}

/* sub-delims    = "!" / "$" / "&" / "'" / "(" / ")"
 *               / "*" / "+" / "," / ";" / "="
 */
static inline BOOL is_subdelim(WCHAR val) {
    return (val == '!' || val == '$' || val == '&' ||
            val == '\'' || val == '(' || val == ')' ||
            val == '*' || val == '+' || val == ',' ||
            val == ';' || val == '=');
}

/* gen-delims  = ":" / "/" / "?" / "#" / "[" / "]" / "@" */
static inline BOOL is_gendelim(WCHAR val) {
    return (val == ':' || val == '/' || val == '?' ||
            val == '#' || val == '[' || val == ']' ||
            val == '@');
}

/* Characters that delimit the end of the authority
 * section of a URI. Sometimes a '\\' is considered
331
 * an authority delimiter.
332 333 334 335 336 337
 */
static inline BOOL is_auth_delim(WCHAR val, BOOL acceptSlash) {
    return (val == '#' || val == '/' || val == '?' ||
            val == '\0' || (acceptSlash && val == '\\'));
}

338 339 340 341 342
/* reserved = gen-delims / sub-delims */
static inline BOOL is_reserved(WCHAR val) {
    return (is_subdelim(val) || is_gendelim(val));
}

343 344 345 346 347 348
static inline BOOL is_hexdigit(WCHAR val) {
    return ((val >= 'a' && val <= 'f') ||
            (val >= 'A' && val <= 'F') ||
            (val >= '0' && val <= '9'));
}

349 350 351 352
static inline BOOL is_path_delim(WCHAR val) {
    return (!val || val == '#' || val == '?');
}

353 354 355 356 357
static inline BOOL is_slash(WCHAR c)
{
    return c == '/' || c == '\\';
}

358 359 360 361 362 363 364 365 366 367 368
static BOOL is_default_port(URL_SCHEME scheme, DWORD port) {
    DWORD i;

    for(i = 0; i < sizeof(default_ports)/sizeof(default_ports[0]); ++i) {
        if(default_ports[i].scheme == scheme && default_ports[i].port)
            return TRUE;
    }

    return FALSE;
}

369 370 371 372 373 374 375 376 377
/* List of schemes types Windows seems to expect to be hierarchical. */
static inline BOOL is_hierarchical_scheme(URL_SCHEME type) {
    return(type == URL_SCHEME_HTTP || type == URL_SCHEME_FTP ||
           type == URL_SCHEME_GOPHER || type == URL_SCHEME_NNTP ||
           type == URL_SCHEME_TELNET || type == URL_SCHEME_WAIS ||
           type == URL_SCHEME_FILE || type == URL_SCHEME_HTTPS ||
           type == URL_SCHEME_RES);
}

378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402
/* Checks if 'flags' contains an invalid combination of Uri_CREATE flags. */
static inline BOOL has_invalid_flag_combination(DWORD flags) {
    return((flags & Uri_CREATE_DECODE_EXTRA_INFO && flags & Uri_CREATE_NO_DECODE_EXTRA_INFO) ||
           (flags & Uri_CREATE_CANONICALIZE && flags & Uri_CREATE_NO_CANONICALIZE) ||
           (flags & Uri_CREATE_CRACK_UNKNOWN_SCHEMES && flags & Uri_CREATE_NO_CRACK_UNKNOWN_SCHEMES) ||
           (flags & Uri_CREATE_PRE_PROCESS_HTML_URI && flags & Uri_CREATE_NO_PRE_PROCESS_HTML_URI) ||
           (flags & Uri_CREATE_IE_SETTINGS && flags & Uri_CREATE_NO_IE_SETTINGS));
}

/* Applies each default Uri_CREATE flags to 'flags' if it
 * doesn't cause a flag conflict.
 */
static void apply_default_flags(DWORD *flags) {
    if(!(*flags & Uri_CREATE_NO_CANONICALIZE))
        *flags |= Uri_CREATE_CANONICALIZE;
    if(!(*flags & Uri_CREATE_NO_DECODE_EXTRA_INFO))
        *flags |= Uri_CREATE_DECODE_EXTRA_INFO;
    if(!(*flags & Uri_CREATE_NO_CRACK_UNKNOWN_SCHEMES))
        *flags |= Uri_CREATE_CRACK_UNKNOWN_SCHEMES;
    if(!(*flags & Uri_CREATE_NO_PRE_PROCESS_HTML_URI))
        *flags |= Uri_CREATE_PRE_PROCESS_HTML_URI;
    if(!(*flags & Uri_CREATE_IE_SETTINGS))
        *flags |= Uri_CREATE_NO_IE_SETTINGS;
}

403 404 405
/* Determines if the URI is hierarchical using the information already parsed into
 * data and using the current location of parsing in the URI string.
 *
406
 * Windows considers a URI hierarchical if one of the following is true:
407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429
 *  A.) It's a wildcard scheme.
 *  B.) It's an implicit file scheme.
 *  C.) It's a known hierarchical scheme and it has two '\\' after the scheme name.
 *      (the '\\' will be converted into "//" during canonicalization).
 *  D.) It's not a relative URI and "//" appears after the scheme name.
 */
static inline BOOL is_hierarchical_uri(const WCHAR **ptr, const parse_data *data) {
    const WCHAR *start = *ptr;

    if(data->scheme_type == URL_SCHEME_WILDCARD)
        return TRUE;
    else if(data->scheme_type == URL_SCHEME_FILE && data->has_implicit_scheme)
        return TRUE;
    else if(is_hierarchical_scheme(data->scheme_type) && (*ptr)[0] == '\\' && (*ptr)[1] == '\\') {
        *ptr += 2;
        return TRUE;
    } else if(!data->is_relative && check_hierarchical(ptr))
        return TRUE;

    *ptr = start;
    return FALSE;
}

430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527
/* Checks if the two Uri's are logically equivalent. It's a simple
 * comparison, since they are both of type Uri, and it can access
 * the properties of each Uri directly without the need to go
 * through the "IUri_Get*" interface calls.
 */
static BOOL are_equal_simple(const Uri *a, const Uri *b) {
    if(a->scheme_type == b->scheme_type) {
        const BOOL known_scheme = a->scheme_type != URL_SCHEME_UNKNOWN;
        const BOOL are_hierarchical =
                (a->authority_start > -1 && b->authority_start > -1);

        if(a->scheme_type == URL_SCHEME_FILE) {
            if(a->canon_len == b->canon_len)
                return !StrCmpIW(a->canon_uri, b->canon_uri);
        }

        /* Only compare the scheme names (if any) if their unknown scheme types. */
        if(!known_scheme) {
            if((a->scheme_start > -1 && b->scheme_start > -1) &&
               (a->scheme_len == b->scheme_len)) {
                /* Make sure the schemes are the same. */
                if(StrCmpNW(a->canon_uri+a->scheme_start, b->canon_uri+b->scheme_start, a->scheme_len))
                    return FALSE;
            } else if(a->scheme_len != b->scheme_len)
                /* One of the Uri's has a scheme name, while the other doesn't. */
                return FALSE;
        }

        /* If they have a userinfo component, perform case sensitive compare. */
        if((a->userinfo_start > -1 && b->userinfo_start > -1) &&
           (a->userinfo_len == b->userinfo_len)) {
            if(StrCmpNW(a->canon_uri+a->userinfo_start, b->canon_uri+b->userinfo_start, a->userinfo_len))
                return FALSE;
        } else if(a->userinfo_len != b->userinfo_len)
            /* One of the Uri's had a userinfo, while the other one doesn't. */
            return FALSE;

        /* Check if they have a host name. */
        if((a->host_start > -1 && b->host_start > -1) &&
           (a->host_len == b->host_len)) {
            /* Perform a case insensitive compare if they are a known scheme type. */
            if(known_scheme) {
                if(StrCmpNIW(a->canon_uri+a->host_start, b->canon_uri+b->host_start, a->host_len))
                    return FALSE;
            } else if(StrCmpNW(a->canon_uri+a->host_start, b->canon_uri+b->host_start, a->host_len))
                return FALSE;
        } else if(a->host_len != b->host_len)
            /* One of the Uri's had a host, while the other one didn't. */
            return FALSE;

        if(a->has_port && b->has_port) {
            if(a->port != b->port)
                return FALSE;
        } else if(a->has_port || b->has_port)
            /* One had a port, while the other one didn't. */
            return FALSE;

        /* Windows is weird with how it handles paths. For example
         * One URI could be "http://google.com" (after canonicalization)
         * and one could be "http://google.com/" and the IsEqual function
         * would still evaluate to TRUE, but, only if they are both hierarchical
         * URIs.
         */
        if((a->path_start > -1 && b->path_start > -1) &&
           (a->path_len == b->path_len)) {
            if(StrCmpNW(a->canon_uri+a->path_start, b->canon_uri+b->path_start, a->path_len))
                return FALSE;
        } else if(are_hierarchical && a->path_len == -1 && b->path_len == 0) {
            if(*(a->canon_uri+a->path_start) != '/')
                return FALSE;
        } else if(are_hierarchical && b->path_len == 1 && a->path_len == 0) {
            if(*(b->canon_uri+b->path_start) != '/')
                return FALSE;
        } else if(a->path_len != b->path_len)
            return FALSE;

        /* Compare the query strings of the two URIs. */
        if((a->query_start > -1 && b->query_start > -1) &&
           (a->query_len == b->query_len)) {
            if(StrCmpNW(a->canon_uri+a->query_start, b->canon_uri+b->query_start, a->query_len))
                return FALSE;
        } else if(a->query_len != b->query_len)
            return FALSE;

        if((a->fragment_start > -1 && b->fragment_start > -1) &&
           (a->fragment_len == b->fragment_len)) {
            if(StrCmpNW(a->canon_uri+a->fragment_start, b->canon_uri+b->fragment_start, a->fragment_len))
                return FALSE;
        } else if(a->fragment_len != b->fragment_len)
            return FALSE;

        /* If we get here, the two URIs are equivalent. */
        return TRUE;
    }

    return FALSE;
}

528
/* Computes the size of the given IPv6 address.
529 530
 * Each h16 component is 16 bits. If there is an IPv4 address, it's
 * 32 bits. If there's an elision it can be 16 to 128 bits, depending
531 532 533 534 535 536 537 538 539 540 541 542 543
 * on the number of other components.
 *
 * Modeled after google-url's CheckIPv6ComponentsSize function
 */
static void compute_ipv6_comps_size(ipv6_address *address) {
    address->components_size = address->h16_count * 2;

    if(address->ipv4)
        /* IPv4 address is 4 bytes. */
        address->components_size += 4;

    if(address->elision) {
        /* An elision can be anywhere from 2 bytes up to 16 bytes.
544
         * Its size depends on the size of the h16 and IPv4 components.
545 546 547 548 549 550 551 552
         */
        address->elision_size = 16 - address->components_size;
        if(address->elision_size < 2)
            address->elision_size = 2;
    } else
        address->elision_size = 0;
}

553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569
/* Taken from dlls/jscript/lex.c */
static int hex_to_int(WCHAR val) {
    if(val >= '0' && val <= '9')
        return val - '0';
    else if(val >= 'a' && val <= 'f')
        return val - 'a' + 10;
    else if(val >= 'A' && val <= 'F')
        return val - 'A' + 10;

    return -1;
}

/* Helper function for converting a percent encoded string
 * representation of a WCHAR value into its actual WCHAR value. If
 * the two characters following the '%' aren't valid hex values then
 * this function returns the NULL character.
 *
570
 * E.g.
571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599
 *  "%2E" will result in '.' being returned by this function.
 */
static WCHAR decode_pct_val(const WCHAR *ptr) {
    WCHAR ret = '\0';

    if(*ptr == '%' && is_hexdigit(*(ptr + 1)) && is_hexdigit(*(ptr + 2))) {
        INT a = hex_to_int(*(ptr + 1));
        INT b = hex_to_int(*(ptr + 2));

        ret = a << 4;
        ret += b;
    }

    return ret;
}

/* Helper function for percent encoding a given character
 * and storing the encoded value into a given buffer (dest).
 *
 * It's up to the calling function to ensure that there is
 * at least enough space in 'dest' for the percent encoded
 * value to be stored (so dest + 3 spaces available).
 */
static inline void pct_encode_val(WCHAR val, WCHAR *dest) {
    dest[0] = '%';
    dest[1] = hexDigits[(val >> 4) & 0xf];
    dest[2] = hexDigits[val & 0xf];
}

600 601 602 603 604 605 606
/* Attempts to parse the domain name from the host.
 *
 * This function also includes the Top-level Domain (TLD) name
 * of the host when it tries to find the domain name. If it finds
 * a valid domain name it will assign 'domain_start' the offset
 * into 'host' where the domain name starts.
 *
607
 * It's implied that if there is a domain name its range is:
608
 * [host+domain_start, host+host_len).
609
 */
610
void find_domain_name(const WCHAR *host, DWORD host_len,
611 612 613 614 615 616 617 618
                             INT *domain_start) {
    const WCHAR *last_tld, *sec_last_tld, *end;

    end = host+host_len-1;

    *domain_start = -1;

    /* There has to be at least enough room for a '.' followed by a
619
     * 3-character TLD for a domain to even exist in the host name.
620 621 622 623
     */
    if(host_len < 4)
        return;

624
    last_tld = memrchrW(host, '.', host_len);
625 626 627 628
    if(!last_tld)
        /* http://hostname -> has no domain name. */
        return;

629
    sec_last_tld = memrchrW(host, '.', last_tld-host);
630 631 632 633 634 635 636 637 638 639 640 641 642
    if(!sec_last_tld) {
        /* If the '.' is at the beginning of the host there
         * has to be at least 3 characters in the TLD for it
         * to be valid.
         *  Ex: .com -> .com as the domain name.
         *      .co  -> has no domain name.
         */
        if(last_tld-host == 0) {
            if(end-(last_tld-1) < 3)
                return;
        } else if(last_tld-host == 3) {
            DWORD i;

643
            /* If there are three characters in front of last_tld and
644 645 646 647 648 649 650 651 652 653 654
             * they are on the list of recognized TLDs, then this
             * host doesn't have a domain (since the host only contains
             * a TLD name.
             *  Ex: edu.uk -> has no domain name.
             *      foo.uk -> foo.uk as the domain name.
             */
            for(i = 0; i < sizeof(recognized_tlds)/sizeof(recognized_tlds[0]); ++i) {
                if(!StrCmpNIW(host, recognized_tlds[i].tld_name, 3))
                    return;
            }
        } else if(last_tld-host < 3)
655
            /* Anything less than 3 characters is considered part
656 657 658 659 660 661 662 663
             * of the TLD name.
             *  Ex: ak.uk -> Has no domain name.
             */
            return;

        /* Otherwise the domain name is the whole host name. */
        *domain_start = 0;
    } else if(end+1-last_tld > 3) {
664
        /* If the last_tld has more than 3 characters, then it's automatically
665 666 667 668 669 670 671 672 673 674 675 676 677 678 679
         * considered the TLD of the domain name.
         *  Ex: www.winehq.org.uk.test -> uk.test as the domain name.
         */
        *domain_start = (sec_last_tld+1)-host;
    } else if(last_tld - (sec_last_tld+1) < 4) {
        DWORD i;
        /* If the sec_last_tld is 3 characters long it HAS to be on the list of
         * recognized to still be considered part of the TLD name, otherwise
         * its considered the domain name.
         *  Ex: www.google.com.uk -> google.com.uk as the domain name.
         *      www.google.foo.uk -> foo.uk as the domain name.
         */
        if(last_tld - (sec_last_tld+1) == 3) {
            for(i = 0; i < sizeof(recognized_tlds)/sizeof(recognized_tlds[0]); ++i) {
                if(!StrCmpNIW(sec_last_tld+1, recognized_tlds[i].tld_name, 3)) {
680
                    const WCHAR *domain = memrchrW(host, '.', sec_last_tld-host);
681 682 683 684 685 686 687 688 689 690 691 692 693

                    if(!domain)
                        *domain_start = 0;
                    else
                        *domain_start = (domain+1) - host;
                    TRACE("Found domain name %s\n", debugstr_wn(host+*domain_start,
                                                        (host+host_len)-(host+*domain_start)));
                    return;
                }
            }

            *domain_start = (sec_last_tld+1)-host;
        } else {
694
            /* Since the sec_last_tld is less than 3 characters it's considered
695 696 697
             * part of the TLD.
             *  Ex: www.google.fo.uk -> google.fo.uk as the domain name.
             */
698
            const WCHAR *domain = memrchrW(host, '.', sec_last_tld-host);
699 700 701 702 703 704 705

            if(!domain)
                *domain_start = 0;
            else
                *domain_start = (domain+1) - host;
        }
    } else {
706
        /* The second to last TLD has more than 3 characters making it
707 708 709 710 711 712 713 714 715 716
         * the domain name.
         *  Ex: www.google.test.us -> test.us as the domain name.
         */
        *domain_start = (sec_last_tld+1)-host;
    }

    TRACE("Found domain name %s\n", debugstr_wn(host+*domain_start,
                                        (host+host_len)-(host+*domain_start)));
}

717
/* Removes the dot segments from a hierarchical URIs path component. This
718 719 720 721 722 723 724 725 726 727 728
 * function performs the removal in place.
 *
 * This function returns the new length of the path string.
 */
static DWORD remove_dot_segments(WCHAR *path, DWORD path_len) {
    WCHAR *out = path;
    const WCHAR *in = out;
    const WCHAR *end = out + path_len;
    DWORD len;

    while(in < end) {
729 730 731
        /* Move the first path segment in the input buffer to the end of
         * the output buffer, and any subsequent characters up to, including
         * the next "/" character (if any) or the end of the input buffer.
732
         */
733 734 735
        while(in < end && !is_slash(*in))
            *out++ = *in++;
        if(in == end)
736
            break;
737
        *out++ = *in++;
738

739 740 741
        while(in < end) {
            if(*in != '.')
                break;
742

743 744 745 746 747
            /* Handle ending "/." */
            if(in + 1 == end) {
                ++in;
                break;
            }
748

749 750 751 752 753
            /* Handle "/./" */
            if(is_slash(in[1])) {
                in += 2;
                continue;
            }
754

755 756 757
            /* If we don't have "/../" or ending "/.." */
            if(in[1] != '.' || (in + 2 != end && !is_slash(in[2])))
                break;
758

759 760 761 762 763 764 765 766 767 768
            /* Find the slash preceding out pointer and move out pointer to it */
            if(out > path+1 && is_slash(*--out))
                --out;
            while(out > path && !is_slash(*(--out)));
            if(is_slash(*out))
                ++out;
            in += 2;
            if(in != end)
                ++in;
        }
769 770 771 772 773 774 775 776
    }

    len = out - path;
    TRACE("(%p %d): Path after dot segments removed %s len=%d\n", path, path_len,
        debugstr_wn(path, len), len);
    return len;
}

777 778 779 780 781 782 783 784 785 786 787 788
/* Attempts to find the file extension in a given path. */
static INT find_file_extension(const WCHAR *path, DWORD path_len) {
    const WCHAR *end;

    for(end = path+path_len-1; end >= path && *end != '/' && *end != '\\'; --end) {
        if(*end == '.')
            return end-path;
    }

    return -1;
}

789 790 791
/* Computes the location where the elision should occur in the IPv6
 * address using the numerical values of each component stored in
 * 'values'. If the address shouldn't contain an elision then 'index'
792
 * is assigned -1 as its value. Otherwise 'index' will contain the
793 794 795 796
 * starting index (into values) where the elision should be, and 'count'
 * will contain the number of cells the elision covers.
 *
 * NOTES:
797
 *  Windows will expand an elision if the elision only represents one h16
798
 *  component of the address.
799 800 801 802
 *
 *  Ex: [1::2:3:4:5:6:7] -> [1:0:2:3:4:5:6:7]
 *
 *  If the IPv6 address contains an IPv4 address, the IPv4 address is also
803
 *  considered for being included as part of an elision if all its components
804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836
 *  are zeros.
 *
 *  Ex: [1:2:3:4:5:6:0.0.0.0] -> [1:2:3:4:5:6::]
 */
static void compute_elision_location(const ipv6_address *address, const USHORT values[8],
                                     INT *index, DWORD *count) {
    DWORD i, max_len, cur_len;
    INT max_index, cur_index;

    max_len = cur_len = 0;
    max_index = cur_index = -1;
    for(i = 0; i < 8; ++i) {
        BOOL check_ipv4 = (address->ipv4 && i == 6);
        BOOL is_end = (check_ipv4 || i == 7);

        if(check_ipv4) {
            /* Check if the IPv4 address contains only zeros. */
            if(values[i] == 0 && values[i+1] == 0) {
                if(cur_index == -1)
                    cur_index = i;

                cur_len += 2;
                ++i;
            }
        } else if(values[i] == 0) {
            if(cur_index == -1)
                cur_index = i;

            ++cur_len;
        }

        if(is_end || values[i] != 0) {
            /* We only consider it for an elision if it's
837
             * more than 1 component long.
838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854
             */
            if(cur_len > 1 && cur_len > max_len) {
                /* Found the new elision location. */
                max_len = cur_len;
                max_index = cur_index;
            }

            /* Reset the current range for the next range of zeros. */
            cur_index = -1;
            cur_len = 0;
        }
    }

    *index = max_index;
    *count = max_len;
}

855 856 857 858 859
/* Removes all the leading and trailing white spaces or
 * control characters from the URI and removes all control
 * characters inside of the URI string.
 */
static BSTR pre_process_uri(LPCWSTR uri) {
860 861
    const WCHAR *start, *end, *ptr;
    WCHAR *ptr2;
862
    DWORD len;
863
    BSTR ret;
864 865 866

    start = uri;
    /* Skip leading controls and whitespace. */
867
    while(*start && (iscntrlW(*start) || isspaceW(*start))) ++start;
868

869 870 871
    /* URI consisted only of control/whitespace. */
    if(!*start)
        return SysAllocStringLen(NULL, 0);
872

873 874
    end = start + strlenW(start);
    while(--end > start && (iscntrlW(*end) || isspaceW(*end)));
875

876 877 878 879 880 881 882 883 884
    len = ++end - start;
    for(ptr = start; ptr < end; ptr++) {
        if(iscntrlW(*ptr))
            len--;
    }

    ret = SysAllocStringLen(NULL, len);
    if(!ret)
        return NULL;
885

886 887 888
    for(ptr = start, ptr2=ret; ptr < end; ptr++) {
        if(!iscntrlW(*ptr))
            *ptr2++ = *ptr;
889 890 891 892 893
    }

    return ret;
}

894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917
/* Converts the specified IPv4 address into an uint value.
 *
 * This function assumes that the IPv4 address has already been validated.
 */
static UINT ipv4toui(const WCHAR *ip, DWORD len) {
    UINT ret = 0;
    DWORD comp_value = 0;
    const WCHAR *ptr;

    for(ptr = ip; ptr < ip+len; ++ptr) {
        if(*ptr == '.') {
            ret <<= 8;
            ret += comp_value;
            comp_value = 0;
        } else
            comp_value = comp_value*10 + (*ptr-'0');
    }

    ret <<= 8;
    ret += comp_value;

    return ret;
}

918
/* Converts an IPv4 address in numerical form into its fully qualified
919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945
 * string form. This function returns the number of characters written
 * to 'dest'. If 'dest' is NULL this function will return the number of
 * characters that would have been written.
 *
 * It's up to the caller to ensure there's enough space in 'dest' for the
 * address.
 */
static DWORD ui2ipv4(WCHAR *dest, UINT address) {
    static const WCHAR formatW[] =
        {'%','u','.','%','u','.','%','u','.','%','u',0};
    DWORD ret = 0;
    UCHAR digits[4];

    digits[0] = (address >> 24) & 0xff;
    digits[1] = (address >> 16) & 0xff;
    digits[2] = (address >> 8) & 0xff;
    digits[3] = address & 0xff;

    if(!dest) {
        WCHAR tmp[16];
        ret = sprintfW(tmp, formatW, digits[0], digits[1], digits[2], digits[3]);
    } else
        ret = sprintfW(dest, formatW, digits[0], digits[1], digits[2], digits[3]);

    return ret;
}

946 947 948 949 950 951 952 953 954 955 956 957 958
static DWORD ui2str(WCHAR *dest, UINT value) {
    static const WCHAR formatW[] = {'%','u',0};
    DWORD ret = 0;

    if(!dest) {
        WCHAR tmp[11];
        ret = sprintfW(tmp, formatW, value);
    } else
        ret = sprintfW(dest, formatW, value);

    return ret;
}

959
/* Converts a h16 component (from an IPv6 address) into its
960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975
 * numerical value.
 *
 * This function assumes that the h16 component has already been validated.
 */
static USHORT h16tous(h16 component) {
    DWORD i;
    USHORT ret = 0;

    for(i = 0; i < component.len; ++i) {
        ret <<= 4;
        ret += hex_to_int(component.str[i]);
    }

    return ret;
}

976
/* Converts an IPv6 address into its 128 bits (16 bytes) numerical value.
977 978 979 980 981 982 983 984 985 986
 *
 * This function assumes that the ipv6_address has already been validated.
 */
static BOOL ipv6_to_number(const ipv6_address *address, USHORT number[8]) {
    DWORD i, cur_component = 0;
    BOOL already_passed_elision = FALSE;

    for(i = 0; i < address->h16_count; ++i) {
        if(address->elision) {
            if(address->components[i].str > address->elision && !already_passed_elision) {
987
                /* Means we just passed the elision and need to add its values to
988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021
                 * 'number' before we do anything else.
                 */
                DWORD j = 0;
                for(j = 0; j < address->elision_size; j+=2)
                    number[cur_component++] = 0;

                already_passed_elision = TRUE;
            }
        }

        number[cur_component++] = h16tous(address->components[i]);
    }

    /* Case when the elision appears after the h16 components. */
    if(!already_passed_elision && address->elision) {
        for(i = 0; i < address->elision_size; i+=2)
            number[cur_component++] = 0;
    }

    if(address->ipv4) {
        UINT value = ipv4toui(address->ipv4, address->ipv4_len);

        if(cur_component != 6) {
            ERR("(%p %p): Failed sanity check with %d\n", address, number, cur_component);
            return FALSE;
        }

        number[cur_component++] = (value >> 16) & 0xffff;
        number[cur_component] = value & 0xffff;
    }

    return TRUE;
}

1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048
/* Checks if the characters pointed to by 'ptr' are
 * a percent encoded data octet.
 *
 * pct-encoded = "%" HEXDIG HEXDIG
 */
static BOOL check_pct_encoded(const WCHAR **ptr) {
    const WCHAR *start = *ptr;

    if(**ptr != '%')
        return FALSE;

    ++(*ptr);
    if(!is_hexdigit(**ptr)) {
        *ptr = start;
        return FALSE;
    }

    ++(*ptr);
    if(!is_hexdigit(**ptr)) {
        *ptr = start;
        return FALSE;
    }

    ++(*ptr);
    return TRUE;
}

1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065
/* dec-octet   = DIGIT                 ; 0-9
 *             / %x31-39 DIGIT         ; 10-99
 *             / "1" 2DIGIT            ; 100-199
 *             / "2" %x30-34 DIGIT     ; 200-249
 *             / "25" %x30-35          ; 250-255
 */
static BOOL check_dec_octet(const WCHAR **ptr) {
    const WCHAR *c1, *c2, *c3;

    c1 = *ptr;
    /* A dec-octet must be at least 1 digit long. */
    if(*c1 < '0' || *c1 > '9')
        return FALSE;

    ++(*ptr);

    c2 = *ptr;
1066
    /* Since the 1-digit requirement was met, it doesn't
1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177
     * matter if this is a DIGIT value, it's considered a
     * dec-octet.
     */
    if(*c2 < '0' || *c2 > '9')
        return TRUE;

    ++(*ptr);

    c3 = *ptr;
    /* Same explanation as above. */
    if(*c3 < '0' || *c3 > '9')
        return TRUE;

    /* Anything > 255 isn't a valid IP dec-octet. */
    if(*c1 >= '2' && *c2 >= '5' && *c3 >= '5') {
        *ptr = c1;
        return FALSE;
    }

    ++(*ptr);
    return TRUE;
}

/* Checks if there is an implicit IPv4 address in the host component of the URI.
 * The max value of an implicit IPv4 address is UINT_MAX.
 *
 *  Ex:
 *      "234567" would be considered an implicit IPv4 address.
 */
static BOOL check_implicit_ipv4(const WCHAR **ptr, UINT *val) {
    const WCHAR *start = *ptr;
    ULONGLONG ret = 0;
    *val = 0;

    while(is_num(**ptr)) {
        ret = ret*10 + (**ptr - '0');

        if(ret > UINT_MAX) {
            *ptr = start;
            return FALSE;
        }
        ++(*ptr);
    }

    if(*ptr == start)
        return FALSE;

    *val = ret;
    return TRUE;
}

/* Checks if the string contains an IPv4 address.
 *
 * This function has a strict mode or a non-strict mode of operation
 * When 'strict' is set to FALSE this function will return TRUE if
 * the string contains at least 'dec-octet "." dec-octet' since partial
 * IPv4 addresses will be normalized out into full IPv4 addresses. When
 * 'strict' is set this function expects there to be a full IPv4 address.
 *
 * IPv4address = dec-octet "." dec-octet "." dec-octet "." dec-octet
 */
static BOOL check_ipv4address(const WCHAR **ptr, BOOL strict) {
    const WCHAR *start = *ptr;

    if(!check_dec_octet(ptr)) {
        *ptr = start;
        return FALSE;
    }

    if(**ptr != '.') {
        *ptr = start;
        return FALSE;
    }

    ++(*ptr);
    if(!check_dec_octet(ptr)) {
        *ptr = start;
        return FALSE;
    }

    if(**ptr != '.') {
        if(strict) {
            *ptr = start;
            return FALSE;
        } else
            return TRUE;
    }

    ++(*ptr);
    if(!check_dec_octet(ptr)) {
        *ptr = start;
        return FALSE;
    }

    if(**ptr != '.') {
        if(strict) {
            *ptr = start;
            return FALSE;
        } else
            return TRUE;
    }

    ++(*ptr);
    if(!check_dec_octet(ptr)) {
        *ptr = start;
        return FALSE;
    }

    /* Found a four digit ip address. */
    return TRUE;
}
1178 1179 1180 1181 1182
/* Tries to parse the scheme name of the URI.
 *
 * scheme = ALPHA *(ALPHA | NUM | '+' | '-' | '.') as defined by RFC 3896.
 * NOTE: Windows accepts a number as the first character of a scheme.
 */
1183
static BOOL parse_scheme_name(const WCHAR **ptr, parse_data *data, DWORD extras) {
1184 1185 1186 1187 1188 1189
    const WCHAR *start = *ptr;

    data->scheme = NULL;
    data->scheme_len = 0;

    while(**ptr) {
1190 1191 1192 1193 1194 1195 1196
        if(**ptr == '*' && *ptr == start) {
            /* Might have found a wildcard scheme. If it is the next
             * char has to be a ':' for it to be a valid URI
             */
            ++(*ptr);
            break;
        } else if(!is_num(**ptr) && !is_alpha(**ptr) && **ptr != '+' &&
1197 1198 1199 1200 1201 1202 1203 1204 1205 1206
           **ptr != '-' && **ptr != '.')
            break;

        (*ptr)++;
    }

    if(*ptr == start)
        return FALSE;

    /* Schemes must end with a ':' */
1207
    if(**ptr != ':' && !((extras & ALLOW_NULL_TERM_SCHEME) && !**ptr)) {
1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218
        *ptr = start;
        return FALSE;
    }

    data->scheme = start;
    data->scheme_len = *ptr - start;

    ++(*ptr);
    return TRUE;
}

1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250
/* Tries to deduce the corresponding URL_SCHEME for the given URI. Stores
 * the deduced URL_SCHEME in data->scheme_type.
 */
static BOOL parse_scheme_type(parse_data *data) {
    /* If there's scheme data then see if it's a recognized scheme. */
    if(data->scheme && data->scheme_len) {
        DWORD i;

        for(i = 0; i < sizeof(recognized_schemes)/sizeof(recognized_schemes[0]); ++i) {
            if(lstrlenW(recognized_schemes[i].scheme_name) == data->scheme_len) {
                /* Has to be a case insensitive compare. */
                if(!StrCmpNIW(recognized_schemes[i].scheme_name, data->scheme, data->scheme_len)) {
                    data->scheme_type = recognized_schemes[i].scheme;
                    return TRUE;
                }
            }
        }

        /* If we get here it means it's not a recognized scheme. */
        data->scheme_type = URL_SCHEME_UNKNOWN;
        return TRUE;
    } else if(data->is_relative) {
        /* Relative URI's have no scheme. */
        data->scheme_type = URL_SCHEME_UNKNOWN;
        return TRUE;
    } else {
        /* Should never reach here! what happened... */
        FIXME("(%p): Unable to determine scheme type for URI %s\n", data, debugstr_w(data->uri));
        return FALSE;
    }
}

1251 1252 1253 1254 1255 1256 1257 1258 1259
/* Tries to parse (or deduce) the scheme_name of a URI. If it can't
 * parse a scheme from the URI it will try to deduce the scheme_name and scheme_type
 * using the flags specified in 'flags' (if any). Flags that affect how this function
 * operates are the Uri_CREATE_ALLOW_* flags.
 *
 * All parsed/deduced information will be stored in 'data' when the function returns.
 *
 * Returns TRUE if it was able to successfully parse the information.
 */
1260
static BOOL parse_scheme(const WCHAR **ptr, parse_data *data, DWORD flags, DWORD extras) {
1261 1262 1263 1264 1265 1266 1267 1268
    static const WCHAR fileW[] = {'f','i','l','e',0};
    static const WCHAR wildcardW[] = {'*',0};

    /* First check to see if the uri could implicitly be a file path. */
    if(is_implicit_file_path(*ptr)) {
        if(flags & Uri_CREATE_ALLOW_IMPLICIT_FILE_SCHEME) {
            data->scheme = fileW;
            data->scheme_len = lstrlenW(fileW);
1269 1270
            data->has_implicit_scheme = TRUE;

1271 1272
            TRACE("(%p %p %x): URI is an implicit file path.\n", ptr, data, flags);
        } else {
1273
            /* Windows does not consider anything that can implicitly be a file
1274 1275 1276 1277 1278 1279
             * path to be a valid URI if the ALLOW_IMPLICIT_FILE_SCHEME flag is not set...
             */
            TRACE("(%p %p %x): URI is implicitly a file path, but, the ALLOW_IMPLICIT_FILE_SCHEME flag wasn't set.\n",
                    ptr, data, flags);
            return FALSE;
        }
1280
    } else if(!parse_scheme_name(ptr, data, extras)) {
1281
        /* No scheme was found, this means it could be:
1282 1283
         *      a) an implicit Wildcard scheme
         *      b) a relative URI
1284
         *      c) an invalid URI.
1285 1286 1287 1288
         */
        if(flags & Uri_CREATE_ALLOW_IMPLICIT_WILDCARD_SCHEME) {
            data->scheme = wildcardW;
            data->scheme_len = lstrlenW(wildcardW);
1289
            data->has_implicit_scheme = TRUE;
1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304

            TRACE("(%p %p %x): URI is an implicit wildcard scheme.\n", ptr, data, flags);
        } else if (flags & Uri_CREATE_ALLOW_RELATIVE) {
            data->is_relative = TRUE;
            TRACE("(%p %p %x): URI is relative.\n", ptr, data, flags);
        } else {
            TRACE("(%p %p %x): Malformed URI found. Unable to deduce scheme name.\n", ptr, data, flags);
            return FALSE;
        }
    }

    if(!data->is_relative)
        TRACE("(%p %p %x): Found scheme=%s scheme_len=%d\n", ptr, data, flags,
                debugstr_wn(data->scheme, data->scheme_len), data->scheme_len);

1305 1306 1307 1308
    if(!parse_scheme_type(data))
        return FALSE;

    TRACE("(%p %p %x): Assigned %d as the URL_SCHEME.\n", ptr, data, flags, data->scheme_type);
1309 1310 1311
    return TRUE;
}

1312
static BOOL parse_username(const WCHAR **ptr, parse_data *data, DWORD flags, DWORD extras) {
1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324
    data->username = *ptr;

    while(**ptr != ':' && **ptr != '@') {
        if(**ptr == '%') {
            if(!check_pct_encoded(ptr)) {
                if(data->scheme_type != URL_SCHEME_UNKNOWN) {
                    *ptr = data->username;
                    data->username = NULL;
                    return FALSE;
                }
            } else
                continue;
1325 1326 1327
        } else if(extras & ALLOW_NULL_TERM_USER_NAME && !**ptr)
            break;
        else if(is_auth_delim(**ptr, data->scheme_type != URL_SCHEME_UNKNOWN)) {
1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339
            *ptr = data->username;
            data->username = NULL;
            return FALSE;
        }

        ++(*ptr);
    }

    data->username_len = *ptr - data->username;
    return TRUE;
}

1340
static BOOL parse_password(const WCHAR **ptr, parse_data *data, DWORD flags, DWORD extras) {
1341 1342 1343 1344 1345 1346
    data->password = *ptr;

    while(**ptr != '@') {
        if(**ptr == '%') {
            if(!check_pct_encoded(ptr)) {
                if(data->scheme_type != URL_SCHEME_UNKNOWN) {
1347
                    *ptr = data->password;
1348 1349 1350 1351 1352
                    data->password = NULL;
                    return FALSE;
                }
            } else
                continue;
1353 1354 1355
        } else if(extras & ALLOW_NULL_TERM_PASSWORD && !**ptr)
            break;
        else if(is_auth_delim(**ptr, data->scheme_type != URL_SCHEME_UNKNOWN)) {
1356
            *ptr = data->password;
1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367
            data->password = NULL;
            return FALSE;
        }

        ++(*ptr);
    }

    data->password_len = *ptr - data->password;
    return TRUE;
}

1368 1369 1370 1371 1372 1373 1374 1375
/* Parses the userinfo part of the URI (if it exists). The userinfo field of
 * a URI can consist of "username:password@", or just "username@".
 *
 * RFC def:
 * userinfo    = *( unreserved / pct-encoded / sub-delims / ":" )
 *
 * NOTES:
 *  1)  If there is more than one ':' in the userinfo part of the URI Windows
1376
 *      uses the first occurrence of ':' to delimit the username and password
1377 1378 1379 1380 1381
 *      components.
 *
 *      ex:
 *          ftp://user:pass:word@winehq.org
 *
1382
 *      would yield "user" as the username and "pass:word" as the password.
1383 1384
 *
 *  2)  Windows allows any character to appear in the "userinfo" part of
1385
 *      a URI, as long as it's not an authority delimiter character set.
1386 1387
 */
static void parse_userinfo(const WCHAR **ptr, parse_data *data, DWORD flags) {
1388
    const WCHAR *start = *ptr;
1389

1390
    if(!parse_username(ptr, data, flags, 0)) {
1391 1392 1393
        TRACE("(%p %p %x): URI contained no userinfo.\n", ptr, data, flags);
        return;
    }
1394

1395 1396 1397 1398 1399 1400 1401 1402 1403
    if(**ptr == ':') {
        ++(*ptr);
        if(!parse_password(ptr, data, flags, 0)) {
            *ptr = start;
            data->username = NULL;
            data->username_len = 0;
            TRACE("(%p %p %x): URI contained no userinfo.\n", ptr, data, flags);
            return;
        }
1404 1405 1406
    }

    if(**ptr != '@') {
1407 1408 1409 1410 1411
        *ptr = start;
        data->username = NULL;
        data->username_len = 0;
        data->password = NULL;
        data->password_len = 0;
1412 1413 1414 1415 1416

        TRACE("(%p %p %x): URI contained no userinfo.\n", ptr, data, flags);
        return;
    }

1417 1418 1419 1420 1421 1422 1423 1424
    if(data->username)
        TRACE("(%p %p %x): Found username %s len=%d.\n", ptr, data, flags,
            debugstr_wn(data->username, data->username_len), data->username_len);

    if(data->password)
        TRACE("(%p %p %x): Found password %s len=%d.\n", ptr, data, flags,
            debugstr_wn(data->password, data->password_len), data->password_len);

1425 1426 1427
    ++(*ptr);
}

1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457
/* Attempts to parse a port from the URI.
 *
 * NOTES:
 *  Windows seems to have a cap on what the maximum value
 *  for a port can be. The max value is USHORT_MAX.
 *
 * port = *DIGIT
 */
static BOOL parse_port(const WCHAR **ptr, parse_data *data, DWORD flags) {
    UINT port = 0;
    data->port = *ptr;

    while(!is_auth_delim(**ptr, data->scheme_type != URL_SCHEME_UNKNOWN)) {
        if(!is_num(**ptr)) {
            *ptr = data->port;
            data->port = NULL;
            return FALSE;
        }

        port = port*10 + (**ptr-'0');

        if(port > USHORT_MAX) {
            *ptr = data->port;
            data->port = NULL;
            return FALSE;
        }

        ++(*ptr);
    }

1458
    data->has_port = TRUE;
1459 1460 1461 1462 1463 1464 1465 1466
    data->port_value = port;
    data->port_len = *ptr - data->port;

    TRACE("(%p %p %x): Found port %s len=%d value=%u\n", ptr, data, flags,
        debugstr_wn(data->port, data->port_len), data->port_len, data->port_value);
    return TRUE;
}

1467 1468 1469
/* Attempts to parse a IPv4 address from the URI.
 *
 * NOTES:
1470
 *  Windows normalizes IPv4 addresses, This means there are three
1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498
 *  possibilities for the URI to contain an IPv4 address.
 *      1)  A well formed address (ex. 192.2.2.2).
 *      2)  A partially formed address. For example "192.0" would
 *          normalize to "192.0.0.0" during canonicalization.
 *      3)  An implicit IPv4 address. For example "256" would
 *          normalize to "0.0.1.0" during canonicalization. Also
 *          note that the maximum value for an implicit IP address
 *          is UINT_MAX, if the value in the URI exceeds this then
 *          it is not considered an IPv4 address.
 */
static BOOL parse_ipv4address(const WCHAR **ptr, parse_data *data, DWORD flags) {
    const BOOL is_unknown = data->scheme_type == URL_SCHEME_UNKNOWN;
    data->host = *ptr;

    if(!check_ipv4address(ptr, FALSE)) {
        if(!check_implicit_ipv4(ptr, &data->implicit_ipv4)) {
            TRACE("(%p %p %x): URI didn't contain anything looking like an IPv4 address.\n",
                ptr, data, flags);
            *ptr = data->host;
            data->host = NULL;
            return FALSE;
        } else
            data->has_implicit_ip = TRUE;
    }

    /* Check if what we found is the only part of the host name (if it isn't
     * we don't have an IPv4 address).
     */
1499 1500 1501 1502 1503 1504 1505 1506
    if(**ptr == ':') {
        ++(*ptr);
        if(!parse_port(ptr, data, flags)) {
            *ptr = data->host;
            data->host = NULL;
            return FALSE;
        }
    } else if(!is_auth_delim(**ptr, !is_unknown)) {
1507
        /* Found more data which belongs to the host, so this isn't an IPv4. */
1508 1509 1510 1511 1512 1513 1514 1515
        *ptr = data->host;
        data->host = NULL;
        data->has_implicit_ip = FALSE;
        return FALSE;
    }

    data->host_len = *ptr - data->host;
    data->host_type = Uri_HOST_IPV4;
1516

1517 1518 1519 1520 1521 1522
    TRACE("(%p %p %x): IPv4 address found. host=%s host_len=%d host_type=%d\n",
        ptr, data, flags, debugstr_wn(data->host, data->host_len),
        data->host_len, data->host_type);
    return TRUE;
}

1523
/* Attempts to parse the reg-name from the URI.
1524 1525 1526
 *
 * Because of the way Windows handles ':' this function also
 * handles parsing the port.
1527 1528 1529 1530 1531
 *
 * reg-name = *( unreserved / pct-encoded / sub-delims )
 *
 * NOTE:
 *  Windows allows everything, but, the characters in "auth_delims" and ':'
1532 1533
 *  to appear in a reg-name, unless it's an unknown scheme type then ':' is
 *  allowed to appear (even if a valid port isn't after it).
1534 1535 1536 1537
 *
 *  Windows doesn't like host names which start with '[' and end with ']'
 *  and don't contain a valid IP literal address in between them.
 *
1538 1539
 *  On Windows if a '[' is encountered in the host name the ':' no longer
 *  counts as a delimiter until you reach the next ']' or an "authority delimiter".
1540 1541 1542
 *
 *  A reg-name CAN be empty.
 */
1543
static BOOL parse_reg_name(const WCHAR **ptr, parse_data *data, DWORD flags, DWORD extras) {
1544 1545
    const BOOL has_start_bracket = **ptr == '[';
    const BOOL known_scheme = data->scheme_type != URL_SCHEME_UNKNOWN;
1546
    const BOOL is_res = data->scheme_type == URL_SCHEME_RES;
1547
    BOOL inside_brackets = has_start_bracket;
1548 1549 1550

    /* res URIs don't have ports. */
    BOOL ignore_col = (extras & IGNORE_PORT_DELIMITER) || is_res;
1551 1552 1553 1554 1555

    /* We have to be careful with file schemes. */
    if(data->scheme_type == URL_SCHEME_FILE) {
        /* This is because an implicit file scheme could be "C:\\test" and it
         * would trick this function into thinking the host is "C", when after
1556 1557
         * canonicalization the host would end up being an empty string. A drive
         * path can also have a '|' instead of a ':' after the drive letter.
1558
         */
1559
        if(is_drive_path(*ptr)) {
1560
            /* Regular old drive paths have no host type (or host name). */
1561 1562 1563 1564
            data->host_type = Uri_HOST_UNKNOWN;
            data->host = *ptr;
            data->host_len = 0;
            return TRUE;
1565
        } else if(is_unc_path(*ptr))
1566 1567 1568 1569 1570 1571
            /* Skip past the "\\" of a UNC path. */
            *ptr += 2;
    }

    data->host = *ptr;

1572 1573 1574 1575 1576
    /* For res URIs, everything before the first '/' is
     * considered the host.
     */
    while((!is_res && !is_auth_delim(**ptr, known_scheme)) ||
          (is_res && **ptr && **ptr != '/')) {
1577
        if(**ptr == ':' && !ignore_col) {
1578
            /* We can ignore ':' if were inside brackets.*/
1579 1580 1581 1582 1583 1584 1585 1586 1587
            if(!inside_brackets) {
                const WCHAR *tmp = (*ptr)++;

                /* Attempt to parse the port. */
                if(!parse_port(ptr, data, flags)) {
                    /* Windows expects there to be a valid port for known scheme types. */
                    if(data->scheme_type != URL_SCHEME_UNKNOWN) {
                        *ptr = data->host;
                        data->host = NULL;
1588
                        TRACE("(%p %p %x %x): Expected valid port\n", ptr, data, flags, extras);
1589 1590 1591
                        return FALSE;
                    } else
                        /* Windows gives up on trying to parse a port when it
1592
                         * encounters an invalid port.
1593 1594 1595 1596 1597 1598 1599
                         */
                        ignore_col = TRUE;
                } else {
                    data->host_len = tmp - data->host;
                    break;
                }
            }
1600
        } else if(**ptr == '%' && (known_scheme && !is_res)) {
1601 1602 1603 1604 1605 1606 1607
            /* Has to be a legit % encoded value. */
            if(!check_pct_encoded(ptr)) {
                *ptr = data->host;
                data->host = NULL;
                return FALSE;
            } else
                continue;
1608 1609 1610 1611
        } else if(is_res && is_forbidden_dos_path_char(**ptr)) {
            *ptr = data->host;
            data->host = NULL;
            return FALSE;
1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622
        } else if(**ptr == ']')
            inside_brackets = FALSE;
        else if(**ptr == '[')
            inside_brackets = TRUE;

        ++(*ptr);
    }

    if(has_start_bracket) {
        /* Make sure the last character of the host wasn't a ']'. */
        if(*(*ptr-1) == ']') {
1623 1624
            TRACE("(%p %p %x %x): Expected an IP literal inside of the host\n",
                ptr, data, flags, extras);
1625 1626 1627 1628 1629 1630
            *ptr = data->host;
            data->host = NULL;
            return FALSE;
        }
    }

1631 1632 1633
    /* Don't overwrite our length if we found a port earlier. */
    if(!data->port)
        data->host_len = *ptr - data->host;
1634 1635

    /* If the host is empty, then it's an unknown host type. */
1636
    if(data->host_len == 0 || is_res)
1637 1638 1639 1640
        data->host_type = Uri_HOST_UNKNOWN;
    else
        data->host_type = Uri_HOST_DNS;

1641
    TRACE("(%p %p %x %x): Parsed reg-name. host=%s len=%d\n", ptr, data, flags, extras,
1642 1643 1644 1645
        debugstr_wn(data->host, data->host_len), data->host_len);
    return TRUE;
}

1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680
/* Attempts to parse an IPv6 address out of the URI.
 *
 * IPv6address =                               6( h16 ":" ) ls32
 *                /                       "::" 5( h16 ":" ) ls32
 *                / [               h16 ] "::" 4( h16 ":" ) ls32
 *                / [ *1( h16 ":" ) h16 ] "::" 3( h16 ":" ) ls32
 *                / [ *2( h16 ":" ) h16 ] "::" 2( h16 ":" ) ls32
 *                / [ *3( h16 ":" ) h16 ] "::"    h16 ":"   ls32
 *                / [ *4( h16 ":" ) h16 ] "::"              ls32
 *                / [ *5( h16 ":" ) h16 ] "::"              h16
 *                / [ *6( h16 ":" ) h16 ] "::"
 *
 * ls32        = ( h16 ":" h16 ) / IPv4address
 *             ; least-significant 32 bits of address.
 *
 * h16         = 1*4HEXDIG
 *             ; 16 bits of address represented in hexadecimal.
 *
 * Modeled after google-url's 'DoParseIPv6' function.
 */
static BOOL parse_ipv6address(const WCHAR **ptr, parse_data *data, DWORD flags) {
    const WCHAR *start, *cur_start;
    ipv6_address ip;

    start = cur_start = *ptr;
    memset(&ip, 0, sizeof(ipv6_address));

    for(;; ++(*ptr)) {
        /* Check if we're on the last character of the host. */
        BOOL is_end = (is_auth_delim(**ptr, data->scheme_type != URL_SCHEME_UNKNOWN)
                        || **ptr == ']');

        BOOL is_split = (**ptr == ':');
        BOOL is_elision = (is_split && !is_end && *(*ptr+1) == ':');

1681
        /* Check if we're at the end of a component, or
1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705
         * if we're at the end of the IPv6 address.
         */
        if(is_split || is_end) {
            DWORD cur_len = 0;

            cur_len = *ptr - cur_start;

            /* h16 can't have a length > 4. */
            if(cur_len > 4) {
                *ptr = start;

                TRACE("(%p %p %x): h16 component to long.\n",
                    ptr, data, flags);
                return FALSE;
            }

            if(cur_len == 0) {
                /* An h16 component can't have the length of 0 unless
                 * the elision is at the beginning of the address, or
                 * at the end of the address.
                 */
                if(!((*ptr == start && is_elision) ||
                    (is_end && (*ptr-2) == ip.elision))) {
                    *ptr = start;
1706
                    TRACE("(%p %p %x): IPv6 component cannot have a length of 0.\n",
1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783
                        ptr, data, flags);
                    return FALSE;
                }
            }

            if(cur_len > 0) {
                /* An IPv6 address can have no more than 8 h16 components. */
                if(ip.h16_count >= 8) {
                    *ptr = start;
                    TRACE("(%p %p %x): Not a IPv6 address, to many h16 components.\n",
                        ptr, data, flags);
                    return FALSE;
                }

                ip.components[ip.h16_count].str = cur_start;
                ip.components[ip.h16_count].len = cur_len;

                TRACE("(%p %p %x): Found h16 component %s, len=%d, h16_count=%d\n",
                    ptr, data, flags, debugstr_wn(cur_start, cur_len), cur_len,
                    ip.h16_count);
                ++ip.h16_count;
            }
        }

        if(is_end)
            break;

        if(is_elision) {
            /* A IPv6 address can only have 1 elision ('::'). */
            if(ip.elision) {
                *ptr = start;

                TRACE("(%p %p %x): IPv6 address cannot have 2 elisions.\n",
                    ptr, data, flags);
                return FALSE;
            }

            ip.elision = *ptr;
            ++(*ptr);
        }

        if(is_split)
            cur_start = *ptr+1;
        else {
            if(!check_ipv4address(ptr, TRUE)) {
                if(!is_hexdigit(**ptr)) {
                    /* Not a valid character for an IPv6 address. */
                    *ptr = start;
                    return FALSE;
                }
            } else {
                /* Found an IPv4 address. */
                ip.ipv4 = cur_start;
                ip.ipv4_len = *ptr - cur_start;

                TRACE("(%p %p %x): Found an attached IPv4 address %s len=%d.\n",
                    ptr, data, flags, debugstr_wn(ip.ipv4, ip.ipv4_len),
                    ip.ipv4_len);

                /* IPv4 addresses can only appear at the end of a IPv6. */
                break;
            }
        }
    }

    compute_ipv6_comps_size(&ip);

    /* Make sure the IPv6 address adds up to 16 bytes. */
    if(ip.components_size + ip.elision_size != 16) {
        *ptr = start;
        TRACE("(%p %p %x): Invalid IPv6 address, did not add up to 16 bytes.\n",
            ptr, data, flags);
        return FALSE;
    }

    if(ip.elision_size == 2) {
        /* For some reason on Windows if an elision that represents
1784
         * only one h16 component is encountered at the very begin or
1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805
         * end of an IPv6 address, Windows does not consider it a
         * valid IPv6 address.
         *
         *  Ex: [::2:3:4:5:6:7] is not valid, even though the sum
         *      of all the components == 128bits.
         */
         if(ip.elision < ip.components[0].str ||
            ip.elision > ip.components[ip.h16_count-1].str) {
            *ptr = start;
            TRACE("(%p %p %x): Invalid IPv6 address. Detected elision of 2 bytes at the beginning or end of the address.\n",
                ptr, data, flags);
            return FALSE;
        }
    }

    data->host_type = Uri_HOST_IPV6;
    data->has_ipv6 = TRUE;
    data->ipv6_address = ip;

    TRACE("(%p %p %x): Found valid IPv6 literal %s len=%d\n",
        ptr, data, flags, debugstr_wn(start, *ptr-start),
1806
        (int)(*ptr-start));
1807 1808 1809
    return TRUE;
}

1810 1811 1812 1813 1814 1815 1816 1817
/*  IPvFuture  = "v" 1*HEXDIG "." 1*( unreserved / sub-delims / ":" ) */
static BOOL parse_ipvfuture(const WCHAR **ptr, parse_data *data, DWORD flags) {
    const WCHAR *start = *ptr;

    /* IPvFuture has to start with a 'v' or 'V'. */
    if(**ptr != 'v' && **ptr != 'V')
        return FALSE;

1818
    /* Following the v there must be at least 1 hex digit. */
1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847
    ++(*ptr);
    if(!is_hexdigit(**ptr)) {
        *ptr = start;
        return FALSE;
    }

    ++(*ptr);
    while(is_hexdigit(**ptr))
        ++(*ptr);

    /* End of the hexdigit sequence must be a '.' */
    if(**ptr != '.') {
        *ptr = start;
        return FALSE;
    }

    ++(*ptr);
    if(!is_unreserved(**ptr) && !is_subdelim(**ptr) && **ptr != ':') {
        *ptr = start;
        return FALSE;
    }

    ++(*ptr);
    while(is_unreserved(**ptr) || is_subdelim(**ptr) || **ptr == ':')
        ++(*ptr);

    data->host_type = Uri_HOST_UNKNOWN;

    TRACE("(%p %p %x): Parsed IPvFuture address %s len=%d\n", ptr, data, flags,
1848
          debugstr_wn(start, *ptr-start), (int)(*ptr-start));
1849 1850 1851 1852

    return TRUE;
}

1853
/* IP-literal = "[" ( IPv6address / IPvFuture  ) "]" */
1854
static BOOL parse_ip_literal(const WCHAR **ptr, parse_data *data, DWORD flags, DWORD extras) {
1855 1856
    data->host = *ptr;

1857
    if(**ptr != '[' && !(extras & ALLOW_BRACKETLESS_IP_LITERAL)) {
1858 1859
        data->host = NULL;
        return FALSE;
1860 1861
    } else if(**ptr == '[')
        ++(*ptr);
1862 1863

    if(!parse_ipv6address(ptr, data, flags)) {
1864
        if(extras & SKIP_IP_FUTURE_CHECK || !parse_ipvfuture(ptr, data, flags)) {
1865 1866 1867 1868
            *ptr = data->host;
            data->host = NULL;
            return FALSE;
        }
1869 1870
    }

1871
    if(**ptr != ']' && !(extras & ALLOW_BRACKETLESS_IP_LITERAL)) {
1872 1873 1874
        *ptr = data->host;
        data->host = NULL;
        return FALSE;
1875 1876 1877 1878 1879 1880
    } else if(!**ptr && extras & ALLOW_BRACKETLESS_IP_LITERAL) {
        /* The IP literal didn't contain brackets and was followed by
         * a NULL terminator, so no reason to even check the port.
         */
        data->host_len = *ptr - data->host;
        return TRUE;
1881 1882 1883
    }

    ++(*ptr);
1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895
    if(**ptr == ':') {
        ++(*ptr);
        /* If a valid port is not found, then let it trickle down to
         * parse_reg_name.
         */
        if(!parse_port(ptr, data, flags)) {
            *ptr = data->host;
            data->host = NULL;
            return FALSE;
        }
    } else
        data->host_len = *ptr - data->host;
1896 1897 1898 1899

    return TRUE;
}

1900 1901 1902 1903
/* Parses the host information from the URI.
 *
 * host = IP-literal / IPv4address / reg-name
 */
1904 1905
static BOOL parse_host(const WCHAR **ptr, parse_data *data, DWORD flags, DWORD extras) {
    if(!parse_ip_literal(ptr, data, flags, extras)) {
1906
        if(!parse_ipv4address(ptr, data, flags)) {
1907 1908 1909
            if(!parse_reg_name(ptr, data, flags, extras)) {
                TRACE("(%p %p %x %x): Malformed URI, Unknown host type.\n",
                    ptr, data, flags, extras);
1910 1911
                return FALSE;
            }
1912
        }
1913 1914 1915 1916 1917
    }

    return TRUE;
}

1918 1919 1920 1921 1922 1923 1924
/* Parses the authority information from the URI.
 *
 * authority   = [ userinfo "@" ] host [ ":" port ]
 */
static BOOL parse_authority(const WCHAR **ptr, parse_data *data, DWORD flags) {
    parse_userinfo(ptr, data, flags);

1925 1926 1927
    /* Parsing the port will happen during one of the host parsing
     * routines (if the URI has a port).
     */
1928
    if(!parse_host(ptr, data, flags, 0))
1929
        return FALSE;
1930 1931 1932 1933

    return TRUE;
}

1934 1935 1936 1937
/* Attempts to parse the path information of a hierarchical URI. */
static BOOL parse_path_hierarchical(const WCHAR **ptr, parse_data *data, DWORD flags) {
    const WCHAR *start = *ptr;
    static const WCHAR slash[] = {'/',0};
1938
    const BOOL is_file = data->scheme_type == URL_SCHEME_FILE;
1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953

    if(is_path_delim(**ptr)) {
        if(data->scheme_type == URL_SCHEME_WILDCARD) {
            /* Wildcard schemes don't get a '/' attached if their path is
             * empty.
             */
            data->path = NULL;
            data->path_len = 0;
        } else if(!(flags & Uri_CREATE_NO_CANONICALIZE)) {
            /* If the path component is empty, then a '/' is added. */
            data->path = slash;
            data->path_len = 1;
        }
    } else {
        while(!is_path_delim(**ptr)) {
1954
            if(**ptr == '%' && data->scheme_type != URL_SCHEME_UNKNOWN && !is_file) {
1955 1956 1957 1958 1959
                if(!check_pct_encoded(ptr)) {
                    *ptr = start;
                    return FALSE;
                } else
                    continue;
1960
            } else if(is_forbidden_dos_path_char(**ptr) && is_file &&
1961 1962 1963 1964 1965 1966
                      (flags & Uri_CREATE_FILE_USE_DOS_PATH)) {
                /* File schemes with USE_DOS_PATH set aren't allowed to have
                 * a '<' or '>' or '\"' appear in them.
                 */
                *ptr = start;
                return FALSE;
1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
            } else if(**ptr == '\\') {
                /* Not allowed to have a backslash if NO_CANONICALIZE is set
                 * and the scheme is known type (but not a file scheme).
                 */
                if(flags & Uri_CREATE_NO_CANONICALIZE) {
                    if(data->scheme_type != URL_SCHEME_FILE &&
                       data->scheme_type != URL_SCHEME_UNKNOWN) {
                        *ptr = start;
                        return FALSE;
                    }
                }
            }

            ++(*ptr);
        }

        /* The only time a URI doesn't have a path is when
         * the NO_CANONICALIZE flag is set and the raw URI
         * didn't contain one.
         */
        if(*ptr == start) {
            data->path = NULL;
            data->path_len = 0;
        } else {
            data->path = start;
            data->path_len = *ptr - start;
        }
    }

    if(data->path)
        TRACE("(%p %p %x): Parsed path %s len=%d\n", ptr, data, flags,
            debugstr_wn(data->path, data->path_len), data->path_len);
    else
        TRACE("(%p %p %x): The URI contained no path\n", ptr, data, flags);

    return TRUE;
}

2005
/* Parses the path of an opaque URI (much less strict then the parser
2006 2007 2008 2009 2010
 * for a hierarchical URI).
 *
 * NOTE:
 *  Windows allows invalid % encoded data to appear in opaque URI paths
 *  for unknown scheme types.
2011 2012 2013
 *
 *  File schemes with USE_DOS_PATH set aren't allowed to have '<', '>', or '\"'
 *  appear in them.
2014 2015 2016
 */
static BOOL parse_path_opaque(const WCHAR **ptr, parse_data *data, DWORD flags) {
    const BOOL known_scheme = data->scheme_type != URL_SCHEME_UNKNOWN;
2017
    const BOOL is_file = data->scheme_type == URL_SCHEME_FILE;
2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028

    data->path = *ptr;

    while(!is_path_delim(**ptr)) {
        if(**ptr == '%' && known_scheme) {
            if(!check_pct_encoded(ptr)) {
                *ptr = data->path;
                data->path = NULL;
                return FALSE;
            } else
                continue;
2029 2030 2031 2032 2033
        } else if(is_forbidden_dos_path_char(**ptr) && is_file &&
                  (flags & Uri_CREATE_FILE_USE_DOS_PATH)) {
            *ptr = data->path;
            data->path = NULL;
            return FALSE;
2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044
        }

        ++(*ptr);
    }

    data->path_len = *ptr - data->path;
    TRACE("(%p %p %x): Parsed opaque URI path %s len=%d\n", ptr, data, flags,
        debugstr_wn(data->path, data->path_len), data->path_len);
    return TRUE;
}

2045 2046
/* Determines how the URI should be parsed after the scheme information.
 *
2047
 * If the scheme is followed by "//", then it is treated as a hierarchical URI
2048 2049 2050 2051
 * which then the authority and path information will be parsed out. Otherwise, the
 * URI will be treated as an opaque URI which the authority information is not parsed
 * out.
 *
2052
 * RFC 3896 definition of hier-part:
2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064
 *
 * hier-part   = "//" authority path-abempty
 *                 / path-absolute
 *                 / path-rootless
 *                 / path-empty
 *
 * MSDN opaque URI definition:
 *  scheme ":" path [ "#" fragment ]
 *
 * NOTES:
 *  If the URI is of an unknown scheme type and has a "//" following the scheme then it
 *  is treated as a hierarchical URI, but, if the CREATE_NO_CRACK_UNKNOWN_SCHEMES flag is
2065
 *  set then it is considered an opaque URI regardless of what follows the scheme information
2066 2067 2068
 *  (per MSDN documentation).
 */
static BOOL parse_hierpart(const WCHAR **ptr, parse_data *data, DWORD flags) {
2069 2070
    const WCHAR *start = *ptr;

2071 2072 2073 2074 2075 2076 2077 2078 2079
    /* For javascript: URIs, simply set everything as a path */
    if(data->scheme_type == URL_SCHEME_JAVASCRIPT) {
        data->path = *ptr;
        data->path_len = strlenW(*ptr);
        data->is_opaque = TRUE;
        *ptr += data->path_len;
        return TRUE;
    }

2080 2081
    /* Checks if the authority information needs to be parsed. */
    if(is_hierarchical_uri(ptr, data)) {
2082 2083 2084 2085 2086 2087 2088 2089 2090
        /* Only treat it as a hierarchical URI if the scheme_type is known or
         * the Uri_CREATE_NO_CRACK_UNKNOWN_SCHEMES flag is not set.
         */
        if(data->scheme_type != URL_SCHEME_UNKNOWN ||
           !(flags & Uri_CREATE_NO_CRACK_UNKNOWN_SCHEMES)) {
            TRACE("(%p %p %x): Treating URI as an hierarchical URI.\n", ptr, data, flags);
            data->is_opaque = FALSE;

            /* TODO: Handle hierarchical URI's, parse authority then parse the path. */
2091 2092 2093
            if(!parse_authority(ptr, data, flags))
                return FALSE;

2094
            return parse_path_hierarchical(ptr, data, flags);
2095
        } else
2096
            /* Reset ptr to its starting position so opaque path parsing
2097 2098 2099
             * begins at the correct location.
             */
            *ptr = start;
2100 2101 2102 2103 2104 2105 2106 2107 2108
    }

    /* If it reaches here, then the URI will be treated as an opaque
     * URI.
     */

    TRACE("(%p %p %x): Treating URI as an opaque URI.\n", ptr, data, flags);

    data->is_opaque = TRUE;
2109 2110 2111
    if(!parse_path_opaque(ptr, data, flags))
        return FALSE;

2112 2113 2114
    return TRUE;
}

2115 2116 2117 2118
/* Attempts to parse the query string from the URI.
 *
 * NOTES:
 *  If NO_DECODE_EXTRA_INFO flag is set, then invalid percent encoded
2119 2120
 *  data is allowed to appear in the query string. For unknown scheme types
 *  invalid percent encoded data is allowed to appear regardless.
2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153
 */
static BOOL parse_query(const WCHAR **ptr, parse_data *data, DWORD flags) {
    const BOOL known_scheme = data->scheme_type != URL_SCHEME_UNKNOWN;

    if(**ptr != '?') {
        TRACE("(%p %p %x): URI didn't contain a query string.\n", ptr, data, flags);
        return TRUE;
    }

    data->query = *ptr;

    ++(*ptr);
    while(**ptr && **ptr != '#') {
        if(**ptr == '%' && known_scheme &&
           !(flags & Uri_CREATE_NO_DECODE_EXTRA_INFO)) {
            if(!check_pct_encoded(ptr)) {
                *ptr = data->query;
                data->query = NULL;
                return FALSE;
            } else
                continue;
        }

        ++(*ptr);
    }

    data->query_len = *ptr - data->query;

    TRACE("(%p %p %x): Parsed query string %s len=%d\n", ptr, data, flags,
        debugstr_wn(data->query, data->query_len), data->query_len);
    return TRUE;
}

2154 2155 2156 2157
/* Attempts to parse the fragment from the URI.
 *
 * NOTES:
 *  If NO_DECODE_EXTRA_INFO flag is set, then invalid percent encoded
2158 2159
 *  data is allowed to appear in the query string. For unknown scheme types
 *  invalid percent encoded data is allowed to appear regardless.
2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192
 */
static BOOL parse_fragment(const WCHAR **ptr, parse_data *data, DWORD flags) {
    const BOOL known_scheme = data->scheme_type != URL_SCHEME_UNKNOWN;

    if(**ptr != '#') {
        TRACE("(%p %p %x): URI didn't contain a fragment.\n", ptr, data, flags);
        return TRUE;
    }

    data->fragment = *ptr;

    ++(*ptr);
    while(**ptr) {
        if(**ptr == '%' && known_scheme &&
           !(flags & Uri_CREATE_NO_DECODE_EXTRA_INFO)) {
            if(!check_pct_encoded(ptr)) {
                *ptr = data->fragment;
                data->fragment = NULL;
                return FALSE;
            } else
                continue;
        }

        ++(*ptr);
    }

    data->fragment_len = *ptr - data->fragment;

    TRACE("(%p %p %x): Parsed fragment %s len=%d\n", ptr, data, flags,
        debugstr_wn(data->fragment, data->fragment_len), data->fragment_len);
    return TRUE;
}

2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206
/* Parses and validates the components of the specified by data->uri
 * and stores the information it parses into 'data'.
 *
 * Returns TRUE if it successfully parsed the URI. False otherwise.
 */
static BOOL parse_uri(parse_data *data, DWORD flags) {
    const WCHAR *ptr;
    const WCHAR **pptr;

    ptr = data->uri;
    pptr = &ptr;

    TRACE("(%p %x): BEGINNING TO PARSE URI %s.\n", data, flags, debugstr_w(data->uri));

2207
    if(!parse_scheme(pptr, data, flags, 0))
2208 2209
        return FALSE;

2210 2211 2212
    if(!parse_hierpart(pptr, data, flags))
        return FALSE;

2213 2214 2215
    if(!parse_query(pptr, data, flags))
        return FALSE;

2216 2217
    if(!parse_fragment(pptr, data, flags))
        return FALSE;
2218

2219 2220 2221 2222
    TRACE("(%p %x): FINISHED PARSING URI.\n", data, flags);
    return TRUE;
}

2223 2224
static BOOL canonicalize_username(const parse_data *data, Uri *uri, DWORD flags, BOOL computeOnly) {
    const WCHAR *ptr;
2225

2226 2227
    if(!data->username) {
        uri->userinfo_start = -1;
2228
        return TRUE;
2229
    }
2230 2231

    uri->userinfo_start = uri->canon_len;
2232 2233
    for(ptr = data->username; ptr < data->username+data->username_len; ++ptr) {
        if(*ptr == '%') {
2234 2235
            /* Only decode % encoded values for known scheme types. */
            if(data->scheme_type != URL_SCHEME_UNKNOWN) {
2236
                /* See if the value really needs decoding. */
2237
                WCHAR val = decode_pct_val(ptr);
2238 2239 2240 2241 2242 2243 2244
                if(is_unreserved(val)) {
                    if(!computeOnly)
                        uri->canon_uri[uri->canon_len] = val;

                    ++uri->canon_len;

                    /* Move pass the hex characters. */
2245
                    ptr += 2;
2246 2247 2248
                    continue;
                }
            }
2249
        } else if(!is_reserved(*ptr) && !is_unreserved(*ptr) && *ptr != '\\') {
2250 2251 2252 2253 2254
            /* Only percent encode forbidden characters if the NO_ENCODE_FORBIDDEN_CHARACTERS flag
             * is NOT set.
             */
            if(!(flags & Uri_CREATE_NO_ENCODE_FORBIDDEN_CHARACTERS)) {
                if(!computeOnly)
2255
                    pct_encode_val(*ptr, uri->canon_uri + uri->canon_len);
2256 2257 2258 2259 2260 2261 2262 2263

                uri->canon_len += 3;
                continue;
            }
        }

        if(!computeOnly)
            /* Nothing special, so just copy the character over. */
2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281
            uri->canon_uri[uri->canon_len] = *ptr;
        ++uri->canon_len;
    }

    return TRUE;
}

static BOOL canonicalize_password(const parse_data *data, Uri *uri, DWORD flags, BOOL computeOnly) {
    const WCHAR *ptr;

    if(!data->password) {
        uri->userinfo_split = -1;
        return TRUE;
    }

    if(uri->userinfo_start == -1)
        /* Has a password, but, doesn't have a username. */
        uri->userinfo_start = uri->canon_len;
2282

2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293
    uri->userinfo_split = uri->canon_len - uri->userinfo_start;

    /* Add the ':' to the userinfo component. */
    if(!computeOnly)
        uri->canon_uri[uri->canon_len] = ':';
    ++uri->canon_len;

    for(ptr = data->password; ptr < data->password+data->password_len; ++ptr) {
        if(*ptr == '%') {
            /* Only decode % encoded values for known scheme types. */
            if(data->scheme_type != URL_SCHEME_UNKNOWN) {
2294
                /* See if the value really needs decoding. */
2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322
                WCHAR val = decode_pct_val(ptr);
                if(is_unreserved(val)) {
                    if(!computeOnly)
                        uri->canon_uri[uri->canon_len] = val;

                    ++uri->canon_len;

                    /* Move pass the hex characters. */
                    ptr += 2;
                    continue;
                }
            }
        } else if(!is_reserved(*ptr) && !is_unreserved(*ptr) && *ptr != '\\') {
            /* Only percent encode forbidden characters if the NO_ENCODE_FORBIDDEN_CHARACTERS flag
             * is NOT set.
             */
            if(!(flags & Uri_CREATE_NO_ENCODE_FORBIDDEN_CHARACTERS)) {
                if(!computeOnly)
                    pct_encode_val(*ptr, uri->canon_uri + uri->canon_len);

                uri->canon_len += 3;
                continue;
            }
        }

        if(!computeOnly)
            /* Nothing special, so just copy the character over. */
            uri->canon_uri[uri->canon_len] = *ptr;
2323 2324 2325
        ++uri->canon_len;
    }

2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350
    return TRUE;
}

/* Canonicalizes the userinfo of the URI represented by the parse_data.
 *
 * Canonicalization of the userinfo is a simple process. If there are any percent
 * encoded characters that fall in the "unreserved" character set, they are decoded
 * to their actual value. If a character is not in the "unreserved" or "reserved" sets
 * then it is percent encoded. Other than that the characters are copied over without
 * change.
 */
static BOOL canonicalize_userinfo(const parse_data *data, Uri *uri, DWORD flags, BOOL computeOnly) {
    uri->userinfo_start = uri->userinfo_split = -1;
    uri->userinfo_len = 0;

    if(!data->username && !data->password)
        /* URI doesn't have userinfo, so nothing to do here. */
        return TRUE;

    if(!canonicalize_username(data, uri, flags, computeOnly))
        return FALSE;

    if(!canonicalize_password(data, uri, flags, computeOnly))
        return FALSE;

2351 2352 2353 2354 2355 2356 2357 2358 2359 2360
    uri->userinfo_len = uri->canon_len - uri->userinfo_start;
    if(!computeOnly)
        TRACE("(%p %p %x %d): Canonicalized userinfo, userinfo_start=%d, userinfo=%s, userinfo_split=%d userinfo_len=%d.\n",
                data, uri, flags, computeOnly, uri->userinfo_start, debugstr_wn(uri->canon_uri + uri->userinfo_start, uri->userinfo_len),
                uri->userinfo_split, uri->userinfo_len);

    /* Now insert the '@' after the userinfo. */
    if(!computeOnly)
        uri->canon_uri[uri->canon_len] = '@';
    ++uri->canon_len;
2361

2362 2363 2364
    return TRUE;
}

2365 2366 2367 2368 2369
/* Attempts to canonicalize a reg_name.
 *
 * Things that happen:
 *  1)  If Uri_CREATE_NO_CANONICALIZE flag is not set, then the reg_name is
 *      lower cased. Unless it's an unknown scheme type, which case it's
2370
 *      no lower cased regardless.
2371 2372 2373 2374 2375 2376 2377 2378 2379
 *
 *  2)  Unreserved % encoded characters are decoded for known
 *      scheme types.
 *
 *  3)  Forbidden characters are % encoded as long as
 *      Uri_CREATE_NO_ENCODE_FORBIDDEN_CHARACTERS flag is not set and
 *      it isn't an unknown scheme type.
 *
 *  4)  If it's a file scheme and the host is "localhost" it's removed.
2380 2381 2382
 *
 *  5)  If it's a file scheme and Uri_CREATE_FILE_USE_DOS_PATH is set,
 *      then the UNC path characters are added before the host name.
2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400
 */
static BOOL canonicalize_reg_name(const parse_data *data, Uri *uri,
                                  DWORD flags, BOOL computeOnly) {
    static const WCHAR localhostW[] =
            {'l','o','c','a','l','h','o','s','t',0};
    const WCHAR *ptr;
    const BOOL known_scheme = data->scheme_type != URL_SCHEME_UNKNOWN;

    if(data->scheme_type == URL_SCHEME_FILE &&
       data->host_len == lstrlenW(localhostW)) {
        if(!StrCmpNIW(data->host, localhostW, data->host_len)) {
            uri->host_start = -1;
            uri->host_len = 0;
            uri->host_type = Uri_HOST_UNKNOWN;
            return TRUE;
        }
    }

2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411
    if(data->scheme_type == URL_SCHEME_FILE && flags & Uri_CREATE_FILE_USE_DOS_PATH) {
        if(!computeOnly) {
            uri->canon_uri[uri->canon_len] = '\\';
            uri->canon_uri[uri->canon_len+1] = '\\';
        }
        uri->canon_len += 2;
        uri->authority_start = uri->canon_len;
    }

    uri->host_start = uri->canon_len;

2412 2413 2414 2415
    for(ptr = data->host; ptr < data->host+data->host_len; ++ptr) {
        if(*ptr == '%' && known_scheme) {
            WCHAR val = decode_pct_val(ptr);
            if(is_unreserved(val)) {
2416
                /* If NO_CANONICALIZE is not set, then windows lower cases the
2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473
                 * decoded value.
                 */
                if(!(flags & Uri_CREATE_NO_CANONICALIZE) && isupperW(val)) {
                    if(!computeOnly)
                        uri->canon_uri[uri->canon_len] = tolowerW(val);
                } else {
                    if(!computeOnly)
                        uri->canon_uri[uri->canon_len] = val;
                }
                ++uri->canon_len;

                /* Skip past the % encoded character. */
                ptr += 2;
                continue;
            } else {
                /* Just copy the % over. */
                if(!computeOnly)
                    uri->canon_uri[uri->canon_len] = *ptr;
                ++uri->canon_len;
            }
        } else if(*ptr == '\\') {
            /* Only unknown scheme types could have made it here with a '\\' in the host name. */
            if(!computeOnly)
                uri->canon_uri[uri->canon_len] = *ptr;
            ++uri->canon_len;
        } else if(!(flags & Uri_CREATE_NO_ENCODE_FORBIDDEN_CHARACTERS) &&
                  !is_unreserved(*ptr) && !is_reserved(*ptr) && known_scheme) {
            if(!computeOnly) {
                pct_encode_val(*ptr, uri->canon_uri+uri->canon_len);

                /* The percent encoded value gets lower cased also. */
                if(!(flags & Uri_CREATE_NO_CANONICALIZE)) {
                    uri->canon_uri[uri->canon_len+1] = tolowerW(uri->canon_uri[uri->canon_len+1]);
                    uri->canon_uri[uri->canon_len+2] = tolowerW(uri->canon_uri[uri->canon_len+2]);
                }
            }

            uri->canon_len += 3;
        } else {
            if(!computeOnly) {
                if(!(flags & Uri_CREATE_NO_CANONICALIZE) && known_scheme)
                    uri->canon_uri[uri->canon_len] = tolowerW(*ptr);
                else
                    uri->canon_uri[uri->canon_len] = *ptr;
            }

            ++uri->canon_len;
        }
    }

    uri->host_len = uri->canon_len - uri->host_start;

    if(!computeOnly)
        TRACE("(%p %p %x %d): Canonicalize reg_name=%s len=%d\n", data, uri, flags,
            computeOnly, debugstr_wn(uri->canon_uri+uri->host_start, uri->host_len),
            uri->host_len);

2474 2475 2476 2477
    if(!computeOnly)
        find_domain_name(uri->canon_uri+uri->host_start, uri->host_len,
            &(uri->domain_offset));

2478 2479 2480
    return TRUE;
}

2481 2482 2483 2484 2485
/* Attempts to canonicalize an implicit IPv4 address. */
static BOOL canonicalize_implicit_ipv4address(const parse_data *data, Uri *uri, DWORD flags, BOOL computeOnly) {
    uri->host_start = uri->canon_len;

    TRACE("%u\n", data->implicit_ipv4);
2486 2487
    /* For unknown scheme types Windows doesn't convert
     * the value into an IP address, but it still considers
2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517
     * it an IPv4 address.
     */
    if(data->scheme_type == URL_SCHEME_UNKNOWN) {
        if(!computeOnly)
            memcpy(uri->canon_uri+uri->canon_len, data->host, data->host_len*sizeof(WCHAR));
        uri->canon_len += data->host_len;
    } else {
        if(!computeOnly)
            uri->canon_len += ui2ipv4(uri->canon_uri+uri->canon_len, data->implicit_ipv4);
        else
            uri->canon_len += ui2ipv4(NULL, data->implicit_ipv4);
    }

    uri->host_len = uri->canon_len - uri->host_start;
    uri->host_type = Uri_HOST_IPV4;

    if(!computeOnly)
        TRACE("%p %p %x %d): Canonicalized implicit IP address=%s len=%d\n",
            data, uri, flags, computeOnly,
            debugstr_wn(uri->canon_uri+uri->host_start, uri->host_len),
            uri->host_len);

    return TRUE;
}

/* Attempts to canonicalize an IPv4 address.
 *
 * If the parse_data represents a URI that has an implicit IPv4 address
 * (ex. http://256/, this function will convert 256 into 0.0.1.0). If
 * the implicit IP address exceeds the value of UINT_MAX (maximum value
2518
 * for an IPv4 address) it's canonicalized as if it were a reg-name.
2519 2520 2521 2522 2523 2524 2525
 *
 * If the parse_data contains a partial or full IPv4 address it normalizes it.
 * A partial IPv4 address is something like "192.0" and would be normalized to
 * "192.0.0.0". With a full (or partial) IPv4 address like "192.002.01.003" would
 * be normalized to "192.2.1.3".
 *
 * NOTES:
2526
 *  Windows ONLY normalizes IPv4 address for known scheme types (one that isn't
2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604
 *  URL_SCHEME_UNKNOWN). For unknown scheme types, it simply copies the data from
 *  the original URI into the canonicalized URI, but, it still recognizes URI's
 *  host type as HOST_IPV4.
 */
static BOOL canonicalize_ipv4address(const parse_data *data, Uri *uri, DWORD flags, BOOL computeOnly) {
    if(data->has_implicit_ip)
        return canonicalize_implicit_ipv4address(data, uri, flags, computeOnly);
    else {
        uri->host_start = uri->canon_len;

        /* Windows only normalizes for known scheme types. */
        if(data->scheme_type != URL_SCHEME_UNKNOWN) {
            /* parse_data contains a partial or full IPv4 address, so normalize it. */
            DWORD i, octetDigitCount = 0, octetCount = 0;
            BOOL octetHasDigit = FALSE;

            for(i = 0; i < data->host_len; ++i) {
                if(data->host[i] == '0' && !octetHasDigit) {
                    /* Can ignore leading zeros if:
                     *  1) It isn't the last digit of the octet.
                     *  2) i+1 != data->host_len
                     *  3) i+1 != '.'
                     */
                    if(octetDigitCount == 2 ||
                       i+1 == data->host_len ||
                       data->host[i+1] == '.') {
                        if(!computeOnly)
                            uri->canon_uri[uri->canon_len] = data->host[i];
                        ++uri->canon_len;
                        TRACE("Adding zero\n");
                    }
                } else if(data->host[i] == '.') {
                    if(!computeOnly)
                        uri->canon_uri[uri->canon_len] = data->host[i];
                    ++uri->canon_len;

                    octetDigitCount = 0;
                    octetHasDigit = FALSE;
                    ++octetCount;
                } else {
                    if(!computeOnly)
                        uri->canon_uri[uri->canon_len] = data->host[i];
                    ++uri->canon_len;

                    ++octetDigitCount;
                    octetHasDigit = TRUE;
                }
            }

            /* Make sure the canonicalized IP address has 4 dec-octets.
             * If doesn't add "0" ones until there is 4;
             */
            for( ; octetCount < 3; ++octetCount) {
                if(!computeOnly) {
                    uri->canon_uri[uri->canon_len] = '.';
                    uri->canon_uri[uri->canon_len+1] = '0';
                }

                uri->canon_len += 2;
            }
        } else {
            /* Windows doesn't normalize addresses in unknown schemes. */
            if(!computeOnly)
                memcpy(uri->canon_uri+uri->canon_len, data->host, data->host_len*sizeof(WCHAR));
            uri->canon_len += data->host_len;
        }

        uri->host_len = uri->canon_len - uri->host_start;
        if(!computeOnly)
            TRACE("(%p %p %x %d): Canonicalized IPv4 address, ip=%s len=%d\n",
                data, uri, flags, computeOnly,
                debugstr_wn(uri->canon_uri+uri->host_start, uri->host_len),
                uri->host_len);
    }

    return TRUE;
}

2605 2606 2607
/* Attempts to canonicalize the IPv6 address of the URI.
 *
 * Multiple things happen during the canonicalization of an IPv6 address:
2608
 *  1)  Any leading zero's in a h16 component are removed.
2609 2610 2611
 *      Ex: [0001:0022::] -> [1:22::]
 *
 *  2)  The longest sequence of zero h16 components are compressed
2612
 *      into a "::" (elision). If there's a tie, the first is chosen.
2613 2614 2615 2616 2617 2618 2619 2620 2621
 *
 *      Ex: [0:0:0:0:1:6:7:8]   -> [::1:6:7:8]
 *          [0:0:0:0:1:2::]     -> [::1:2:0:0]
 *          [0:0:1:2:0:0:7:8]   -> [::1:2:0:0:7:8]
 *
 *  3)  If an IPv4 address is attached to the IPv6 address, it's
 *      also normalized.
 *      Ex: [::001.002.022.000] -> [::1.2.22.0]
 *
2622
 *  4)  If an elision is present, but, only represents one h16 component
2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751
 *      it's expanded.
 *
 *      Ex: [1::2:3:4:5:6:7] -> [1:0:2:3:4:5:6:7]
 *
 *  5)  If the IPv6 address contains an IPv4 address and there exists
 *      at least 1 non-zero h16 component the IPv4 address is converted
 *      into two h16 components, otherwise it's normalized and kept as is.
 *
 *      Ex: [::192.200.003.4]       -> [::192.200.3.4]
 *          [ffff::192.200.003.4]   -> [ffff::c0c8:3041]
 *
 * NOTE:
 *  For unknown scheme types Windows simply copies the address over without any
 *  changes.
 *
 *  IPv4 address can be included in an elision if all its components are 0's.
 */
static BOOL canonicalize_ipv6address(const parse_data *data, Uri *uri,
                                     DWORD flags, BOOL computeOnly) {
    uri->host_start = uri->canon_len;

    if(data->scheme_type == URL_SCHEME_UNKNOWN) {
        if(!computeOnly)
            memcpy(uri->canon_uri+uri->canon_len, data->host, data->host_len*sizeof(WCHAR));
        uri->canon_len += data->host_len;
    } else {
        USHORT values[8];
        INT elision_start;
        DWORD i, elision_len;

        if(!ipv6_to_number(&(data->ipv6_address), values)) {
            TRACE("(%p %p %x %d): Failed to compute numerical value for IPv6 address.\n",
                data, uri, flags, computeOnly);
            return FALSE;
        }

        if(!computeOnly)
            uri->canon_uri[uri->canon_len] = '[';
        ++uri->canon_len;

        /* Find where the elision should occur (if any). */
        compute_elision_location(&(data->ipv6_address), values, &elision_start, &elision_len);

        TRACE("%p %p %x %d): Elision starts at %d, len=%u\n", data, uri, flags,
            computeOnly, elision_start, elision_len);

        for(i = 0; i < 8; ++i) {
            BOOL in_elision = (elision_start > -1 && i >= elision_start &&
                               i < elision_start+elision_len);
            BOOL do_ipv4 = (i == 6 && data->ipv6_address.ipv4 && !in_elision &&
                            data->ipv6_address.h16_count == 0);

            if(i == elision_start) {
                if(!computeOnly) {
                    uri->canon_uri[uri->canon_len] = ':';
                    uri->canon_uri[uri->canon_len+1] = ':';
                }
                uri->canon_len += 2;
            }

            /* We can ignore the current component if we're in the elision. */
            if(in_elision)
                continue;

            /* We only add a ':' if we're not at i == 0, or when we're at
             * the very end of elision range since the ':' colon was handled
             * earlier. Otherwise we would end up with ":::" after elision.
             */
            if(i != 0 && !(elision_start > -1 && i == elision_start+elision_len)) {
                if(!computeOnly)
                    uri->canon_uri[uri->canon_len] = ':';
                ++uri->canon_len;
            }

            if(do_ipv4) {
                UINT val;
                DWORD len;

                /* Combine the two parts of the IPv4 address values. */
                val = values[i];
                val <<= 16;
                val += values[i+1];

                if(!computeOnly)
                    len = ui2ipv4(uri->canon_uri+uri->canon_len, val);
                else
                    len = ui2ipv4(NULL, val);

                uri->canon_len += len;
                ++i;
            } else {
                /* Write a regular h16 component to the URI. */

                /* Short circuit for the trivial case. */
                if(values[i] == 0) {
                    if(!computeOnly)
                        uri->canon_uri[uri->canon_len] = '0';
                    ++uri->canon_len;
                } else {
                    static const WCHAR formatW[] = {'%','x',0};

                    if(!computeOnly)
                        uri->canon_len += sprintfW(uri->canon_uri+uri->canon_len,
                                            formatW, values[i]);
                    else {
                        WCHAR tmp[5];
                        uri->canon_len += sprintfW(tmp, formatW, values[i]);
                    }
                }
            }
        }

        /* Add the closing ']'. */
        if(!computeOnly)
            uri->canon_uri[uri->canon_len] = ']';
        ++uri->canon_len;
    }

    uri->host_len = uri->canon_len - uri->host_start;

    if(!computeOnly)
        TRACE("(%p %p %x %d): Canonicalized IPv6 address %s, len=%d\n", data, uri, flags,
            computeOnly, debugstr_wn(uri->canon_uri+uri->host_start, uri->host_len),
            uri->host_len);

    return TRUE;
}

/* Attempts to canonicalize the host of the URI (if any). */
2752 2753 2754
static BOOL canonicalize_host(const parse_data *data, Uri *uri, DWORD flags, BOOL computeOnly) {
    uri->host_start = -1;
    uri->host_len = 0;
2755
    uri->domain_offset = -1;
2756 2757 2758

    if(data->host) {
        switch(data->host_type) {
2759 2760 2761 2762 2763 2764
        case Uri_HOST_DNS:
            uri->host_type = Uri_HOST_DNS;
            if(!canonicalize_reg_name(data, uri, flags, computeOnly))
                return FALSE;

            break;
2765 2766 2767 2768 2769
        case Uri_HOST_IPV4:
            uri->host_type = Uri_HOST_IPV4;
            if(!canonicalize_ipv4address(data, uri, flags, computeOnly))
                return FALSE;

2770 2771 2772 2773 2774 2775
            break;
        case Uri_HOST_IPV6:
            if(!canonicalize_ipv6address(data, uri, flags, computeOnly))
                return FALSE;

            uri->host_type = Uri_HOST_IPV6;
2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787
            break;
        case Uri_HOST_UNKNOWN:
            if(data->host_len > 0 || data->scheme_type != URL_SCHEME_FILE) {
                uri->host_start = uri->canon_len;

                /* Nothing happens to unknown host types. */
                if(!computeOnly)
                    memcpy(uri->canon_uri+uri->canon_len, data->host, data->host_len*sizeof(WCHAR));
                uri->canon_len += data->host_len;
                uri->host_len = data->host_len;
            }

2788
            uri->host_type = Uri_HOST_UNKNOWN;
2789 2790
            break;
        default:
2791 2792 2793
            FIXME("(%p %p %x %d): Canonicalization for host type %d not supported.\n", data,
                    uri, flags, computeOnly, data->host_type);
            return FALSE;
2794 2795 2796 2797 2798 2799
       }
   }

   return TRUE;
}

2800 2801 2802 2803 2804
static BOOL canonicalize_port(const parse_data *data, Uri *uri, DWORD flags, BOOL computeOnly) {
    BOOL has_default_port = FALSE;
    USHORT default_port = 0;
    DWORD i;

2805 2806
    uri->port_offset = -1;

2807 2808 2809 2810 2811 2812 2813 2814 2815
    /* Check if the scheme has a default port. */
    for(i = 0; i < sizeof(default_ports)/sizeof(default_ports[0]); ++i) {
        if(default_ports[i].scheme == data->scheme_type) {
            has_default_port = TRUE;
            default_port = default_ports[i].port;
            break;
        }
    }

2816
    uri->has_port = data->has_port || has_default_port;
2817 2818 2819 2820 2821 2822 2823

    /* Possible cases:
     *  1)  Has a port which is the default port.
     *  2)  Has a port (not the default).
     *  3)  Doesn't have a port, but, scheme has a default port.
     *  4)  No port.
     */
2824
    if(has_default_port && data->has_port && data->port_value == default_port) {
2825 2826
        /* If it's the default port and this flag isn't set, don't do anything. */
        if(flags & Uri_CREATE_NO_CANONICALIZE) {
2827
            uri->port_offset = uri->canon_len-uri->authority_start;
2828
            if(!computeOnly)
2829
                uri->canon_uri[uri->canon_len] = ':';
2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841
            ++uri->canon_len;

            if(data->port) {
                /* Copy the original port over. */
                if(!computeOnly)
                    memcpy(uri->canon_uri+uri->canon_len, data->port, data->port_len*sizeof(WCHAR));
                uri->canon_len += data->port_len;
            } else {
                if(!computeOnly)
                    uri->canon_len += ui2str(uri->canon_uri+uri->canon_len, data->port_value);
                else
                    uri->canon_len += ui2str(NULL, data->port_value);
2842 2843 2844 2845
            }
        }

        uri->port = default_port;
2846
    } else if(data->has_port) {
2847
        uri->port_offset = uri->canon_len-uri->authority_start;
2848 2849 2850 2851
        if(!computeOnly)
            uri->canon_uri[uri->canon_len] = ':';
        ++uri->canon_len;

2852
        if(flags & Uri_CREATE_NO_CANONICALIZE && data->port) {
2853 2854 2855 2856 2857 2858
            /* Copy the original over without changes. */
            if(!computeOnly)
                memcpy(uri->canon_uri+uri->canon_len, data->port, data->port_len*sizeof(WCHAR));
            uri->canon_len += data->port_len;
        } else {
            if(!computeOnly)
2859 2860 2861
                uri->canon_len += ui2str(uri->canon_uri+uri->canon_len, data->port_value);
            else
                uri->canon_len += ui2str(NULL, data->port_value);
2862 2863 2864 2865 2866 2867 2868 2869 2870
        }

        uri->port = data->port_value;
    } else if(has_default_port)
        uri->port = default_port;

    return TRUE;
}

2871 2872
/* Canonicalizes the authority of the URI represented by the parse_data. */
static BOOL canonicalize_authority(const parse_data *data, Uri *uri, DWORD flags, BOOL computeOnly) {
2873 2874 2875
    uri->authority_start = uri->canon_len;
    uri->authority_len = 0;

2876 2877 2878
    if(!canonicalize_userinfo(data, uri, flags, computeOnly))
        return FALSE;

2879 2880 2881
    if(!canonicalize_host(data, uri, flags, computeOnly))
        return FALSE;

2882 2883
    if(!canonicalize_port(data, uri, flags, computeOnly))
        return FALSE;
2884

2885
    if(uri->host_start != -1 || (data->is_relative && (data->password || data->username)))
2886 2887 2888 2889
        uri->authority_len = uri->canon_len - uri->authority_start;
    else
        uri->authority_start = -1;

2890 2891 2892
    return TRUE;
}

2893 2894 2895 2896 2897
/* Attempts to canonicalize the path of a hierarchical URI.
 *
 * Things that happen:
 *  1). Forbidden characters are percent encoded, unless the NO_ENCODE_FORBIDDEN
 *      flag is set or it's a file URI. Forbidden characters are always encoded
2898
 *      for file schemes regardless and forbidden characters are never encoded
2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917
 *      for unknown scheme types.
 *
 *  2). For known scheme types '\\' are changed to '/'.
 *
 *  3). Percent encoded, unreserved characters are decoded to their actual values.
 *      Unless the scheme type is unknown. For file schemes any percent encoded
 *      character in the unreserved or reserved set is decoded.
 *
 *  4). For File schemes if the path is starts with a drive letter and doesn't
 *      start with a '/' then one is appended.
 *      Ex: file://c:/test.mp3 -> file:///c:/test.mp3
 *
 *  5). Dot segments are removed from the path for all scheme types
 *      unless NO_CANONICALIZE flag is set. Dot segments aren't removed
 *      for wildcard scheme types.
 *
 * NOTES:
 *      file://c:/test%20test   -> file:///c:/test%2520test
 *      file://c:/test%3Etest   -> file:///c:/test%253Etest
2918
 * if Uri_CREATE_FILE_USE_DOS_PATH is not set:
2919 2920 2921 2922 2923 2924 2925 2926
 *      file:///c:/test%20test  -> file:///c:/test%20test
 *      file:///c:/test%test    -> file:///c:/test%25test
 */
static BOOL canonicalize_path_hierarchical(const parse_data *data, Uri *uri,
                                           DWORD flags, BOOL computeOnly) {
    const WCHAR *ptr;
    const BOOL known_scheme = data->scheme_type != URL_SCHEME_UNKNOWN;
    const BOOL is_file = data->scheme_type == URL_SCHEME_FILE;
2927
    const BOOL is_res = data->scheme_type == URL_SCHEME_RES;
2928 2929 2930 2931 2932 2933 2934 2935 2936 2937

    BOOL escape_pct = FALSE;

    if(!data->path) {
        uri->path_start = -1;
        uri->path_len = 0;
        return TRUE;
    }

    uri->path_start = uri->canon_len;
2938
    ptr = data->path;
2939

2940 2941 2942
    if(is_file && uri->host_start == -1) {
        /* Check if a '/' needs to be appended for the file scheme. */
        if(data->path_len > 1 && is_drive_path(ptr) && !(flags & Uri_CREATE_FILE_USE_DOS_PATH)) {
2943 2944 2945 2946
            if(!computeOnly)
                uri->canon_uri[uri->canon_len] = '/';
            uri->canon_len++;
            escape_pct = TRUE;
2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959
        } else if(*ptr == '/') {
            if(!(flags & Uri_CREATE_FILE_USE_DOS_PATH)) {
                /* Copy the extra '/' over. */
                if(!computeOnly)
                    uri->canon_uri[uri->canon_len] = '/';
                ++uri->canon_len;
            }
            ++ptr;
        }

        if(is_drive_path(ptr)) {
            if(!computeOnly) {
                uri->canon_uri[uri->canon_len] = *ptr;
2960
                /* If there's a '|' after the drive letter, convert it to a ':'. */
2961 2962 2963 2964
                uri->canon_uri[uri->canon_len+1] = ':';
            }
            ptr += 2;
            uri->canon_len += 2;
2965 2966 2967
        }
    }

2968 2969 2970 2971 2972 2973 2974
    if(!is_file && *(data->path) && *(data->path) != '/') {
        /* Prepend a '/' to the path if it doesn't have one. */
        if(!computeOnly)
            uri->canon_uri[uri->canon_len] = '/';
        ++uri->canon_len;
    }

2975
    for(; ptr < data->path+data->path_len; ++ptr) {
2976 2977
        BOOL do_default_action = TRUE;

2978
        if(*ptr == '%' && !is_res) {
2979 2980 2981
            const WCHAR *tmp = ptr;
            WCHAR val;

2982
            /* Check if the % represents a valid encoded char, or if it needs encoding. */
2983
            BOOL force_encode = !check_pct_encoded(&tmp) && is_file && !(flags&Uri_CREATE_FILE_USE_DOS_PATH);
2984 2985 2986 2987 2988 2989 2990
            val = decode_pct_val(ptr);

            if(force_encode || escape_pct) {
                /* Escape the percent sign in the file URI. */
                if(!computeOnly)
                    pct_encode_val(*ptr, uri->canon_uri+uri->canon_len);
                uri->canon_len += 3;
2991
                do_default_action = FALSE;
2992
            } else if((is_unreserved(val) && known_scheme) ||
2993 2994
                      (is_file && (is_unreserved(val) || is_reserved(val) ||
                      (val && flags&Uri_CREATE_FILE_USE_DOS_PATH && !is_forbidden_dos_path_char(val))))) {
2995 2996 2997 2998 2999 3000 3001
                if(!computeOnly)
                    uri->canon_uri[uri->canon_len] = val;
                ++uri->canon_len;

                ptr += 2;
                continue;
            }
3002 3003
        } else if(*ptr == '/' && is_file && (flags & Uri_CREATE_FILE_USE_DOS_PATH)) {
            /* Convert the '/' back to a '\\'. */
3004
            if(!computeOnly)
3005
                uri->canon_uri[uri->canon_len] = '\\';
3006
            ++uri->canon_len;
3007
            do_default_action = FALSE;
3008
        } else if(*ptr == '\\' && known_scheme) {
3009 3010
            if(!(is_file && (flags & Uri_CREATE_FILE_USE_DOS_PATH))) {
                /* Convert '\\' into a '/'. */
3011 3012 3013
                if(!computeOnly)
                    uri->canon_uri[uri->canon_len] = '/';
                ++uri->canon_len;
3014
                do_default_action = FALSE;
3015
            }
3016
        } else if(known_scheme && !is_res && !is_unreserved(*ptr) && !is_reserved(*ptr) &&
3017
                  (!(flags & Uri_CREATE_NO_ENCODE_FORBIDDEN_CHARACTERS) || is_file)) {
3018
            if(!(is_file && (flags & Uri_CREATE_FILE_USE_DOS_PATH))) {
3019 3020 3021 3022
                /* Escape the forbidden character. */
                if(!computeOnly)
                    pct_encode_val(*ptr, uri->canon_uri+uri->canon_len);
                uri->canon_len += 3;
3023
                do_default_action = FALSE;
3024
            }
3025 3026 3027
        }

        if(do_default_action) {
3028 3029 3030 3031 3032 3033 3034 3035
            if(!computeOnly)
                uri->canon_uri[uri->canon_len] = *ptr;
            ++uri->canon_len;
        }
    }

    uri->path_len = uri->canon_len - uri->path_start;

3036
    /* Removing the dot segments only happens when it's not in
3037 3038
     * computeOnly mode and it's not a wildcard scheme. File schemes
     * with USE_DOS_PATH set don't get dot segments removed.
3039
     */
3040 3041 3042
    if(!(is_file && (flags & Uri_CREATE_FILE_USE_DOS_PATH)) &&
       data->scheme_type != URL_SCHEME_WILDCARD) {
        if(!(flags & Uri_CREATE_NO_CANONICALIZE) && !computeOnly) {
3043 3044 3045 3046 3047 3048 3049 3050 3051
            /* Remove the dot segments (if any) and reset everything to the new
             * correct length.
             */
            DWORD new_len = remove_dot_segments(uri->canon_uri+uri->path_start, uri->path_len);
            uri->canon_len -= uri->path_len-new_len;
            uri->path_len = new_len;
        }
    }

3052 3053 3054 3055 3056 3057 3058 3059
    if(!computeOnly)
        TRACE("Canonicalized path %s len=%d\n",
            debugstr_wn(uri->canon_uri+uri->path_start, uri->path_len),
            uri->path_len);

    return TRUE;
}

3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070
/* Attempts to canonicalize the path for an opaque URI.
 *
 * For known scheme types:
 *  1)  forbidden characters are percent encoded if
 *      NO_ENCODE_FORBIDDEN_CHARACTERS isn't set.
 *
 *  2)  Percent encoded, unreserved characters are decoded
 *      to their actual values, for known scheme types.
 *
 *  3)  '\\' are changed to '/' for known scheme types
 *      except for mailto schemes.
3071 3072 3073 3074 3075 3076
 *
 *  4)  For file schemes, if USE_DOS_PATH is set all '/'
 *      are converted to backslashes.
 *
 *  5)  For file schemes, if USE_DOS_PATH isn't set all '\'
 *      are converted to forward slashes.
3077 3078 3079 3080
 */
static BOOL canonicalize_path_opaque(const parse_data *data, Uri *uri, DWORD flags, BOOL computeOnly) {
    const WCHAR *ptr;
    const BOOL known_scheme = data->scheme_type != URL_SCHEME_UNKNOWN;
3081
    const BOOL is_file = data->scheme_type == URL_SCHEME_FILE;
3082 3083 3084 3085 3086 3087 3088 3089 3090

    if(!data->path) {
        uri->path_start = -1;
        uri->path_len = 0;
        return TRUE;
    }

    uri->path_start = uri->canon_len;

3091
    /* For javascript: URIs, simply copy path part without any canonicalization */
3092 3093 3094 3095 3096 3097 3098 3099
    if(data->scheme_type == URL_SCHEME_JAVASCRIPT) {
        if(!computeOnly)
            memcpy(uri->canon_uri+uri->canon_len, data->path, data->path_len*sizeof(WCHAR));
        uri->path_len = data->path_len;
        uri->canon_len += data->path_len;
        return TRUE;
    }

3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113
    /* Windows doesn't allow a "//" to appear after the scheme
     * of a URI, if it's an opaque URI.
     */
    if(data->scheme && *(data->path) == '/' && *(data->path+1) == '/') {
        /* So it inserts a "/." before the "//" if it exists. */
        if(!computeOnly) {
            uri->canon_uri[uri->canon_len] = '/';
            uri->canon_uri[uri->canon_len+1] = '.';
        }

        uri->canon_len += 2;
    }

    for(ptr = data->path; ptr < data->path+data->path_len; ++ptr) {
3114 3115
        BOOL do_default_action = TRUE;

3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126
        if(*ptr == '%' && known_scheme) {
            WCHAR val = decode_pct_val(ptr);

            if(is_unreserved(val)) {
                if(!computeOnly)
                    uri->canon_uri[uri->canon_len] = val;
                ++uri->canon_len;

                ptr += 2;
                continue;
            }
3127 3128 3129 3130
        } else if(*ptr == '/' && is_file && (flags & Uri_CREATE_FILE_USE_DOS_PATH)) {
            if(!computeOnly)
                uri->canon_uri[uri->canon_len] = '\\';
            ++uri->canon_len;
3131 3132 3133
            do_default_action = FALSE;
        } else if(*ptr == '\\') {
            if(is_file && !(flags & Uri_CREATE_FILE_USE_DOS_PATH)) {
3134 3135 3136 3137
                /* Convert to a '/'. */
                if(!computeOnly)
                    uri->canon_uri[uri->canon_len] = '/';
                ++uri->canon_len;
3138
                do_default_action = FALSE;
3139
            }
3140 3141
        } else if(known_scheme && !is_unreserved(*ptr) && !is_reserved(*ptr) &&
                  !(flags & Uri_CREATE_NO_ENCODE_FORBIDDEN_CHARACTERS)) {
3142
            if(!(is_file && (flags & Uri_CREATE_FILE_USE_DOS_PATH))) {
3143 3144 3145
                if(!computeOnly)
                    pct_encode_val(*ptr, uri->canon_uri+uri->canon_len);
                uri->canon_len += 3;
3146
                do_default_action = FALSE;
3147
            }
3148 3149 3150
        }

        if(do_default_action) {
3151 3152 3153 3154 3155 3156
            if(!computeOnly)
                uri->canon_uri[uri->canon_len] = *ptr;
            ++uri->canon_len;
        }
    }

3157 3158 3159 3160 3161 3162
    if(data->scheme_type == URL_SCHEME_MK && !computeOnly && !(flags & Uri_CREATE_NO_CANONICALIZE)) {
        DWORD new_len = remove_dot_segments(uri->canon_uri + uri->path_start,
                                            uri->canon_len - uri->path_start);
        uri->canon_len = uri->path_start + new_len;
    }

3163 3164
    uri->path_len = uri->canon_len - uri->path_start;

3165 3166 3167
    if(!computeOnly)
        TRACE("(%p %p %x %d): Canonicalized opaque URI path %s len=%d\n", data, uri, flags, computeOnly,
            debugstr_wn(uri->canon_uri+uri->path_start, uri->path_len), uri->path_len);
3168 3169 3170
    return TRUE;
}

3171 3172 3173 3174 3175 3176 3177
/* Determines how the URI represented by the parse_data should be canonicalized.
 *
 * Essentially, if the parse_data represents an hierarchical URI then it calls
 * canonicalize_authority and the canonicalization functions for the path. If the
 * URI is opaque it canonicalizes the path of the URI.
 */
static BOOL canonicalize_hierpart(const parse_data *data, Uri *uri, DWORD flags, BOOL computeOnly) {
3178 3179 3180 3181 3182 3183 3184 3185 3186
    if(!data->is_opaque || (data->is_relative && (data->password || data->username))) {
        /* "//" is only added for non-wildcard scheme types.
         *
         * A "//" is only added to a relative URI if it has a
         * host or port component (this only happens if a IUriBuilder
         * is generating an IUri).
         */
        if((data->is_relative && (data->host || data->has_port)) ||
           (!data->is_relative && data->scheme_type != URL_SCHEME_WILDCARD)) {
3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198
            if(!computeOnly) {
                INT pos = uri->canon_len;

                uri->canon_uri[pos] = '/';
                uri->canon_uri[pos+1] = '/';
           }
           uri->canon_len += 2;
        }

        if(!canonicalize_authority(data, uri, flags, computeOnly))
            return FALSE;

3199 3200 3201 3202 3203 3204 3205
        if(data->is_relative && (data->password || data->username)) {
            if(!canonicalize_path_opaque(data, uri, flags, computeOnly))
                return FALSE;
        } else {
            if(!canonicalize_path_hierarchical(data, uri, flags, computeOnly))
                return FALSE;
        }
3206
    } else {
3207
        /* Opaque URI's don't have an authority. */
3208 3209
        uri->userinfo_start = uri->userinfo_split = -1;
        uri->userinfo_len = 0;
3210 3211
        uri->host_start = -1;
        uri->host_len = 0;
3212
        uri->host_type = Uri_HOST_UNKNOWN;
3213
        uri->has_port = FALSE;
3214 3215
        uri->authority_start = -1;
        uri->authority_len = 0;
3216
        uri->domain_offset = -1;
3217
        uri->port_offset = -1;
3218

3219 3220 3221
        if(is_hierarchical_scheme(data->scheme_type)) {
            DWORD i;

3222 3223 3224
            /* Absolute URIs aren't displayed for known scheme types
             * which should be hierarchical URIs.
             */
3225
            uri->display_modifiers |= URI_DISPLAY_NO_ABSOLUTE_URI;
3226

3227 3228 3229 3230 3231 3232 3233 3234 3235 3236
            /* Windows also sets the port for these (if they have one). */
            for(i = 0; i < sizeof(default_ports)/sizeof(default_ports[0]); ++i) {
                if(data->scheme_type == default_ports[i].scheme) {
                    uri->has_port = TRUE;
                    uri->port = default_ports[i].port;
                    break;
                }
            }
        }

3237 3238
        if(!canonicalize_path_opaque(data, uri, flags, computeOnly))
            return FALSE;
3239 3240
    }

3241 3242 3243 3244 3245 3246
    if(uri->path_start > -1 && !computeOnly)
        /* Finding file extensions happens for both types of URIs. */
        uri->extension_offset = find_file_extension(uri->canon_uri+uri->path_start, uri->path_len);
    else
        uri->extension_offset = -1;

3247 3248 3249
    return TRUE;
}

3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309
/* Attempts to canonicalize the query string of the URI.
 *
 * Things that happen:
 *  1)  For known scheme types forbidden characters
 *      are percent encoded, unless the NO_DECODE_EXTRA_INFO flag is set
 *      or NO_ENCODE_FORBIDDEN_CHARACTERS is set.
 *
 *  2)  For known scheme types, percent encoded, unreserved characters
 *      are decoded as long as the NO_DECODE_EXTRA_INFO flag isn't set.
 */
static BOOL canonicalize_query(const parse_data *data, Uri *uri, DWORD flags, BOOL computeOnly) {
    const WCHAR *ptr, *end;
    const BOOL known_scheme = data->scheme_type != URL_SCHEME_UNKNOWN;

    if(!data->query) {
        uri->query_start = -1;
        uri->query_len = 0;
        return TRUE;
    }

    uri->query_start = uri->canon_len;

    end = data->query+data->query_len;
    for(ptr = data->query; ptr < end; ++ptr) {
        if(*ptr == '%') {
            if(known_scheme && !(flags & Uri_CREATE_NO_DECODE_EXTRA_INFO)) {
                WCHAR val = decode_pct_val(ptr);
                if(is_unreserved(val)) {
                    if(!computeOnly)
                        uri->canon_uri[uri->canon_len] = val;
                    ++uri->canon_len;

                    ptr += 2;
                    continue;
                }
            }
        } else if(known_scheme && !is_unreserved(*ptr) && !is_reserved(*ptr)) {
            if(!(flags & Uri_CREATE_NO_ENCODE_FORBIDDEN_CHARACTERS) &&
               !(flags & Uri_CREATE_NO_DECODE_EXTRA_INFO)) {
                if(!computeOnly)
                    pct_encode_val(*ptr, uri->canon_uri+uri->canon_len);
                uri->canon_len += 3;
                continue;
            }
        }

        if(!computeOnly)
            uri->canon_uri[uri->canon_len] = *ptr;
        ++uri->canon_len;
    }

    uri->query_len = uri->canon_len - uri->query_start;

    if(!computeOnly)
        TRACE("(%p %p %x %d): Canonicalized query string %s len=%d\n", data, uri, flags,
            computeOnly, debugstr_wn(uri->canon_uri+uri->query_start, uri->query_len),
            uri->query_len);
    return TRUE;
}

3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359
static BOOL canonicalize_fragment(const parse_data *data, Uri *uri, DWORD flags, BOOL computeOnly) {
    const WCHAR *ptr, *end;
    const BOOL known_scheme = data->scheme_type != URL_SCHEME_UNKNOWN;

    if(!data->fragment) {
        uri->fragment_start = -1;
        uri->fragment_len = 0;
        return TRUE;
    }

    uri->fragment_start = uri->canon_len;

    end = data->fragment + data->fragment_len;
    for(ptr = data->fragment; ptr < end; ++ptr) {
        if(*ptr == '%') {
            if(known_scheme && !(flags & Uri_CREATE_NO_DECODE_EXTRA_INFO)) {
                WCHAR val = decode_pct_val(ptr);
                if(is_unreserved(val)) {
                    if(!computeOnly)
                        uri->canon_uri[uri->canon_len] = val;
                    ++uri->canon_len;

                    ptr += 2;
                    continue;
                }
            }
        } else if(known_scheme && !is_unreserved(*ptr) && !is_reserved(*ptr)) {
            if(!(flags & Uri_CREATE_NO_ENCODE_FORBIDDEN_CHARACTERS) &&
               !(flags & Uri_CREATE_NO_DECODE_EXTRA_INFO)) {
                if(!computeOnly)
                    pct_encode_val(*ptr, uri->canon_uri+uri->canon_len);
                uri->canon_len += 3;
                continue;
            }
        }

        if(!computeOnly)
            uri->canon_uri[uri->canon_len] = *ptr;
        ++uri->canon_len;
    }

    uri->fragment_len = uri->canon_len - uri->fragment_start;

    if(!computeOnly)
        TRACE("(%p %p %x %d): Canonicalized fragment %s len=%d\n", data, uri, flags,
            computeOnly, debugstr_wn(uri->canon_uri+uri->fragment_start, uri->fragment_len),
            uri->fragment_len);
    return TRUE;
}

3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390
/* Canonicalizes the scheme information specified in the parse_data using the specified flags. */
static BOOL canonicalize_scheme(const parse_data *data, Uri *uri, DWORD flags, BOOL computeOnly) {
    uri->scheme_start = -1;
    uri->scheme_len = 0;

    if(!data->scheme) {
        /* The only type of URI that doesn't have to have a scheme is a relative
         * URI.
         */
        if(!data->is_relative) {
            FIXME("(%p %p %x): Unable to determine the scheme type of %s.\n", data,
                    uri, flags, debugstr_w(data->uri));
            return FALSE;
        }
    } else {
        if(!computeOnly) {
            DWORD i;
            INT pos = uri->canon_len;

            for(i = 0; i < data->scheme_len; ++i) {
                /* Scheme name must be lower case after canonicalization. */
                uri->canon_uri[i + pos] = tolowerW(data->scheme[i]);
            }

            uri->canon_uri[i + pos] = ':';
            uri->scheme_start = pos;

            TRACE("(%p %p %x): Canonicalized scheme=%s, len=%d.\n", data, uri, flags,
                    debugstr_wn(uri->canon_uri,  uri->scheme_len), data->scheme_len);
        }

3391
        /* This happens in both computation modes. */
3392 3393 3394 3395 3396 3397
        uri->canon_len += data->scheme_len + 1;
        uri->scheme_len = data->scheme_len;
    }
    return TRUE;
}

3398
/* Computes what the length of the URI specified by the parse_data will be
3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416
 * after canonicalization occurs using the specified flags.
 *
 * This function will return a non-zero value indicating the length of the canonicalized
 * URI, or -1 on error.
 */
static int compute_canonicalized_length(const parse_data *data, DWORD flags) {
    Uri uri;

    memset(&uri, 0, sizeof(Uri));

    TRACE("(%p %x): Beginning to compute canonicalized length for URI %s\n", data, flags,
            debugstr_w(data->uri));

    if(!canonicalize_scheme(data, &uri, flags, TRUE)) {
        ERR("(%p %x): Failed to compute URI scheme length.\n", data, flags);
        return -1;
    }

3417 3418 3419 3420 3421
    if(!canonicalize_hierpart(data, &uri, flags, TRUE)) {
        ERR("(%p %x): Failed to compute URI hierpart length.\n", data, flags);
        return -1;
    }

3422 3423 3424 3425 3426
    if(!canonicalize_query(data, &uri, flags, TRUE)) {
        ERR("(%p %x): Failed to compute query string length.\n", data, flags);
        return -1;
    }

3427 3428 3429 3430 3431
    if(!canonicalize_fragment(data, &uri, flags, TRUE)) {
        ERR("(%p %x): Failed to compute fragment length.\n", data, flags);
        return -1;
    }

3432 3433 3434 3435 3436 3437
    TRACE("(%p %x): Finished computing canonicalized URI length. length=%d\n", data, flags, uri.canon_len);

    return uri.canon_len;
}

/* Canonicalizes the URI data specified in the parse_data, using the given flags. If the
3438
 * canonicalization succeeds it will store all the canonicalization information
3439 3440 3441
 * in the pointer to the Uri.
 *
 * To canonicalize a URI this function first computes what the length of the URI
3442
 * specified by the parse_data will be. Once this is done it will then perform the actual
3443 3444 3445 3446 3447 3448
 * canonicalization of the URI.
 */
static HRESULT canonicalize_uri(const parse_data *data, Uri *uri, DWORD flags) {
    INT len;

    uri->canon_uri = NULL;
3449
    uri->canon_size = uri->canon_len = 0;
3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464

    TRACE("(%p %p %x): beginning to canonicalize URI %s.\n", data, uri, flags, debugstr_w(data->uri));

    /* First try to compute the length of the URI. */
    len = compute_canonicalized_length(data, flags);
    if(len == -1) {
        ERR("(%p %p %x): Could not compute the canonicalized length of %s.\n", data, uri, flags,
                debugstr_w(data->uri));
        return E_INVALIDARG;
    }

    uri->canon_uri = heap_alloc((len+1)*sizeof(WCHAR));
    if(!uri->canon_uri)
        return E_OUTOFMEMORY;

3465
    uri->canon_size = len;
3466 3467 3468 3469 3470 3471
    if(!canonicalize_scheme(data, uri, flags, FALSE)) {
        ERR("(%p %p %x): Unable to canonicalize the scheme of the URI.\n", data, uri, flags);
        return E_INVALIDARG;
    }
    uri->scheme_type = data->scheme_type;

3472 3473 3474 3475 3476
    if(!canonicalize_hierpart(data, uri, flags, FALSE)) {
        ERR("(%p %p %x): Unable to canonicalize the heirpart of the URI\n", data, uri, flags);
        return E_INVALIDARG;
    }

3477 3478 3479 3480 3481 3482
    if(!canonicalize_query(data, uri, flags, FALSE)) {
        ERR("(%p %p %x): Unable to canonicalize query string of the URI.\n",
            data, uri, flags);
        return E_INVALIDARG;
    }

3483 3484 3485 3486 3487 3488
    if(!canonicalize_fragment(data, uri, flags, FALSE)) {
        ERR("(%p %p %x): Unable to canonicalize fragment of the URI.\n",
            data, uri, flags);
        return E_INVALIDARG;
    }

3489 3490 3491 3492 3493
    /* There's a possibility we didn't use all the space we allocated
     * earlier.
     */
    if(uri->canon_len < uri->canon_size) {
        /* This happens if the URI is hierarchical and dot
3494
         * segments were removed from its path.
3495 3496 3497 3498 3499 3500 3501 3502 3503
         */
        WCHAR *tmp = heap_realloc(uri->canon_uri, (uri->canon_len+1)*sizeof(WCHAR));
        if(!tmp)
            return E_OUTOFMEMORY;

        uri->canon_uri = tmp;
        uri->canon_size = uri->canon_len;
    }

3504
    uri->canon_uri[uri->canon_len] = '\0';
3505
    TRACE("(%p %p %x): finished canonicalizing the URI. uri=%s\n", data, uri, flags, debugstr_w(uri->canon_uri));
3506 3507 3508 3509

    return S_OK;
}

3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544 3545
static HRESULT get_builder_component(LPWSTR *component, DWORD *component_len,
                                     LPCWSTR source, DWORD source_len,
                                     LPCWSTR *output, DWORD *output_len)
{
    if(!output_len) {
        if(output)
            *output = NULL;
        return E_POINTER;
    }

    if(!output) {
        *output_len = 0;
        return E_POINTER;
    }

    if(!(*component) && source) {
        /* Allocate 'component', and copy the contents from 'source'
         * into the new allocation.
         */
        *component = heap_alloc((source_len+1)*sizeof(WCHAR));
        if(!(*component))
            return E_OUTOFMEMORY;

        memcpy(*component, source, source_len*sizeof(WCHAR));
        (*component)[source_len] = '\0';
        *component_len = source_len;
    }

    *output = *component;
    *output_len = *component_len;
    return *output ? S_OK : S_FALSE;
}

/* Allocates 'component' and copies the string from 'new_value' into 'component'.
 * If 'prefix' is set and 'new_value' isn't NULL, then it checks if 'new_value'
 * starts with 'prefix'. If it doesn't then 'prefix' is prepended to 'component'.
3546 3547
 *
 * If everything is successful, then will set 'success_flag' in 'flags'.
3548 3549
 */
static HRESULT set_builder_component(LPWSTR *component, DWORD *component_len, LPCWSTR new_value,
3550
                                     WCHAR prefix, DWORD *flags, DWORD success_flag)
3551
{
3552
    heap_free(*component);
3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577

    if(!new_value) {
        *component = NULL;
        *component_len = 0;
    } else {
        BOOL add_prefix = FALSE;
        DWORD len = lstrlenW(new_value);
        DWORD pos = 0;

        if(prefix && *new_value != prefix) {
            add_prefix = TRUE;
            *component = heap_alloc((len+2)*sizeof(WCHAR));
        } else
            *component = heap_alloc((len+1)*sizeof(WCHAR));

        if(!(*component))
            return E_OUTOFMEMORY;

        if(add_prefix)
            (*component)[pos++] = prefix;

        memcpy(*component+pos, new_value, (len+1)*sizeof(WCHAR));
        *component_len = len+pos;
    }

3578
    *flags |= success_flag;
3579 3580 3581
    return S_OK;
}

3582 3583
static void reset_builder(UriBuilder *builder) {
    if(builder->uri)
3584
        IUri_Release(&builder->uri->IUri_iface);
3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619
    builder->uri = NULL;

    heap_free(builder->fragment);
    builder->fragment = NULL;
    builder->fragment_len = 0;

    heap_free(builder->host);
    builder->host = NULL;
    builder->host_len = 0;

    heap_free(builder->password);
    builder->password = NULL;
    builder->password_len = 0;

    heap_free(builder->path);
    builder->path = NULL;
    builder->path_len = 0;

    heap_free(builder->query);
    builder->query = NULL;
    builder->query_len = 0;

    heap_free(builder->scheme);
    builder->scheme = NULL;
    builder->scheme_len = 0;

    heap_free(builder->username);
    builder->username = NULL;
    builder->username_len = 0;

    builder->has_port = FALSE;
    builder->port = 0;
    builder->modified_props = 0;
}

3620
static HRESULT validate_scheme_name(const UriBuilder *builder, parse_data *data, DWORD flags) {
3621
    const WCHAR *component;
3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637
    const WCHAR *ptr;
    const WCHAR **pptr;
    DWORD expected_len;

    if(builder->scheme) {
        ptr = builder->scheme;
        expected_len = builder->scheme_len;
    } else if(builder->uri && builder->uri->scheme_start > -1) {
        ptr = builder->uri->canon_uri+builder->uri->scheme_start;
        expected_len = builder->uri->scheme_len;
    } else {
        static const WCHAR nullW[] = {0};
        ptr = nullW;
        expected_len = 0;
    }

3638
    component = ptr;
3639 3640 3641 3642
    pptr = &ptr;
    if(parse_scheme(pptr, data, flags, ALLOW_NULL_TERM_SCHEME) &&
       data->scheme_len == expected_len) {
        if(data->scheme)
3643 3644
            TRACE("(%p %p %x): Found valid scheme component %s len=%d.\n", builder, data, flags,
               debugstr_wn(data->scheme, data->scheme_len), data->scheme_len);
3645 3646
    } else {
        TRACE("(%p %p %x): Invalid scheme component found %s.\n", builder, data, flags,
3647
            debugstr_wn(component, expected_len));
3648 3649 3650 3651 3652 3653
        return INET_E_INVALID_URL;
   }

    return S_OK;
}

3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674
static HRESULT validate_username(const UriBuilder *builder, parse_data *data, DWORD flags) {
    const WCHAR *ptr;
    const WCHAR **pptr;
    DWORD expected_len;

    if(builder->username) {
        ptr = builder->username;
        expected_len = builder->username_len;
    } else if(!(builder->modified_props & Uri_HAS_USER_NAME) && builder->uri &&
              builder->uri->userinfo_start > -1 && builder->uri->userinfo_split != 0) {
        /* Just use the username from the base Uri. */
        data->username = builder->uri->canon_uri+builder->uri->userinfo_start;
        data->username_len = (builder->uri->userinfo_split > -1) ?
                                        builder->uri->userinfo_split : builder->uri->userinfo_len;
        ptr = NULL;
    } else {
        ptr = NULL;
        expected_len = 0;
    }

    if(ptr) {
3675
        const WCHAR *component = ptr;
3676 3677 3678
        pptr = &ptr;
        if(parse_username(pptr, data, flags, ALLOW_NULL_TERM_USER_NAME) &&
           data->username_len == expected_len)
3679 3680
            TRACE("(%p %p %x): Found valid username component %s len=%d.\n", builder, data, flags,
                debugstr_wn(data->username, data->username_len), data->username_len);
3681 3682
        else {
            TRACE("(%p %p %x): Invalid username component found %s.\n", builder, data, flags,
3683
                debugstr_wn(component, expected_len));
3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709
            return INET_E_INVALID_URL;
        }
    }

    return S_OK;
}

static HRESULT validate_password(const UriBuilder *builder, parse_data *data, DWORD flags) {
    const WCHAR *ptr;
    const WCHAR **pptr;
    DWORD expected_len;

    if(builder->password) {
        ptr = builder->password;
        expected_len = builder->password_len;
    } else if(!(builder->modified_props & Uri_HAS_PASSWORD) && builder->uri &&
              builder->uri->userinfo_split > -1) {
        data->password = builder->uri->canon_uri+builder->uri->userinfo_start+builder->uri->userinfo_split+1;
        data->password_len = builder->uri->userinfo_len-builder->uri->userinfo_split-1;
        ptr = NULL;
    } else {
        ptr = NULL;
        expected_len = 0;
    }

    if(ptr) {
3710
        const WCHAR *component = ptr;
3711 3712 3713
        pptr = &ptr;
        if(parse_password(pptr, data, flags, ALLOW_NULL_TERM_PASSWORD) &&
           data->password_len == expected_len)
3714 3715
            TRACE("(%p %p %x): Found valid password component %s len=%d.\n", builder, data, flags,
                debugstr_wn(data->password, data->password_len), data->password_len);
3716 3717
        else {
            TRACE("(%p %p %x): Invalid password component found %s.\n", builder, data, flags,
3718
                debugstr_wn(component, expected_len));
3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739
            return INET_E_INVALID_URL;
        }
    }

    return S_OK;
}

static HRESULT validate_userinfo(const UriBuilder *builder, parse_data *data, DWORD flags) {
    HRESULT hr;

    hr = validate_username(builder, data, flags);
    if(FAILED(hr))
        return hr;

    hr = validate_password(builder, data, flags);
    if(FAILED(hr))
        return hr;

    return S_OK;
}

3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754
static HRESULT validate_host(const UriBuilder *builder, parse_data *data, DWORD flags) {
    const WCHAR *ptr;
    const WCHAR **pptr;
    DWORD expected_len;

    if(builder->host) {
        ptr = builder->host;
        expected_len = builder->host_len;
    } else if(!(builder->modified_props & Uri_HAS_HOST) && builder->uri && builder->uri->host_start > -1) {
        ptr = builder->uri->canon_uri + builder->uri->host_start;
        expected_len = builder->uri->host_len;
    } else
        ptr = NULL;

    if(ptr) {
3755
        const WCHAR *component = ptr;
3756 3757 3758 3759 3760 3761 3762
        DWORD extras = ALLOW_BRACKETLESS_IP_LITERAL|IGNORE_PORT_DELIMITER|SKIP_IP_FUTURE_CHECK;
        pptr = &ptr;

        if(parse_host(pptr, data, flags, extras) && data->host_len == expected_len)
            TRACE("(%p %p %x): Found valid host name %s len=%d type=%d.\n", builder, data, flags,
                debugstr_wn(data->host, data->host_len), data->host_len, data->host_type);
        else {
3763 3764
            TRACE("(%p %p %x): Invalid host name found %s.\n", builder, data, flags,
                debugstr_wn(component, expected_len));
3765 3766 3767 3768 3769 3770 3771
            return INET_E_INVALID_URL;
        }
    }

    return S_OK;
}

3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786
static void setup_port(const UriBuilder *builder, parse_data *data, DWORD flags) {
    if(builder->modified_props & Uri_HAS_PORT) {
        if(builder->has_port) {
            data->has_port = TRUE;
            data->port_value = builder->port;
        }
    } else if(builder->uri && builder->uri->has_port) {
        data->has_port = TRUE;
        data->port_value = builder->uri->port;
    }

    if(data->has_port)
        TRACE("(%p %p %x): Using %u as port for IUri.\n", builder, data, flags, data->port_value);
}

3787 3788
static HRESULT validate_path(const UriBuilder *builder, parse_data *data, DWORD flags) {
    const WCHAR *ptr = NULL;
3789
    const WCHAR *component;
3790 3791
    const WCHAR **pptr;
    DWORD expected_len;
3792 3793
    BOOL check_len = TRUE;
    BOOL valid = FALSE;
3794 3795 3796 3797 3798 3799 3800 3801

    if(builder->path) {
        ptr = builder->path;
        expected_len = builder->path_len;
    } else if(!(builder->modified_props & Uri_HAS_PATH) &&
              builder->uri && builder->uri->path_start > -1) {
        ptr = builder->uri->canon_uri+builder->uri->path_start;
        expected_len = builder->uri->path_len;
3802 3803 3804 3805
    } else {
        static const WCHAR nullW[] = {0};
        ptr = nullW;
        check_len = FALSE;
3806
        expected_len = -1;
3807 3808
    }

3809 3810
    component = ptr;
    pptr = &ptr;
3811

3812 3813 3814 3815 3816
    /* How the path is validated depends on what type of
     * URI it is.
     */
    valid = data->is_opaque ?
        parse_path_opaque(pptr, data, flags) : parse_path_hierarchical(pptr, data, flags);
3817

3818
    if(!valid || (check_len && expected_len != data->path_len)) {
3819
        TRACE("(%p %p %x): Invalid path component %s.\n", builder, data, flags,
3820
            debugstr_wn(component, expected_len) );
3821
        return INET_E_INVALID_URL;
3822 3823
    }

3824 3825 3826
    TRACE("(%p %p %x): Valid path component %s len=%d.\n", builder, data, flags,
        debugstr_wn(data->path, data->path_len), data->path_len);

3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878
    return S_OK;
}

static HRESULT validate_query(const UriBuilder *builder, parse_data *data, DWORD flags) {
    const WCHAR *ptr = NULL;
    const WCHAR **pptr;
    DWORD expected_len;

    if(builder->query) {
        ptr = builder->query;
        expected_len = builder->query_len;
    } else if(!(builder->modified_props & Uri_HAS_QUERY) && builder->uri &&
              builder->uri->query_start > -1) {
        ptr = builder->uri->canon_uri+builder->uri->query_start;
        expected_len = builder->uri->query_len;
    }

    if(ptr) {
        const WCHAR *component = ptr;
        pptr = &ptr;

        if(parse_query(pptr, data, flags) && expected_len == data->query_len)
            TRACE("(%p %p %x): Valid query component %s len=%d.\n", builder, data, flags,
                debugstr_wn(data->query, data->query_len), data->query_len);
        else {
            TRACE("(%p %p %x): Invalid query component %s.\n", builder, data, flags,
                debugstr_wn(component, expected_len));
            return INET_E_INVALID_URL;
        }
    }

    return S_OK;
}

static HRESULT validate_fragment(const UriBuilder *builder, parse_data *data, DWORD flags) {
    const WCHAR *ptr = NULL;
    const WCHAR **pptr;
    DWORD expected_len;

    if(builder->fragment) {
        ptr = builder->fragment;
        expected_len = builder->fragment_len;
    } else if(!(builder->modified_props & Uri_HAS_FRAGMENT) && builder->uri &&
              builder->uri->fragment_start > -1) {
        ptr = builder->uri->canon_uri+builder->uri->fragment_start;
        expected_len = builder->uri->fragment_len;
    }

    if(ptr) {
        const WCHAR *component = ptr;
        pptr = &ptr;

3879
        if(parse_fragment(pptr, data, flags) && expected_len == data->fragment_len)
3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891
            TRACE("(%p %p %x): Valid fragment component %s len=%d.\n", builder, data, flags,
                debugstr_wn(data->fragment, data->fragment_len), data->fragment_len);
        else {
            TRACE("(%p %p %x): Invalid fragment component %s.\n", builder, data, flags,
                debugstr_wn(component, expected_len));
            return INET_E_INVALID_URL;
        }
    }

    return S_OK;
}

3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902
static HRESULT validate_components(const UriBuilder *builder, parse_data *data, DWORD flags) {
    HRESULT hr;

    memset(data, 0, sizeof(parse_data));

    TRACE("(%p %p %x): Beginning to validate builder components.\n", builder, data, flags);

    hr = validate_scheme_name(builder, data, flags);
    if(FAILED(hr))
        return hr;

3903 3904 3905 3906 3907 3908 3909 3910 3911 3912
    /* Extra validation for file schemes. */
    if(data->scheme_type == URL_SCHEME_FILE) {
        if((builder->password || (builder->uri && builder->uri->userinfo_split > -1)) ||
           (builder->username || (builder->uri && builder->uri->userinfo_start > -1))) {
            TRACE("(%p %p %x): File schemes can't contain a username or password.\n",
                builder, data, flags);
            return INET_E_INVALID_URL;
        }
    }

3913 3914 3915 3916
    hr = validate_userinfo(builder, data, flags);
    if(FAILED(hr))
        return hr;

3917 3918 3919 3920
    hr = validate_host(builder, data, flags);
    if(FAILED(hr))
        return hr;

3921 3922
    setup_port(builder, data, flags);

3923 3924 3925 3926 3927 3928
    /* The URI is opaque if it doesn't have an authority component. */
    if(!data->is_relative)
        data->is_opaque = !data->username && !data->password && !data->host && !data->has_port;
    else
        data->is_opaque = !data->host && !data->has_port;

3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943
    hr = validate_path(builder, data, flags);
    if(FAILED(hr))
        return hr;

    hr = validate_query(builder, data, flags);
    if(FAILED(hr))
        return hr;

    hr = validate_fragment(builder, data, flags);
    if(FAILED(hr))
        return hr;

    TRACE("(%p %p %x): Finished validating builder components.\n", builder, data, flags);

    return S_OK;
3944 3945
}

3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967
static void convert_to_dos_path(const WCHAR *path, DWORD path_len,
                                WCHAR *output, DWORD *output_len)
{
    const WCHAR *ptr = path;

    if(path_len > 3 && *ptr == '/' && is_drive_path(path+1))
        /* Skip over the leading / before the drive path. */
        ++ptr;

    for(; ptr < path+path_len; ++ptr) {
        if(*ptr == '/') {
            if(output)
                *output++ = '\\';
            (*output_len)++;
        } else {
            if(output)
                *output++ = *ptr;
            (*output_len)++;
        }
    }
}

3968
/* Generates a raw uri string using the parse_data. */
3969
static DWORD generate_raw_uri(const parse_data *data, BSTR uri, DWORD flags) {
3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986
    DWORD length = 0;

    if(data->scheme) {
        if(uri) {
            memcpy(uri, data->scheme, data->scheme_len*sizeof(WCHAR));
            uri[data->scheme_len] = ':';
        }
        length += data->scheme_len+1;
    }

    if(!data->is_opaque) {
        /* For the "//" which appears before the authority component. */
        if(uri) {
            uri[length] = '/';
            uri[length+1] = '/';
        }
        length += 2;
3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998

        /* Check if we need to add the "\\" before the host name
         * of a UNC server name in a DOS path.
         */
        if(flags & RAW_URI_CONVERT_TO_DOS_PATH &&
           data->scheme_type == URL_SCHEME_FILE && data->host) {
            if(uri) {
                uri[length] = '\\';
                uri[length+1] = '\\';
            }
            length += 2;
        }
3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055
    }

    if(data->username) {
        if(uri)
            memcpy(uri+length, data->username, data->username_len*sizeof(WCHAR));
        length += data->username_len;
    }

    if(data->password) {
        if(uri) {
            uri[length] = ':';
            memcpy(uri+length+1, data->password, data->password_len*sizeof(WCHAR));
        }
        length += data->password_len+1;
    }

    if(data->password || data->username) {
        if(uri)
            uri[length] = '@';
        ++length;
    }

    if(data->host) {
        /* IPv6 addresses get the brackets added around them if they don't already
         * have them.
         */
        const BOOL add_brackets = data->host_type == Uri_HOST_IPV6 && *(data->host) != '[';
        if(add_brackets) {
            if(uri)
                uri[length] = '[';
            ++length;
        }

        if(uri)
            memcpy(uri+length, data->host, data->host_len*sizeof(WCHAR));
        length += data->host_len;

        if(add_brackets) {
            if(uri)
                uri[length] = ']';
            length++;
        }
    }

    if(data->has_port) {
        /* The port isn't included in the raw uri if it's the default
         * port for the scheme type.
         */
        DWORD i;
        BOOL is_default = FALSE;

        for(i = 0; i < sizeof(default_ports)/sizeof(default_ports[0]); ++i) {
            if(data->scheme_type == default_ports[i].scheme &&
               data->port_value == default_ports[i].port)
                is_default = TRUE;
        }

4056
        if(!is_default || flags & RAW_URI_FORCE_PORT_DISP) {
4057 4058 4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075
            if(uri)
                uri[length] = ':';
            ++length;

            if(uri)
                length += ui2str(uri+length, data->port_value);
            else
                length += ui2str(NULL, data->port_value);
        }
    }

    /* Check if a '/' should be added before the path for hierarchical URIs. */
    if(!data->is_opaque && data->path && *(data->path) != '/') {
        if(uri)
            uri[length] = '/';
        ++length;
    }

    if(data->path) {
4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090
        if(!data->is_opaque && data->scheme_type == URL_SCHEME_FILE &&
           flags & RAW_URI_CONVERT_TO_DOS_PATH) {
            DWORD len = 0;

            if(uri)
                convert_to_dos_path(data->path, data->path_len, uri+length, &len);
            else
                convert_to_dos_path(data->path, data->path_len, NULL, &len);

            length += len;
        } else {
            if(uri)
                memcpy(uri+length, data->path, data->path_len*sizeof(WCHAR));
            length += data->path_len;
        }
4091 4092 4093 4094 4095 4096 4097 4098 4099 4100 4101 4102 4103 4104 4105 4106 4107 4108 4109 4110 4111 4112 4113
    }

    if(data->query) {
        if(uri)
            memcpy(uri+length, data->query, data->query_len*sizeof(WCHAR));
        length += data->query_len;
    }

    if(data->fragment) {
        if(uri)
            memcpy(uri+length, data->fragment, data->fragment_len*sizeof(WCHAR));
        length += data->fragment_len;
    }

    if(uri)
        TRACE("(%p %p): Generated raw uri=%s len=%d\n", data, uri, debugstr_wn(uri, length), length);
    else
        TRACE("(%p %p): Computed raw uri len=%d\n", data, uri, length);

    return length;
}

static HRESULT generate_uri(const UriBuilder *builder, const parse_data *data, Uri *uri, DWORD flags) {
4114
    HRESULT hr;
4115
    DWORD length = generate_raw_uri(data, NULL, 0);
4116 4117 4118 4119
    uri->raw_uri = SysAllocStringLen(NULL, length);
    if(!uri->raw_uri)
        return E_OUTOFMEMORY;

4120
    generate_raw_uri(data, uri->raw_uri, 0);
4121

4122
    hr = canonicalize_uri(data, uri, flags);
4123
    if(FAILED(hr)) {
4124 4125
        if(hr == E_INVALIDARG)
            return INET_E_INVALID_URL;
4126 4127 4128
        return hr;
    }

4129
    uri->create_flags = flags;
4130
    return S_OK;
4131 4132
}

4133 4134 4135 4136
static inline Uri* impl_from_IUri(IUri *iface)
{
    return CONTAINING_RECORD(iface, Uri, IUri_iface);
}
4137

4138 4139 4140 4141 4142 4143 4144
static inline void destory_uri_obj(Uri *This)
{
    SysFreeString(This->raw_uri);
    heap_free(This->canon_uri);
    heap_free(This);
}

4145 4146
static HRESULT WINAPI Uri_QueryInterface(IUri *iface, REFIID riid, void **ppv)
{
4147
    Uri *This = impl_from_IUri(iface);
4148 4149 4150

    if(IsEqualGUID(&IID_IUnknown, riid)) {
        TRACE("(%p)->(IID_IUnknown %p)\n", This, ppv);
4151
        *ppv = &This->IUri_iface;
4152 4153
    }else if(IsEqualGUID(&IID_IUri, riid)) {
        TRACE("(%p)->(IID_IUri %p)\n", This, ppv);
4154
        *ppv = &This->IUri_iface;
4155 4156 4157
    }else if(IsEqualGUID(&IID_IUriBuilderFactory, riid)) {
        TRACE("(%p)->(IID_IUriBuilderFactory %p)\n", This, riid);
        *ppv = &This->IUriBuilderFactory_iface;
4158 4159 4160 4161
    }else if(IsEqualGUID(&IID_IUriObj, riid)) {
        TRACE("(%p)->(IID_IUriObj %p)\n", This, ppv);
        *ppv = This;
        return S_OK;
4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173
    }else {
        TRACE("(%p)->(%s %p)\n", This, debugstr_guid(riid), ppv);
        *ppv = NULL;
        return E_NOINTERFACE;
    }

    IUnknown_AddRef((IUnknown*)*ppv);
    return S_OK;
}

static ULONG WINAPI Uri_AddRef(IUri *iface)
{
4174
    Uri *This = impl_from_IUri(iface);
4175 4176 4177 4178 4179 4180 4181 4182 4183
    LONG ref = InterlockedIncrement(&This->ref);

    TRACE("(%p) ref=%d\n", This, ref);

    return ref;
}

static ULONG WINAPI Uri_Release(IUri *iface)
{
4184
    Uri *This = impl_from_IUri(iface);
4185 4186 4187 4188
    LONG ref = InterlockedDecrement(&This->ref);

    TRACE("(%p) ref=%d\n", This, ref);

4189 4190
    if(!ref)
        destory_uri_obj(This);
4191 4192 4193 4194 4195 4196

    return ref;
}

static HRESULT WINAPI Uri_GetPropertyBSTR(IUri *iface, Uri_PROPERTY uriProp, BSTR *pbstrProperty, DWORD dwFlags)
{
4197
    Uri *This = impl_from_IUri(iface);
4198 4199
    HRESULT hres;
    TRACE("(%p)->(%d %p %x)\n", This, uriProp, pbstrProperty, dwFlags);
4200 4201 4202 4203 4204 4205 4206

    if(!pbstrProperty)
        return E_POINTER;

    if(uriProp > Uri_PROPERTY_STRING_LAST) {
        /* Windows allocates an empty BSTR for invalid Uri_PROPERTY's. */
        *pbstrProperty = SysAllocStringLen(NULL, 0);
4207 4208
        if(!(*pbstrProperty))
            return E_OUTOFMEMORY;
4209 4210 4211 4212 4213 4214 4215 4216

        /* It only returns S_FALSE for the ZONE property... */
        if(uriProp == Uri_PROPERTY_ZONE)
            return S_FALSE;
        else
            return S_OK;
    }

4217 4218 4219 4220 4221 4222 4223
    /* Don't have support for flags yet. */
    if(dwFlags) {
        FIXME("(%p)->(%d %p %x)\n", This, uriProp, pbstrProperty, dwFlags);
        return E_NOTIMPL;
    }

    switch(uriProp) {
4224
    case Uri_PROPERTY_ABSOLUTE_URI:
4225
        if(This->display_modifiers & URI_DISPLAY_NO_ABSOLUTE_URI) {
4226 4227 4228
            *pbstrProperty = SysAllocStringLen(NULL, 0);
            hres = S_FALSE;
        } else {
4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258
            if(This->scheme_type != URL_SCHEME_UNKNOWN && This->userinfo_start > -1) {
                if(This->userinfo_len == 0) {
                    /* Don't include the '@' after the userinfo component. */
                    *pbstrProperty = SysAllocStringLen(NULL, This->canon_len-1);
                    hres = S_OK;
                    if(*pbstrProperty) {
                        /* Copy everything before it. */
                        memcpy(*pbstrProperty, This->canon_uri, This->userinfo_start*sizeof(WCHAR));

                        /* And everything after it. */
                        memcpy(*pbstrProperty+This->userinfo_start, This->canon_uri+This->userinfo_start+1,
                               (This->canon_len-This->userinfo_start-1)*sizeof(WCHAR));
                    }
                } else if(This->userinfo_split == 0 && This->userinfo_len == 1) {
                    /* Don't include the ":@" */
                    *pbstrProperty = SysAllocStringLen(NULL, This->canon_len-2);
                    hres = S_OK;
                    if(*pbstrProperty) {
                        memcpy(*pbstrProperty, This->canon_uri, This->userinfo_start*sizeof(WCHAR));
                        memcpy(*pbstrProperty+This->userinfo_start, This->canon_uri+This->userinfo_start+2,
                               (This->canon_len-This->userinfo_start-2)*sizeof(WCHAR));
                    }
                } else {
                    *pbstrProperty = SysAllocString(This->canon_uri);
                    hres = S_OK;
                }
            } else {
                *pbstrProperty = SysAllocString(This->canon_uri);
                hres = S_OK;
            }
4259
        }
4260 4261 4262 4263 4264

        if(!(*pbstrProperty))
            hres = E_OUTOFMEMORY;

        break;
4265 4266
    case Uri_PROPERTY_AUTHORITY:
        if(This->authority_start > -1) {
4267 4268 4269 4270 4271 4272
            if(This->port_offset > -1 && is_default_port(This->scheme_type, This->port) &&
               This->display_modifiers & URI_DISPLAY_NO_DEFAULT_PORT_AUTH)
                /* Don't include the port in the authority component. */
                *pbstrProperty = SysAllocStringLen(This->canon_uri+This->authority_start, This->port_offset);
            else
                *pbstrProperty = SysAllocStringLen(This->canon_uri+This->authority_start, This->authority_len);
4273 4274 4275 4276 4277 4278
            hres = S_OK;
        } else {
            *pbstrProperty = SysAllocStringLen(NULL, 0);
            hres = S_FALSE;
        }

4279 4280 4281
        if(!(*pbstrProperty))
            hres = E_OUTOFMEMORY;

4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305
        break;
    case Uri_PROPERTY_DISPLAY_URI:
        /* The Display URI contains everything except for the userinfo for known
         * scheme types.
         */
        if(This->scheme_type != URL_SCHEME_UNKNOWN && This->userinfo_start > -1) {
            *pbstrProperty = SysAllocStringLen(NULL, This->canon_len-This->userinfo_len);

            if(*pbstrProperty) {
                /* Copy everything before the userinfo over. */
                memcpy(*pbstrProperty, This->canon_uri, This->userinfo_start*sizeof(WCHAR));
                /* Copy everything after the userinfo over. */
                memcpy(*pbstrProperty+This->userinfo_start,
                   This->canon_uri+This->userinfo_start+This->userinfo_len+1,
                   (This->canon_len-(This->userinfo_start+This->userinfo_len+1))*sizeof(WCHAR));
            }
        } else
            *pbstrProperty = SysAllocString(This->canon_uri);

        if(!(*pbstrProperty))
            hres = E_OUTOFMEMORY;
        else
            hres = S_OK;

4306 4307 4308 4309 4310 4311 4312 4313 4314 4315 4316
        break;
    case Uri_PROPERTY_DOMAIN:
        if(This->domain_offset > -1) {
            *pbstrProperty = SysAllocStringLen(This->canon_uri+This->host_start+This->domain_offset,
                                               This->host_len-This->domain_offset);
            hres = S_OK;
        } else {
            *pbstrProperty = SysAllocStringLen(NULL, 0);
            hres = S_FALSE;
        }

4317 4318 4319 4320 4321 4322 4323 4324 4325 4326 4327 4328 4329 4330
        if(!(*pbstrProperty))
            hres = E_OUTOFMEMORY;

        break;
    case Uri_PROPERTY_EXTENSION:
        if(This->extension_offset > -1) {
            *pbstrProperty = SysAllocStringLen(This->canon_uri+This->path_start+This->extension_offset,
                                               This->path_len-This->extension_offset);
            hres = S_OK;
        } else {
            *pbstrProperty = SysAllocStringLen(NULL, 0);
            hres = S_FALSE;
        }

4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343
        if(!(*pbstrProperty))
            hres = E_OUTOFMEMORY;

        break;
    case Uri_PROPERTY_FRAGMENT:
        if(This->fragment_start > -1) {
            *pbstrProperty = SysAllocStringLen(This->canon_uri+This->fragment_start, This->fragment_len);
            hres = S_OK;
        } else {
            *pbstrProperty = SysAllocStringLen(NULL, 0);
            hres = S_FALSE;
        }

4344 4345 4346 4347
        if(!(*pbstrProperty))
            hres = E_OUTOFMEMORY;

        break;
4348 4349
    case Uri_PROPERTY_HOST:
        if(This->host_start > -1) {
4350 4351 4352 4353 4354 4355
            /* The '[' and ']' aren't included for IPv6 addresses. */
            if(This->host_type == Uri_HOST_IPV6)
                *pbstrProperty = SysAllocStringLen(This->canon_uri+This->host_start+1, This->host_len-2);
            else
                *pbstrProperty = SysAllocStringLen(This->canon_uri+This->host_start, This->host_len);

4356 4357
            hres = S_OK;
        } else {
4358 4359
            *pbstrProperty = SysAllocStringLen(NULL, 0);
            hres = S_FALSE;
4360 4361 4362 4363 4364 4365
        }

        if(!(*pbstrProperty))
            hres = E_OUTOFMEMORY;

        break;
4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379
    case Uri_PROPERTY_PASSWORD:
        if(This->userinfo_split > -1) {
            *pbstrProperty = SysAllocStringLen(
                This->canon_uri+This->userinfo_start+This->userinfo_split+1,
                This->userinfo_len-This->userinfo_split-1);
            hres = S_OK;
        } else {
            *pbstrProperty = SysAllocStringLen(NULL, 0);
            hres = S_FALSE;
        }

        if(!(*pbstrProperty))
            return E_OUTOFMEMORY;

4380 4381 4382 4383 4384 4385 4386 4387 4388 4389
        break;
    case Uri_PROPERTY_PATH:
        if(This->path_start > -1) {
            *pbstrProperty = SysAllocStringLen(This->canon_uri+This->path_start, This->path_len);
            hres = S_OK;
        } else {
            *pbstrProperty = SysAllocStringLen(NULL, 0);
            hres = S_FALSE;
        }

4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 4400 4401 4402 4403 4404 4405
        if(!(*pbstrProperty))
            hres = E_OUTOFMEMORY;

        break;
    case Uri_PROPERTY_PATH_AND_QUERY:
        if(This->path_start > -1) {
            *pbstrProperty = SysAllocStringLen(This->canon_uri+This->path_start, This->path_len+This->query_len);
            hres = S_OK;
        } else if(This->query_start > -1) {
            *pbstrProperty = SysAllocStringLen(This->canon_uri+This->query_start, This->query_len);
            hres = S_OK;
        } else {
            *pbstrProperty = SysAllocStringLen(NULL, 0);
            hres = S_FALSE;
        }

4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418
        if(!(*pbstrProperty))
            hres = E_OUTOFMEMORY;

        break;
    case Uri_PROPERTY_QUERY:
        if(This->query_start > -1) {
            *pbstrProperty = SysAllocStringLen(This->canon_uri+This->query_start, This->query_len);
            hres = S_OK;
        } else {
            *pbstrProperty = SysAllocStringLen(NULL, 0);
            hres = S_FALSE;
        }

4419 4420 4421
        if(!(*pbstrProperty))
            hres = E_OUTOFMEMORY;

4422
        break;
4423 4424 4425 4426 4427 4428 4429
    case Uri_PROPERTY_RAW_URI:
        *pbstrProperty = SysAllocString(This->raw_uri);
        if(!(*pbstrProperty))
            hres = E_OUTOFMEMORY;
        else
            hres = S_OK;
        break;
4430 4431 4432 4433 4434 4435 4436 4437 4438
    case Uri_PROPERTY_SCHEME_NAME:
        if(This->scheme_start > -1) {
            *pbstrProperty = SysAllocStringLen(This->canon_uri + This->scheme_start, This->scheme_len);
            hres = S_OK;
        } else {
            *pbstrProperty = SysAllocStringLen(NULL, 0);
            hres = S_FALSE;
        }

4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451
        if(!(*pbstrProperty))
            hres = E_OUTOFMEMORY;

        break;
    case Uri_PROPERTY_USER_INFO:
        if(This->userinfo_start > -1) {
            *pbstrProperty = SysAllocStringLen(This->canon_uri+This->userinfo_start, This->userinfo_len);
            hres = S_OK;
        } else {
            *pbstrProperty = SysAllocStringLen(NULL, 0);
            hres = S_FALSE;
        }

4452 4453 4454
        if(!(*pbstrProperty))
            hres = E_OUTOFMEMORY;

4455 4456
        break;
    case Uri_PROPERTY_USER_NAME:
4457
        if(This->userinfo_start > -1 && This->userinfo_split != 0) {
4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475
            /* If userinfo_split is set, that means a password exists
             * so the username is only from userinfo_start to userinfo_split.
             */
            if(This->userinfo_split > -1) {
                *pbstrProperty = SysAllocStringLen(This->canon_uri + This->userinfo_start, This->userinfo_split);
                hres = S_OK;
            } else {
                *pbstrProperty = SysAllocStringLen(This->canon_uri + This->userinfo_start, This->userinfo_len);
                hres = S_OK;
            }
        } else {
            *pbstrProperty = SysAllocStringLen(NULL, 0);
            hres = S_FALSE;
        }

        if(!(*pbstrProperty))
            return E_OUTOFMEMORY;

4476
        break;
4477 4478 4479 4480 4481 4482
    default:
        FIXME("(%p)->(%d %p %x)\n", This, uriProp, pbstrProperty, dwFlags);
        hres = E_NOTIMPL;
    }

    return hres;
4483 4484 4485 4486
}

static HRESULT WINAPI Uri_GetPropertyLength(IUri *iface, Uri_PROPERTY uriProp, DWORD *pcchProperty, DWORD dwFlags)
{
4487
    Uri *This = impl_from_IUri(iface);
4488 4489
    HRESULT hres;
    TRACE("(%p)->(%d %p %x)\n", This, uriProp, pcchProperty, dwFlags);
4490 4491 4492 4493 4494 4495 4496 4497

    if(!pcchProperty)
        return E_INVALIDARG;

    /* Can only return a length for a property if it's a string. */
    if(uriProp > Uri_PROPERTY_STRING_LAST)
        return E_INVALIDARG;

4498 4499 4500 4501 4502 4503 4504
    /* Don't have support for flags yet. */
    if(dwFlags) {
        FIXME("(%p)->(%d %p %x)\n", This, uriProp, pcchProperty, dwFlags);
        return E_NOTIMPL;
    }

    switch(uriProp) {
4505
    case Uri_PROPERTY_ABSOLUTE_URI:
4506
        if(This->display_modifiers & URI_DISPLAY_NO_ABSOLUTE_URI) {
4507 4508 4509
            *pcchProperty = 0;
            hres = S_FALSE;
        } else {
4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522
            if(This->scheme_type != URL_SCHEME_UNKNOWN) {
                if(This->userinfo_start > -1 && This->userinfo_len == 0)
                    /* Don't include the '@' in the length. */
                    *pcchProperty = This->canon_len-1;
                else if(This->userinfo_start > -1 && This->userinfo_len == 1 &&
                        This->userinfo_split == 0)
                    /* Don't include the ":@" in the length. */
                    *pcchProperty = This->canon_len-2;
                else
                    *pcchProperty = This->canon_len;
            } else
                *pcchProperty = This->canon_len;

4523 4524 4525
            hres = S_OK;
        }

4526
        break;
4527
    case Uri_PROPERTY_AUTHORITY:
4528 4529 4530 4531 4532 4533 4534
        if(This->port_offset > -1 &&
           This->display_modifiers & URI_DISPLAY_NO_DEFAULT_PORT_AUTH &&
           is_default_port(This->scheme_type, This->port))
            /* Only count up until the port in the authority. */
            *pcchProperty = This->port_offset;
        else
            *pcchProperty = This->authority_len;
4535 4536
        hres = (This->authority_start > -1) ? S_OK : S_FALSE;
        break;
4537 4538 4539 4540 4541 4542 4543 4544
    case Uri_PROPERTY_DISPLAY_URI:
        if(This->scheme_type != URL_SCHEME_UNKNOWN && This->userinfo_start > -1)
            *pcchProperty = This->canon_len-This->userinfo_len-1;
        else
            *pcchProperty = This->canon_len;

        hres = S_OK;
        break;
4545 4546 4547 4548 4549 4550 4551
    case Uri_PROPERTY_DOMAIN:
        if(This->domain_offset > -1)
            *pcchProperty = This->host_len - This->domain_offset;
        else
            *pcchProperty = 0;

        hres = (This->domain_offset > -1) ? S_OK : S_FALSE;
4552 4553 4554 4555 4556 4557 4558 4559 4560 4561
        break;
    case Uri_PROPERTY_EXTENSION:
        if(This->extension_offset > -1) {
            *pcchProperty = This->path_len - This->extension_offset;
            hres = S_OK;
        } else {
            *pcchProperty = 0;
            hres = S_FALSE;
        }

4562 4563 4564 4565
        break;
    case Uri_PROPERTY_FRAGMENT:
        *pcchProperty = This->fragment_len;
        hres = (This->fragment_start > -1) ? S_OK : S_FALSE;
4566
        break;
4567 4568
    case Uri_PROPERTY_HOST:
        *pcchProperty = This->host_len;
4569 4570 4571 4572 4573

        /* '[' and ']' aren't included in the length. */
        if(This->host_type == Uri_HOST_IPV6)
            *pcchProperty -= 2;

4574 4575
        hres = (This->host_start > -1) ? S_OK : S_FALSE;
        break;
4576 4577 4578 4579
    case Uri_PROPERTY_PASSWORD:
        *pcchProperty = (This->userinfo_split > -1) ? This->userinfo_len-This->userinfo_split-1 : 0;
        hres = (This->userinfo_split > -1) ? S_OK : S_FALSE;
        break;
4580 4581 4582 4583
    case Uri_PROPERTY_PATH:
        *pcchProperty = This->path_len;
        hres = (This->path_start > -1) ? S_OK : S_FALSE;
        break;
4584 4585 4586 4587
    case Uri_PROPERTY_PATH_AND_QUERY:
        *pcchProperty = This->path_len+This->query_len;
        hres = (This->path_start > -1 || This->query_start > -1) ? S_OK : S_FALSE;
        break;
4588 4589 4590 4591
    case Uri_PROPERTY_QUERY:
        *pcchProperty = This->query_len;
        hres = (This->query_start > -1) ? S_OK : S_FALSE;
        break;
4592 4593 4594 4595
    case Uri_PROPERTY_RAW_URI:
        *pcchProperty = SysStringLen(This->raw_uri);
        hres = S_OK;
        break;
4596 4597 4598 4599
    case Uri_PROPERTY_SCHEME_NAME:
        *pcchProperty = This->scheme_len;
        hres = (This->scheme_start > -1) ? S_OK : S_FALSE;
        break;
4600 4601 4602 4603
    case Uri_PROPERTY_USER_INFO:
        *pcchProperty = This->userinfo_len;
        hres = (This->userinfo_start > -1) ? S_OK : S_FALSE;
        break;
4604 4605
    case Uri_PROPERTY_USER_NAME:
        *pcchProperty = (This->userinfo_split > -1) ? This->userinfo_split : This->userinfo_len;
4606 4607 4608 4609
        if(This->userinfo_split == 0)
            hres = S_FALSE;
        else
            hres = (This->userinfo_start > -1) ? S_OK : S_FALSE;
4610
        break;
4611 4612 4613 4614 4615 4616
    default:
        FIXME("(%p)->(%d %p %x)\n", This, uriProp, pcchProperty, dwFlags);
        hres = E_NOTIMPL;
    }

    return hres;
4617 4618 4619 4620
}

static HRESULT WINAPI Uri_GetPropertyDWORD(IUri *iface, Uri_PROPERTY uriProp, DWORD *pcchProperty, DWORD dwFlags)
{
4621
    Uri *This = impl_from_IUri(iface);
4622 4623
    HRESULT hres;

4624
    TRACE("(%p)->(%d %p %x)\n", This, uriProp, pcchProperty, dwFlags);
4625 4626 4627 4628 4629 4630

    if(!pcchProperty)
        return E_INVALIDARG;

    /* Microsoft's implementation for the ZONE property of a URI seems to be lacking...
     * From what I can tell, instead of checking which URLZONE the URI belongs to it
4631
     * simply assigns URLZONE_INVALID and returns E_NOTIMPL. This also applies to the GetZone
4632 4633 4634 4635 4636 4637 4638
     * function.
     */
    if(uriProp == Uri_PROPERTY_ZONE) {
        *pcchProperty = URLZONE_INVALID;
        return E_NOTIMPL;
    }

4639 4640 4641 4642 4643
    if(uriProp < Uri_PROPERTY_DWORD_START) {
        *pcchProperty = 0;
        return E_INVALIDARG;
    }

4644
    switch(uriProp) {
4645 4646
    case Uri_PROPERTY_HOST_TYPE:
        *pcchProperty = This->host_type;
4647
        hres = S_OK;
4648 4649 4650 4651 4652 4653 4654 4655 4656 4657
        break;
    case Uri_PROPERTY_PORT:
        if(!This->has_port) {
            *pcchProperty = 0;
            hres = S_FALSE;
        } else {
            *pcchProperty = This->port;
            hres = S_OK;
        }

4658 4659 4660 4661
        break;
    case Uri_PROPERTY_SCHEME:
        *pcchProperty = This->scheme_type;
        hres = S_OK;
4662 4663 4664 4665 4666 4667 4668
        break;
    default:
        FIXME("(%p)->(%d %p %x)\n", This, uriProp, pcchProperty, dwFlags);
        hres = E_NOTIMPL;
    }

    return hres;
4669 4670 4671 4672
}

static HRESULT WINAPI Uri_HasProperty(IUri *iface, Uri_PROPERTY uriProp, BOOL *pfHasProperty)
{
4673
    Uri *This = impl_from_IUri(iface);
4674
    TRACE("(%p)->(%d %p)\n", This, uriProp, pfHasProperty);
4675 4676 4677 4678

    if(!pfHasProperty)
        return E_INVALIDARG;

4679 4680
    switch(uriProp) {
    case Uri_PROPERTY_ABSOLUTE_URI:
4681
        *pfHasProperty = !(This->display_modifiers & URI_DISPLAY_NO_ABSOLUTE_URI);
4682 4683 4684 4685 4686 4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701 4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714 4715 4716 4717 4718 4719 4720 4721
        break;
    case Uri_PROPERTY_AUTHORITY:
        *pfHasProperty = This->authority_start > -1;
        break;
    case Uri_PROPERTY_DISPLAY_URI:
        *pfHasProperty = TRUE;
        break;
    case Uri_PROPERTY_DOMAIN:
        *pfHasProperty = This->domain_offset > -1;
        break;
    case Uri_PROPERTY_EXTENSION:
        *pfHasProperty = This->extension_offset > -1;
        break;
    case Uri_PROPERTY_FRAGMENT:
        *pfHasProperty = This->fragment_start > -1;
        break;
    case Uri_PROPERTY_HOST:
        *pfHasProperty = This->host_start > -1;
        break;
    case Uri_PROPERTY_PASSWORD:
        *pfHasProperty = This->userinfo_split > -1;
        break;
    case Uri_PROPERTY_PATH:
        *pfHasProperty = This->path_start > -1;
        break;
    case Uri_PROPERTY_PATH_AND_QUERY:
        *pfHasProperty = (This->path_start > -1 || This->query_start > -1);
        break;
    case Uri_PROPERTY_QUERY:
        *pfHasProperty = This->query_start > -1;
        break;
    case Uri_PROPERTY_RAW_URI:
        *pfHasProperty = TRUE;
        break;
    case Uri_PROPERTY_SCHEME_NAME:
        *pfHasProperty = This->scheme_start > -1;
        break;
    case Uri_PROPERTY_USER_INFO:
        *pfHasProperty = This->userinfo_start > -1;
        break;
4722 4723 4724 4725 4726 4727
    case Uri_PROPERTY_USER_NAME:
        if(This->userinfo_split == 0)
            *pfHasProperty = FALSE;
        else
            *pfHasProperty = This->userinfo_start > -1;
        break;
4728 4729 4730 4731 4732 4733 4734 4735 4736 4737 4738 4739 4740 4741 4742 4743 4744 4745
    case Uri_PROPERTY_HOST_TYPE:
        *pfHasProperty = TRUE;
        break;
    case Uri_PROPERTY_PORT:
        *pfHasProperty = This->has_port;
        break;
    case Uri_PROPERTY_SCHEME:
        *pfHasProperty = TRUE;
        break;
    case Uri_PROPERTY_ZONE:
        *pfHasProperty = FALSE;
        break;
    default:
        FIXME("(%p)->(%d %p): Unsupported property type.\n", This, uriProp, pfHasProperty);
        return E_NOTIMPL;
    }

    return S_OK;
4746 4747 4748 4749
}

static HRESULT WINAPI Uri_GetAbsoluteUri(IUri *iface, BSTR *pstrAbsoluteUri)
{
4750
    TRACE("(%p)->(%p)\n", iface, pstrAbsoluteUri);
4751
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_ABSOLUTE_URI, pstrAbsoluteUri, 0);
4752 4753 4754 4755
}

static HRESULT WINAPI Uri_GetAuthority(IUri *iface, BSTR *pstrAuthority)
{
4756
    TRACE("(%p)->(%p)\n", iface, pstrAuthority);
4757
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_AUTHORITY, pstrAuthority, 0);
4758 4759 4760 4761
}

static HRESULT WINAPI Uri_GetDisplayUri(IUri *iface, BSTR *pstrDisplayUri)
{
4762
    TRACE("(%p)->(%p)\n", iface, pstrDisplayUri);
4763
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_DISPLAY_URI, pstrDisplayUri, 0);
4764 4765 4766 4767
}

static HRESULT WINAPI Uri_GetDomain(IUri *iface, BSTR *pstrDomain)
{
4768
    TRACE("(%p)->(%p)\n", iface, pstrDomain);
4769
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_DOMAIN, pstrDomain, 0);
4770 4771 4772 4773
}

static HRESULT WINAPI Uri_GetExtension(IUri *iface, BSTR *pstrExtension)
{
4774
    TRACE("(%p)->(%p)\n", iface, pstrExtension);
4775
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_EXTENSION, pstrExtension, 0);
4776 4777 4778 4779
}

static HRESULT WINAPI Uri_GetFragment(IUri *iface, BSTR *pstrFragment)
{
4780
    TRACE("(%p)->(%p)\n", iface, pstrFragment);
4781
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_FRAGMENT, pstrFragment, 0);
4782 4783 4784 4785
}

static HRESULT WINAPI Uri_GetHost(IUri *iface, BSTR *pstrHost)
{
4786
    TRACE("(%p)->(%p)\n", iface, pstrHost);
4787
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_HOST, pstrHost, 0);
4788 4789 4790 4791
}

static HRESULT WINAPI Uri_GetPassword(IUri *iface, BSTR *pstrPassword)
{
4792
    TRACE("(%p)->(%p)\n", iface, pstrPassword);
4793
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_PASSWORD, pstrPassword, 0);
4794 4795 4796 4797
}

static HRESULT WINAPI Uri_GetPath(IUri *iface, BSTR *pstrPath)
{
4798
    TRACE("(%p)->(%p)\n", iface, pstrPath);
4799
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_PATH, pstrPath, 0);
4800 4801 4802 4803
}

static HRESULT WINAPI Uri_GetPathAndQuery(IUri *iface, BSTR *pstrPathAndQuery)
{
4804
    TRACE("(%p)->(%p)\n", iface, pstrPathAndQuery);
4805
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_PATH_AND_QUERY, pstrPathAndQuery, 0);
4806 4807 4808 4809
}

static HRESULT WINAPI Uri_GetQuery(IUri *iface, BSTR *pstrQuery)
{
4810
    TRACE("(%p)->(%p)\n", iface, pstrQuery);
4811
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_QUERY, pstrQuery, 0);
4812 4813 4814 4815
}

static HRESULT WINAPI Uri_GetRawUri(IUri *iface, BSTR *pstrRawUri)
{
4816
    TRACE("(%p)->(%p)\n", iface, pstrRawUri);
4817
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_RAW_URI, pstrRawUri, 0);
4818 4819 4820 4821
}

static HRESULT WINAPI Uri_GetSchemeName(IUri *iface, BSTR *pstrSchemeName)
{
4822
    TRACE("(%p)->(%p)\n", iface, pstrSchemeName);
4823
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_SCHEME_NAME, pstrSchemeName, 0);
4824 4825 4826 4827
}

static HRESULT WINAPI Uri_GetUserInfo(IUri *iface, BSTR *pstrUserInfo)
{
4828
    TRACE("(%p)->(%p)\n", iface, pstrUserInfo);
4829
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_USER_INFO, pstrUserInfo, 0);
4830 4831 4832 4833
}

static HRESULT WINAPI Uri_GetUserName(IUri *iface, BSTR *pstrUserName)
{
4834
    TRACE("(%p)->(%p)\n", iface, pstrUserName);
4835
    return IUri_GetPropertyBSTR(iface, Uri_PROPERTY_USER_NAME, pstrUserName, 0);
4836 4837 4838 4839
}

static HRESULT WINAPI Uri_GetHostType(IUri *iface, DWORD *pdwHostType)
{
4840
    TRACE("(%p)->(%p)\n", iface, pdwHostType);
4841
    return IUri_GetPropertyDWORD(iface, Uri_PROPERTY_HOST_TYPE, pdwHostType, 0);
4842 4843 4844 4845
}

static HRESULT WINAPI Uri_GetPort(IUri *iface, DWORD *pdwPort)
{
4846
    TRACE("(%p)->(%p)\n", iface, pdwPort);
4847
    return IUri_GetPropertyDWORD(iface, Uri_PROPERTY_PORT, pdwPort, 0);
4848 4849 4850 4851
}

static HRESULT WINAPI Uri_GetScheme(IUri *iface, DWORD *pdwScheme)
{
4852
    TRACE("(%p)->(%p)\n", iface, pdwScheme);
4853
    return IUri_GetPropertyDWORD(iface, Uri_PROPERTY_SCHEME, pdwScheme, 0);
4854 4855 4856 4857
}

static HRESULT WINAPI Uri_GetZone(IUri *iface, DWORD *pdwZone)
{
4858
    TRACE("(%p)->(%p)\n", iface, pdwZone);
4859
    return IUri_GetPropertyDWORD(iface, Uri_PROPERTY_ZONE,pdwZone, 0);
4860 4861 4862 4863
}

static HRESULT WINAPI Uri_GetProperties(IUri *iface, DWORD *pdwProperties)
{
4864
    Uri *This = impl_from_IUri(iface);
4865
    TRACE("(%p)->(%p)\n", This, pdwProperties);
4866 4867 4868 4869

    if(!pdwProperties)
        return E_INVALIDARG;

4870
    /* All URIs have these. */
4871 4872
    *pdwProperties = Uri_HAS_DISPLAY_URI|Uri_HAS_RAW_URI|Uri_HAS_SCHEME|Uri_HAS_HOST_TYPE;

4873
    if(!(This->display_modifiers & URI_DISPLAY_NO_ABSOLUTE_URI))
4874
        *pdwProperties |= Uri_HAS_ABSOLUTE_URI;
4875 4876 4877 4878 4879 4880

    if(This->scheme_start > -1)
        *pdwProperties |= Uri_HAS_SCHEME_NAME;

    if(This->authority_start > -1) {
        *pdwProperties |= Uri_HAS_AUTHORITY;
4881 4882 4883 4884 4885
        if(This->userinfo_start > -1) {
            *pdwProperties |= Uri_HAS_USER_INFO;
            if(This->userinfo_split != 0)
                *pdwProperties |= Uri_HAS_USER_NAME;
        }
4886 4887 4888 4889 4890 4891 4892 4893
        if(This->userinfo_split > -1)
            *pdwProperties |= Uri_HAS_PASSWORD;
        if(This->host_start > -1)
            *pdwProperties |= Uri_HAS_HOST;
        if(This->domain_offset > -1)
            *pdwProperties |= Uri_HAS_DOMAIN;
    }

4894 4895
    if(This->has_port)
        *pdwProperties |= Uri_HAS_PORT;
4896 4897 4898 4899 4900 4901 4902 4903 4904 4905 4906 4907
    if(This->path_start > -1)
        *pdwProperties |= Uri_HAS_PATH|Uri_HAS_PATH_AND_QUERY;
    if(This->query_start > -1)
        *pdwProperties |= Uri_HAS_QUERY|Uri_HAS_PATH_AND_QUERY;

    if(This->extension_offset > -1)
        *pdwProperties |= Uri_HAS_EXTENSION;

    if(This->fragment_start > -1)
        *pdwProperties |= Uri_HAS_FRAGMENT;

    return S_OK;
4908 4909 4910 4911
}

static HRESULT WINAPI Uri_IsEqual(IUri *iface, IUri *pUri, BOOL *pfEqual)
{
4912
    Uri *This = impl_from_IUri(iface);
4913 4914
    Uri *other;

4915 4916 4917 4918 4919 4920 4921 4922 4923 4924 4925 4926
    TRACE("(%p)->(%p %p)\n", This, pUri, pfEqual);

    if(!pfEqual)
        return E_POINTER;

    if(!pUri) {
        *pfEqual = FALSE;

        /* For some reason Windows returns S_OK here... */
        return S_OK;
    }

4927 4928 4929 4930 4931 4932 4933 4934 4935 4936
    /* Try to convert it to a Uri (allows for a more simple comparison). */
    if((other = get_uri_obj(pUri)))
        *pfEqual = are_equal_simple(This, other);
    else {
        /* Do it the hard way. */
        FIXME("(%p)->(%p %p) No support for unknown IUri's yet.\n", iface, pUri, pfEqual);
        return E_NOTIMPL;
    }

    return S_OK;
4937 4938 4939 4940 4941 4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969
}

static const IUriVtbl UriVtbl = {
    Uri_QueryInterface,
    Uri_AddRef,
    Uri_Release,
    Uri_GetPropertyBSTR,
    Uri_GetPropertyLength,
    Uri_GetPropertyDWORD,
    Uri_HasProperty,
    Uri_GetAbsoluteUri,
    Uri_GetAuthority,
    Uri_GetDisplayUri,
    Uri_GetDomain,
    Uri_GetExtension,
    Uri_GetFragment,
    Uri_GetHost,
    Uri_GetPassword,
    Uri_GetPath,
    Uri_GetPathAndQuery,
    Uri_GetQuery,
    Uri_GetRawUri,
    Uri_GetSchemeName,
    Uri_GetUserInfo,
    Uri_GetUserName,
    Uri_GetHostType,
    Uri_GetPort,
    Uri_GetScheme,
    Uri_GetZone,
    Uri_GetProperties,
    Uri_IsEqual
};

4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986 4987 4988 4989 4990 4991 4992 4993 4994 4995 4996 4997 4998 4999 5000 5001 5002 5003 5004 5005 5006 5007 5008 5009 5010 5011 5012 5013 5014 5015 5016 5017 5018 5019 5020
static inline Uri* impl_from_IUriBuilderFactory(IUriBuilderFactory *iface)
{
    return CONTAINING_RECORD(iface, Uri, IUriBuilderFactory_iface);
}

static HRESULT WINAPI UriBuilderFactory_QueryInterface(IUriBuilderFactory *iface, REFIID riid, void **ppv)
{
    Uri *This = impl_from_IUriBuilderFactory(iface);

    if(IsEqualGUID(&IID_IUnknown, riid)) {
        TRACE("(%p)->(IID_IUnknown %p)\n", This, ppv);
        *ppv = &This->IUriBuilderFactory_iface;
    }else if(IsEqualGUID(&IID_IUriBuilderFactory, riid)) {
        TRACE("(%p)->(IID_IUriBuilderFactory %p)\n", This, ppv);
        *ppv = &This->IUriBuilderFactory_iface;
    }else if(IsEqualGUID(&IID_IUri, riid)) {
        TRACE("(%p)->(IID_IUri %p)\n", This, ppv);
        *ppv = &This->IUri_iface;
    }else {
        TRACE("(%p)->(%s %p)\n", This, debugstr_guid(riid), ppv);
        *ppv = NULL;
        return E_NOINTERFACE;
    }

    IUnknown_AddRef((IUnknown*)*ppv);
    return S_OK;
}

static ULONG WINAPI UriBuilderFactory_AddRef(IUriBuilderFactory *iface)
{
    Uri *This = impl_from_IUriBuilderFactory(iface);
    LONG ref = InterlockedIncrement(&This->ref);

    TRACE("(%p) ref=%d\n", This, ref);

    return ref;
}

static ULONG WINAPI UriBuilderFactory_Release(IUriBuilderFactory *iface)
{
    Uri *This = impl_from_IUriBuilderFactory(iface);
    LONG ref = InterlockedDecrement(&This->ref);

    TRACE("(%p) ref=%d\n", This, ref);

    if(!ref)
        destory_uri_obj(This);

    return ref;
}

5021 5022 5023 5024
static HRESULT WINAPI UriBuilderFactory_CreateIUriBuilder(IUriBuilderFactory *iface,
                                                          DWORD dwFlags,
                                                          DWORD_PTR dwReserved,
                                                          IUriBuilder **ppIUriBuilder)
5025 5026 5027 5028 5029 5030 5031 5032 5033 5034 5035 5036 5037 5038 5039
{
    Uri *This = impl_from_IUriBuilderFactory(iface);
    TRACE("(%p)->(%08x %08x %p)\n", This, dwFlags, (DWORD)dwReserved, ppIUriBuilder);

    if(!ppIUriBuilder)
        return E_POINTER;

    if(dwFlags || dwReserved) {
        *ppIUriBuilder = NULL;
        return E_INVALIDARG;
    }

    return CreateIUriBuilder(NULL, 0, 0, ppIUriBuilder);
}

5040 5041 5042 5043
static HRESULT WINAPI UriBuilderFactory_CreateInitializedIUriBuilder(IUriBuilderFactory *iface,
                                                                     DWORD dwFlags,
                                                                     DWORD_PTR dwReserved,
                                                                     IUriBuilder **ppIUriBuilder)
5044 5045 5046 5047 5048 5049 5050 5051 5052 5053 5054 5055 5056 5057 5058 5059 5060 5061 5062
{
    Uri *This = impl_from_IUriBuilderFactory(iface);
    TRACE("(%p)->(%08x %08x %p)\n", This, dwFlags, (DWORD)dwReserved, ppIUriBuilder);

    if(!ppIUriBuilder)
        return E_POINTER;

    if(dwFlags || dwReserved) {
        *ppIUriBuilder = NULL;
        return E_INVALIDARG;
    }

    return CreateIUriBuilder(&This->IUri_iface, 0, 0, ppIUriBuilder);
}

static const IUriBuilderFactoryVtbl UriBuilderFactoryVtbl = {
    UriBuilderFactory_QueryInterface,
    UriBuilderFactory_AddRef,
    UriBuilderFactory_Release,
5063 5064
    UriBuilderFactory_CreateIUriBuilder,
    UriBuilderFactory_CreateInitializedIUriBuilder
5065 5066
};

5067 5068 5069
static Uri* create_uri_obj(void) {
    Uri *ret = heap_alloc_zero(sizeof(Uri));
    if(ret) {
5070
        ret->IUri_iface.lpVtbl = &UriVtbl;
5071
        ret->IUriBuilderFactory_iface.lpVtbl = &UriBuilderFactoryVtbl;
5072 5073 5074 5075 5076 5077
        ret->ref = 1;
    }

    return ret;
}

5078 5079
/***********************************************************************
 *           CreateUri (urlmon.@)
5080 5081 5082 5083 5084 5085 5086 5087 5088 5089 5090 5091 5092
 *
 * Creates a new IUri object using the URI represented by pwzURI. This function
 * parses and validates the components of pwzURI and then canonicalizes the
 * parsed components.
 *
 * PARAMS
 *  pwzURI      [I] The URI to parse, validate, and canonicalize.
 *  dwFlags     [I] Flags which can affect how the parsing/canonicalization is performed.
 *  dwReserved  [I] Reserved (not used).
 *  ppURI       [O] The resulting IUri after parsing/canonicalization occurs.
 *
 * RETURNS
 *  Success: Returns S_OK. ppURI contains the pointer to the newly allocated IUri.
5093 5094
 *  Failure: E_INVALIDARG if there are invalid flag combinations in dwFlags, or an
 *           invalid parameter, or pwzURI doesn't represent a valid URI.
5095 5096 5097 5098 5099 5100
 *           E_OUTOFMEMORY if any memory allocation fails.
 *
 * NOTES
 *  Default flags:
 *      Uri_CREATE_CANONICALIZE, Uri_CREATE_DECODE_EXTRA_INFO, Uri_CREATE_CRACK_UNKNOWN_SCHEMES,
 *      Uri_CREATE_PRE_PROCESS_HTML_URI, Uri_CREATE_NO_IE_SETTINGS.
5101 5102 5103
 */
HRESULT WINAPI CreateUri(LPCWSTR pwzURI, DWORD dwFlags, DWORD_PTR dwReserved, IUri **ppURI)
{
5104 5105 5106 5107
    const DWORD supported_flags = Uri_CREATE_ALLOW_RELATIVE|Uri_CREATE_ALLOW_IMPLICIT_WILDCARD_SCHEME|
        Uri_CREATE_ALLOW_IMPLICIT_FILE_SCHEME|Uri_CREATE_NO_CANONICALIZE|Uri_CREATE_CANONICALIZE|
        Uri_CREATE_DECODE_EXTRA_INFO|Uri_CREATE_NO_DECODE_EXTRA_INFO|Uri_CREATE_CRACK_UNKNOWN_SCHEMES|
        Uri_CREATE_NO_CRACK_UNKNOWN_SCHEMES|Uri_CREATE_PRE_PROCESS_HTML_URI|Uri_CREATE_NO_PRE_PROCESS_HTML_URI|
5108
        Uri_CREATE_NO_IE_SETTINGS|Uri_CREATE_NO_ENCODE_FORBIDDEN_CHARACTERS|Uri_CREATE_FILE_USE_DOS_PATH;
5109
    Uri *ret;
5110
    HRESULT hr;
5111
    parse_data data;
5112 5113 5114

    TRACE("(%s %x %x %p)\n", debugstr_w(pwzURI), dwFlags, (DWORD)dwReserved, ppURI);

5115 5116 5117
    if(!ppURI)
        return E_INVALIDARG;

5118
    if(!pwzURI) {
5119 5120 5121 5122
        *ppURI = NULL;
        return E_INVALIDARG;
    }

5123
    /* Check for invalid flags. */
5124
    if(has_invalid_flag_combination(dwFlags)) {
5125 5126 5127 5128
        *ppURI = NULL;
        return E_INVALIDARG;
    }

5129 5130
    /* Currently unsupported. */
    if(dwFlags & ~supported_flags)
5131
        FIXME("Ignoring unsupported flag(s) %x\n", dwFlags & ~supported_flags);
5132

5133 5134 5135
    ret = create_uri_obj();
    if(!ret) {
        *ppURI = NULL;
5136
        return E_OUTOFMEMORY;
5137
    }
5138

5139 5140 5141
    /* Explicitly set the default flags if it doesn't cause a flag conflict. */
    apply_default_flags(&dwFlags);

5142 5143 5144 5145 5146 5147
    /* Pre process the URI, unless told otherwise. */
    if(!(dwFlags & Uri_CREATE_NO_PRE_PROCESS_HTML_URI))
        ret->raw_uri = pre_process_uri(pwzURI);
    else
        ret->raw_uri = SysAllocString(pwzURI);

5148 5149 5150 5151 5152
    if(!ret->raw_uri) {
        heap_free(ret);
        return E_OUTOFMEMORY;
    }

5153 5154 5155
    memset(&data, 0, sizeof(parse_data));
    data.uri = ret->raw_uri;

5156
    /* Validate and parse the URI into it's components. */
5157 5158
    if(!parse_uri(&data, dwFlags)) {
        /* Encountered an unsupported or invalid URI */
5159
        IUri_Release(&ret->IUri_iface);
5160 5161 5162 5163
        *ppURI = NULL;
        return E_INVALIDARG;
    }

5164 5165 5166
    /* Canonicalize the URI. */
    hr = canonicalize_uri(&data, ret, dwFlags);
    if(FAILED(hr)) {
5167
        IUri_Release(&ret->IUri_iface);
5168 5169 5170 5171
        *ppURI = NULL;
        return hr;
    }

5172 5173
    ret->create_flags = dwFlags;

5174
    *ppURI = &ret->IUri_iface;
5175 5176
    return S_OK;
}
5177

5178 5179
/***********************************************************************
 *           CreateUriWithFragment (urlmon.@)
5180 5181 5182 5183 5184 5185
 *
 * Creates a new IUri object. This is almost the same as CreateUri, expect that
 * it allows you to explicitly specify a fragment (pwzFragment) for pwzURI.
 *
 * PARAMS
 *  pwzURI      [I] The URI to parse and perform canonicalization on.
5186
 *  pwzFragment [I] The explicit fragment string which should be added to pwzURI.
5187 5188 5189 5190 5191 5192 5193 5194
 *  dwFlags     [I] The flags which will be passed to CreateUri.
 *  dwReserved  [I] Reserved (not used).
 *  ppURI       [O] The resulting IUri after parsing/canonicalization.
 *
 * RETURNS
 *  Success: S_OK. ppURI contains the pointer to the newly allocated IUri.
 *  Failure: E_INVALIDARG if pwzURI already contains a fragment and pwzFragment
 *           isn't NULL. Will also return E_INVALIDARG for the same reasons as
5195
 *           CreateUri will. E_OUTOFMEMORY if any allocation fails.
5196 5197 5198 5199
 */
HRESULT WINAPI CreateUriWithFragment(LPCWSTR pwzURI, LPCWSTR pwzFragment, DWORD dwFlags,
                                     DWORD_PTR dwReserved, IUri **ppURI)
{
5200 5201 5202 5203 5204 5205 5206 5207 5208 5209 5210 5211 5212 5213 5214 5215 5216 5217 5218 5219 5220 5221 5222 5223 5224 5225 5226 5227 5228 5229 5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241 5242 5243 5244 5245 5246 5247 5248 5249
    HRESULT hres;
    TRACE("(%s %s %x %x %p)\n", debugstr_w(pwzURI), debugstr_w(pwzFragment), dwFlags, (DWORD)dwReserved, ppURI);

    if(!ppURI)
        return E_INVALIDARG;

    if(!pwzURI) {
        *ppURI = NULL;
        return E_INVALIDARG;
    }

    /* Check if a fragment should be appended to the URI string. */
    if(pwzFragment) {
        WCHAR *uriW;
        DWORD uri_len, frag_len;
        BOOL add_pound;

        /* Check if the original URI already has a fragment component. */
        if(StrChrW(pwzURI, '#')) {
            *ppURI = NULL;
            return E_INVALIDARG;
        }

        uri_len = lstrlenW(pwzURI);
        frag_len = lstrlenW(pwzFragment);

        /* If the fragment doesn't start with a '#', one will be added. */
        add_pound = *pwzFragment != '#';

        if(add_pound)
            uriW = heap_alloc((uri_len+frag_len+2)*sizeof(WCHAR));
        else
            uriW = heap_alloc((uri_len+frag_len+1)*sizeof(WCHAR));

        if(!uriW)
            return E_OUTOFMEMORY;

        memcpy(uriW, pwzURI, uri_len*sizeof(WCHAR));
        if(add_pound)
            uriW[uri_len++] = '#';
        memcpy(uriW+uri_len, pwzFragment, (frag_len+1)*sizeof(WCHAR));

        hres = CreateUri(uriW, dwFlags, 0, ppURI);

        heap_free(uriW);
    } else
        /* A fragment string wasn't specified, so just forward the call. */
        hres = CreateUri(pwzURI, dwFlags, 0, ppURI);

    return hres;
5250 5251
}

5252 5253 5254 5255 5256 5257 5258 5259 5260 5261 5262 5263 5264 5265 5266 5267 5268 5269 5270 5271 5272 5273 5274 5275 5276 5277 5278 5279 5280 5281
static HRESULT build_uri(const UriBuilder *builder, IUri **uri, DWORD create_flags,
                         DWORD use_orig_flags, DWORD encoding_mask)
{
    HRESULT hr;
    parse_data data;
    Uri *ret;

    if(!uri)
        return E_POINTER;

    if(encoding_mask && (!builder->uri || builder->modified_props)) {
        *uri = NULL;
        return E_NOTIMPL;
    }

    /* Decide what flags should be used when creating the Uri. */
    if((use_orig_flags & UriBuilder_USE_ORIGINAL_FLAGS) && builder->uri)
        create_flags = builder->uri->create_flags;
    else {
        if(has_invalid_flag_combination(create_flags)) {
            *uri = NULL;
            return E_INVALIDARG;
        }

        /* Set the default flags if they don't cause a conflict. */
        apply_default_flags(&create_flags);
    }

    /* Return the base IUri if no changes have been made and the create_flags match. */
    if(builder->uri && !builder->modified_props && builder->uri->create_flags == create_flags) {
5282
        *uri = &builder->uri->IUri_iface;
5283 5284 5285 5286 5287 5288 5289 5290 5291 5292 5293 5294 5295 5296 5297 5298 5299 5300
        IUri_AddRef(*uri);
        return S_OK;
    }

    hr = validate_components(builder, &data, create_flags);
    if(FAILED(hr)) {
        *uri = NULL;
        return hr;
    }

    ret = create_uri_obj();
    if(!ret) {
        *uri = NULL;
        return E_OUTOFMEMORY;
    }

    hr = generate_uri(builder, &data, ret, create_flags);
    if(FAILED(hr)) {
5301
        IUri_Release(&ret->IUri_iface);
5302 5303 5304 5305
        *uri = NULL;
        return hr;
    }

5306
    *uri = &ret->IUri_iface;
5307 5308 5309
    return S_OK;
}

5310 5311 5312 5313
static inline UriBuilder* impl_from_IUriBuilder(IUriBuilder *iface)
{
    return CONTAINING_RECORD(iface, UriBuilder, IUriBuilder_iface);
}
5314 5315 5316

static HRESULT WINAPI UriBuilder_QueryInterface(IUriBuilder *iface, REFIID riid, void **ppv)
{
5317
    UriBuilder *This = impl_from_IUriBuilder(iface);
5318 5319 5320

    if(IsEqualGUID(&IID_IUnknown, riid)) {
        TRACE("(%p)->(IID_IUnknown %p)\n", This, ppv);
5321
        *ppv = &This->IUriBuilder_iface;
5322
    }else if(IsEqualGUID(&IID_IUriBuilder, riid)) {
5323 5324
        TRACE("(%p)->(IID_IUriBuilder %p)\n", This, ppv);
        *ppv = &This->IUriBuilder_iface;
5325 5326 5327 5328 5329 5330 5331 5332 5333 5334 5335 5336
    }else {
        TRACE("(%p)->(%s %p)\n", This, debugstr_guid(riid), ppv);
        *ppv = NULL;
        return E_NOINTERFACE;
    }

    IUnknown_AddRef((IUnknown*)*ppv);
    return S_OK;
}

static ULONG WINAPI UriBuilder_AddRef(IUriBuilder *iface)
{
5337
    UriBuilder *This = impl_from_IUriBuilder(iface);
5338 5339 5340 5341 5342 5343 5344 5345 5346
    LONG ref = InterlockedIncrement(&This->ref);

    TRACE("(%p) ref=%d\n", This, ref);

    return ref;
}

static ULONG WINAPI UriBuilder_Release(IUriBuilder *iface)
{
5347
    UriBuilder *This = impl_from_IUriBuilder(iface);
5348 5349 5350 5351
    LONG ref = InterlockedDecrement(&This->ref);

    TRACE("(%p) ref=%d\n", This, ref);

5352
    if(!ref) {
5353
        if(This->uri) IUri_Release(&This->uri->IUri_iface);
5354
        heap_free(This->fragment);
5355
        heap_free(This->host);
5356
        heap_free(This->password);
5357
        heap_free(This->path);
5358
        heap_free(This->query);
5359
        heap_free(This->scheme);
5360
        heap_free(This->username);
5361
        heap_free(This);
5362
    }
5363 5364 5365 5366 5367 5368 5369 5370 5371

    return ref;
}

static HRESULT WINAPI UriBuilder_CreateUriSimple(IUriBuilder *iface,
                                                 DWORD        dwAllowEncodingPropertyMask,
                                                 DWORD_PTR    dwReserved,
                                                 IUri       **ppIUri)
{
5372
    UriBuilder *This = impl_from_IUriBuilder(iface);
5373
    HRESULT hr;
5374 5375
    TRACE("(%p)->(%d %d %p)\n", This, dwAllowEncodingPropertyMask, (DWORD)dwReserved, ppIUri);

5376 5377 5378 5379
    hr = build_uri(This, ppIUri, 0, UriBuilder_USE_ORIGINAL_FLAGS, dwAllowEncodingPropertyMask);
    if(hr == E_NOTIMPL)
        FIXME("(%p)->(%d %d %p)\n", This, dwAllowEncodingPropertyMask, (DWORD)dwReserved, ppIUri);
    return hr;
5380 5381 5382 5383 5384 5385 5386 5387
}

static HRESULT WINAPI UriBuilder_CreateUri(IUriBuilder *iface,
                                           DWORD        dwCreateFlags,
                                           DWORD        dwAllowEncodingPropertyMask,
                                           DWORD_PTR    dwReserved,
                                           IUri       **ppIUri)
{
5388
    UriBuilder *This = impl_from_IUriBuilder(iface);
5389
    HRESULT hr;
5390 5391
    TRACE("(%p)->(0x%08x %d %d %p)\n", This, dwCreateFlags, dwAllowEncodingPropertyMask, (DWORD)dwReserved, ppIUri);

5392 5393 5394 5395
    if(dwCreateFlags == -1)
        hr = build_uri(This, ppIUri, 0, UriBuilder_USE_ORIGINAL_FLAGS, dwAllowEncodingPropertyMask);
    else
        hr = build_uri(This, ppIUri, dwCreateFlags, 0, dwAllowEncodingPropertyMask);
5396

5397 5398 5399
    if(hr == E_NOTIMPL)
        FIXME("(%p)->(0x%08x %d %d %p)\n", This, dwCreateFlags, dwAllowEncodingPropertyMask, (DWORD)dwReserved, ppIUri);
    return hr;
5400 5401 5402 5403 5404 5405 5406 5407 5408
}

static HRESULT WINAPI UriBuilder_CreateUriWithFlags(IUriBuilder *iface,
                                         DWORD        dwCreateFlags,
                                         DWORD        dwUriBuilderFlags,
                                         DWORD        dwAllowEncodingPropertyMask,
                                         DWORD_PTR    dwReserved,
                                         IUri       **ppIUri)
{
5409
    UriBuilder *This = impl_from_IUriBuilder(iface);
5410
    HRESULT hr;
5411 5412 5413
    TRACE("(%p)->(0x%08x 0x%08x %d %d %p)\n", This, dwCreateFlags, dwUriBuilderFlags,
        dwAllowEncodingPropertyMask, (DWORD)dwReserved, ppIUri);

5414 5415 5416 5417 5418
    hr = build_uri(This, ppIUri, dwCreateFlags, dwUriBuilderFlags, dwAllowEncodingPropertyMask);
    if(hr == E_NOTIMPL)
        FIXME("(%p)->(0x%08x 0x%08x %d %d %p)\n", This, dwCreateFlags, dwUriBuilderFlags,
            dwAllowEncodingPropertyMask, (DWORD)dwReserved, ppIUri);
    return hr;
5419 5420 5421 5422
}

static HRESULT WINAPI  UriBuilder_GetIUri(IUriBuilder *iface, IUri **ppIUri)
{
5423
    UriBuilder *This = impl_from_IUriBuilder(iface);
5424 5425 5426 5427 5428
    TRACE("(%p)->(%p)\n", This, ppIUri);

    if(!ppIUri)
        return E_POINTER;

5429
    if(This->uri) {
5430
        IUri *uri = &This->uri->IUri_iface;
5431 5432 5433 5434 5435 5436
        IUri_AddRef(uri);
        *ppIUri = uri;
    } else
        *ppIUri = NULL;

    return S_OK;
5437 5438 5439 5440
}

static HRESULT WINAPI UriBuilder_SetIUri(IUriBuilder *iface, IUri *pIUri)
{
5441
    UriBuilder *This = impl_from_IUriBuilder(iface);
5442 5443 5444 5445 5446 5447 5448 5449 5450 5451 5452 5453 5454 5455 5456 5457 5458 5459 5460 5461 5462 5463 5464 5465 5466 5467 5468
    TRACE("(%p)->(%p)\n", This, pIUri);

    if(pIUri) {
        Uri *uri;

        if((uri = get_uri_obj(pIUri))) {
            /* Only reset the builder if it's Uri isn't the same as
             * the Uri passed to the function.
             */
            if(This->uri != uri) {
                reset_builder(This);

                This->uri = uri;
                if(uri->has_port)
                    This->port = uri->port;

                IUri_AddRef(pIUri);
            }
        } else {
            FIXME("(%p)->(%p) Unknown IUri types not supported yet.\n", This, pIUri);
            return E_NOTIMPL;
        }
    } else if(This->uri)
        /* Only reset the builder if it's Uri isn't NULL. */
        reset_builder(This);

    return S_OK;
5469 5470 5471 5472
}

static HRESULT WINAPI UriBuilder_GetFragment(IUriBuilder *iface, DWORD *pcchFragment, LPCWSTR *ppwzFragment)
{
5473
    UriBuilder *This = impl_from_IUriBuilder(iface);
5474 5475
    TRACE("(%p)->(%p %p)\n", This, pcchFragment, ppwzFragment);

5476 5477 5478 5479 5480
    if(!This->uri || This->uri->fragment_start == -1 || This->modified_props & Uri_HAS_FRAGMENT)
        return get_builder_component(&This->fragment, &This->fragment_len, NULL, 0, ppwzFragment, pcchFragment);
    else
        return get_builder_component(&This->fragment, &This->fragment_len, This->uri->canon_uri+This->uri->fragment_start,
                                     This->uri->fragment_len, ppwzFragment, pcchFragment);
5481 5482 5483 5484
}

static HRESULT WINAPI UriBuilder_GetHost(IUriBuilder *iface, DWORD *pcchHost, LPCWSTR *ppwzHost)
{
5485
    UriBuilder *This = impl_from_IUriBuilder(iface);
5486 5487
    TRACE("(%p)->(%p %p)\n", This, pcchHost, ppwzHost);

5488 5489 5490 5491 5492 5493 5494 5495 5496 5497
    if(!This->uri || This->uri->host_start == -1 || This->modified_props & Uri_HAS_HOST)
        return get_builder_component(&This->host, &This->host_len, NULL, 0, ppwzHost, pcchHost);
    else {
        if(This->uri->host_type == Uri_HOST_IPV6)
            /* Don't include the '[' and ']' around the address. */
            return get_builder_component(&This->host, &This->host_len, This->uri->canon_uri+This->uri->host_start+1,
                                         This->uri->host_len-2, ppwzHost, pcchHost);
        else
            return get_builder_component(&This->host, &This->host_len, This->uri->canon_uri+This->uri->host_start,
                                         This->uri->host_len, ppwzHost, pcchHost);
5498
    }
5499 5500 5501 5502
}

static HRESULT WINAPI UriBuilder_GetPassword(IUriBuilder *iface, DWORD *pcchPassword, LPCWSTR *ppwzPassword)
{
5503
    UriBuilder *This = impl_from_IUriBuilder(iface);
5504 5505
    TRACE("(%p)->(%p %p)\n", This, pcchPassword, ppwzPassword);

5506 5507 5508 5509 5510 5511
    if(!This->uri || This->uri->userinfo_split == -1 || This->modified_props & Uri_HAS_PASSWORD)
        return get_builder_component(&This->password, &This->password_len, NULL, 0, ppwzPassword, pcchPassword);
    else {
        const WCHAR *start = This->uri->canon_uri+This->uri->userinfo_start+This->uri->userinfo_split+1;
        DWORD len = This->uri->userinfo_len-This->uri->userinfo_split-1;
        return get_builder_component(&This->password, &This->password_len, start, len, ppwzPassword, pcchPassword);
5512
    }
5513 5514 5515 5516
}

static HRESULT WINAPI UriBuilder_GetPath(IUriBuilder *iface, DWORD *pcchPath, LPCWSTR *ppwzPath)
{
5517
    UriBuilder *This = impl_from_IUriBuilder(iface);
5518 5519
    TRACE("(%p)->(%p %p)\n", This, pcchPath, ppwzPath);

5520 5521 5522 5523 5524
    if(!This->uri || This->uri->path_start == -1 || This->modified_props & Uri_HAS_PATH)
        return get_builder_component(&This->path, &This->path_len, NULL, 0, ppwzPath, pcchPath);
    else
        return get_builder_component(&This->path, &This->path_len, This->uri->canon_uri+This->uri->path_start,
                                     This->uri->path_len, ppwzPath, pcchPath);
5525 5526 5527 5528
}

static HRESULT WINAPI UriBuilder_GetPort(IUriBuilder *iface, BOOL *pfHasPort, DWORD *pdwPort)
{
5529
    UriBuilder *This = impl_from_IUriBuilder(iface);
5530 5531 5532 5533 5534 5535 5536 5537 5538 5539 5540 5541 5542
    TRACE("(%p)->(%p %p)\n", This, pfHasPort, pdwPort);

    if(!pfHasPort) {
        if(pdwPort)
            *pdwPort = 0;
        return E_POINTER;
    }

    if(!pdwPort) {
        *pfHasPort = FALSE;
        return E_POINTER;
    }

5543 5544 5545
    *pfHasPort = This->has_port;
    *pdwPort = This->port;
    return S_OK;
5546 5547 5548 5549
}

static HRESULT WINAPI UriBuilder_GetQuery(IUriBuilder *iface, DWORD *pcchQuery, LPCWSTR *ppwzQuery)
{
5550
    UriBuilder *This = impl_from_IUriBuilder(iface);
5551 5552
    TRACE("(%p)->(%p %p)\n", This, pcchQuery, ppwzQuery);

5553 5554 5555 5556 5557
    if(!This->uri || This->uri->query_start == -1 || This->modified_props & Uri_HAS_QUERY)
        return get_builder_component(&This->query, &This->query_len, NULL, 0, ppwzQuery, pcchQuery);
    else
        return get_builder_component(&This->query, &This->query_len, This->uri->canon_uri+This->uri->query_start,
                                     This->uri->query_len, ppwzQuery, pcchQuery);
5558 5559 5560 5561
}

static HRESULT WINAPI UriBuilder_GetSchemeName(IUriBuilder *iface, DWORD *pcchSchemeName, LPCWSTR *ppwzSchemeName)
{
5562
    UriBuilder *This = impl_from_IUriBuilder(iface);
5563 5564
    TRACE("(%p)->(%p %p)\n", This, pcchSchemeName, ppwzSchemeName);

5565 5566 5567 5568 5569
    if(!This->uri || This->uri->scheme_start == -1 || This->modified_props & Uri_HAS_SCHEME_NAME)
        return get_builder_component(&This->scheme, &This->scheme_len, NULL, 0, ppwzSchemeName, pcchSchemeName);
    else
        return get_builder_component(&This->scheme, &This->scheme_len, This->uri->canon_uri+This->uri->scheme_start,
                                     This->uri->scheme_len, ppwzSchemeName, pcchSchemeName);
5570 5571 5572 5573
}

static HRESULT WINAPI UriBuilder_GetUserName(IUriBuilder *iface, DWORD *pcchUserName, LPCWSTR *ppwzUserName)
{
5574
    UriBuilder *This = impl_from_IUriBuilder(iface);
5575 5576
    TRACE("(%p)->(%p %p)\n", This, pcchUserName, ppwzUserName);

5577
    if(!This->uri || This->uri->userinfo_start == -1 || This->uri->userinfo_split == 0 ||
5578 5579 5580 5581
       This->modified_props & Uri_HAS_USER_NAME)
        return get_builder_component(&This->username, &This->username_len, NULL, 0, ppwzUserName, pcchUserName);
    else {
        const WCHAR *start = This->uri->canon_uri+This->uri->userinfo_start;
5582

5583 5584 5585 5586 5587 5588 5589 5590
        /* Check if there's a password in the userinfo section. */
        if(This->uri->userinfo_split > -1)
            /* Don't include the password. */
            return get_builder_component(&This->username, &This->username_len, start,
                                         This->uri->userinfo_split, ppwzUserName, pcchUserName);
        else
            return get_builder_component(&This->username, &This->username_len, start,
                                         This->uri->userinfo_len, ppwzUserName, pcchUserName);
5591
    }
5592 5593 5594 5595
}

static HRESULT WINAPI UriBuilder_SetFragment(IUriBuilder *iface, LPCWSTR pwzNewValue)
{
5596
    UriBuilder *This = impl_from_IUriBuilder(iface);
5597
    TRACE("(%p)->(%s)\n", This, debugstr_w(pwzNewValue));
5598 5599
    return set_builder_component(&This->fragment, &This->fragment_len, pwzNewValue, '#',
                                 &This->modified_props, Uri_HAS_FRAGMENT);
5600 5601 5602 5603
}

static HRESULT WINAPI UriBuilder_SetHost(IUriBuilder *iface, LPCWSTR pwzNewValue)
{
5604
    UriBuilder *This = impl_from_IUriBuilder(iface);
5605
    TRACE("(%p)->(%s)\n", This, debugstr_w(pwzNewValue));
5606 5607 5608 5609 5610

    /* Host name can't be set to NULL. */
    if(!pwzNewValue)
        return E_INVALIDARG;

5611 5612
    return set_builder_component(&This->host, &This->host_len, pwzNewValue, 0,
                                 &This->modified_props, Uri_HAS_HOST);
5613 5614 5615 5616
}

static HRESULT WINAPI UriBuilder_SetPassword(IUriBuilder *iface, LPCWSTR pwzNewValue)
{
5617
    UriBuilder *This = impl_from_IUriBuilder(iface);
5618
    TRACE("(%p)->(%s)\n", This, debugstr_w(pwzNewValue));
5619 5620
    return set_builder_component(&This->password, &This->password_len, pwzNewValue, 0,
                                 &This->modified_props, Uri_HAS_PASSWORD);
5621 5622 5623 5624
}

static HRESULT WINAPI UriBuilder_SetPath(IUriBuilder *iface, LPCWSTR pwzNewValue)
{
5625
    UriBuilder *This = impl_from_IUriBuilder(iface);
5626
    TRACE("(%p)->(%s)\n", This, debugstr_w(pwzNewValue));
5627 5628
    return set_builder_component(&This->path, &This->path_len, pwzNewValue, 0,
                                 &This->modified_props, Uri_HAS_PATH);
5629 5630 5631 5632
}

static HRESULT WINAPI UriBuilder_SetPort(IUriBuilder *iface, BOOL fHasPort, DWORD dwNewValue)
{
5633
    UriBuilder *This = impl_from_IUriBuilder(iface);
5634 5635 5636 5637 5638 5639
    TRACE("(%p)->(%d %d)\n", This, fHasPort, dwNewValue);

    This->has_port = fHasPort;
    This->port = dwNewValue;
    This->modified_props |= Uri_HAS_PORT;
    return S_OK;
5640 5641 5642 5643
}

static HRESULT WINAPI UriBuilder_SetQuery(IUriBuilder *iface, LPCWSTR pwzNewValue)
{
5644
    UriBuilder *This = impl_from_IUriBuilder(iface);
5645 5646 5647
    TRACE("(%p)->(%s)\n", This, debugstr_w(pwzNewValue));
    return set_builder_component(&This->query, &This->query_len, pwzNewValue, '?',
                                 &This->modified_props, Uri_HAS_QUERY);
5648 5649 5650 5651
}

static HRESULT WINAPI UriBuilder_SetSchemeName(IUriBuilder *iface, LPCWSTR pwzNewValue)
{
5652
    UriBuilder *This = impl_from_IUriBuilder(iface);
5653
    TRACE("(%p)->(%s)\n", This, debugstr_w(pwzNewValue));
5654 5655

    /* Only set the scheme name if it's not NULL or empty. */
5656
    if(!pwzNewValue || !*pwzNewValue)
5657
        return E_INVALIDARG;
5658 5659 5660

    return set_builder_component(&This->scheme, &This->scheme_len, pwzNewValue, 0,
                                 &This->modified_props, Uri_HAS_SCHEME_NAME);
5661 5662 5663 5664
}

static HRESULT WINAPI UriBuilder_SetUserName(IUriBuilder *iface, LPCWSTR pwzNewValue)
{
5665
    UriBuilder *This = impl_from_IUriBuilder(iface);
5666 5667 5668
    TRACE("(%p)->(%s)\n", This, debugstr_w(pwzNewValue));
    return set_builder_component(&This->username, &This->username_len, pwzNewValue, 0,
                                 &This->modified_props, Uri_HAS_USER_NAME);
5669 5670 5671 5672
}

static HRESULT WINAPI UriBuilder_RemoveProperties(IUriBuilder *iface, DWORD dwPropertyMask)
{
5673 5674
    const DWORD accepted_flags = Uri_HAS_AUTHORITY|Uri_HAS_DOMAIN|Uri_HAS_EXTENSION|Uri_HAS_FRAGMENT|Uri_HAS_HOST|
                                 Uri_HAS_PASSWORD|Uri_HAS_PATH|Uri_HAS_PATH_AND_QUERY|Uri_HAS_QUERY|
5675
                                 Uri_HAS_USER_INFO|Uri_HAS_USER_NAME;
5676

5677
    UriBuilder *This = impl_from_IUriBuilder(iface);
5678 5679 5680 5681 5682 5683 5684 5685
    TRACE("(%p)->(0x%08x)\n", This, dwPropertyMask);

    if(dwPropertyMask & ~accepted_flags)
        return E_INVALIDARG;

    if(dwPropertyMask & Uri_HAS_FRAGMENT)
        UriBuilder_SetFragment(iface, NULL);

5686 5687 5688
    /* Even though you can't set the host name to NULL or an
     * empty string, you can still remove it... for some reason.
     */
5689
    if(dwPropertyMask & Uri_HAS_HOST)
5690 5691
        set_builder_component(&This->host, &This->host_len, NULL, 0,
                              &This->modified_props, Uri_HAS_HOST);
5692 5693 5694 5695 5696 5697 5698 5699 5700 5701 5702 5703 5704 5705 5706 5707 5708

    if(dwPropertyMask & Uri_HAS_PASSWORD)
        UriBuilder_SetPassword(iface, NULL);

    if(dwPropertyMask & Uri_HAS_PATH)
        UriBuilder_SetPath(iface, NULL);

    if(dwPropertyMask & Uri_HAS_PORT)
        UriBuilder_SetPort(iface, FALSE, 0);

    if(dwPropertyMask & Uri_HAS_QUERY)
        UriBuilder_SetQuery(iface, NULL);

    if(dwPropertyMask & Uri_HAS_USER_NAME)
        UriBuilder_SetUserName(iface, NULL);

    return S_OK;
5709 5710 5711 5712
}

static HRESULT WINAPI UriBuilder_HasBeenModified(IUriBuilder *iface, BOOL *pfModified)
{
5713
    UriBuilder *This = impl_from_IUriBuilder(iface);
5714 5715 5716 5717 5718
    TRACE("(%p)->(%p)\n", This, pfModified);

    if(!pfModified)
        return E_POINTER;

5719 5720
    *pfModified = This->modified_props > 0;
    return S_OK;
5721 5722 5723 5724 5725 5726 5727 5728 5729 5730 5731 5732 5733 5734 5735 5736 5737 5738 5739 5740 5741 5742 5743 5744 5745 5746 5747 5748 5749 5750 5751 5752 5753 5754 5755 5756 5757 5758 5759 5760
}

static const IUriBuilderVtbl UriBuilderVtbl = {
    UriBuilder_QueryInterface,
    UriBuilder_AddRef,
    UriBuilder_Release,
    UriBuilder_CreateUriSimple,
    UriBuilder_CreateUri,
    UriBuilder_CreateUriWithFlags,
    UriBuilder_GetIUri,
    UriBuilder_SetIUri,
    UriBuilder_GetFragment,
    UriBuilder_GetHost,
    UriBuilder_GetPassword,
    UriBuilder_GetPath,
    UriBuilder_GetPort,
    UriBuilder_GetQuery,
    UriBuilder_GetSchemeName,
    UriBuilder_GetUserName,
    UriBuilder_SetFragment,
    UriBuilder_SetHost,
    UriBuilder_SetPassword,
    UriBuilder_SetPath,
    UriBuilder_SetPort,
    UriBuilder_SetQuery,
    UriBuilder_SetSchemeName,
    UriBuilder_SetUserName,
    UriBuilder_RemoveProperties,
    UriBuilder_HasBeenModified,
};

/***********************************************************************
 *           CreateIUriBuilder (urlmon.@)
 */
HRESULT WINAPI CreateIUriBuilder(IUri *pIUri, DWORD dwFlags, DWORD_PTR dwReserved, IUriBuilder **ppIUriBuilder)
{
    UriBuilder *ret;

    TRACE("(%p %x %x %p)\n", pIUri, dwFlags, (DWORD)dwReserved, ppIUriBuilder);

5761 5762 5763
    if(!ppIUriBuilder)
        return E_POINTER;

5764
    ret = heap_alloc_zero(sizeof(UriBuilder));
5765 5766 5767
    if(!ret)
        return E_OUTOFMEMORY;

5768
    ret->IUriBuilder_iface.lpVtbl = &UriBuilderVtbl;
5769 5770
    ret->ref = 1;

5771 5772 5773 5774 5775 5776
    if(pIUri) {
        Uri *uri;

        if((uri = get_uri_obj(pIUri))) {
            IUri_AddRef(pIUri);
            ret->uri = uri;
5777 5778 5779 5780 5781

            if(uri->has_port)
                /* Windows doesn't set 'has_port' to TRUE in this case. */
                ret->port = uri->port;

5782 5783 5784 5785 5786 5787 5788 5789
        } else {
            heap_free(ret);
            *ppIUriBuilder = NULL;
            FIXME("(%p %x %x %p): Unknown IUri types not supported yet.\n", pIUri, dwFlags,
                (DWORD)dwReserved, ppIUriBuilder);
            return E_NOTIMPL;
        }
    }
5790

5791
    *ppIUriBuilder = &ret->IUriBuilder_iface;
5792 5793
    return S_OK;
}
5794

5795 5796 5797 5798 5799 5800 5801 5802 5803 5804 5805
/* Merges the base path with the relative path and stores the resulting path
 * and path len in 'result' and 'result_len'.
 */
static HRESULT merge_paths(parse_data *data, const WCHAR *base, DWORD base_len, const WCHAR *relative,
                           DWORD relative_len, WCHAR **result, DWORD *result_len, DWORD flags)
{
    const WCHAR *end = NULL;
    DWORD base_copy_len = 0;
    WCHAR *ptr;

    if(base_len) {
5806
        /* Find the characters that will be copied over from
5807 5808
         * the base path.
         */
5809
        end = memrchrW(base, '/', base_len);
5810 5811
        if(!end && data->scheme_type == URL_SCHEME_FILE)
            /* Try looking for a '\\'. */
5812
            end = memrchrW(base, '\\', base_len);
5813 5814 5815 5816 5817 5818 5819 5820 5821 5822 5823 5824 5825 5826 5827 5828 5829 5830 5831 5832 5833 5834 5835 5836 5837 5838 5839
    }

    if(end) {
        base_copy_len = (end+1)-base;
        *result = heap_alloc((base_copy_len+relative_len+1)*sizeof(WCHAR));
    } else
        *result = heap_alloc((relative_len+1)*sizeof(WCHAR));

    if(!(*result)) {
        *result_len = 0;
        return E_OUTOFMEMORY;
    }

    ptr = *result;
    if(end) {
        memcpy(ptr, base, base_copy_len*sizeof(WCHAR));
        ptr += base_copy_len;
    }

    memcpy(ptr, relative, relative_len*sizeof(WCHAR));
    ptr += relative_len;
    *ptr = '\0';

    *result_len = (ptr-*result);
    return S_OK;
}

5840
static HRESULT combine_uri(Uri *base, Uri *relative, DWORD flags, IUri **result, DWORD extras) {
5841 5842 5843
    Uri *ret;
    HRESULT hr;
    parse_data data;
5844 5845 5846
    DWORD create_flags = 0, len = 0;

    memset(&data, 0, sizeof(parse_data));
5847 5848 5849 5850 5851 5852 5853 5854 5855 5856 5857 5858 5859 5860 5861 5862 5863 5864 5865 5866

    /* Base case is when the relative Uri has a scheme name,
     * if it does, then 'result' will contain the same data
     * as the relative Uri.
     */
    if(relative->scheme_start > -1) {
        data.uri = SysAllocString(relative->raw_uri);
        if(!data.uri) {
            *result = NULL;
            return E_OUTOFMEMORY;
        }

        parse_uri(&data, 0);

        ret = create_uri_obj();
        if(!ret) {
            *result = NULL;
            return E_OUTOFMEMORY;
        }

5867 5868 5869 5870 5871 5872 5873
        if(extras & COMBINE_URI_FORCE_FLAG_USE) {
            if(flags & URL_DONT_SIMPLIFY)
                create_flags |= Uri_CREATE_NO_CANONICALIZE;
            if(flags & URL_DONT_UNESCAPE_EXTRA_INFO)
                create_flags |= Uri_CREATE_NO_DECODE_EXTRA_INFO;
        }

5874
        ret->raw_uri = data.uri;
5875
        hr = canonicalize_uri(&data, ret, create_flags);
5876
        if(FAILED(hr)) {
5877
            IUri_Release(&ret->IUri_iface);
5878 5879 5880 5881 5882 5883 5884
            *result = NULL;
            return hr;
        }

        apply_default_flags(&create_flags);
        ret->create_flags = create_flags;

5885
        *result = &ret->IUri_iface;
5886
    } else {
5887
        WCHAR *path = NULL;
5888 5889 5890 5891 5892 5893 5894 5895 5896 5897 5898 5899 5900 5901 5902 5903 5904 5905 5906 5907 5908 5909 5910 5911 5912 5913 5914 5915 5916 5917 5918 5919 5920
        DWORD raw_flags = 0;

        if(base->scheme_start > -1) {
            data.scheme = base->canon_uri+base->scheme_start;
            data.scheme_len = base->scheme_len;
            data.scheme_type = base->scheme_type;
        } else {
            data.is_relative = TRUE;
            data.scheme_type = URL_SCHEME_UNKNOWN;
            create_flags |= Uri_CREATE_ALLOW_RELATIVE;
        }

        if(base->authority_start > -1) {
            if(base->userinfo_start > -1 && base->userinfo_split != 0) {
                data.username = base->canon_uri+base->userinfo_start;
                data.username_len = (base->userinfo_split > -1) ? base->userinfo_split : base->userinfo_len;
            }

            if(base->userinfo_split > -1) {
                data.password = base->canon_uri+base->userinfo_start+base->userinfo_split+1;
                data.password_len = base->userinfo_len-base->userinfo_split-1;
            }

            if(base->host_start > -1) {
                data.host = base->canon_uri+base->host_start;
                data.host_len = base->host_len;
                data.host_type = base->host_type;
            }

            if(base->has_port) {
                data.has_port = TRUE;
                data.port_value = base->port;
            }
5921
        } else if(base->scheme_type != URL_SCHEME_FILE)
5922 5923 5924 5925 5926 5927 5928 5929 5930 5931 5932 5933 5934 5935 5936 5937 5938 5939 5940 5941 5942 5943 5944
            data.is_opaque = TRUE;

        if(relative->path_start == -1 || !relative->path_len) {
            if(base->path_start > -1) {
                data.path = base->canon_uri+base->path_start;
                data.path_len = base->path_len;
            } else if((base->path_start == -1 || !base->path_len) && !data.is_opaque) {
                /* Just set the path as a '/' if the base didn't have
                 * one and if it's an hierarchical URI.
                 */
                static const WCHAR slashW[] = {'/',0};
                data.path = slashW;
                data.path_len = 1;
            }

            if(relative->query_start > -1) {
                data.query = relative->canon_uri+relative->query_start;
                data.query_len = relative->query_len;
            } else if(base->query_start > -1) {
                data.query = base->canon_uri+base->query_start;
                data.query_len = base->query_len;
            }
        } else {
5945
            const WCHAR *ptr, **pptr;
5946 5947
            DWORD path_offset = 0, path_len = 0;

5948 5949 5950 5951 5952 5953
            /* There's two possibilities on what will happen to the path component
             * of the result IUri. First, if the relative path begins with a '/'
             * then the resulting path will just be the relative path. Second, if
             * relative path doesn't begin with a '/' then the base path and relative
             * path are merged together.
             */
5954 5955 5956 5957 5958 5959 5960 5961 5962 5963 5964 5965 5966 5967 5968 5969 5970 5971 5972 5973 5974 5975 5976 5977 5978 5979 5980 5981 5982 5983 5984 5985 5986 5987 5988
            if(relative->path_len && *(relative->canon_uri+relative->path_start) == '/') {
                WCHAR *tmp = NULL;
                BOOL copy_drive_path = FALSE;

                /* If the relative IUri's path starts with a '/', then we
                 * don't use the base IUri's path. Unless the base IUri
                 * is a file URI, in which case it uses the drive path of
                 * the base IUri (if it has any) in the new path.
                 */
                if(base->scheme_type == URL_SCHEME_FILE) {
                    if(base->path_len > 3 && *(base->canon_uri+base->path_start) == '/' &&
                       is_drive_path(base->canon_uri+base->path_start+1)) {
                        path_len += 3;
                        copy_drive_path = TRUE;
                    }
                }

                path_len += relative->path_len;

                path = heap_alloc((path_len+1)*sizeof(WCHAR));
                if(!path) {
                    *result = NULL;
                    return E_OUTOFMEMORY;
                }

                tmp = path;

                /* Copy the base paths, drive path over. */
                if(copy_drive_path) {
                    memcpy(tmp, base->canon_uri+base->path_start, 3*sizeof(WCHAR));
                    tmp += 3;
                }

                memcpy(tmp, relative->canon_uri+relative->path_start, relative->path_len*sizeof(WCHAR));
                path[path_len] = '\0';
5989 5990 5991 5992 5993 5994
            } else {
                /* Merge the base path with the relative path. */
                hr = merge_paths(&data, base->canon_uri+base->path_start, base->path_len,
                                 relative->canon_uri+relative->path_start, relative->path_len,
                                 &path, &path_len, flags);
                if(FAILED(hr)) {
5995
                    *result = NULL;
5996 5997 5998 5999 6000 6001 6002 6003 6004 6005 6006
                    return hr;
                }

                /* If the resulting IUri is a file URI, the drive path isn't
                 * reduced out when the dot segments are removed.
                 */
                if(path_len >= 3 && data.scheme_type == URL_SCHEME_FILE && !data.host) {
                    if(*path == '/' && is_drive_path(path+1))
                        path_offset = 2;
                    else if(is_drive_path(path))
                        path_offset = 1;
6007 6008 6009 6010 6011 6012 6013 6014 6015
                }
            }

            /* Check if the dot segments need to be removed from the path. */
            if(!(flags & URL_DONT_SIMPLIFY) && !data.is_opaque) {
                DWORD offset = (path_offset > 0) ? path_offset+1 : 0;
                DWORD new_len = remove_dot_segments(path+offset,path_len-offset);

                if(new_len != path_len) {
6016
                    WCHAR *tmp = heap_realloc(path, (offset+new_len+1)*sizeof(WCHAR));
6017 6018 6019 6020 6021 6022
                    if(!tmp) {
                        heap_free(path);
                        *result = NULL;
                        return E_OUTOFMEMORY;
                    }

6023
                    tmp[new_len+offset] = '\0';
6024
                    path = tmp;
6025
                    path_len = new_len+offset;
6026 6027
                }
            }
6028

6029 6030 6031 6032 6033
            if(relative->query_start > -1) {
                data.query = relative->canon_uri+relative->query_start;
                data.query_len = relative->query_len;
            }

6034 6035 6036 6037 6038 6039 6040 6041 6042
            /* Make sure the path component is valid. */
            ptr = path;
            pptr = &ptr;
            if((data.is_opaque && !parse_path_opaque(pptr, &data, 0)) ||
               (!data.is_opaque && !parse_path_hierarchical(pptr, &data, 0))) {
                heap_free(path);
                *result = NULL;
                return E_INVALIDARG;
            }
6043 6044 6045 6046 6047 6048 6049 6050 6051
        }

        if(relative->fragment_start > -1) {
            data.fragment = relative->canon_uri+relative->fragment_start;
            data.fragment_len = relative->fragment_len;
        }

        if(flags & URL_DONT_SIMPLIFY)
            raw_flags |= RAW_URI_FORCE_PORT_DISP;
6052 6053
        if(flags & URL_FILE_USE_PATHURL)
            raw_flags |= RAW_URI_CONVERT_TO_DOS_PATH;
6054 6055 6056 6057

        len = generate_raw_uri(&data, data.uri, raw_flags);
        data.uri = SysAllocStringLen(NULL, len);
        if(!data.uri) {
6058
            heap_free(path);
6059 6060 6061 6062 6063 6064 6065 6066 6067
            *result = NULL;
            return E_OUTOFMEMORY;
        }

        generate_raw_uri(&data, data.uri, raw_flags);

        ret = create_uri_obj();
        if(!ret) {
            SysFreeString(data.uri);
6068
            heap_free(path);
6069 6070 6071 6072
            *result = NULL;
            return E_OUTOFMEMORY;
        }

6073 6074 6075 6076 6077
        if(flags & URL_DONT_SIMPLIFY)
            create_flags |= Uri_CREATE_NO_CANONICALIZE;
        if(flags & URL_FILE_USE_PATHURL)
            create_flags |= Uri_CREATE_FILE_USE_DOS_PATH;

6078 6079 6080
        ret->raw_uri = data.uri;
        hr = canonicalize_uri(&data, ret, create_flags);
        if(FAILED(hr)) {
6081
            IUri_Release(&ret->IUri_iface);
6082 6083 6084 6085
            *result = NULL;
            return hr;
        }

6086 6087 6088
        if(flags & URL_DONT_SIMPLIFY)
            ret->display_modifiers |= URI_DISPLAY_NO_DEFAULT_PORT_AUTH;

6089 6090
        apply_default_flags(&create_flags);
        ret->create_flags = create_flags;
6091
        *result = &ret->IUri_iface;
6092 6093

        heap_free(path);
6094 6095 6096 6097 6098
    }

    return S_OK;
}

6099 6100 6101 6102 6103 6104
/***********************************************************************
 *           CoInternetCombineIUri (urlmon.@)
 */
HRESULT WINAPI CoInternetCombineIUri(IUri *pBaseUri, IUri *pRelativeUri, DWORD dwCombineFlags,
                                     IUri **ppCombinedUri, DWORD_PTR dwReserved)
{
6105
    HRESULT hr;
6106
    IInternetProtocolInfo *info;
6107
    Uri *relative, *base;
6108 6109 6110 6111 6112 6113 6114 6115 6116 6117
    TRACE("(%p %p %x %p %x)\n", pBaseUri, pRelativeUri, dwCombineFlags, ppCombinedUri, (DWORD)dwReserved);

    if(!ppCombinedUri)
        return E_INVALIDARG;

    if(!pBaseUri || !pRelativeUri) {
        *ppCombinedUri = NULL;
        return E_INVALIDARG;
    }

6118 6119 6120 6121 6122 6123 6124 6125 6126
    relative = get_uri_obj(pRelativeUri);
    base = get_uri_obj(pBaseUri);
    if(!relative || !base) {
        *ppCombinedUri = NULL;
        FIXME("(%p %p %x %p %x) Unknown IUri types not supported yet.\n",
            pBaseUri, pRelativeUri, dwCombineFlags, ppCombinedUri, (DWORD)dwReserved);
        return E_NOTIMPL;
    }

6127 6128 6129 6130 6131 6132 6133 6134 6135 6136 6137 6138 6139 6140 6141
    info = get_protocol_info(base->canon_uri);
    if(info) {
        WCHAR result[INTERNET_MAX_URL_LENGTH+1];
        DWORD result_len = 0;

        hr = IInternetProtocolInfo_CombineUrl(info, base->canon_uri, relative->canon_uri, dwCombineFlags,
                                              result, INTERNET_MAX_URL_LENGTH+1, &result_len, 0);
        IInternetProtocolInfo_Release(info);
        if(SUCCEEDED(hr)) {
            hr = CreateUri(result, Uri_CREATE_ALLOW_RELATIVE, 0, ppCombinedUri);
            if(SUCCEEDED(hr))
                return hr;
        }
    }

6142
    return combine_uri(base, relative, dwCombineFlags, ppCombinedUri, 0);
6143
}
6144 6145 6146 6147 6148 6149 6150

/***********************************************************************
 *           CoInternetCombineUrlEx (urlmon.@)
 */
HRESULT WINAPI CoInternetCombineUrlEx(IUri *pBaseUri, LPCWSTR pwzRelativeUrl, DWORD dwCombineFlags,
                                      IUri **ppCombinedUri, DWORD_PTR dwReserved)
{
6151 6152 6153
    IUri *relative;
    Uri *base;
    HRESULT hr;
6154
    IInternetProtocolInfo *info;
6155

6156 6157 6158 6159 6160 6161 6162 6163 6164 6165 6166 6167 6168 6169 6170 6171
    TRACE("(%p %s %x %p %x) stub\n", pBaseUri, debugstr_w(pwzRelativeUrl), dwCombineFlags,
        ppCombinedUri, (DWORD)dwReserved);

    if(!ppCombinedUri)
        return E_POINTER;

    if(!pwzRelativeUrl) {
        *ppCombinedUri = NULL;
        return E_UNEXPECTED;
    }

    if(!pBaseUri) {
        *ppCombinedUri = NULL;
        return E_INVALIDARG;
    }

6172 6173 6174 6175 6176 6177 6178 6179
    base = get_uri_obj(pBaseUri);
    if(!base) {
        *ppCombinedUri = NULL;
        FIXME("(%p %s %x %p %x) Unknown IUri's not supported yet.\n", pBaseUri, debugstr_w(pwzRelativeUrl),
            dwCombineFlags, ppCombinedUri, (DWORD)dwReserved);
        return E_NOTIMPL;
    }

6180 6181 6182 6183 6184 6185 6186 6187 6188 6189 6190 6191 6192 6193 6194
    info = get_protocol_info(base->canon_uri);
    if(info) {
        WCHAR result[INTERNET_MAX_URL_LENGTH+1];
        DWORD result_len = 0;

        hr = IInternetProtocolInfo_CombineUrl(info, base->canon_uri, pwzRelativeUrl, dwCombineFlags,
                                              result, INTERNET_MAX_URL_LENGTH+1, &result_len, 0);
        IInternetProtocolInfo_Release(info);
        if(SUCCEEDED(hr)) {
            hr = CreateUri(result, Uri_CREATE_ALLOW_RELATIVE, 0, ppCombinedUri);
            if(SUCCEEDED(hr))
                return hr;
        }
    }

6195 6196 6197 6198 6199 6200 6201 6202 6203 6204
    hr = CreateUri(pwzRelativeUrl, Uri_CREATE_ALLOW_RELATIVE, 0, &relative);
    if(FAILED(hr)) {
        *ppCombinedUri = NULL;
        return hr;
    }

    hr = combine_uri(base, get_uri_obj(relative), dwCombineFlags, ppCombinedUri, COMBINE_URI_FORCE_FLAG_USE);

    IUri_Release(relative);
    return hr;
6205
}
6206

6207 6208 6209 6210 6211 6212 6213 6214 6215 6216 6217 6218 6219 6220 6221
static HRESULT parse_canonicalize(const Uri *uri, DWORD flags, LPWSTR output,
                                  DWORD output_len, DWORD *result_len)
{
    const WCHAR *ptr = NULL;
    WCHAR *path = NULL;
    const WCHAR **pptr;
    WCHAR buffer[INTERNET_MAX_URL_LENGTH+1];
    DWORD len = 0;
    BOOL reduce_path;

    /* URL_UNESCAPE only has effect if none of the URL_ESCAPE flags are set. */
    const BOOL allow_unescape = !(flags & URL_ESCAPE_UNSAFE) &&
                                !(flags & URL_ESCAPE_SPACES_ONLY) &&
                                !(flags & URL_ESCAPE_PERCENT);

6222

6223 6224 6225 6226 6227 6228 6229 6230 6231 6232 6233 6234 6235 6236 6237 6238 6239 6240 6241 6242 6243 6244 6245 6246 6247 6248 6249 6250 6251 6252 6253 6254 6255 6256 6257 6258 6259 6260 6261 6262 6263 6264 6265 6266 6267 6268 6269 6270 6271 6272 6273 6274 6275 6276 6277 6278 6279 6280 6281 6282 6283 6284 6285 6286 6287 6288 6289 6290 6291 6292 6293 6294 6295 6296 6297 6298 6299 6300
    /* Check if the dot segments need to be removed from the
     * path component.
     */
    if(uri->scheme_start > -1 && uri->path_start > -1) {
        ptr = uri->canon_uri+uri->scheme_start+uri->scheme_len+1;
        pptr = &ptr;
    }
    reduce_path = !(flags & URL_NO_META) &&
                  !(flags & URL_DONT_SIMPLIFY) &&
                  ptr && check_hierarchical(pptr);

    for(ptr = uri->canon_uri; ptr < uri->canon_uri+uri->canon_len; ++ptr) {
        BOOL do_default_action = TRUE;

        /* Keep track of the path if we need to remove dot segments from
         * it later.
         */
        if(reduce_path && !path && ptr == uri->canon_uri+uri->path_start)
            path = buffer+len;

        /* Check if it's time to reduce the path. */
        if(reduce_path && ptr == uri->canon_uri+uri->path_start+uri->path_len) {
            DWORD current_path_len = (buffer+len) - path;
            DWORD new_path_len = remove_dot_segments(path, current_path_len);

            /* Update the current length. */
            len -= (current_path_len-new_path_len);
            reduce_path = FALSE;
        }

        if(*ptr == '%') {
            const WCHAR decoded = decode_pct_val(ptr);
            if(decoded) {
                if(allow_unescape && (flags & URL_UNESCAPE)) {
                    buffer[len++] = decoded;
                    ptr += 2;
                    do_default_action = FALSE;
                }
            }

            /* See if %'s needed to encoded. */
            if(do_default_action && (flags & URL_ESCAPE_PERCENT)) {
                pct_encode_val(*ptr, buffer+len);
                len += 3;
                do_default_action = FALSE;
            }
        } else if(*ptr == ' ') {
            if((flags & URL_ESCAPE_SPACES_ONLY) &&
               !(flags & URL_ESCAPE_UNSAFE)) {
                pct_encode_val(*ptr, buffer+len);
                len += 3;
                do_default_action = FALSE;
            }
        } else if(!is_reserved(*ptr) && !is_unreserved(*ptr)) {
            if(flags & URL_ESCAPE_UNSAFE) {
                pct_encode_val(*ptr, buffer+len);
                len += 3;
                do_default_action = FALSE;
            }
        }

        if(do_default_action)
            buffer[len++] = *ptr;
    }

    /* Sometimes the path is the very last component of the IUri, so
     * see if the dot segments need to be reduced now.
     */
    if(reduce_path && path) {
        DWORD current_path_len = (buffer+len) - path;
        DWORD new_path_len = remove_dot_segments(path, current_path_len);

        /* Update the current length. */
        len -= (current_path_len-new_path_len);
    }

    buffer[len++] = 0;

6301
    /* The null terminator isn't included in the length. */
6302 6303 6304 6305 6306 6307 6308 6309 6310
    *result_len = len-1;
    if(len > output_len)
        return STRSAFE_E_INSUFFICIENT_BUFFER;
    else
        memcpy(output, buffer, len*sizeof(WCHAR));

    return S_OK;
}

6311 6312 6313 6314 6315 6316 6317 6318 6319 6320 6321 6322 6323 6324 6325 6326 6327 6328 6329 6330 6331 6332 6333 6334 6335 6336 6337 6338
static HRESULT parse_friendly(IUri *uri, LPWSTR output, DWORD output_len,
                              DWORD *result_len)
{
    HRESULT hr;
    DWORD display_len;
    BSTR display;

    hr = IUri_GetPropertyLength(uri, Uri_PROPERTY_DISPLAY_URI, &display_len, 0);
    if(FAILED(hr)) {
        *result_len = 0;
        return hr;
    }

    *result_len = display_len;
    if(display_len+1 > output_len)
        return STRSAFE_E_INSUFFICIENT_BUFFER;

    hr = IUri_GetDisplayUri(uri, &display);
    if(FAILED(hr)) {
        *result_len = 0;
        return hr;
    }

    memcpy(output, display, (display_len+1)*sizeof(WCHAR));
    SysFreeString(display);
    return S_OK;
}

6339 6340 6341 6342 6343 6344 6345 6346 6347 6348 6349 6350 6351 6352 6353 6354 6355 6356 6357 6358 6359 6360 6361 6362 6363 6364 6365 6366 6367 6368 6369 6370 6371 6372 6373 6374 6375 6376 6377 6378 6379 6380 6381 6382 6383 6384 6385 6386 6387 6388 6389
static HRESULT parse_rootdocument(const Uri *uri, LPWSTR output, DWORD output_len,
                                  DWORD *result_len)
{
    static const WCHAR colon_slashesW[] = {':','/','/'};

    WCHAR *ptr;
    DWORD len = 0;

    /* Windows only returns the root document if the URI has an authority
     * and it's not an unknown scheme type or a file scheme type.
     */
    if(uri->authority_start == -1 ||
       uri->scheme_type == URL_SCHEME_UNKNOWN ||
       uri->scheme_type == URL_SCHEME_FILE) {
        *result_len = 0;
        if(!output_len)
            return STRSAFE_E_INSUFFICIENT_BUFFER;

        output[0] = 0;
        return S_OK;
    }

    len = uri->scheme_len+uri->authority_len;
    /* For the "://" and '/' which will be added. */
    len += 4;

    if(len+1 > output_len) {
        *result_len = len;
        return STRSAFE_E_INSUFFICIENT_BUFFER;
    }

    ptr = output;
    memcpy(ptr, uri->canon_uri+uri->scheme_start, uri->scheme_len*sizeof(WCHAR));

    /* Add the "://". */
    ptr += uri->scheme_len;
    memcpy(ptr, colon_slashesW, sizeof(colon_slashesW));

    /* Add the authority. */
    ptr += sizeof(colon_slashesW)/sizeof(WCHAR);
    memcpy(ptr, uri->canon_uri+uri->authority_start, uri->authority_len*sizeof(WCHAR));

    /* Add the '/' after the authority. */
    ptr += uri->authority_len;
    *ptr = '/';
    ptr[1] = 0;

    *result_len = len;
    return S_OK;
}

6390 6391 6392 6393 6394 6395 6396 6397 6398 6399 6400 6401 6402 6403 6404 6405 6406 6407 6408 6409 6410 6411 6412 6413 6414 6415 6416 6417 6418 6419 6420 6421 6422
static HRESULT parse_document(const Uri *uri, LPWSTR output, DWORD output_len,
                              DWORD *result_len)
{
    DWORD len = 0;

    /* It has to be a known scheme type, but, it can't be a file
     * scheme. It also has to hierarchical.
     */
    if(uri->scheme_type == URL_SCHEME_UNKNOWN ||
       uri->scheme_type == URL_SCHEME_FILE ||
       uri->authority_start == -1) {
        *result_len = 0;
        if(output_len < 1)
            return STRSAFE_E_INSUFFICIENT_BUFFER;

        output[0] = 0;
        return S_OK;
    }

    if(uri->fragment_start > -1)
        len = uri->fragment_start;
    else
        len = uri->canon_len;

    *result_len = len;
    if(len+1 > output_len)
        return STRSAFE_E_INSUFFICIENT_BUFFER;

    memcpy(output, uri->canon_uri, len*sizeof(WCHAR));
    output[len] = 0;
    return S_OK;
}

6423 6424 6425 6426 6427 6428 6429 6430 6431 6432 6433 6434 6435 6436 6437 6438 6439 6440 6441 6442 6443 6444 6445 6446 6447 6448 6449 6450 6451 6452 6453 6454 6455 6456 6457 6458 6459 6460 6461 6462 6463 6464 6465 6466 6467 6468 6469 6470 6471 6472 6473 6474 6475 6476 6477 6478 6479 6480
static HRESULT parse_path_from_url(const Uri *uri, LPWSTR output, DWORD output_len,
                                   DWORD *result_len)
{
    const WCHAR *path_ptr;
    WCHAR buffer[INTERNET_MAX_URL_LENGTH+1];
    WCHAR *ptr;

    if(uri->scheme_type != URL_SCHEME_FILE) {
        *result_len = 0;
        if(output_len > 0)
            output[0] = 0;
        return E_INVALIDARG;
    }

    ptr = buffer;
    if(uri->host_start > -1) {
        static const WCHAR slash_slashW[] = {'\\','\\'};

        memcpy(ptr, slash_slashW, sizeof(slash_slashW));
        ptr += sizeof(slash_slashW)/sizeof(WCHAR);
        memcpy(ptr, uri->canon_uri+uri->host_start, uri->host_len*sizeof(WCHAR));
        ptr += uri->host_len;
    }

    path_ptr = uri->canon_uri+uri->path_start;
    if(uri->path_len > 3 && *path_ptr == '/' && is_drive_path(path_ptr+1))
        /* Skip past the '/' in front of the drive path. */
        ++path_ptr;

    for(; path_ptr < uri->canon_uri+uri->path_start+uri->path_len; ++path_ptr, ++ptr) {
        BOOL do_default_action = TRUE;

        if(*path_ptr == '%') {
            const WCHAR decoded = decode_pct_val(path_ptr);
            if(decoded) {
                *ptr = decoded;
                path_ptr += 2;
                do_default_action = FALSE;
            }
        } else if(*path_ptr == '/') {
            *ptr = '\\';
            do_default_action = FALSE;
        }

        if(do_default_action)
            *ptr = *path_ptr;
    }

    *ptr = 0;

    *result_len = ptr-buffer;
    if(*result_len+1 > output_len)
        return STRSAFE_E_INSUFFICIENT_BUFFER;

    memcpy(output, buffer, (*result_len+1)*sizeof(WCHAR));
    return S_OK;
}

6481 6482 6483 6484 6485 6486 6487 6488 6489 6490 6491 6492 6493 6494 6495 6496 6497 6498 6499 6500 6501 6502 6503 6504 6505 6506 6507 6508 6509
static HRESULT parse_url_from_path(IUri *uri, LPWSTR output, DWORD output_len,
                                   DWORD *result_len)
{
    HRESULT hr;
    BSTR received;
    DWORD len = 0;

    hr = IUri_GetPropertyLength(uri, Uri_PROPERTY_ABSOLUTE_URI, &len, 0);
    if(FAILED(hr)) {
        *result_len = 0;
        return hr;
    }

    *result_len = len;
    if(len+1 > output_len)
        return STRSAFE_E_INSUFFICIENT_BUFFER;

    hr = IUri_GetAbsoluteUri(uri, &received);
    if(FAILED(hr)) {
        *result_len = 0;
        return hr;
    }

    memcpy(output, received, (len+1)*sizeof(WCHAR));
    SysFreeString(received);

    return S_OK;
}

6510 6511 6512 6513 6514 6515 6516 6517 6518 6519 6520 6521 6522 6523 6524 6525 6526 6527 6528 6529 6530 6531 6532 6533 6534 6535 6536 6537 6538
static HRESULT parse_schema(IUri *uri, LPWSTR output, DWORD output_len,
                            DWORD *result_len)
{
    HRESULT hr;
    DWORD len;
    BSTR received;

    hr = IUri_GetPropertyLength(uri, Uri_PROPERTY_SCHEME_NAME, &len, 0);
    if(FAILED(hr)) {
        *result_len = 0;
        return hr;
    }

    *result_len = len;
    if(len+1 > output_len)
        return STRSAFE_E_INSUFFICIENT_BUFFER;

    hr = IUri_GetSchemeName(uri, &received);
    if(FAILED(hr)) {
        *result_len = 0;
        return hr;
    }

    memcpy(output, received, (len+1)*sizeof(WCHAR));
    SysFreeString(received);

    return S_OK;
}

6539 6540 6541 6542 6543 6544 6545 6546 6547 6548 6549 6550 6551 6552 6553 6554 6555 6556 6557 6558 6559 6560 6561 6562 6563 6564 6565 6566
static HRESULT parse_site(IUri *uri, LPWSTR output, DWORD output_len, DWORD *result_len)
{
    HRESULT hr;
    DWORD len;
    BSTR received;

    hr = IUri_GetPropertyLength(uri, Uri_PROPERTY_HOST, &len, 0);
    if(FAILED(hr)) {
        *result_len = 0;
        return hr;
    }

    *result_len = len;
    if(len+1 > output_len)
        return STRSAFE_E_INSUFFICIENT_BUFFER;

    hr = IUri_GetHost(uri, &received);
    if(FAILED(hr)) {
        *result_len = 0;
        return hr;
    }

    memcpy(output, received, (len+1)*sizeof(WCHAR));
    SysFreeString(received);

    return S_OK;
}

6567 6568 6569 6570 6571 6572 6573 6574 6575 6576 6577 6578 6579 6580 6581 6582 6583 6584 6585 6586 6587 6588 6589 6590 6591 6592 6593 6594
static HRESULT parse_domain(IUri *uri, LPWSTR output, DWORD output_len, DWORD *result_len)
{
    HRESULT hr;
    DWORD len;
    BSTR received;

    hr = IUri_GetPropertyLength(uri, Uri_PROPERTY_DOMAIN, &len, 0);
    if(FAILED(hr)) {
        *result_len = 0;
        return hr;
    }

    *result_len = len;
    if(len+1 > output_len)
        return STRSAFE_E_INSUFFICIENT_BUFFER;

    hr = IUri_GetDomain(uri, &received);
    if(FAILED(hr)) {
        *result_len = 0;
        return hr;
    }

    memcpy(output, received, (len+1)*sizeof(WCHAR));
    SysFreeString(received);

    return S_OK;
}

6595 6596 6597 6598 6599 6600 6601 6602 6603 6604 6605 6606 6607 6608 6609 6610 6611 6612 6613 6614 6615 6616 6617 6618 6619 6620 6621 6622
static HRESULT parse_anchor(IUri *uri, LPWSTR output, DWORD output_len, DWORD *result_len)
{
    HRESULT hr;
    DWORD len;
    BSTR received;

    hr = IUri_GetPropertyLength(uri, Uri_PROPERTY_FRAGMENT, &len, 0);
    if(FAILED(hr)) {
        *result_len = 0;
        return hr;
    }

    *result_len = len;
    if(len+1 > output_len)
        return STRSAFE_E_INSUFFICIENT_BUFFER;

    hr = IUri_GetFragment(uri, &received);
    if(FAILED(hr)) {
        *result_len = 0;
        return hr;
    }

    memcpy(output, received, (len+1)*sizeof(WCHAR));
    SysFreeString(received);

    return S_OK;
}

6623 6624 6625 6626 6627 6628 6629 6630
/***********************************************************************
 *           CoInternetParseIUri (urlmon.@)
 */
HRESULT WINAPI CoInternetParseIUri(IUri *pIUri, PARSEACTION ParseAction, DWORD dwFlags,
                                   LPWSTR pwzResult, DWORD cchResult, DWORD *pcchResult,
                                   DWORD_PTR dwReserved)
{
    HRESULT hr;
6631
    Uri *uri;
6632
    IInternetProtocolInfo *info;
6633 6634 6635 6636 6637 6638 6639 6640 6641 6642 6643 6644

    TRACE("(%p %d %x %p %d %p %x)\n", pIUri, ParseAction, dwFlags, pwzResult,
        cchResult, pcchResult, (DWORD)dwReserved);

    if(!pcchResult)
        return E_POINTER;

    if(!pwzResult || !pIUri) {
        *pcchResult = 0;
        return E_INVALIDARG;
    }

6645 6646 6647 6648 6649 6650 6651 6652 6653 6654 6655 6656 6657 6658 6659
    if(!(uri = get_uri_obj(pIUri))) {
        *pcchResult = 0;
        FIXME("(%p %d %x %p %d %p %x) Unknown IUri's not supported for this action.\n",
            pIUri, ParseAction, dwFlags, pwzResult, cchResult, pcchResult, (DWORD)dwReserved);
        return E_NOTIMPL;
    }

    info = get_protocol_info(uri->canon_uri);
    if(info) {
        hr = IInternetProtocolInfo_ParseUrl(info, uri->canon_uri, ParseAction, dwFlags,
                                            pwzResult, cchResult, pcchResult, 0);
        IInternetProtocolInfo_Release(info);
        if(SUCCEEDED(hr)) return hr;
    }

6660
    switch(ParseAction) {
6661 6662 6663
    case PARSE_CANONICALIZE:
        hr = parse_canonicalize(uri, dwFlags, pwzResult, cchResult, pcchResult);
        break;
6664 6665 6666
    case PARSE_FRIENDLY:
        hr = parse_friendly(pIUri, pwzResult, cchResult, pcchResult);
        break;
6667 6668 6669
    case PARSE_ROOTDOCUMENT:
        hr = parse_rootdocument(uri, pwzResult, cchResult, pcchResult);
        break;
6670 6671 6672
    case PARSE_DOCUMENT:
        hr = parse_document(uri, pwzResult, cchResult, pcchResult);
        break;
6673 6674 6675
    case PARSE_PATH_FROM_URL:
        hr = parse_path_from_url(uri, pwzResult, cchResult, pcchResult);
        break;
6676 6677 6678
    case PARSE_URL_FROM_PATH:
        hr = parse_url_from_path(pIUri, pwzResult, cchResult, pcchResult);
        break;
6679 6680 6681
    case PARSE_SCHEMA:
        hr = parse_schema(pIUri, pwzResult, cchResult, pcchResult);
        break;
6682 6683 6684
    case PARSE_SITE:
        hr = parse_site(pIUri, pwzResult, cchResult, pcchResult);
        break;
6685 6686 6687
    case PARSE_DOMAIN:
        hr = parse_domain(pIUri, pwzResult, cchResult, pcchResult);
        break;
6688 6689 6690 6691
    case PARSE_LOCATION:
    case PARSE_ANCHOR:
        hr = parse_anchor(pIUri, pwzResult, cchResult, pcchResult);
        break;
6692 6693 6694 6695 6696 6697 6698 6699 6700 6701 6702 6703 6704 6705 6706 6707
    case PARSE_SECURITY_URL:
    case PARSE_MIME:
    case PARSE_SERVER:
    case PARSE_SECURITY_DOMAIN:
        *pcchResult = 0;
        hr = E_FAIL;
        break;
    default:
        *pcchResult = 0;
        hr = E_NOTIMPL;
        FIXME("(%p %d %x %p %d %p %x) Partial stub.\n", pIUri, ParseAction, dwFlags,
            pwzResult, cchResult, pcchResult, (DWORD)dwReserved);
    }

    return hr;
}