content/browser/appcache/appcache_manifest_parser.cc

   1 // Copyright 2014 The Chromium Authors. All rights reserved.
   2 // Use of this source code is governed by a BSD-style license that can be
   3 // found in the LICENSE file.
   4 //
   5 // This is a port of ManifestParser.cc from WebKit/WebCore/loader/appcache.
   6
   7 /*
   8  * Copyright (C) 2008 Apple Inc. All Rights Reserved.
   9  *
  10  * Redistribution and use in source and binary forms, with or without
  11  * modification, are permitted provided that the following conditions
  12  * are met:
  13  * 1. Redistributions of source code must retain the above copyright
  14  *    notice, this list of conditions and the following disclaimer.
  15  * 2. Redistributions in binary form must reproduce the above copyright
  16  *    notice, this list of conditions and the following disclaimer in the
  17  *    documentation and/or other materials provided with the distribution.
  18  *
  19  * THIS SOFTWARE IS PROVIDED BY APPLE INC. ``AS IS'' AND ANY
  20  * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  21  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
  22  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL APPLE INC. OR
  23  * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
  24  * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
  25  * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
  26  * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
  27  * OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
  28  * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
  29  * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  30  */
  31
  32 #include "content/browser/appcache/appcache_manifest_parser.h"
  33
  34 #include "base/command_line.h"
  35 #include "base/i18n/icu_string_conversions.h"
  36 #include "base/logging.h"
  37 #include "base/strings/utf_string_conversions.h"
  38 #include "url/gurl.h"
  39
  40 namespace content {
  41
  42 namespace {
  43
  44 // Helper function used to identify 'isPattern' annotations.
  45 bool HasPatternMatchingAnnotation(const wchar_t* line_p,
  46                                   const wchar_t* line_end) {
  47   // Skip whitespace separating the resource url from the annotation.
  48   // Note: trailing whitespace has already been trimmed from the line.
  49   while (line_p < line_end && (*line_p == '\t' || *line_p == ' '))
  50     ++line_p;
  51   if (line_p == line_end)
  52     return false;
  53   std::wstring annotation(line_p, line_end - line_p);
  54   return annotation == L"isPattern";
  55 }
  56
  57 }
  58
  59 enum Mode {
  60   EXPLICIT,
  61   INTERCEPT,
  62   FALLBACK,
  63   ONLINE_WHITELIST,
  64   UNKNOWN_MODE,
  65 };
  66
  67 enum InterceptVerb {
  68   RETURN,
  69   EXECUTE,
  70   UNKNOWN_VERB,
  71 };
  72
  73 AppCacheManifest::AppCacheManifest()
  74     : online_whitelist_all(false),
  75       did_ignore_intercept_namespaces(false) {
  76 }
  77
  78 AppCacheManifest::~AppCacheManifest() {}
  79
  80 bool ParseManifest(const GURL& manifest_url, const char* data, int length,
  81                    ParseMode parse_mode, AppCacheManifest& manifest) {
  82   // This is an implementation of the parsing algorithm specified in
  83   // the HTML5 offline web application docs:
  84   //   http://www.w3.org/TR/html5/offline.html
  85   // Do not modify it without consulting those docs.
  86   // Though you might be tempted to convert these wstrings to UTF-8 or
  87   // base::string16, this implementation seems simpler given the constraints.
  88
  89   const wchar_t kSignature[] = L"CACHE MANIFEST";
  90   const size_t kSignatureLength = arraysize(kSignature) - 1;
  91   const wchar_t kChromiumSignature[] = L"CHROMIUM CACHE MANIFEST";
  92   const size_t kChromiumSignatureLength = arraysize(kChromiumSignature) - 1;
  93
  94   DCHECK(manifest.explicit_urls.empty());
  95   DCHECK(manifest.fallback_namespaces.empty());
  96   DCHECK(manifest.online_whitelist_namespaces.empty());
  97   DCHECK(!manifest.online_whitelist_all);
  98   DCHECK(!manifest.did_ignore_intercept_namespaces);
  99
 100   Mode mode = EXPLICIT;
 101
 102   std::wstring data_string;
 103   // TODO(jennb): cannot do UTF8ToWide(data, length, &data_string);
 104   // until UTF8ToWide uses 0xFFFD Unicode replacement character.
 105   base::CodepageToWide(std::string(data, length), base::kCodepageUTF8,
 106                        base::OnStringConversionError::SUBSTITUTE, &data_string);
 107   const wchar_t* p = data_string.c_str();
 108   const wchar_t* end = p + data_string.length();
 109
 110   // Look for the magic signature: "^\xFEFF?CACHE MANIFEST[ \t]?"
 111   // Example: "CACHE MANIFEST #comment" is a valid signature.
 112   // Example: "CACHE MANIFEST;V2" is not.
 113
 114   // When the input data starts with a UTF-8 Byte-Order-Mark
 115   // (0xEF, 0xBB, 0xBF), the UTF8ToWide() function converts it to a
 116   // Unicode BOM (U+FEFF). Skip a converted Unicode BOM if it exists.
 117   int bom_offset = 0;
 118   if (!data_string.empty() && data_string[0] == 0xFEFF) {
 119     bom_offset = 1;
 120     ++p;
 121   }
 122
 123   if (p >= end)
 124     return false;
 125
 126   // Check for a supported signature and skip p past it.
 127   if (0 == data_string.compare(bom_offset, kSignatureLength,
 128                                kSignature)) {
 129     p += kSignatureLength;
 130   } else if (0 == data_string.compare(bom_offset, kChromiumSignatureLength,
 131                                       kChromiumSignature)) {
 132     p += kChromiumSignatureLength;
 133   } else {
 134     return false;
 135   }
 136
 137   // Character after "CACHE MANIFEST" must be whitespace.
 138   if (p < end && *p != ' ' && *p != '\t' && *p != '\n' && *p != '\r')
 139     return false;
 140
 141   // Skip to the end of the line.
 142   while (p < end && *p != '\r' && *p != '\n')
 143     ++p;
 144
 145   while (1) {
 146     // Skip whitespace
 147     while (p < end && (*p == '\n' || *p == '\r' || *p == ' ' || *p == '\t'))
 148       ++p;
 149
 150     if (p == end)
 151       break;
 152
 153     const wchar_t* line_start = p;
 154
 155     // Find the end of the line
 156     while (p < end && *p != '\r' && *p != '\n')
 157       ++p;
 158
 159     // Check if we have a comment
 160     if (*line_start == '#')
 161       continue;
 162
 163     // Get rid of trailing whitespace
 164     const wchar_t* tmp = p - 1;
 165     while (tmp > line_start && (*tmp == ' ' || *tmp == '\t'))
 166       --tmp;
 167
 168     std::wstring line(line_start, tmp - line_start + 1);
 169
 170     if (line == L"CACHE:") {
 171       mode = EXPLICIT;
 172     } else if (line == L"FALLBACK:") {
 173       mode = FALLBACK;
 174     } else if (line == L"NETWORK:") {
 175       mode = ONLINE_WHITELIST;
 176     } else if (line == L"CHROMIUM-INTERCEPT:") {
 177       mode = INTERCEPT;
 178     } else if (*(line.end() - 1) == ':') {
 179       mode = UNKNOWN_MODE;
 180     } else if (mode == UNKNOWN_MODE) {
 181       continue;
 182     } else if (line == L"*" && mode == ONLINE_WHITELIST) {
 183       manifest.online_whitelist_all = true;
 184       continue;
 185     } else if (mode == EXPLICIT || mode == ONLINE_WHITELIST) {
 186       const wchar_t *line_p = line.c_str();
 187       const wchar_t *line_end = line_p + line.length();
 188
 189       // Look for whitespace separating the URL from subsequent ignored tokens.
 190       while (line_p < line_end && *line_p != '\t' && *line_p != ' ')
 191         ++line_p;
 192
 193       base::string16 url16;
 194       base::WideToUTF16(line.c_str(), line_p - line.c_str(), &url16);
 195       GURL url = manifest_url.Resolve(url16);
 196       if (!url.is_valid())
 197         continue;
 198       if (url.has_ref()) {
 199         GURL::Replacements replacements;
 200         replacements.ClearRef();
 201         url = url.ReplaceComponents(replacements);
 202       }
 203
 204       // Scheme component must be the same as the manifest URL's.
 205       if (url.scheme() != manifest_url.scheme()) {
 206         continue;
 207       }
 208
 209       // See http://code.google.com/p/chromium/issues/detail?id=69594
 210       // We willfully violate the HTML5 spec at this point in order
 211       // to support the appcaching of cross-origin HTTPS resources.
 212       // Per the spec, EXPLICIT cross-origin HTTS resources should be
 213       // ignored here. We've opted for a milder constraint and allow
 214       // caching unless the resource has a "no-store" header. That
 215       // condition is enforced in AppCacheUpdateJob.
 216
 217       if (mode == EXPLICIT) {
 218         manifest.explicit_urls.insert(url.spec());
 219       } else {
 220         bool is_pattern = HasPatternMatchingAnnotation(line_p, line_end);
 221         manifest.online_whitelist_namespaces.push_back(
 222             AppCacheNamespace(APPCACHE_NETWORK_NAMESPACE, url, GURL(),
 223                 is_pattern));
 224       }
 225     } else if (mode == INTERCEPT) {
 226       if (parse_mode != PARSE_MANIFEST_ALLOWING_INTERCEPTS) {
 227         manifest.did_ignore_intercept_namespaces = true;
 228         continue;
 229       }
 230
 231       // Lines of the form,
 232       // <urlnamespace> <intercept_type> <targeturl>
 233       const wchar_t* line_p = line.c_str();
 234       const wchar_t* line_end = line_p + line.length();
 235
 236       // Look for first whitespace separating the url namespace from
 237       // the intercept type.
 238       while (line_p < line_end && *line_p != '\t' && *line_p != ' ')
 239         ++line_p;
 240
 241       if (line_p == line_end)
 242         continue;  // There was no whitespace separating the URLs.
 243
 244       base::string16 namespace_url16;
 245       base::WideToUTF16(line.c_str(), line_p - line.c_str(), &namespace_url16);
 246       GURL namespace_url = manifest_url.Resolve(namespace_url16);
 247       if (!namespace_url.is_valid())
 248         continue;
 249       if (namespace_url.has_ref()) {
 250         GURL::Replacements replacements;
 251         replacements.ClearRef();
 252         namespace_url = namespace_url.ReplaceComponents(replacements);
 253       }
 254
 255       // The namespace URL must have the same scheme, host and port
 256       // as the manifest's URL.
 257       if (manifest_url.GetOrigin() != namespace_url.GetOrigin())
 258         continue;
 259
 260       // Skip whitespace separating namespace from the type.
 261       while (line_p < line_end && (*line_p == '\t' || *line_p == ' '))
 262         ++line_p;
 263
 264       // Look for whitespace separating the type from the target url.
 265       const wchar_t* type_start = line_p;
 266       while (line_p < line_end && *line_p != '\t' && *line_p != ' ')
 267         ++line_p;
 268
 269       // Look for a type value we understand, otherwise skip the line.
 270       InterceptVerb verb = UNKNOWN_VERB;
 271       std::wstring type(type_start, line_p - type_start);
 272       if (type == L"return") {
 273         verb = RETURN;
 274       } else if (type == L"execute" &&
 275                  base::CommandLine::ForCurrentProcess()->HasSwitch(
 276                     kEnableExecutableHandlers)) {
 277         verb = EXECUTE;
 278       }
 279       if (verb == UNKNOWN_VERB)
 280         continue;
 281
 282       // Skip whitespace separating type from the target_url.
 283       while (line_p < line_end && (*line_p == '\t' || *line_p == ' '))
 284         ++line_p;
 285
 286       // Look for whitespace separating the URL from subsequent ignored tokens.
 287       const wchar_t* target_url_start = line_p;
 288       while (line_p < line_end && *line_p != '\t' && *line_p != ' ')
 289         ++line_p;
 290
 291       base::string16 target_url16;
 292       base::WideToUTF16(target_url_start, line_p - target_url_start,
 293                         &target_url16);
 294       GURL target_url = manifest_url.Resolve(target_url16);
 295       if (!target_url.is_valid())
 296         continue;
 297
 298       if (target_url.has_ref()) {
 299         GURL::Replacements replacements;
 300         replacements.ClearRef();
 301         target_url = target_url.ReplaceComponents(replacements);
 302       }
 303       if (manifest_url.GetOrigin() != target_url.GetOrigin())
 304         continue;
 305
 306       bool is_pattern = HasPatternMatchingAnnotation(line_p, line_end);
 307       manifest.intercept_namespaces.push_back(
 308           AppCacheNamespace(APPCACHE_INTERCEPT_NAMESPACE, namespace_url,
 309                     target_url, is_pattern, verb == EXECUTE));
 310     } else if (mode == FALLBACK) {
 311       const wchar_t* line_p = line.c_str();
 312       const wchar_t* line_end = line_p + line.length();
 313
 314       // Look for whitespace separating the two URLs
 315       while (line_p < line_end && *line_p != '\t' && *line_p != ' ')
 316         ++line_p;
 317
 318       if (line_p == line_end) {
 319         // There was no whitespace separating the URLs.
 320         continue;
 321       }
 322
 323       base::string16 namespace_url16;
 324       base::WideToUTF16(line.c_str(), line_p - line.c_str(), &namespace_url16);
 325       GURL namespace_url = manifest_url.Resolve(namespace_url16);
 326       if (!namespace_url.is_valid())
 327         continue;
 328       if (namespace_url.has_ref()) {
 329         GURL::Replacements replacements;
 330         replacements.ClearRef();
 331         namespace_url = namespace_url.ReplaceComponents(replacements);
 332       }
 333
 334       // Fallback namespace URL must have the same scheme, host and port
 335       // as the manifest's URL.
 336       if (manifest_url.GetOrigin() != namespace_url.GetOrigin()) {
 337         continue;
 338       }
 339
 340       // Skip whitespace separating fallback namespace from URL.
 341       while (line_p < line_end && (*line_p == '\t' || *line_p == ' '))
 342         ++line_p;
 343
 344       // Look for whitespace separating the URL from subsequent ignored tokens.
 345       const wchar_t* fallback_start = line_p;
 346       while (line_p < line_end && *line_p != '\t' && *line_p != ' ')
 347         ++line_p;
 348
 349       base::string16 fallback_url16;
 350       base::WideToUTF16(fallback_start, line_p - fallback_start,
 351                         &fallback_url16);
 352       GURL fallback_url = manifest_url.Resolve(fallback_url16);
 353       if (!fallback_url.is_valid())
 354         continue;
 355       if (fallback_url.has_ref()) {
 356         GURL::Replacements replacements;
 357         replacements.ClearRef();
 358         fallback_url = fallback_url.ReplaceComponents(replacements);
 359       }
 360
 361       // Fallback entry URL must have the same scheme, host and port
 362       // as the manifest's URL.
 363       if (manifest_url.GetOrigin() != fallback_url.GetOrigin()) {
 364         continue;
 365       }
 366
 367       bool is_pattern = HasPatternMatchingAnnotation(line_p, line_end);
 368
 369       // Store regardless of duplicate namespace URL. Only first match
 370       // will ever be used.
 371       manifest.fallback_namespaces.push_back(
 372           AppCacheNamespace(APPCACHE_FALLBACK_NAMESPACE, namespace_url,
 373                     fallback_url, is_pattern));
 374     } else {
 375       NOTREACHED();
 376     }
 377   }
 378
 379   return true;
 380 }
 381
 382 }  // namespace content