LimaStatusCode SyntacticAnalysisXmlLogger::process(
AnalysisContent& analysis) const
{
TimeUtils::updateCurrentTime();
LinguisticMetaData* metadata=static_cast<LinguisticMetaData*>(analysis.getData("LinguisticMetaData"));
if (metadata == 0) {
SALOGINIT;
LERROR << "no LinguisticMetaData ! abort" << LENDL;
return MISSING_DATA;
}
DumperStream* dstream=initialize(analysis);
std::ostream& outputStream=dstream->out();
/*std::ofstream outputStream;
if (!openLogFile(outputStream,metadata->getMetaData("FileName")))
{
SALOGINIT;
LERROR << "Can't open log file " << LENDL;
return CANNOT_OPEN_FILE_ERROR;
}*/
SALOGINIT;
const SyntacticData* syntacticData=static_cast<const SyntacticData*>(analysis.getData("SyntacticData"));
if (syntacticData==0)
{
LERROR << "no SyntacticData ! abort" << LENDL;
return MISSING_DATA;
}
AnalysisGraph* anagraph=static_cast<AnalysisGraph*>(analysis.getData("PosGraph"));
if (anagraph==0)
{
LERROR << "no AnalysisGraph ! abort" << LENDL;
return MISSING_DATA;
}
SegmentationData* sb=static_cast<SegmentationData*>(analysis.getData("SentenceBoundaries"));
if (sb==0)
{
LERROR << "no SentenceBounds ! abort" << LENDL;
return MISSING_DATA;
}
// LinguisticGraph* graph=anagraph->getGraph();
std::set< std::pair<size_t, size_t> > alreadyDumped;
outputStream << "<?xml version='1.0' encoding='UTF-8'?>" << std::endl;
outputStream << "<syntactic_analysis_dump>" << std::endl;
// ??OME2 SegmentationData::iterator sbItr=sb->begin();
std::vector<Segment>::iterator sbItr=(sb->getSegments()).begin();
while (sbItr!=(sb->getSegments()).end())
{
LinguisticGraphVertex beginSentence=sbItr->getFirstVertex();
LinguisticGraphVertex endSentence=sbItr->getLastVertex();
dumpLimaData(outputStream,
beginSentence,
endSentence,
anagraph,
syntacticData);
sbItr++;
}
outputStream << "</syntactic_analysis_dump>" << std::endl;
delete dstream;
TimeUtils::logElapsedTime("SyntacticAnalysisXmlLogger");
return SUCCESS_ID;
}
LimaStatusCode EasyXmlDumper::process(AnalysisContent& analysis) const
{
TimeUtils::updateCurrentTime();
DUMPERLOGINIT;
LinguisticMetaData* metadata = static_cast<LinguisticMetaData*>(analysis.getData("LinguisticMetaData"));
if (metadata == 0) {
LERROR << "EasyXmlDumper::process no LinguisticMetaData ! abort";
return MISSING_DATA;
}
string filename = metadata->getMetaData("FileName");
LDEBUG << "EasyXmlDumper::process Filename: " << filename;
LDEBUG << "handler will be: " << m_handler;
// MediaId langid = static_cast<const Common::MediaticData::LanguageData&>(Common::MediaticData::MediaticData::single().mediaData(metadata->getMetaData("Lang"))).getMedia();
AnalysisHandlerContainer* h = static_cast<AnalysisHandlerContainer*>(analysis.getData("AnalysisHandlerContainer"));
AbstractTextualAnalysisHandler* handler = static_cast<AbstractTextualAnalysisHandler*>(h->getHandler(m_handler));
if (handler==0)
{
LERROR << "EasyXmlDumper::process: handler " << m_handler << " has not been given to the core client";
return MISSING_DATA;
}
AnalysisGraph* graph = static_cast<AnalysisGraph*>(analysis.getData(m_graph));
if (graph == 0)
{
graph = new AnalysisGraph(m_graph,m_language,true,true);
analysis.setData(m_graph,graph);
}
SyntacticData* syntacticData = static_cast<SyntacticData*>(analysis.getData("SyntacticData"));
if (syntacticData == 0)
{
syntacticData = new SyntacticAnalysis::SyntacticData(static_cast<AnalysisGraph*>(analysis.getData(m_graph)),0);
syntacticData->setupDependencyGraph();
analysis.setData("SyntacticData",syntacticData);
}
AnnotationData* annotationData = static_cast< AnnotationData* >(analysis.getData("AnnotationData"));
if (annotationData == 0)
{
annotationData = new AnnotationData();
if (static_cast<AnalysisGraph*>(analysis.getData("AnalysisGraph")) != 0)
{
static_cast<AnalysisGraph*>(analysis.getData("AnalysisGraph"))->populateAnnotationGraph(annotationData, "AnalysisGraph");
}
analysis.setData("AnnotationData",annotationData);
}
handler->startAnalysis();
HandlerStreamBuf hsb(handler);
std::ostream outputStream(&hsb);
LDEBUG << "EasyXmlDumper:: process before printing heading";
AnalysisGraph* anaGraph = static_cast<AnalysisGraph*>(analysis.getData("AnalysisGraph"));
AnalysisGraph* posGraph = static_cast<AnalysisGraph*>(analysis.getData("PosGraph"));
if (anaGraph != 0 && posGraph != 0)
{
LDEBUG << "EasyXmlDumper:: begin of posgraph";
std::vector< bool > alreadyDumpedTokens;
std::map< LinguisticAnalysisStructure::Token*, uint64_t > fullTokens;
LinguisticGraphVertexIt i, i_end;
uint64_t id = 0;
alreadyDumpedTokens.resize(num_vertices(*posGraph->getGraph()));
for (boost::tie(i, i_end) = vertices(*posGraph->getGraph()); i != i_end; ++i)
{
LDEBUG << "EasyXmlDumper:: examine posgraph for " << id;
alreadyDumpedTokens[id] = false;
fullTokens[get(vertex_token, *posGraph->getGraph(), *i)] = id;
id++;
}
/* No need for sentence boundaries in Easy input
LinguisticGraphVertex sentenceBegin = sb->getStartVertex();
SegmentationData::iterator sbItr = sb->begin();
LinguisticGraphVertex sentenceBegin = sb->getStartVertex();
SegmentationData::iterator sbItr = sb->begin();
*/
LinguisticGraphVertex sentenceBegin = posGraph->firstVertex();
LinguisticGraphVertex sentenceEnd = posGraph->lastVertex();
string sentIdPrefix;
try {
sentIdPrefix = metadata->getMetaData("docid");
LDEBUG << "EasyXmlDumper:: retrieve sentence id " << sentIdPrefix;
}catch (LinguisticProcessingException& ) {
sentIdPrefix = "";
}
if(sentIdPrefix.length() <= 0)
sentIdPrefix = "E";
/* No need for sentence boundaries in Easy input
while (sbItr != sb->end())
{
LinguisticGraphVertex sentenceEnd = *sbItr;
*/
LDEBUG << "EasyXmlDumper:: inside posgraph while ";
dumpLimaData(outputStream,
sentenceBegin,
sentenceEnd,
*anaGraph,
*posGraph,
*annotationData,
*syntacticData,
"PosGraph",
alreadyDumpedTokens,
fullTokens,
sentIdPrefix);
/* No need for sentence boundaries in Easy input
sentenceBegin = sentenceEnd;
sbItr++;
}
*/
LDEBUG << "EasyXmlDumper:: end of posgraph";
}
return SUCCESS_ID;
}
