size_t IsoffMainParser::parseSegmentBase(Node * segmentBaseNode, SegmentInformation *info)
{
SegmentBase *base;
if(!segmentBaseNode || !(base = new (std::nothrow) SegmentBase(info)))
return 0;
if(segmentBaseNode->hasAttribute("indexRange"))
{
size_t start = 0, end = 0;
if (std::sscanf(segmentBaseNode->getAttributeValue("indexRange").c_str(), "%zu-%zu", &start, &end) == 2)
{
IndexSegment *index = new (std::nothrow) DashIndexSegment(info);
if(index)
{
index->setByteRange(start, end);
base->indexSegment.Set(index);
/* index must be before data, so data starts at index end */
base->setByteRange(end + 1, 0);
}
}
}
parseInitSegment(DOMHelper::getFirstChildElementByName(segmentBaseNode, "Initialization"), base, info);
info->setSegmentBase(base);
return 1;
}
virtual bool cut(Unicode::const_iterator begin, Unicode::const_iterator end, vector<string>& res)const
{
if(begin == end)
{
return false;
}
vector<Unicode> uRes;
uRes.reserve(end - begin);
if (!cut(begin, end, uRes))
{
return false;
}
size_t offset = res.size();
res.resize(res.size() + uRes.size());
for(size_t i = 0; i < uRes.size(); i ++, offset++)
{
if(!TransCode::encode(uRes[i], res[offset]))
{
Rcout<<"encode failed."<<std::endl;
}
}
return true;
}
bool cut(Unicode::const_iterator begin, Unicode::const_iterator end, vector<string>& res) const
{
assert(_getInitFlag());
if (begin >= end)
{
LogError("begin >= end");
return false;
}
vector<Unicode> uRes;
if (!cut(begin, end, uRes))
{
LogError("get unicode cut result error.");
return false;
}
string tmp;
for (vector<Unicode>::const_iterator uItr = uRes.begin(); uItr != uRes.end(); uItr++)
{
if (TransCode::encode(*uItr, tmp))
{
res.push_back(tmp);
}
else
{
LogError("encode failed.");
}
}
return true;
}
virtual bool cut(Unicode::const_iterator begin, Unicode::const_iterator end, vector<string>& res)const
{
assert(_getInitFlag());
if(begin == end)
{
return false;
}
vector<Unicode> words;
words.reserve(end - begin);
if(!cut(begin, end, words))
{
return false;
}
size_t offset = res.size();
res.resize(res.size() + words.size());
for(size_t i = 0; i < words.size(); i++)
{
if(!TransCode::encode(words[i], res[i + offset]))
{
LogError("encode failed.");
res[i + offset].clear();
}
}
return true;
}
virtual bool cut(Unicode::const_iterator begin, Unicode::const_iterator end, vector<string>& res)const
{
assert(_getInitFlag());
if(begin == end)
{
return false;
}
vector<Unicode> words;
if(!cut(begin, end, words))
{
return false;
}
string word;
for(size_t i = 0; i < words.size(); i++)
{
if(TransCode::encode(words[i], word))
{
res.push_back(word);
}
else
{
LogError("encode failed.");
}
}
return true;
}
bool cut(const string& sentence,
vector<string>& words,
size_t max_word_len) const {
Unicode unicode;
if (!TransCode::decode(sentence, unicode)) {
return false;
}
vector<Unicode> unicodeWords;
cut(unicode.begin(), unicode.end(),
unicodeWords, max_word_len);
words.resize(unicodeWords.size());
for (size_t i = 0; i < words.size(); i++) {
TransCode::encode(unicodeWords[i], words[i]);
}
return true;
}
virtual bool cut(Unicode::const_iterator begin, Unicode::const_iterator end, vector<string>& res)const {
if(begin == end) {
return false;
}
vector<Unicode> words;
words.reserve(end - begin);
if(!cut(begin, end, words)) {
return false;
}
size_t offset = res.size();
res.resize(res.size() + words.size());
for(size_t i = 0; i < words.size(); i++) {
TransCode::encode(words[i], res[offset + i]);
}
return true;
}
virtual bool cut(Unicode::const_iterator begin, Unicode::const_iterator end, vector<string>& res)const
{
assert(_getInitFlag());
vector<TrieNodeInfo> segWordInfos;
if(!cut(begin, end, segWordInfos))
{
return false;
}
string tmp;
for(uint i = 0; i < segWordInfos.size(); i++)
{
if(TransCode::encode(segWordInfos[i].word, tmp))
{
res.push_back(tmp);
}
else
{
LogError("encode failed.");
}
}
return true;
}
