def search(self, query=None, sort_by=None, start_offset=0,
max_offset=None, fields=None, **kwargs):
query = self.parse_query(query)
if xapian.Query.empty(query):
return 0, lambda: []
database = self.database
enquire = xapian.Enquire(database)
enquire.set_query(query)
if 'collapse_valueno' in self.schema:
enquire.set_collapse_key(self.schema['collapse_valueno'])
if sort_by:
sorter = xapian.MultiValueSorter()
for sort_field in sort_by:
if sort_field.startswith('-'):
reverse = True
sort_field = sort_field[1:] # Strip the '-'
else:
reverse = False # Reverse is inverted in Xapian -- http://trac.xapian.org/ticket/311
sorter.add(self._value_column(sort_field), reverse)
enquire.set_sort_by_key(sorter, True)
if not max_offset:
max_offset = database.get_doccount() - start_offset
matches = self._get_enquire_mset(database, enquire, start_offset, max_offset)
def result_generator():
for match in matches:
r = msgpack.unpackb(self._get_document_data(database, match.document))
if fields is not None: # 如果fields为[], 这情况下，不返回任何一项
item = {}
if isinstance(fields, list):
for field in fields:
if field == 'terms':
item['terms'] = dict([(term.term[5:], term.wdf) for term in match.document.termlist() if term.term.startswith('XTEXT')])
else:
item[field] = r.get(field)
else:
item = r
yield item
return self._get_hit_count(database, enquire), result_generator
def make_enq1(db):
enq = xapian.Enquire(db)
sorter = xapian.MultiValueSorter()
enq.set_sort_by_key(sorter, True)
del sorter
return enq
def make_enq3(db):
enq = xapian.Enquire(db)
sorter = xapian.MultiValueSorter()
enq.set_sort_by_relevance_then_key(sorter, True)
del sorter
return enq
def search(self,
query,
sort_by=None,
start_offset=0,
end_offset=None,
fields='',
highlight=False,
facets=None,
date_facets=None,
query_facets=None,
narrow_queries=None,
spelling_query=None,
limit_to_registered_models=True,
result_class=None,
**kwargs):
"""
Executes the Xapian::query as defined in `query`.
Required arguments:
`query` -- Search query to execute
Optional arguments:
`sort_by` -- Sort results by specified field (default = None)
`start_offset` -- Slice results from `start_offset` (default = 0)
`end_offset` -- Slice results at `end_offset` (default = None), if None, then all documents
`fields` -- Filter results on `fields` (default = '')
`highlight` -- Highlight terms in results (default = False)
`facets` -- Facet results on fields (default = None)
`date_facets` -- Facet results on date ranges (default = None)
`query_facets` -- Facet results on queries (default = None)
`narrow_queries` -- Narrow queries (default = None)
`spelling_query` -- An optional query to execute spelling suggestion on
`limit_to_registered_models` -- Limit returned results to models registered in the current `SearchSite` (default = True)
Returns:
A dictionary with the following keys:
`results` -- A list of `SearchResult`
`hits` -- The total available results
`facets` - A dictionary of facets with the following keys:
`fields` -- A list of field facets
`dates` -- A list of date facets
`queries` -- A list of query facets
If faceting was not used, the `facets` key will not be present
If `query` is None, returns no results.
If `INCLUDE_SPELLING` was enabled in the connection options, the
extra flag `FLAG_SPELLING_CORRECTION` will be passed to the query parser
and any suggestions for spell correction will be returned as well as
the results.
"""
if xapian.Query.empty(query):
return {
'results': [],
'hits': 0,
}
database = self._database()
if result_class is None:
result_class = SearchResult
if self.include_spelling is True:
spelling_suggestion = self._do_spelling_suggestion(
database, query, spelling_query)
else:
spelling_suggestion = ''
if narrow_queries is not None:
query = xapian.Query(
xapian.Query.OP_AND, query,
xapian.Query(xapian.Query.OP_AND, [
self.parse_query(narrow_query)
for narrow_query in narrow_queries
]))
if limit_to_registered_models:
registered_models = self.build_models_list()
if len(registered_models) > 0:
query = xapian.Query(
xapian.Query.OP_AND, query,
xapian.Query(xapian.Query.OP_OR, [
xapian.Query('%s%s' % (DOCUMENT_CT_TERM_PREFIX, model))
for model in registered_models
]))
enquire = xapian.Enquire(database)
if hasattr(settings, 'HAYSTACK_XAPIAN_WEIGHTING_SCHEME'):
enquire.set_weighting_scheme(
xapian.BM25Weight(*settings.HAYSTACK_XAPIAN_WEIGHTING_SCHEME))
enquire.set_query(query)
if sort_by:
sorter = xapian.MultiValueSorter()
for sort_field in sort_by:
if sort_field.startswith('-'):
reverse = True
sort_field = sort_field[1:] # Strip the '-'
else:
reverse = False # Reverse is inverted in Xapian -- http://trac.xapian.org/ticket/311
sorter.add(self._value_column(sort_field), reverse)
enquire.set_sort_by_key_then_relevance(sorter, True)
results = []
facets_dict = {
'fields': {},
'dates': {},
'queries': {},
}
if not end_offset:
end_offset = database.get_doccount() - start_offset
matches = self._get_enquire_mset(database, enquire, start_offset,
end_offset)
for match in matches:
app_label, module_name, pk, model_data = pickle.loads(
self._get_document_data(database, match.document))
if highlight:
model_data['highlighted'] = {
self.content_field_name:
self._do_highlight(model_data.get(self.content_field_name),
query)
}
results.append(
result_class(app_label, module_name, pk, match.percent,
**model_data))
if facets:
facets_dict['fields'] = self._do_field_facets(results, facets)
if date_facets:
facets_dict['dates'] = self._do_date_facets(results, date_facets)
if query_facets:
facets_dict['queries'] = self._do_query_facets(
results, query_facets)
return {
'results': results,
'hits': self._get_hit_count(database, enquire),
'facets': facets_dict,
'spelling_suggestion': spelling_suggestion,
}