def _update_doc_in_index(self, index_writer, doc):
last_mod = datetime.datetime.fromtimestamp(doc.last_mod)
docid = unicode(doc.docid)
txt = u""
for page in doc.pages:
for line in page.text:
txt += unicode(line) + u"\n"
for label in doc.labels:
txt += u" " + unicode(label.name)
txt = txt.strip()
txt = strip_accents(txt)
if txt == u"":
# make sure the text field is not empty. Whoosh doesn't like that
txt = u"empty"
labels = u",".join([strip_accents(unicode(label.name))
for label in doc.labels])
index_writer.update_document(
docid=docid,
doctype=doc.doctype,
content=txt,
label=labels,
last_read=last_mod
)
return True
def _update_doc_in_index(index_writer, doc):
"""
Add/Update a document in the index
"""
last_mod = datetime.datetime.fromtimestamp(doc.last_mod)
docid = unicode(doc.docid)
txt = u""
for page in doc.pages:
txt += u"\n".join([unicode(line) for line in page.text])
extra_txt = doc.extra_text
if extra_txt != u"":
txt += extra_txt + u"\n"
for label in doc.labels:
txt += u" " + unicode(label.name)
txt = txt.strip()
txt = strip_accents(txt)
if txt == u"":
# make sure the text field is not empty. Whoosh doesn't like that
txt = u"empty"
labels = u",".join(
[strip_accents(unicode(label.name)) for label in doc.labels])
index_writer.update_document(docid=docid,
doctype=doc.doctype,
content=txt,
label=labels,
date=doc.date,
last_read=last_mod)
return True
def find_suggestions(self, sentence):
"""
Search all possible suggestions. Suggestions returned always have at
least one document matching.
Arguments:
sentence --- keywords (single strings) for which we want
suggestions
Return:
An array of sets of keywords. Each set of keywords (-> one string)
is a suggestion.
"""
keywords = sentence.split(" ")
final_suggestions = []
corrector = self.__searcher.corrector("content")
for keyword_idx in range(0, len(keywords)):
keyword = strip_accents(keywords[keyword_idx])
if (len(keyword) <= MIN_KEYWORD_LEN):
continue
keyword_suggestions = corrector.suggest(keyword, limit=5)[:]
for keyword_suggestion in keyword_suggestions:
new_suggestion = keywords[:]
new_suggestion[keyword_idx] = keyword_suggestion
new_suggestion = u" ".join(new_suggestion)
if len(self.find_documents(new_suggestion)) <= 0:
continue
final_suggestions.append(new_suggestion)
final_suggestions.sort()
return final_suggestions
def find_suggestions(self, sentence):
"""
Search all possible suggestions. Suggestions returned always have at
least one document matching.
Arguments:
sentence --- keywords (single strings) for which we want
suggestions
Return:
An array of sets of keywords. Each set of keywords (-> one string)
is a suggestion.
"""
keywords = sentence.split(" ")
final_suggestions = []
corrector = self.__searcher.corrector("content")
for keyword_idx in range(0, len(keywords)):
keyword = strip_accents(keywords[keyword_idx])
if (len(keyword) <= MIN_KEYWORD_LEN):
continue
keyword_suggestions = corrector.suggest(keyword, limit=5)[:]
for keyword_suggestion in keyword_suggestions:
new_suggestion = keywords[:]
new_suggestion[keyword_idx] = keyword_suggestion
new_suggestion = u" ".join(new_suggestion)
if len(self.find_documents(new_suggestion)) <= 0:
continue
final_suggestions.append(new_suggestion)
final_suggestions.sort()
return final_suggestions
def find_documents(self, sentence):
"""
Returns all the documents matching the given keywords
Arguments:
keywords --- keywords (single string)
Returns:
An array of document id (strings)
"""
sentence = sentence.strip()
if sentence == u"":
return self.docs
sentence = strip_accents(sentence)
query = self.__qparser.parse(sentence)
return self.__find_documents(query)
def find_documents(self, sentence):
"""
Returns all the documents matching the given keywords
Arguments:
keywords --- keywords (single string)
Returns:
An array of document id (strings)
"""
sentence = sentence.strip()
if sentence == u"":
return self.docs
sentence = strip_accents(sentence)
query = self.__qparser.parse(sentence)
return self.__find_documents(query)
