chrome/browser/drive/drive_api_util.cc

   1 // Copyright (c) 2012 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 #include "chrome/browser/drive/drive_api_util.h"
   6
   7 #include <string>
   8
   9 #include "base/files/file.h"
  10 #include "base/logging.h"
  11 #include "base/md5.h"
  12 #include "base/strings/string16.h"
  13 #include "base/strings/string_util.h"
  14 #include "base/strings/stringprintf.h"
  15 #include "base/strings/utf_string_conversions.h"
  16 #include "base/values.h"
  17 #include "google_apis/drive/drive_api_parser.h"
  18 #include "net/base/escape.h"
  19 #include "net/base/io_buffer.h"
  20 #include "net/base/net_errors.h"
  21 #include "storage/browser/blob/file_stream_reader.h"
  22 #include "third_party/re2/re2/re2.h"
  23 #include "url/gurl.h"
  24
  25 namespace drive {
  26 namespace util {
  27 namespace {
  28
  29 struct HostedDocumentKind {
  30   const char* mime_type;
  31   const char* extension;
  32 };
  33
  34 const HostedDocumentKind kHostedDocumentKinds[] = {
  35     {kGoogleDocumentMimeType,     ".gdoc"},
  36     {kGoogleSpreadsheetMimeType,  ".gsheet"},
  37     {kGooglePresentationMimeType, ".gslides"},
  38     {kGoogleDrawingMimeType,      ".gdraw"},
  39     {kGoogleTableMimeType,        ".gtable"},
  40     {kGoogleFormMimeType,         ".gform"},
  41     {kGoogleMapMimeType,          ".gmaps"},
  42 };
  43
  44 const char kUnknownHostedDocumentExtension[] = ".glink";
  45
  46 const int kMd5DigestBufferSize = 512 * 1024;  // 512 kB.
  47
  48 }  // namespace
  49
  50 std::string EscapeQueryStringValue(const std::string& str) {
  51   std::string result;
  52   result.reserve(str.size());
  53   for (size_t i = 0; i < str.size(); ++i) {
  54     if (str[i] == '\\' || str[i] == '\'') {
  55       result.push_back('\\');
  56     }
  57     result.push_back(str[i]);
  58   }
  59   return result;
  60 }
  61
  62 std::string TranslateQuery(const std::string& original_query) {
  63   // In order to handle non-ascii white spaces correctly, convert to UTF16.
  64   base::string16 query = base::UTF8ToUTF16(original_query);
  65   const base::string16 kDelimiter(
  66       base::kWhitespaceUTF16 + base::ASCIIToUTF16("\""));
  67
  68   std::string result;
  69   for (size_t index = query.find_first_not_of(base::kWhitespaceUTF16);
  70        index != base::string16::npos;
  71        index = query.find_first_not_of(base::kWhitespaceUTF16, index)) {
  72     bool is_exclusion = (query[index] == '-');
  73     if (is_exclusion)
  74       ++index;
  75     if (index == query.length()) {
  76       // Here, the token is '-' and it should be ignored.
  77       continue;
  78     }
  79
  80     size_t begin_token = index;
  81     base::string16 token;
  82     if (query[begin_token] == '"') {
  83       // Quoted query.
  84       ++begin_token;
  85       size_t end_token = query.find('"', begin_token);
  86       if (end_token == base::string16::npos) {
  87         // This is kind of syntax error, since quoted string isn't finished.
  88         // However, the query is built by user manually, so here we treat
  89         // whole remaining string as a token as a fallback, by appending
  90         // a missing double-quote character.
  91         end_token = query.length();
  92         query.push_back('"');
  93       }
  94
  95       token = query.substr(begin_token, end_token - begin_token);
  96       index = end_token + 1;  // Consume last '"', too.
  97     } else {
  98       size_t end_token = query.find_first_of(kDelimiter, begin_token);
  99       if (end_token == base::string16::npos) {
 100         end_token = query.length();
 101       }
 102
 103       token = query.substr(begin_token, end_token - begin_token);
 104       index = end_token;
 105     }
 106
 107     if (token.empty()) {
 108       // Just ignore an empty token.
 109       continue;
 110     }
 111
 112     if (!result.empty()) {
 113       // If there are two or more tokens, need to connect with "and".
 114       result.append(" and ");
 115     }
 116
 117     // The meaning of "fullText" should include title, description and content.
 118     base::StringAppendF(
 119         &result,
 120         "%sfullText contains \'%s\'",
 121         is_exclusion ? "not " : "",
 122         EscapeQueryStringValue(base::UTF16ToUTF8(token)).c_str());
 123   }
 124
 125   return result;
 126 }
 127
 128 std::string CanonicalizeResourceId(const std::string& resource_id) {
 129   // If resource ID is in the old WAPI format starting with a prefix like
 130   // "document:", strip it and return the remaining part.
 131   std::string stripped_resource_id;
 132   if (RE2::FullMatch(resource_id, "^[a-z-]+(?::|%3A)([\\w-]+)$",
 133                      &stripped_resource_id))
 134     return stripped_resource_id;
 135   return resource_id;
 136 }
 137
 138 std::string GetMd5Digest(const base::FilePath& file_path) {
 139   base::File file(file_path, base::File::FLAG_OPEN | base::File::FLAG_READ);
 140   if (!file.IsValid())
 141     return std::string();
 142
 143   base::MD5Context context;
 144   base::MD5Init(&context);
 145
 146   int64 offset = 0;
 147   scoped_ptr<char[]> buffer(new char[kMd5DigestBufferSize]);
 148   while (true) {
 149     int result = file.Read(offset, buffer.get(), kMd5DigestBufferSize);
 150     if (result < 0) {
 151       // Found an error.
 152       return std::string();
 153     }
 154
 155     if (result == 0) {
 156       // End of file.
 157       break;
 158     }
 159
 160     offset += result;
 161     base::MD5Update(&context, base::StringPiece(buffer.get(), result));
 162   }
 163
 164   base::MD5Digest digest;
 165   base::MD5Final(&digest, &context);
 166   return MD5DigestToBase16(digest);
 167 }
 168
 169 FileStreamMd5Digester::FileStreamMd5Digester()
 170     : buffer_(new net::IOBuffer(kMd5DigestBufferSize)) {
 171 }
 172
 173 FileStreamMd5Digester::~FileStreamMd5Digester() {
 174 }
 175
 176 void FileStreamMd5Digester::GetMd5Digest(
 177     scoped_ptr<storage::FileStreamReader> stream_reader,
 178     const ResultCallback& callback) {
 179   reader_ = stream_reader.Pass();
 180   callback_ = callback;
 181   base::MD5Init(&md5_context_);
 182
 183   // Start the read/hash.
 184   ReadNextChunk();
 185 }
 186
 187 void FileStreamMd5Digester::ReadNextChunk() {
 188   const int result = reader_->Read(
 189       buffer_.get(), kMd5DigestBufferSize,
 190       base::Bind(&FileStreamMd5Digester::OnChunkRead, base::Unretained(this)));
 191   if (result != net::ERR_IO_PENDING)
 192     OnChunkRead(result);
 193 }
 194
 195 void FileStreamMd5Digester::OnChunkRead(int bytesRead) {
 196   if (bytesRead < 0) {
 197     // Error - just return empty string.
 198     callback_.Run("");
 199     return;
 200   } else if (bytesRead == 0) {
 201     // EOF.
 202     base::MD5Digest digest;
 203     base::MD5Final(&digest, &md5_context_);
 204     std::string result = MD5DigestToBase16(digest);
 205     callback_.Run(result);
 206     return;
 207   }
 208
 209   // Read data and digest it.
 210   base::MD5Update(&md5_context_, base::StringPiece(buffer_->data(), bytesRead));
 211
 212   // Kick off the next read.
 213   ReadNextChunk();
 214 }
 215
 216 std::string GetHostedDocumentExtension(const std::string& mime_type) {
 217   for (size_t i = 0; i < arraysize(kHostedDocumentKinds); ++i) {
 218     if (mime_type == kHostedDocumentKinds[i].mime_type)
 219       return kHostedDocumentKinds[i].extension;
 220   }
 221   return kUnknownHostedDocumentExtension;
 222 }
 223
 224 bool IsKnownHostedDocumentMimeType(const std::string& mime_type) {
 225   for (size_t i = 0; i < arraysize(kHostedDocumentKinds); ++i) {
 226     if (mime_type == kHostedDocumentKinds[i].mime_type)
 227       return true;
 228   }
 229   return false;
 230 }
 231
 232 bool HasHostedDocumentExtension(const base::FilePath& path) {
 233   const std::string extension = base::FilePath(path.Extension()).AsUTF8Unsafe();
 234   for (size_t i = 0; i < arraysize(kHostedDocumentKinds); ++i) {
 235     if (extension == kHostedDocumentKinds[i].extension)
 236       return true;
 237   }
 238   return extension == kUnknownHostedDocumentExtension;
 239 }
 240
 241 }  // namespace util
 242 }  // namespace drive