Method GetOneScriptSpan

cld2/internal/getonescriptspan.cc:801–1029

Copy next run of same-script non-tag letters to buffer [NUL terminated] Buffer ALWAYS has leading space and trailing space space space NUL

// Copy next run of same-script non-tag letters to buffer [NUL terminated]
// Buffer ALWAYS has leading space and trailing space space space NUL
bool ScriptScanner::GetOneScriptSpan(LangSpan* span) {
  if (!letters_marks_only_) {
    // Return non-tag text, including punctuation and digits
    return GetOneTextSpan(span);
  }

  span->text = script_buffer_;
  span->text_bytes = 0;
  span->offset = next_byte_ - start_byte_;
  span->ulscript = UNKNOWN_ULSCRIPT;
  span->lang = UNKNOWN_LANGUAGE;
  span->truncated = false;

  // struct timeval script_start, script_mid, script_end;

  int put_soft_limit = kMaxScriptBytes - kWithinScriptTail;
  if ((kMaxScriptBytes <= byte_length_) &&
      (byte_length_ < (2 * kMaxScriptBytes))) {
    // Try to split the last two fragments in half
    put_soft_limit = byte_length_ / 2;
  }


  int spanscript;           // The script of this span
  int sc = UNKNOWN_ULSCRIPT;  // The script of next character
  int tlen = 0;
  int plen = 0;

  script_buffer_[0] = ' ';  // Always a space at front of output
  script_buffer_[1] = '\0';
  int take = 0;
  int put = 1;              // Start after the initial space

  // Build offsets from span->text back to start_byte_ + span->offset
  // This mapping reflects deletion of non-letters, expansion of
  // entities, etc.
  map2original_.Clear();
  map2original_.Delete(span->offset);   // So that MapBack(0) gives offset

  // Get to the first real non-tag letter or entity that is a letter
  int skip = SkipToFrontOfSpan(next_byte_, byte_length_, &spanscript);
  next_byte_ += skip;
  byte_length_ -= skip;

  if (skip != 1) {
    map2original_.Delete(skip);
    map2original_.Insert(1);
  } else {
    map2original_.Copy(1);
  }
  if (byte_length_ <= 0) {
    map2original_.Reset();
    return false;               // No more letters to be found
  }

  // There is at least one letter, so we know the script for this span
  span->ulscript = (ULScript)spanscript;