-
Notifications
You must be signed in to change notification settings - Fork 2
/
pysolr.py
335 lines (279 loc) · 10.5 KB
/
pysolr.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
# -*- coding: utf-8 -*-
"""
All we need to create a Solr connection is a url.
>>> #conn = Solr('http://127.0.0.1:8983/solr/')
>>> conn = Solr('http://127.0.0.1:8080/solr/default/')
First, completely clear the index.
>>> conn.delete(q='*:*')
For now, we can only index python dictionaries. Each key in the dictionary
will correspond to a field in Solr.
>>> docs = [
... {'id': 'testdoc.1', 'order_i': 1, 'name': 'document 1', 'text': u'Paul Verlaine'},
... {'id': 'testdoc.2', 'order_i': 2, 'name': 'document 2', 'text': u'Владимир Маякoвский'},
... {'id': 'testdoc.3', 'order_i': 3, 'name': 'document 3', 'text': u'test'},
... {'id': 'testdoc.4', 'order_i': 4, 'name': 'document 4', 'text': u'test'}
... ]
We can add documents to the index by passing a list of docs to the connection's
add method.
>>> conn.add(docs)
>>> results = conn.search('Verlaine')
>>> len(results)
1
>>> results = conn.search(u'Владимир')
>>> len(results)
1
Simple tests for searching. We can optionally sort the results using Solr's
sort syntax, that is, the field name and either asc or desc.
>>> results = conn.search('test', sort='order_i asc')
>>> for result in results:
... print result['name']
document 3
document 4
>>> results = conn.search('test', sort='order_i desc')
>>> for result in results:
... print result['name']
document 4
document 3
To update documents, we just use the add method.
>>> docs = [
... {'id': 'testdoc.4', 'order_i': 4, 'name': 'document 4', 'text': u'blah'}
... ]
>>> conn.add(docs)
>>> len(conn.search('blah'))
1
>>> len(conn.search('test'))
1
We can delete documents from the index by id, or by supplying a query.
>>> conn.delete(id='testdoc.1')
>>> conn.delete(q='name:"document 2"')
>>> results = conn.search('Verlaine')
>>> len(results)
0
Docs can also have multiple values for any particular key. This lets us use
Solr's multiValue fields.
>>> docs = [
... {'id': 'testdoc.5', 'cat': ['poetry', 'science'], 'name': 'document 5', 'text': u''},
... {'id': 'testdoc.6', 'cat': ['science-fiction',], 'name': 'document 6', 'text': u''},
... ]
>>> conn.add(docs)
>>> results = conn.search('cat:"poetry"')
>>> for result in results:
... print result['name']
document 5
>>> results = conn.search('cat:"science-fiction"')
>>> for result in results:
... print result['name']
document 6
>>> results = conn.search('cat:"science"')
>>> for result in results:
... print result['name']
document 5
"""
# TODO: unicode support is pretty sloppy. define it better.
from httplib import HTTPConnection
from urllib import urlencode
from urlparse import urlsplit
from datetime import datetime, date
from time import strptime
try:
# for python 2.5
from xml.etree import cElementTree as ET
except ImportError:
try:
# use cElementTree if available
import cElementTree as ET
except ImportError:
from elementtree import ElementTree as ET
__all__ = ['Solr']
class SolrError(Exception):
pass
class Results(object):
def __init__(self, docs, hits):
self.docs = docs
self.hits = hits
def __len__(self):
return len(self.docs)
def __iter__(self):
return iter(self.docs)
class Solr(object):
def __init__(self, url):
self.url = url
scheme, netloc, path, query, fragment = urlsplit(url)
netloc = netloc.split(':')
self.host = netloc[0]
if len(netloc) == 1:
self.host, self.port = netloc[0], None
else:
self.host, self.port = netloc
self.path = path.rstrip('/')
def _select(self, params):
# encode the query as utf-8 so urlencode can handle it
params['q'] = params['q'].encode('utf-8')
path = '%s/select/?%s' % (self.path, urlencode(params))
conn = HTTPConnection(self.host, self.port)
conn.request('GET', path)
return conn.getresponse()
def _update(self, message):
"""
Posts the given xml message to http://<host>:<port>/solr/update and
returns the result.
"""
path = '%s/update/' % self.path
conn = HTTPConnection(self.host, self.port)
conn.request('POST', path, message, {'Content-type': 'text/xml'})
return conn.getresponse()
def _extract_error(self, response):
"""
Extract the actual error message from a solr response. Unfortunately,
this means scraping the html.
"""
et = ET.parse(response)
return et.findtext('body/pre')
# Converters #############################################################
def _from_python(self, value):
"""
Converts python values to a form suitable for insertion into the xml
we send to solr.
"""
if isinstance(value, datetime):
value = value.strftime('%Y-%m-%dT%H:%M:%S.000Z')
elif isinstance(value, date):
value = value.strftime('%Y-%m-%dT00:00:00.000Z')
elif isinstance(value, bool):
if value:
value = 'true'
else:
value = 'false'
else:
value = unicode(value)
return value
def bool_to_python(self, value):
"""
Convert a 'bool' field from solr's xml format to python and return it.
"""
if value == 'true':
return True
elif value == 'false':
return False
def str_to_python(self, value):
"""
Convert an 'str' field from solr's xml format to python and return it.
"""
return unicode(value)
def int_to_python(self, value):
"""
Convert an 'int' field from solr's xml format to python and return it.
"""
return int(value)
def date_to_python(self, value):
"""
Convert a 'date' field from solr's xml format to python and return it.
"""
# this throws away fractions of a second
return datetime(*strptime(value[:-5], "%Y-%m-%dT%H:%M:%S")[0:6])
def float_to_python(self, value):
"""
Convert a 'float' field from solr's xml format to python and return it.
"""
return float(value)
def double_to_python(self, value):
"""
Convert a 'double' field from solr's xml format to python and return
it. Since Python does not have separate type for double, this is the
same as float.
"""
return self.float_to_python(value)
# API Methods ############################################################
def search(self, q, sort=None, start=None, rows=None):
"""Performs a search and returns the results."""
params = {'q': q}
if start:
params['start'] = start
if rows:
params['rows'] = rows
if sort:
params['sort'] = sort
response = self._select(params)
if response.status != 200:
raise SolrError(self._extract_error(response))
# TODO: make result retrieval lazy and allow custom result objects
# also, this has become rather ugly and definitely needs some cleanup.
et = ET.parse(response)
result = et.find('result')
hits = int(result.get('numFound'))
docs = result.findall('doc')
results = []
for doc in docs:
result = {}
for element in doc.getchildren():
if element.tag == 'arr':
result_val = []
for array_element in element.getchildren():
converter_name = '%s_to_python' % array_element.tag
converter = getattr(self, converter_name)
result_val.append(converter(array_element.text))
else:
converter_name = '%s_to_python' % element.tag
converter = getattr(self, converter_name)
result_val = converter(element.text)
result[element.get('name')] = result_val
results.append(result)
return Results(results, hits)
def add(self, docs, commit=True):
"""Adds or updates documents. For now, docs is a list of dictionaies
where each key is the field name and each value is the value to index.
"""
message = ET.Element('add')
for doc in docs:
d = ET.Element('doc')
for key, value in doc.items():
# handle lists, tuples, and other iterabes
if hasattr(value, '__iter__'):
for v in value:
f = ET.Element('field', name=key)
f.text = self._from_python(v)
d.append(f)
# handle strings and unicode
else:
f = ET.Element('field', name=key)
f.text = self._from_python(value)
d.append(f)
message.append(d)
m = ET.tostring(message)
response = self._update(m)
if response.status != 200:
raise SolrError(self._extract_error(response))
# TODO: Supposedly, we can put a <commit /> element in the same post body
# as the add element. That isn't working for some reason, and it would save us
# an extra trip to the server. This works for now.
if commit:
self.commit()
def delete(self, id=None, q=None, commit=True, fromPending=True, fromCommitted=True):
"""Deletes documents."""
if id is None and q is None:
raise ValueError('You must specify "id" or "q".')
elif id is not None and q is not None:
raise ValueError('You many only specify "id" OR "q", not both.')
elif id is not None:
m = '<delete><id>%s</id></delete>' % id
elif q is not None:
m = '<delete><query>%s</query></delete>' % q
response = self._update(m)
if response.status != 200:
raise SolrError(self._extract_error(response))
# TODO: Supposedly, we can put a <commit /> element in the same post body
# as the delete element. That isn't working for some reason, and it would save us
# an extra trip to the server. This works for now.
if commit:
self.commit()
def commit(self):
response = self._update('<commit />')
if response.status != 200:
raise SolrError(self._extract_error(response))
def optimize(self):
response = self._update('<optimize />')
if response.status != 200:
raise SolrError(self._extract_error(response))
if __name__ == "__main__":
import doctest
doctest.testmod()