def __init__(self, spec):
if isinstance(spec, Spec):
self.spec = spec
else:
self.spec = Spec(spec)
self.discodex_client = DiscodexClient()
self.docset = Docset(self.spec.docset_name)
class FreequeryClient(object):
def __init__(self, spec):
if isinstance(spec, Spec):
self.spec = spec
else:
self.spec = Spec(spec)
self.discodex_client = DiscodexClient()
self.docset = Docset(self.spec.docset_name)
def query(self, q, ranked=True):
"""Return a ranked list of matching `Document` instances."""
qq = Query.parse(q)
res = self.discodex_client.query(self.spec.invindex_name, qq)
res = map(TfIdf.undemux, res)
if not res:
return []
pageranks = None
if ranked:
scoredb = ScoreDB(self.spec.scoredb_path)
uris = [e[0] for e in res]
pageranks = dict(scoredb.rank(uris))
if not pageranks:
raise Exception("no ranks available")
docs = []
for uri,scores in res:
doc = self.docset.get(uri)
doc.score = Score(**scores)
if pageranks:
doc.score['pagerank'] = pageranks[uri]
doc.excerpt = doc.excerpt(qq)
docs.append(doc)
return docs
def index(self, **kwargs):
if not self.docset.exists():
print "fq: cannot index `%s': no such docset" % self.spec.docset_name
exit(1)
job = IndexJob(self.spec, self.discodex_client, **kwargs)
job.start()
def linkparse(self, **kwargs):
job = LinkParseJob(self.spec, **kwargs)
job.start()
def rank(self, **kwargs):
job = PagerankJob(self.spec, **kwargs)
job.start()
def client(self):
from discodex.client import DiscodexClient
return DiscodexClient(self.options.host, self.options.port)
