void WebVTTParser::parseBytes(const char* data, unsigned length)
{
// 4.8.10.13.3 WHATWG WebVTT Parser algorithm.
// 1-3 - Initial setup.
unsigned position = 0;
while (position < length) {
String line = collectNextLine(data, length, &position);
switch (m_state) {
case Initial:
// Buffer up at least 9 bytes before proceeding with checking for the file identifier.
m_identifierData.append(data, length);
if (m_identifierData.size() < bomLength + fileIdentifierLength)
return;
// 4-12 - Collect the first line and check for "WEBVTT".
if (!hasRequiredFileIdentifier()) {
if (m_client)
m_client->fileFailedToParse();
return;
}
m_state = Header;
m_identifierData.clear();
break;
case Header:
// 13-18 - Allow a header (comment area) under the WEBVTT line.
if (line.isEmpty())
m_state = Id;
break;
case Id:
// 19-29 - Allow any number of line terminators, then initialize new cue values.
if (line.isEmpty())
break;
resetCueValues();
// 30-39 - Check if this line contains an optional identifier or timing data.
m_state = collectCueId(line);
break;
case TimingsAndSettings:
// 40 - Collect cue timings and settings.
m_state = collectTimingsAndSettings(line);
break;
case CueText:
// 41-53 - Collect the cue text, create a cue, and add it to the output.
m_state = collectCueText(line, length, position);
break;
case BadCue:
// 54-62 - Collect and discard the remaining cue.
m_state = ignoreBadCue(line);
break;
}
}
}
bool WebVTTParser::hasRequiredFileIdentifier()
{
// A WebVTT file identifier consists of an optional BOM character,
// the string "WEBVTT" followed by an optional space or tab character,
// and any number of characters that are not line terminators ...
unsigned position = 0;
if (m_identifierData.size() >= bomLength && m_identifierData[0] == '\xEF' && m_identifierData[1] == '\xBB' && m_identifierData[2] == '\xBF')
position += bomLength;
String line = collectNextLine(m_identifierData.data(), m_identifierData.size(), &position);
if (line.length() < fileIdentifierLength)
return false;
if (line.substring(0, fileIdentifierLength) != "WEBVTT")
return false;
if (line.length() > fileIdentifierLength && line[fileIdentifierLength] != ' ' && line[fileIdentifierLength] != '\t')
return false;
return true;
}
bool WebVTTParser::hasRequiredFileIdentifier(const char* data, unsigned length)
{
// A WebVTT file identifier consists of an optional BOM character,
// the string "WEBVTT" followed by an optional space or tab character,
// and any number of characters that are not line terminators ...
unsigned position = 0;
if (length >= bomLength && data[0] == '\xEF' && data[1] == '\xBB' && data[2] == '\xBF')
position += bomLength;
String line = collectNextLine(data, length, &position);
if (line.length() < fileIdentiferLength)
return false;
if (line.length() == fileIdentiferLength && line != "WEBVTT")
return false;
if (!hasLongWebVTTIdentifier(line))
return false;
return true;
}
