gistfile1.py

from struct import pack, unpack, calcsize
import cStringIO

def sphinx_unpack(fmt, fp):
    return unpack(fmt, fp.read(calcsize(fmt)))

def parse_query_result(data, query_count):
    data = cStringIO.StringIO(data)
    result = []

    for i in xrange(0, query_count):
        res = {'error': None,
               'warning': None,
               'fields': [],
               'attrs': [],
               'matches': [],
               'total': 0,
               'total_found': 0,
               'time': None,
               'words': [], }
        result.append(res)
        status = sphinx_unpack('>L', data)[0]
        if status != SEARCHD_OK:
            length = sphinx_unpack('>L', data)[0]
            message = data.read(length)
            if status == SEARCHD_WARNING:
                res['warning'] = message
            else:
                res['error'] = message
                continue

        fields_count = sphinx_unpack('>L', data)[0]
        while  fields_count and data:
            fields_count -= 1
            length = sphinx_unpack('>L',  data)[0]
            field = data.read(length)
            res['fields'].append(field)

        attrs_count = sphinx_unpack('>L', data)[0]
        while attrs_count and data:
            attrs_count -= 1
            length = sphinx_unpack('>L', data)[0]
            attr, data = data.read(length)
            attr_type = sphinx_unpack('>L', data)[0]
            res['attrs'].append([attr, attr_type])

        count = sphinx_unpack('>L', data)[0]
        id64 = sphinx_unpack('>L', data)[0]

        while count and data:
            count -= 1
            if id64:
                doc, weight = sphinx_unpack('>QL', data)
            else:
                doc, weight = sphinx_unpack('>2L', data)

            match = {'id': doc, 'weight': weight, 'attrs': {}}
            for i, attr in enumerate(res['attrs']):
                if  attr[1] == SPH_ATTR_FLOAT:
                    match['attrs'][attr[0]] = sphinx_unpack('>f', data)[0]
                elif attr[1] == SPH_ATTR_BIGINT:
                    match['attrs'][attr[0]] = sphinx_unpack('>q', data)[0]
                elif attr[1] == SPH_ATTR_STRING:
                    slen = sphinx_unpack('>L', data)[0]
                    match['attrs'][attr[0]] = data.read(slen) if slen else ''
                elif attr[1] == (SPH_ATTR_MULTI | SPH_ATTR_INTEGER):
                    values_count = sphinx_unpack('>L', data)[0]
                    vals = []
                    for n in xrange(0, values_count):
                        vals.append(sphinx_unpack('>L', data)[0])
                    match['attrs'][attr[0]] = vals
                else:
                    match['attrs'][attr[0]] = sphinx_unpack('>L', data)[0]
            res['matches'].append(match)

        (res['total'], res['total_found'],
         res['time'], words) = sphinx_unpack('>4L', data)
        while words:
            words -= 1
            length = sphinx_unpack('>L', data)[0]
            word = data.read(length)
            docs, hits = sphinx_unpack('>2L', data)
            res['words'].append(dict(word=word, docs=docs, hits=hits))
    return result