def testScore(self):
reactor = CallTrace('reactor')
settings = LuceneSettings(commitCount=1, similarity=TermFrequencySimilarity(), verbose=False)
lucene = Lucene(join(self.tempdir, 'lucene'), reactor=reactor, settings=settings)
document = Document()
document.add(TextField('field', 'x '*100, Field.Store.NO))
returnValueFromGenerator(lucene.addDocument(identifier="identifier", document=document))
q = TermQuery(Term("field", 'x'))
result = returnValueFromGenerator(lucene.executeQuery(q))
self.assertAlmostEqual(0.1, result.hits[0].score)
q.setBoost(10.0)
result = returnValueFromGenerator(lucene.executeQuery(q))
self.assertAlmostEqual(1, result.hits[0].score)
class DeDupFilterCollectorTest(SeecrTestCase):
def setUp(self):
super(DeDupFilterCollectorTest, self).setUp()
self._reactor = CallTrace('reactor')
settings = LuceneSettings(commitCount=1, verbose=False)
self.lucene = Lucene(self.tempdir,
reactor=self._reactor,
settings=settings)
def tearDown(self):
self.lucene.close()
super(DeDupFilterCollectorTest, self).tearDown()
def testCollectorTransparentlyDelegatesToNextCollector(self):
self._addDocument("urn:1", 2)
tc = TopScoreDocSuperCollector(100, True)
c = DeDupFilterSuperCollector("__isformatof__", "__sort__", tc)
self.lucene.search(query=MatchAllDocsQuery(), collector=c)
self.assertEquals(1, tc.topDocs(0).totalHits)
def _addDocument(self, identifier, isformatof, sort=None):
doc = Document()
if isformatof:
doc.add(NumericDocValuesField("__isformatof__", long(isformatof)))
if sort:
doc.add(NumericDocValuesField("__sort__", long(sort)))
consume(self.lucene.addDocument(identifier, doc))
self.lucene.commit() # Explicitly, not required: since commitCount=1.
def testCollectorFiltersTwoSimilar(self):
self._addDocument("urn:1", 2, 1)
self._addDocument("urn:2", 2, 2)
tc = TopScoreDocSuperCollector(100, True)
c = DeDupFilterSuperCollector("__isformatof__", "__sort__", tc)
self.lucene.search(query=MatchAllDocsQuery(), collector=c)
topDocsResult = tc.topDocs(0)
self.assertEquals(1, topDocsResult.totalHits)
self.assertEquals(1, len(topDocsResult.scoreDocs))
docId = topDocsResult.scoreDocs[0].doc
key = c.keyForDocId(docId)
identifier = self.lucene._index.getDocument(
key.getDocId()).get(IDFIELD)
self.assertEquals('urn:2', identifier)
self.assertEquals(2, key.count)
def testCollectorFiltersTwoTimesTwoSimilarOneNot(self):
self._addDocument("urn:1", 1, 2001)
self._addDocument("urn:2", 3, 2009) # result 2x
self._addDocument("urn:3", 50, 2010) # result 1x
self._addDocument("urn:4", 3, 2001)
self._addDocument("urn:5", 1, 2009) # result 2x
#expected: "urn:2', "urn:3" and "urn:5" in no particular order
tc = TopScoreDocSuperCollector(100, True)
c = DeDupFilterSuperCollector("__isformatof__", "__sort__", tc)
self.lucene.search(query=MatchAllDocsQuery(), collector=c)
topDocsResult = tc.topDocs(0)
self.assertEquals(3, topDocsResult.totalHits)
self.assertEquals(3, len(topDocsResult.scoreDocs))
rawDocIds = [scoreDoc.doc for scoreDoc in topDocsResult.scoreDocs]
netDocIds = [c.keyForDocId(rawDocId).docId for rawDocId in rawDocIds]
identifiers = set(
self.lucene._index.getDocument(doc).get(IDFIELD)
for doc in netDocIds)
self.assertEquals(set(["urn:2", "urn:3", "urn:5"]), identifiers)
self.assertEquals(
[1, 2, 2], list(sorted(c.keyForDocId(d).count for d in netDocIds)))
def testSilentyYieldsWrongResultWhenFieldNameDoesNotMatch(self):
self._addDocument("urn:1", 2)
tc = TopScoreDocSuperCollector(100, True)
c = DeDupFilterSuperCollector("__wrong_field__", "__sort__", tc)
self.lucene.search(query=MatchAllDocsQuery(), collector=c)
self.assertEquals(1, tc.topDocs(0).totalHits)
def testShouldAddResultsWithoutIsFormatOf(self):
self._addDocument("urn:1", 2)
self._addDocument("urn:2", None)
self._addDocument("urn:3", 2)
self._addDocument("urn:4", None)
self._addDocument("urn:5", None)
self._addDocument("urn:6", None)
self._addDocument("urn:7", None)
self._addDocument("urn:8", None)
self._addDocument("urn:9", None)
self._addDocument("urn:A", None)
self._addDocument("urn:B", None) # trigger a merge
tc = TopScoreDocSuperCollector(100, True)
c = DeDupFilterSuperCollector("__isformatof__", "__sort__", tc)
self.lucene.search(query=MatchAllDocsQuery(), collector=c)
self.assertEquals(10, tc.topDocs(0).totalHits)
class LuceneTest(SeecrTestCase):
def setUp(self):
SeecrTestCase.setUp(self)
self.setUpLucene()
def setUpLucene(self, **kwargs):
self._lucene = Lucene(host="localhost", port=1234, name='lucene', settings=LuceneSettings(), **kwargs)
self.post = []
self.response = ""
connect = self._lucene._connect()
def mockPost(data, path, **kwargs):
self.post.append(dict(data=data, path=path))
raise StopIteration(self.response)
yield
connect._post = mockPost
self.read = []
self.response = ""
def mockRead(path, **kwargs):
self.read.append(path)
raise StopIteration(self.response)
yield
connect.read = mockRead
self._lucene._connect = lambda: connect
def testPostSettingsAddObserverInit(self):
self.assertEqual([], self.post)
self._lucene.observer_init()
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/settings/', self.post[0]['path'])
self.assertEquals(DEFAULTS, loads(self.post[0]['data']))
def testInitialize(self):
self.assertEqual([], self.post)
consume(self._lucene.initialize())
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/settings/', self.post[0]['path'])
self.assertEquals(DEFAULTS, loads(self.post[0]['data']))
def testAdd(self):
registry = FieldRegistry()
fields = [registry.createField("id", "id1")]
consume(self._lucene.addDocument(identifier='id1', fields=fields))
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/update/?identifier=id1', self.post[0]['path'])
self.assertEqual('[{"type": "TextField", "name": "id", "value": "id1"}]', self.post[0]['data'])
def testAddWithoutIdentifier(self):
registry = FieldRegistry()
fields = [registry.createField("id", "id1")]
consume(self._lucene.addDocument(fields=fields))
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/update/?', self.post[0]['path'])
self.assertEqual('[{"type": "TextField", "name": "id", "value": "id1"}]', self.post[0]['data'])
def testDelete(self):
consume(self._lucene.delete(identifier='id1'))
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/delete/?identifier=id1', self.post[0]['path'])
self.assertEqual(None, self.post[0]['data'])
def testDeleteByQuery(self):
query = QueryExpressionToLuceneQueryDict([], LuceneSettings()).convert(cqlToExpression("field=value"))
consume(self._lucene.delete(luceneQuery=query))
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/delete/', self.post[0]['path'])
self.assertEqual('{"query": {"term": {"field": "field", "value": "value"}, "type": "TermQuery"}}', self.post[0]['data'])
def testExecuteQuery(self):
self.response = JsonDict({
"total": 887,
"queryTime": 6,
"times": {"searchTime": 3},
"hits": [{
"id": "record:1", "score": 0.1234,
"duplicateCount": {"__key__": 2},
"duplicates": {"__grouping_key__": [{"id": 'record:1'}, {"id": 'record:2'}]}
}],
"drilldownData": [
{"fieldname": "facet", "path": [], "terms": [{"term": "term", "count": 1}]}
],
"suggestions": {
"valeu": ["value"]
}
}).dumps()
query = QueryExpressionToLuceneQueryDict([], LuceneSettings()).convert(cqlToExpression("field=value"))
response = retval(self._lucene.executeQuery(
luceneQuery=query, start=1, stop=5,
facets=[dict(maxTerms=10, fieldname='facet')],
sortKeys=[dict(sortBy='field', sortDescending=False)],
suggestionRequest=dict(suggests=['valeu'], count=2, field='field1'),
dedupField="__key__",
clustering=True,
storedFields=["field"]
))
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/query/', self.post[0]['path'])
self.assertEqual({
"start": 1, "stop": 5,
"storedFields": ["field"],
"query": {"term": {"field": "field", "value": "value"}, "type": "TermQuery"},
"facets": [{"fieldname": "facet", "maxTerms": 10}],
"sortKeys": [{"sortBy": "field", "sortDescending": False, "type": "String", 'missingValue': 'STRING_LAST'}],
"suggestionRequest": dict(suggests=['valeu'], count=2, field='field1'),
"dedupField": "__key__",
"dedupSortField": None,
"clustering": True,
}, loads(self.post[0]['data']))
self.assertEqual(887, response.total)
self.assertEqual(6, response.queryTime)
self.assertEqual({'searchTime': 3}, response.times)
self.assertEqual(1, len(response.hits))
self.assertEqual("record:1", response.hits[0].id)
self.assertEqual(0.1234, response.hits[0].score)
self.assertEqual(dict(__key__=2), response.hits[0].duplicateCount)
self.assertEqual([
{"fieldname": "facet", "path": [], "terms": [{"term": "term", "count": 1}]}
], response.drilldownData)
self.assertEqual({'valeu': ['value']}, response.suggestions)
def testPrefixSearch(self):
self.response = JsonList([["value0", 1], ["value1", 2]]).dumps()
response = retval(self._lucene.prefixSearch(fieldname='field1', prefix='valu'))
self.assertEquals(['value1', 'value0'], response.hits)
response = retval(self._lucene.prefixSearch(fieldname='field1', prefix='valu', showCount=True))
self.assertEquals([('value1', 2), ('value0', 1)], response.hits)
def testNumDocs(self):
self.response = "150"
result = retval(self._lucene.numDocs())
self.assertEqual(150, result)
self.assertEqual([{'data': None, 'path': '/lucene/numDocs/'}], self.post)
def testFieldnames(self):
self.response = '["field1", "field2"]'
result = retval(self._lucene.fieldnames())
self.assertEqual(["field1", "field2"], result.hits)
self.assertEqual([{"data": None, "path": "/lucene/fieldnames/"}], self.post)
def testDrilldownFieldnames(self):
self.response = '["field1", "field2"]'
result = retval(self._lucene.drilldownFieldnames())
self.assertEqual(["field1", "field2"], result.hits)
self.assertEqual([{"data": None, "path": "/lucene/drilldownFieldnames/?limit=50"}], self.post)
result = retval(self._lucene.drilldownFieldnames(limit=1, path=['field']))
self.assertEqual(["field1", "field2"], result.hits)
self.assertEqual({"data": None, "path": "/lucene/drilldownFieldnames/?dim=field&limit=1"}, self.post[-1])
result = retval(self._lucene.drilldownFieldnames(limit=1, path=['xyz', 'abc', 'field']))
self.assertEqual(["field1", "field2"], result.hits)
self.assertEqual({"data": None, "path": "/lucene/drilldownFieldnames/?dim=xyz&limit=1&path=abc&path=field"}, self.post[-1])
def testUpdateSettings(self):
self.response = JsonDict(numberOfConcurrentTasks=6, similarity="BM25(k1=1.2,b=0.75)", clustering=JsonDict(clusterMoreRecords=100, clusteringEps=0.4, clusteringMinPoints=1))
settings = retval(self._lucene.getSettings())
self.assertEqual(['/settings/'], self.read)
self.assertEquals({'numberOfConcurrentTasks': 6, 'similarity': u'BM25(k1=1.2,b=0.75)', 'clustering': {'clusterMoreRecords': 100, 'clusteringEps': 0.4, 'clusteringMinPoints': 1}}, settings)
clusterFields = [
{"filterValue": None, "fieldname": "untokenized.dcterms:isFormatOf.uri", "weight": 0}
]
self.response = ""
consume(self._lucene.setSettings(similarity=dict(name="bm25", k1=1.0, b=2.0), numberOfConcurrentTasks=10, clustering=dict(clusterMoreRecords=200, clusteringEps=1.0, clusteringMinPoints=2, fields=clusterFields)))
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/settings/', self.post[0]['path'])
self.assertEqual({
"numberOfConcurrentTasks": 10,
"similarity": dict(type="BM25Similarity", k1=1.0, b=2.0),
"clustering": {
"clusterMoreRecords": 200,
"clusteringEps": 1.0,
"clusteringMinPoints": 2,
"fields": [
{"filterValue": None, "fieldname": "untokenized.dcterms:isFormatOf.uri", "weight": 0}
]
}
}, loads(self.post[0]['data']))
consume(self._lucene.setSettings(numberOfConcurrentTasks=5, similarity=None, clustering=None))
self.assertEqual(2, len(self.post))
self.assertEqual('/lucene/settings/', self.post[1]['path'])
self.assertEqual({
"numberOfConcurrentTasks": 5,
}, loads(self.post[1]['data']))
def testSimilarDocs(self):
self.response = JsonDict({
"total": 887,
"queryTime": 6,
"times": {"searchTime": 3},
"hits": [
{"id": "record:1", "score": 0.1234},
{"id": "record:2", "score": 0.1234},
],
}).dumps()
response = retval(self._lucene.similarDocuments(identifier='record:3'))
self.assertEqual(887, response.total)
self.assertEqual(2, len(response.hits))
def testLuceneReadonly(self):
self.setUpLucene(readonly=True)
self._lucene.observer_init()
self.assertEqual([], self.post)
self.assertRaises(RuntimeError, lambda: consume(self._lucene.setSettings()))
self.assertRaises(RuntimeError, lambda: consume(self._lucene.addDocument(fields=[])))
self.assertRaises(RuntimeError, lambda: consume(self._lucene.delete('identifier')))
def testLuceneServerHostPortDynamic(self):
lucene = Lucene(name='lucene', settings=LuceneSettings(), readonly=True)
def httprequest1_1Mock(**kwargs):
raise StopIteration(parseResponse(HTTP_RESPONSE))
yield
observer = CallTrace(
'observer',
returnValues=dict(luceneServer=('example.org', 1234)),
methods=dict(httprequest1_1=httprequest1_1Mock))
lucene.addObserver(observer)
query = QueryExpressionToLuceneQueryDict([], LuceneSettings()).convert(cqlToExpression("field=value"))
response = retval(lucene.executeQuery(
luceneQuery=query, start=1, stop=5,
))
self.assertEquals(887, response.total)
self.assertEquals(['luceneServer', 'httprequest1_1'], observer.calledMethodNames())
class LuceneTest(SeecrTestCase):
def setUp(self):
SeecrTestCase.setUp(self)
self._lucene = Lucene(host="localhost", port=1234, name='lucene', settings=LuceneSettings())
self.post = []
self.response = ""
def mockPost(data, path, **kwargs):
self.post.append(dict(data=data, path=path))
raise StopIteration(self.response)
yield
self._lucene._connect._post = mockPost
self.read = []
self.response = ""
def mockRead(path, **kwargs):
self.read.append(path)
raise StopIteration(self.response)
yield
self._lucene._connect.read = mockRead
def testPostSettingsAddObserverInit(self):
self.assertEqual([], self.post)
self._lucene.observer_init()
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/settings/', self.post[0]['path'])
self.assertEquals(DEFAULTS, loads(self.post[0]['data']))
def testInitialize(self):
self.assertEqual([], self.post)
consume(self._lucene.initialize())
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/settings/', self.post[0]['path'])
self.assertEquals(DEFAULTS, loads(self.post[0]['data']))
def testAdd(self):
registry = FieldRegistry()
fields = [registry.createField("id", "id1")]
consume(self._lucene.addDocument(identifier='id1', fields=fields))
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/update/?identifier=id1', self.post[0]['path'])
self.assertEqual('[{"type": "TextField", "name": "id", "value": "id1"}]', self.post[0]['data'])
def testAddWithoutIdentifier(self):
registry = FieldRegistry()
fields = [registry.createField("id", "id1")]
consume(self._lucene.addDocument(fields=fields))
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/update/?', self.post[0]['path'])
self.assertEqual('[{"type": "TextField", "name": "id", "value": "id1"}]', self.post[0]['data'])
def testDelete(self):
consume(self._lucene.delete(identifier='id1'))
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/delete/?identifier=id1', self.post[0]['path'])
self.assertEqual(None, self.post[0]['data'])
def testExecuteQuery(self):
self.response = JsonDict({
"total": 887,
"queryTime": 6,
"times": {"searchTime": 3},
"hits": [{
"id": "record:1", "score": 0.1234,
"duplicateCount": {"__key__": 2},
"duplicates": {"__grouping_key__": [{"id": 'record:1'}, {"id": 'record:2'}]}
}],
"drilldownData": [
{"fieldname": "facet", "path": [], "terms": [{"term": "term", "count": 1}]}
],
"suggestions": {
"valeu": ["value"]
}
}).dumps()
query = QueryExpressionToLuceneQueryDict([], LuceneSettings()).convert(cqlToExpression("field=value"))
response = retval(self._lucene.executeQuery(
luceneQuery=query, start=1, stop=5,
facets=[dict(maxTerms=10, fieldname='facet')],
sortKeys=[dict(sortBy='field', sortDescending=False)],
suggestionRequest=dict(suggests=['valeu'], count=2, field='field1'),
dedupField="__key__",
clustering=True,
storedFields=["field"]
))
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/query/', self.post[0]['path'])
self.assertEqual({
"start": 1, "stop": 5,
"storedFields": ["field"],
"query": {"term": {"field": "field", "value": "value"}, "type": "TermQuery"},
"facets": [{"fieldname": "facet", "maxTerms": 10}],
"sortKeys": [{"sortBy": "field", "sortDescending": False, "type": "String", 'missingValue': 'STRING_LAST'}],
"suggestionRequest": dict(suggests=['valeu'], count=2, field='field1'),
"dedupField": "__key__",
"dedupSortField": None,
"clustering": True,
}, loads(self.post[0]['data']))
self.assertEqual(887, response.total)
self.assertEqual(6, response.queryTime)
self.assertEqual({'searchTime': 3}, response.times)
self.assertEqual(1, len(response.hits))
self.assertEqual("record:1", response.hits[0].id)
self.assertEqual(0.1234, response.hits[0].score)
self.assertEqual(dict(__key__=2), response.hits[0].duplicateCount)
self.assertEqual([
{"fieldname": "facet", "path": [], "terms": [{"term": "term", "count": 1}]}
], response.drilldownData)
self.assertEqual({'valeu': ['value']}, response.suggestions)
def testPrefixSearch(self):
self.response = JsonList([["value0", 1], ["value1", 2]]).dumps()
response = retval(self._lucene.prefixSearch(fieldname='field1', prefix='valu'))
self.assertEquals(['value1', 'value0'], response.hits)
response = retval(self._lucene.prefixSearch(fieldname='field1', prefix='valu', showCount=True))
self.assertEquals([('value1', 2), ('value0', 1)], response.hits)
def testNumDocs(self):
self.response = "150"
result = retval(self._lucene.numDocs())
self.assertEqual(150, result)
self.assertEqual([{'data': None, 'path': '/lucene/numDocs/'}], self.post)
def testFieldnames(self):
self.response = '["field1", "field2"]'
result = retval(self._lucene.fieldnames())
self.assertEqual(["field1", "field2"], result.hits)
self.assertEqual([{"data": None, "path": "/lucene/fieldnames/"}], self.post)
def testDrilldownFieldnames(self):
self.response = '["field1", "field2"]'
result = retval(self._lucene.drilldownFieldnames())
self.assertEqual(["field1", "field2"], result.hits)
self.assertEqual([{"data": None, "path": "/lucene/drilldownFieldnames/?limit=50"}], self.post)
result = retval(self._lucene.drilldownFieldnames(limit=1, path=['field']))
self.assertEqual(["field1", "field2"], result.hits)
self.assertEqual({"data": None, "path": "/lucene/drilldownFieldnames/?dim=field&limit=1"}, self.post[-1])
result = retval(self._lucene.drilldownFieldnames(limit=1, path=['xyz', 'abc', 'field']))
self.assertEqual(["field1", "field2"], result.hits)
self.assertEqual({"data": None, "path": "/lucene/drilldownFieldnames/?dim=xyz&limit=1&path=abc&path=field"}, self.post[-1])
def testUpdateSettings(self):
self.response = JsonDict(numberOfConcurrentTasks=6, similarity="BM25(k1=1.2,b=0.75)", clustering=JsonDict(clusterMoreRecords=100, clusteringEps=0.4, clusteringMinPoints=1))
settings = retval(self._lucene.getSettings())
self.assertEqual(['/settings/'], self.read)
self.assertEquals({'numberOfConcurrentTasks': 6, 'similarity': u'BM25(k1=1.2,b=0.75)', 'clustering': {'clusterMoreRecords': 100, 'clusteringEps': 0.4, 'clusteringMinPoints': 1}}, settings)
clusterFields = [
{"filterValue": None, "fieldname": "untokenized.dcterms:isFormatOf.uri", "weight": 0}
]
self.response = ""
consume(self._lucene.setSettings(similarity=dict(name="bm25", k1=1.0, b=2.0), numberOfConcurrentTasks=10, clustering=dict(clusterMoreRecords=200, clusteringEps=1.0, clusteringMinPoints=2, fields=clusterFields)))
self.assertEqual(1, len(self.post))
self.assertEqual('/lucene/settings/', self.post[0]['path'])
self.assertEqual({
"numberOfConcurrentTasks": 10,
"similarity": dict(type="BM25Similarity", k1=1.0, b=2.0),
"clustering": {
"clusterMoreRecords": 200,
"clusteringEps": 1.0,
"clusteringMinPoints": 2,
"fields": [
{"filterValue": None, "fieldname": "untokenized.dcterms:isFormatOf.uri", "weight": 0}
]
}
}, loads(self.post[0]['data']))
consume(self._lucene.setSettings(numberOfConcurrentTasks=5, similarity=None, clustering=None))
self.assertEqual(2, len(self.post))
self.assertEqual('/lucene/settings/', self.post[1]['path'])
self.assertEqual({
"numberOfConcurrentTasks": 5,
}, loads(self.post[1]['data']))
def testSimilarDocs(self):
self.response = JsonDict({
"total": 887,
"queryTime": 6,
"times": {"searchTime": 3},
"hits": [
{"id": "record:1", "score": 0.1234},
{"id": "record:2", "score": 0.1234},
],
}).dumps()
response = retval(self._lucene.similarDocuments(identifier='record:3'))
self.assertEqual(887, response.total)
self.assertEqual(2, len(response.hits))
class DeDupFilterCollectorTest(SeecrTestCase):
def setUp(self):
super(DeDupFilterCollectorTest, self).setUp()
self._reactor = CallTrace('reactor')
settings = LuceneSettings(commitCount=1, verbose=False)
self.lucene = Lucene(self.tempdir, reactor=self._reactor, settings=settings)
def tearDown(self):
self.lucene.close()
super(DeDupFilterCollectorTest, self).tearDown()
def testCollectorTransparentlyDelegatesToNextCollector(self):
self._addDocument("urn:1", 2)
tc = TopScoreDocSuperCollector(100, True)
c = DeDupFilterSuperCollector("__isformatof__", "__sort__", tc)
self.lucene.search(query=MatchAllDocsQuery(), collector=c)
self.assertEquals(1, tc.topDocs(0).totalHits)
def _addDocument(self, identifier, isformatof, sort=None):
doc = Document()
if isformatof:
doc.add(NumericDocValuesField("__isformatof__", long(isformatof)))
if sort:
doc.add(NumericDocValuesField("__sort__", long(sort)))
consume(self.lucene.addDocument(identifier, doc))
self.lucene.commit() # Explicitly, not required: since commitCount=1.
def testCollectorFiltersTwoSimilar(self):
self._addDocument("urn:1", 2, 1)
self._addDocument("urn:2", 2, 2)
tc = TopScoreDocSuperCollector(100, True)
c = DeDupFilterSuperCollector("__isformatof__", "__sort__", tc)
self.lucene.search(query=MatchAllDocsQuery(), collector=c)
topDocsResult = tc.topDocs(0)
self.assertEquals(1, topDocsResult.totalHits)
self.assertEquals(1, len(topDocsResult.scoreDocs))
docId = topDocsResult.scoreDocs[0].doc
key = c.keyForDocId(docId)
identifier = self.lucene._index.getDocument(key.getDocId()).get(IDFIELD)
self.assertEquals('urn:2', identifier)
self.assertEquals(2, key.count)
def testCollectorFiltersTwoTimesTwoSimilarOneNot(self):
self._addDocument("urn:1", 1, 2001)
self._addDocument("urn:2", 3, 2009) # result 2x
self._addDocument("urn:3", 50, 2010) # result 1x
self._addDocument("urn:4", 3, 2001)
self._addDocument("urn:5", 1, 2009) # result 2x
#expected: "urn:2', "urn:3" and "urn:5" in no particular order
tc = TopScoreDocSuperCollector(100, True)
c = DeDupFilterSuperCollector("__isformatof__", "__sort__", tc)
self.lucene.search(query=MatchAllDocsQuery(), collector=c)
topDocsResult = tc.topDocs(0)
self.assertEquals(3, topDocsResult.totalHits)
self.assertEquals(3, len(topDocsResult.scoreDocs))
rawDocIds = [scoreDoc.doc for scoreDoc in topDocsResult.scoreDocs]
netDocIds = [c.keyForDocId(rawDocId).docId for rawDocId in rawDocIds]
identifiers = set(self.lucene._index.getDocument(doc).get(IDFIELD) for doc in netDocIds)
self.assertEquals(set(["urn:2", "urn:3", "urn:5"]), identifiers)
self.assertEquals([1,2,2], list(sorted(c.keyForDocId(d).count for d in netDocIds)))
def testSilentyYieldsWrongResultWhenFieldNameDoesNotMatch(self):
self._addDocument("urn:1", 2)
tc = TopScoreDocSuperCollector(100, True)
c = DeDupFilterSuperCollector("__wrong_field__", "__sort__", tc)
self.lucene.search(query=MatchAllDocsQuery(), collector=c)
self.assertEquals(1, tc.topDocs(0).totalHits)
def testShouldAddResultsWithoutIsFormatOf(self):
self._addDocument("urn:1", 2)
self._addDocument("urn:2", None)
self._addDocument("urn:3", 2)
self._addDocument("urn:4", None)
self._addDocument("urn:5", None)
self._addDocument("urn:6", None)
self._addDocument("urn:7", None)
self._addDocument("urn:8", None)
self._addDocument("urn:9", None)
self._addDocument("urn:A", None)
self._addDocument("urn:B", None) # trigger a merge
tc = TopScoreDocSuperCollector(100, True)
c = DeDupFilterSuperCollector("__isformatof__", "__sort__", tc)
self.lucene.search(query=MatchAllDocsQuery(), collector=c)
self.assertEquals(10, tc.topDocs(0).totalHits)
