from django.conf import settings

import lucene

lucene.initVM(lucene.CLASSPATH)

STORE = lucene.SimpleFSDirectory(lucene.File(settings.INDEX_PATH))

ANALYZER = lucene.PerFieldAnalyzerWrapper(lucene.StandardAnalyzer(lucene.Version.LUCENE_CURRENT))

ANALYZER.addAnalyzer("tags", lucene.FrenchAnalyzer(lucene.Version.LUCENE_CURRENT))

ANALYZER.addAnalyzer("title", lucene.FrenchAnalyzer(lucene.Version.LUCENE_CURRENT))

ANALYZER.addAnalyzer("abstract", lucene.FrenchAnalyzer(lucene.Version.LUCENE_CURRENT))

ANALYZER.addAnalyzer("all", lucene.FrenchAnalyzer(lucene.Version.LUCENE_CURRENT))

def get_results_with_context(field, query):

    res = get_results_list(field, query)

    searcher = get_searcher()

    query = get_query_parser(field).parse(query)

    contexts = [] 

    for i in res:

        doc = searcher.doc(i.doc)  

        ids = {"iri_id":doc.get("iri_id"), "ensemble_id":doc.get("ensemble_id"), "decoupage_id":doc.get("decoupage_id"), "element_id":doc.get("element_id"), "project_id":doc.get("project_id")}     

        score = i.score

        title = doc.getField('title').stringValue()

        desc = doc.getField('abstract').stringValue()

        tags = doc.getField('tags').stringValue()

        begin = doc.getField('begin').stringValue()

        duration = doc.getField('duration').stringValue()

        ids['context'] = desc

        ids['title'] = title

        ids['tags'] = tags

        ids['score'] = score

        ids['lucene_id'] = i.doc

        ids['begin'] = begin

        ids['duration'] = duration

        contexts.append(ids)     

    searcher.close()

    return contexts

def get_results_list(field, query):

    indexSearcher = get_searcher()

    queryParser = get_query_parser(field)

    queryObj = queryParser.parse(query)

    hits = indexSearcher.search(queryObj, settings.LDT_MAX_SEARCH_NUMBER)

    return hits.scoreDocs

def highlight_documents(results_list, query, field):

    searcher = get_searcher()

    analyzer = lucene.FrenchAnalyzer(lucene.Version.LUCENE_CURRENT)

    formatter = lucene.SimpleHTMLFormatter('<span class="highlight">', '</span>')

    query = get_query_parser(field).parse(query)

    highlighter = lucene.Highlighter(formatter, lucene.QueryScorer (query))

    for project in results_list:

        for segment in project['list']:

            lucene_doc = searcher.doc(segment.lucene_id)

            segment.context = get_highlighted_text(lucene_doc, analyzer, highlighter, 'abstract')

            tags = get_highlighted_text(lucene_doc, analyzer, highlighter, 'tags')

            segment.title = get_highlighted_text(lucene_doc, analyzer, highlighter, 'title')

            if segment.context == u'':

                segment.context = lucene_doc.getField('abstract').stringValue()

            if tags == u'':

                tags = lucene_doc.getField('tags').stringValue()

            if segment.title == u'':

                segment.title = lucene_doc.getField('title').stringValue() 

            segment.context_tags = tags[tags.find(';')+1:]          

    return results_list    

def get_highlighted_text(doc, analyzer, highlighter, field):

    res = doc.getField(field).stringValue()

    ts = analyzer.tokenStream("body", lucene.StringReader(res)) 

    res = highlighter.getBestFragments(ts, res, settings.LDT_MAX_FRAGMENT_PER_SEARCH, "...")

    return res    

def get_writer(new=False):

    lucene.getVMEnv().attachCurrentThread()

    return lucene.IndexWriter(STORE, ANALYZER, new, lucene.IndexWriter.MaxFieldLength.UNLIMITED)

def get_searcher():

    lucene.getVMEnv().attachCurrentThread()

    return lucene.IndexSearcher(STORE)

def get_query_parser(field):

    queryParser = lucene.QueryParser(lucene.Version.LUCENE_30, field, lucene.FrenchAnalyzer(lucene.Version.LUCENE_30))

    queryParser.setDefaultOperator(lucene.QueryParser.Operator.AND)

    return queryParser