static void
RemoveControlCharactersFrom(nsAString& aStr, TextRangeArray* aRanges)
{
size_t firstControlCharOffset = FindFirstControlCharacter(aStr);
if (firstControlCharOffset == (size_t)-1) {
return;
}
nsAutoString copy(aStr);
const char16_t* sourceBegin = copy.BeginReading();
const char16_t* sourceEnd = copy.EndReading();
char16_t* dest = aStr.BeginWriting();
if (NS_WARN_IF(!dest)) {
return;
}
char16_t* curDest = dest + firstControlCharOffset;
size_t i = firstControlCharOffset;
for (const char16_t* source = sourceBegin + firstControlCharOffset;
source < sourceEnd; ++source) {
if (*source == '\t' || *source == '\n' || !IsControlChar(*source)) {
*curDest = *source;
++curDest;
++i;
} else if (aRanges) {
aRanges->RemoveCharacter(i);
}
}
aStr.SetLength(curDest - dest);
}
void CPDF_TextPage::AddCharInfoByRLDirection(wchar_t wChar,
const PAGECHAR_INFO& info) {
PAGECHAR_INFO info2 = info;
if (IsControlChar(info2)) {
info2.m_Index = -1;
m_CharList.push_back(info2);
return;
}
info2.m_Index = m_TextBuf.GetLength();
wChar = FX_GetMirrorChar(wChar);
size_t nCount = Unicode_GetNormalization(wChar, nullptr);
if (nCount == 0) {
info2.m_Unicode = wChar;
m_TextBuf.AppendChar(info2.m_Unicode);
m_CharList.push_back(info2);
return;
}
std::unique_ptr<wchar_t, FxFreeDeleter> pDst(FX_Alloc(wchar_t, nCount));
Unicode_GetNormalization(wChar, pDst.get());
for (size_t nIndex = 0; nIndex < nCount; ++nIndex) {
info2.m_Unicode = pDst.get()[nIndex];
info2.m_Flag = FPDFTEXT_CHAR_PIECE;
m_TextBuf.AppendChar(info2.m_Unicode);
m_CharList.push_back(info2);
}
}
static size_t
FindFirstControlCharacter(const nsAString& aStr)
{
const char16_t* sourceBegin = aStr.BeginReading();
const char16_t* sourceEnd = aStr.EndReading();
for (const char16_t* source = sourceBegin; source < sourceEnd; ++source) {
if (*source != '\t' && IsControlChar(*source)) {
return source - sourceBegin;
}
}
return -1;
}
int Sc_strlen(const char *str, int maxlen1, int maxlen2, bool accept_color_chars, bool accept_control_chars)
{
int a = maxlen1 < maxlen2 ? maxlen1 : maxlen2;
int len = 0;
while (a-- > 0)
{
len++;
if (*str == 0)
return len;
if (!accept_control_chars && IsColorChar(*str))
return 0;
if (!accept_control_chars && IsControlChar(*str))
return 0;
str++;
}
return 0;
}
void CPDF_TextPage::ParseTextPage() {
m_bIsParsed = false;
m_TextBuf.Clear();
m_CharList.clear();
m_pPreTextObj = nullptr;
ProcessObject();
m_bIsParsed = true;
m_CharIndex.clear();
const int nCount = CountChars();
if (nCount)
m_CharIndex.push_back(0);
for (int i = 0; i < nCount; ++i) {
int indexSize = pdfium::CollectionSize<int>(m_CharIndex);
const PAGECHAR_INFO& charinfo = m_CharList[i];
if (charinfo.m_Flag == FPDFTEXT_CHAR_GENERATED ||
(charinfo.m_Unicode != 0 && !IsControlChar(charinfo)) ||
(charinfo.m_Unicode == 0 && charinfo.m_CharCode != 0)) {
if (indexSize % 2) {
m_CharIndex.push_back(1);
} else {
if (indexSize <= 0)
continue;
m_CharIndex[indexSize - 1] += 1;
}
} else {
if (indexSize % 2) {
if (indexSize <= 0)
continue;
m_CharIndex[indexSize - 1] = i + 1;
} else {
m_CharIndex.push_back(i + 1);
}
}
}
int indexSize = pdfium::CollectionSize<int>(m_CharIndex);
if (indexSize % 2)
m_CharIndex.erase(m_CharIndex.begin() + indexSize - 1);
}
