tor-browser

The Tor Browser
git clone https://git.dasho.dev/tor-browser.git
Log | Files | Refs | README | LICENSE

mozEnglishWordUtils.cpp (3285B)


      1 /* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
      2 /* This Source Code Form is subject to the terms of the Mozilla Public
      3 * License, v. 2.0. If a copy of the MPL was not distributed with this
      4 * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
      5 
      6 #include "mozEnglishWordUtils.h"
      7 #include "nsComponentManagerUtils.h"
      8 #include "nsReadableUtils.h"
      9 #include "nsUnicharUtils.h"
     10 #include "nsUnicodeProperties.h"
     11 #include "nsCRT.h"
     12 
     13 NS_IMPL_CYCLE_COLLECTION(mozEnglishWordUtils, mURLDetector)
     14 
     15 mozEnglishWordUtils::mozEnglishWordUtils() {
     16  mURLDetector = do_CreateInstance(MOZ_TXTTOHTMLCONV_CONTRACTID);
     17 }
     18 
     19 mozEnglishWordUtils::~mozEnglishWordUtils() {}
     20 
     21 // This needs vast improvement
     22 
     23 // static
     24 bool mozEnglishWordUtils::ucIsAlpha(char16_t aChar) {
     25  // XXX we have to fix callers to handle the full Unicode range
     26  return nsUGenCategory::kLetter == mozilla::unicode::GetGenCategory(aChar);
     27 }
     28 
     29 bool mozEnglishWordUtils::FindNextWord(const nsAString& aWord, uint32_t offset,
     30                                       int32_t* begin, int32_t* end) {
     31  if (offset >= aWord.Length()) {
     32    *begin = -1;
     33    *end = -1;
     34    return false;
     35  }
     36 
     37  const char16_t* word = aWord.BeginReading();
     38  uint32_t length = aWord.Length();
     39  const char16_t* p = word + offset;
     40  const char16_t* endbuf = word + length;
     41  const char16_t* startWord = p;
     42 
     43  // XXX These loops should be modified to handle non-BMP characters.
     44  // if previous character is a word character, need to advance out of the
     45  // word
     46  if (offset > 0 && ucIsAlpha(*(p - 1))) {
     47    while (p < endbuf && ucIsAlpha(*p)) {
     48      p++;
     49    }
     50  }
     51  while ((p < endbuf) && (!ucIsAlpha(*p))) {
     52    p++;
     53  }
     54  startWord = p;
     55  while ((p < endbuf) && ((ucIsAlpha(*p)) || (*p == '\''))) {
     56    p++;
     57  }
     58 
     59  // we could be trying to break down a url, we don't want to break a url into
     60  // parts, instead we want to find out if it really is a url and if so, skip
     61  // it, advancing startWord to a point after the url.
     62 
     63  // before we spend more time looking to see if the word is a url, look for a
     64  // url identifer and make sure that identifer isn't the last character in
     65  // the word fragment.
     66  if ((p < endbuf - 1) && (*p == ':' || *p == '@' || *p == '.')) {
     67    // ok, we have a possible url...do more research to find out if we really
     68    // have one and determine the length of the url so we can skip over it.
     69 
     70    if (mURLDetector) {
     71      int32_t startPos = -1;
     72      int32_t endPos = -1;
     73 
     74      mURLDetector->FindURLInPlaintext(startWord, endbuf - startWord,
     75                                       p - startWord, &startPos, &endPos);
     76 
     77      // ok, if we got a url, adjust the array bounds, skip the current url
     78      // text and find the next word again
     79      if (startPos != -1 && endPos != -1) {
     80        startWord = p + endPos + 1;  // skip over the url
     81 
     82        // now recursively call FindNextWord to search for the next word now
     83        // that we have skipped the url
     84        return FindNextWord(aWord, startWord - word, begin, end);
     85      }
     86    }
     87  }
     88 
     89  while ((p > startWord) && (*(p - 1) == '\'')) {  // trim trailing apostrophes
     90    p--;
     91  }
     92 
     93  if (startWord == endbuf) {
     94    *begin = -1;
     95    *end = -1;
     96    return false;
     97  }
     98  *begin = startWord - word;
     99  *end = p - word;
    100  return true;
    101 }