public String processing2(String input, boolean consolidateHeader, boolean consolidateCitations)
throws Exception {
if (input == null) {
throw new GrobidResourceException("Cannot process pdf file, because input file was null.");
}
File inputFile = new File(input);
if (!inputFile.exists()) {
throw new GrobidResourceException(
"Cannot process pdf file, because input file '"
+ inputFile.getAbsolutePath()
+ "' does not exists.");
}
if (tmpPath == null) {
throw new GrobidResourceException("Cannot process pdf file, because temp path is null.");
}
if (!tmpPath.exists()) {
throw new GrobidResourceException(
"Cannot process pdf file, because temp path '"
+ tmpPath.getAbsolutePath()
+ "' does not exists.");
}
doc = new Document(input, tmpPath.getAbsolutePath());
try {
int startPage = -1;
int endPage = -1;
pathXML =
doc.pdf2xml(true, false, startPage, endPage, input, tmpPath.getAbsolutePath(), true);
// with timeout,
// no force pdf reloading
// input is the pdf absolute path, tmpPath is the temp. directory for the temp. lxml file,
// path is the resource path
// and we process images in the pdf file
if (pathXML == null) {
throw new Exception("PDF parsing fails");
}
doc.setPathXML(pathXML);
ArrayList<String> tokenizations = doc.addFeaturesDocument();
if (doc.getBlocks() == null) {
throw new Exception("PDF parsing resulted in empty content");
}
String fulltext = doc.getFulltextFeatured(true, true);
StringTokenizer st = new StringTokenizer(fulltext, "\n");
feedTaggerAndParse(st);
StringBuilder res = new StringBuilder();
for (int i = 0; i < tagger.size(); i++) {
for (int j = 0; j < tagger.xsize(); j++) {
res.append(tagger.x(i, j)).append("\t");
}
res.append(tagger.y2(i));
res.append("\n");
}
// buffer for the fulltext block
String rese = res.toString();
// set the different sections of the Document object
doc = BasicStructureBuilder.resultSegmentation(doc, rese, tokenizations);
// header processing
if (headerParser == null) {
headerParser = new HeaderParser();
}
resHeader = new BiblioItem();
headerParser.processingHeaderBlock(consolidateHeader, doc, resHeader);
// the language identification is normally done during the header parsing, but only
// based on header information.
// LanguageUtilities languageUtilities = LanguageUtilities.getInstance();
Language langu = languageUtilities.runLanguageId(resHeader.getTitle() + "\n" + doc.getBody());
if (langu != null) {
String lang = langu.getLangId();
doc.setLanguage(lang);
resHeader.setLanguage(lang);
}
// citation processing
if (citationParser == null) {
citationParser = new CitationParser();
}
ArrayList<BibDataSet> resCitations;
// ArrayList<String> tokenizationsRef = doc.getTokenizationsReferences();
// System.out.println(tokenizationsRef.toString());
// resCitations = BasicStructureBuilder.giveReferenceSegments(doc);
resCitations = doc.bibDataSets;
if (resCitations != null) {
for (BibDataSet bds : resCitations) {
String marker = bds.getRefSymbol();
if (marker != null) {
marker = marker.replace(".", "");
marker = marker.replace(" ", "");
bds.setRefSymbol(marker);
}
BiblioItem bib = citationParser.processing(bds.getRawBib(), consolidateCitations);
bds.setResBib(bib);
}
}
// final combination
return toTEI(doc, rese, tokenizations, resHeader, false, null, false);
} catch (Exception e) {
throw new GrobidException("An exception occured while running Grobid.", e);
} finally {
// keep it clean when leaving...
doc.cleanLxmlFile(pathXML, false);
}
}
