chrome/tools/convert_dict/hunspell_reader.cc

   1 // Copyright (c) 2006-2008 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/tools/convert_dict/hunspell_reader.h"
   6
   7 #include "base/strings/string_util.h"
   8
   9 namespace convert_dict {
  10
  11 // This silly 64K buffer is just copied from Hunspell's way of parsing.
  12 const int kLineBufferLen = 65535;
  13 char line_buffer[kLineBufferLen];
  14
  15 // Shortcut for trimming whitespace from both ends of the line.
  16 void TrimLine(std::string* line) {
  17   if (line->size() > 3 &&
  18       static_cast<unsigned char>((*line)[0]) == 0xef &&
  19       static_cast<unsigned char>((*line)[1]) == 0xbb &&
  20       static_cast<unsigned char>((*line)[2]) == 0xbf)
  21     *line = line->substr(3);
  22
  23   // Treat this text as an ASCII text and trim whitespace characters as
  24   // hunspell does. The returned text is to be converted into UTF-8 text with
  25   // the encoding defined in an affix file.
  26   base::TrimWhitespace(*line, base::TRIM_ALL, line);
  27 }
  28
  29 std::string ReadLine(FILE* file) {
  30   const char* line = fgets(line_buffer, kLineBufferLen - 1, file);
  31   if (!line)
  32     return std::string();
  33
  34   std::string str = line;
  35   TrimLine(&str);
  36   return str;
  37 }
  38
  39 void StripComment(std::string* line) {
  40   for (size_t i = 0; i < line->size(); i++) {
  41     if ((*line)[i] == '#') {
  42       line->resize(i);
  43       TrimLine(line);
  44       return;
  45     }
  46   }
  47 }
  48
  49 }  // namespace convert_dict