def indexDocs(self, img_url, toi, tid):
try:
t1 = FieldType()
t1.setIndexed(True)
t1.setStored(True)
t1.setTokenized(True)
t1.setIndexOptions(
FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
print("Adding", img_url)
name = "Pictures/1.jpg"
conn = urllib.urlopen(img_url)
f = open(name, 'wb')
f.write(conn.read())
f.close()
img = cv2.imread(name)
sdf = img_search_color(img)
storeDir = 'Picture_new/' + sdf.strs
if not os.path.exists(storeDir):
os.mkdir(storeDir)
cv2.imwrite(storeDir + '/' + str(toi) + '___' + str(tid) + '.jpg',
img)
'''storeDir2 = 'Picture_user/'+str(tid)
if not os.path.exists(storeDir2):
n = 0
os.mkdir(storeDir2)
else :
n = len(os.listdir(storeDir2))
cv2.imwrite(storeDir2+'/'+str(toi)+ '_'+ str(n) + '.jpg',img)'''
except Exception, e:
print("Failed in indexDocs:", e)
def __init__(self, indexDir, doClear=True, computeLengthNorm=False):
# if not jpype.isJVMStarted():
# lucene.initVM()
lucene.getVMEnv().attachCurrentThread()
self.analyzer = StandardAnalyzer(Version.LUCENE_CURRENT)
# self.analyzer = LimitTokenCountAnalyzer(self.analyzer, 100678)#is here?
self.config = IndexWriterConfig(Version.LUCENE_CURRENT, self.analyzer)
self.config.setRAMBufferSizeMB(256.0) # 设置自动提交的最大RAM为256MB
self.config.setMaxBufferedDocs(10000) # 设置自动提交的最大Docs个数为10000
if not computeLengthNorm:
sim = CustomSimilarity()
self.config.setSimilarity(sim)
self.path = os.path.join(INDEX_PATH, indexDir)
# print self.path
# path.mkdir(self.path)
# if doClear:
# self.clearExistingIndex()
self.store = SimpleFSDirectory(File(self.path))
self.writer = IndexWriter(self.store, self.config)
self.t1 = FieldType() # 域t1
self.t1.setIndexed(True)
self.t1.setStored(True)
self.t1.setTokenized(False)
self.t1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
self.t2 = FieldType() # 域t2
self.t2.setIndexed(True)
self.t2.setStored(False)
self.t2.setTokenized(True)
self.t2.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
def run(self):
print "Booting lucene driver worker...."
lucene.initVM()
self.fieldType1 = FieldType()
self.fieldType1.setIndexed(True)
self.fieldType1.setStored(False)
self.fieldType1.setTokenized(True)
self.fieldType2 = FieldType()
self.fieldType2.setIndexed(True)
self.fieldType2.setStored(True)
self.fieldType2.setTokenized(False)
while(True):
data = self.queue.get()
da = data[1]
response = None
try:
self.fil = File(da['data']['indexdir'])
self.d = NIOFSDirectory(self.fil)
self.analyzer = StandardAnalyzer(Version.LUCENE_CURRENT)
self.conf = IndexWriterConfig(
Version.LUCENE_CURRENT,
self.analyzer)
response = getattr(self, da['action'])(da['data'])
self.d.close()
except Exception as e:
print e
if response is None:
response = {}
self.ret[data[0]] = response
def __init__(self, folder=None, fields=[], similarity="tfidf"):
self.jcc = lucene.initVM()
if folder:
self.directory = SimpleFSDirectory(File(folder))
else:
self.directory = RAMDirectory()
self.fields = {}
for field in fields:
ft = FieldType()
for pname, pvalue in field.props.items():
setter = getattr(ft, "set" + pname.capitalize())
setter(pvalue)
ft.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
# ft.setOmitNorms(True)
self.fields[field.name] = ft
self.similarity = similarity.lower()
self.analyzer = StandardAnalyzer(Version.LUCENE_CURRENT)
self.writer = None
self.searcher = None
def index(self, root):
t = FieldType()
t.setIndexed(True)
t.setStored(True)
t.setTokenized(True)
t.setStoreTermVectors(True)
for path, dirs, files in os.walk(root):
for file in files:
filePath = os.path.join(path, file)
fd = open(filePath)
content = unicode(fd.read(), 'iso-8859-1')
fd.close()
doc = Document()
doc.add(Field('name', file, StringField.TYPE_STORED))
parent = os.path.split(path)[1]
doc.add(Field('parent', parent, StringField.TYPE_STORED))
if len(content) > 0:
doc.add(Field('content', content, t))
print 'Indexing %s' % file
self.mWriter.addDocument(doc)
self.mWriter.commit()
self.mWriter.close()
def create_minidoc(termstring, field='text'):
# To store term vectors (used for query expansion) we have to use a custom fieldtype
customfield = FieldType()
customfield.setIndexOptions(IndexOptions.DOCS_AND_FREQS)
customfield.setStored(True)
customfield.setTokenized(True)
customfield.setStoreTermVectors(True)
doc = Document()
doc.add(Field(field, termstring, customfield))
return doc
def init_writing(self):
self.__field_type_searching = FieldType(TextField.TYPE_STORED)
self.__field_type_frequency = FieldType(StringField.TYPE_STORED)
self.__field_type_frequency.setStored(True)
self.__field_type_frequency.setTokenized(False)
self.__field_type_frequency.setStoreTermVectors(True)
self.__field_type_frequency.setStoreTermVectorPositions(True)
self.__field_type_frequency.\
setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS)
fs_directory = SimpleFSDirectory(Paths.get(self.directory))
self.__writer = IndexWriter(fs_directory, IndexWriterConfig())
def _set_fieldtypes(self):
self._ft1 = FieldType()
self._ft1.setStored(True)
self._ft1.setTokenized(False)
self._ft1.setIndexOptions(IndexOptions.DOCS_AND_FREQS)
self._ft2 = FieldType()
self._ft2.setStored(True)
self._ft2.setTokenized(True)
self._ft1.setStoreTermVectors(True)
self._ft1.setStoreTermVectorOffsets(True)
self._ft1.setStoreTermVectorPositions(True)
self._ft2.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
def indexer(self, root, writer):
t1 = FieldType()
t1.setStored(True)
t1.setTokenized(True)
t1.setIndexOptions(IndexOptions.DOCS_AND_FREQS)
def repalcer(text):
chars = '\\`*_{}[]()>#+-.!$‘'
for c in chars:
if c in text:
text = text.replace(c, ' ')
return text
for root, dirnames, filenames in os.walk(root):
i = 0
for filename in filenames:
i += 1
with open(os.path.join(root, filename)) as f:
for line in f.readlines():
line = line.split(' ', 2)
docname = line[0] + ' ' + line[1]
name = repalcer(line[0])
contents = line[2]
doc = Document()
doc.add(Field('docname', docname, t1))
doc.add(Field('name', name, t1))
doc.add(Field('contents', contents, t1))
writer.addDocument(doc)
print('File %d done indexing' % i)
def indexDocs(self, root, writer):
t1 = FieldType()
t1.setStored(True)
t1.setTokenized(True)
t1.setIndexOptions(IndexOptions.DOCS_AND_FREQS)
wikiFile = ZipFile(root, 'r')
files = wikiFile.namelist()
i = 0
for file in files[1:]:
i += 1
wiki = wikiFile.open(file, 'r')
for line in wiki:
for line in codecs.iterdecode(wiki, 'utf8'):
normailized = unicodedata.normalize('NFD',
line).split(' ', 2)
if not normailized[1].isdigit(): continue
docname = normailized[0] + ' ' + normailized[1]
name = re.sub(r'[^a-zA-Z0-9]', ' ', normailized[0])
contents = normailized[2]
doc = Document()
doc.add(Field('docname', docname, t1))
doc.add(Field('name', name, t1))
doc.add(Field('contents', contents, t1))
writer.addDocument(doc)
print('File %d done indexing' % i, file)
def main():
INDEX_DIR = "full_index1"
DOCUMENTS_DIR = "/media/joseph/Windows8_OS/Users/Joseph/AppData/Local/lxss/home/jwymbs23/data_science_projects/french_pamphlets/frc-data-master/OCR_text/"
# Initialize lucene and JVM
lucene.initVM(vmargs=['-Djava.awt.headless=true'])
print("lucene version is:", lucene.VERSION, '\n')
store = getStore(INDEX_DIR)
analyzer = getAnalyzer()
writer = getWriter(store=store, analyzer=analyzer, create=True)
#get list of documents
doc_list = getDoclist(DOCUMENTS_DIR)
ftype = FieldType()
ftype.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
ftype.setTokenized(True)
ftype.setStoreTermVectors(True)
ftype.freeze()
for cd, doc_name in enumerate(doc_list):
if not cd % 1000:
print(cd, '--', len(doc_list))
with open(doc_name, 'r') as d:
doc_lines = d.readlines()
full_text = ''.join([i.strip() for i in doc_lines]).lower()
try:
# create a document that would we added to the index
doc = Document()
# Add fields to this document
#could process fname here instead of in the dataframe later
doc.add(
Field("identifier",
doc_name.split('/')[-1], TextField.TYPE_STORED)
) #Store.YES))#, Field.Index.ANALYZED))
doc.add(
Field("vectext", full_text, ftype)
) #TextField.TYPE_STORED, TermVector.YES, ))#Store.YES))#, Field.Index.ANALYZED))
doc.add(Field("text", full_text, TextField.TYPE_STORED))
# Add the document to the index
writer.addDocument(doc)
except:
print("Failed in indexDocs: ", doc_name)
#writer.optimize()
writer.commit()
def get_field_type(self, stored, indexed):
'''
Generate a `FieldType` according to the parameters.
Input: `stored`: whether the field is stored
`indexed`: whether the field is indexed
Output: `FieldType` with the correct parameters
'''
t = FieldType()
t.setStored(stored)
t.setTokenized(indexed)
if indexed: # Indexes documents, frequencies and positions
t.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
else: # Not indexed
t.setIndexOptions(IndexOptions.NONE)
return t
def create_field_types(self):
""" Create the field types that will be used to specify
what actions lucene should take on the various fields
supplied to index.
"""
self.field_clean = FieldType()
self.field_clean.setIndexed(True)
self.field_clean.setStored(True)
self.field_clean.setTokenized(False)
self.field_clean.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
self.field_dirty = FieldType()
self.field_dirty.setIndexed(True)
self.field_dirty.setStored(False)
self.field_dirty.setTokenized(True)
self.field_dirty.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
def __init__(self, indexDir="", debug=False, verbose=False):
"""
:Parameters:
- `indexDir`: Path where the Index will be saved. (Str)
- `debug`: Create the Index in RAM Memory (indexDir will be ignored). (Boolean)
- `verbose`: Provide additional information about the initialization process. (Boolean)
"""
self.__verbose = verbose
if indexDir != "":
INDEX_DIR = indexDir
else:
INDEX_DIR = os.path.dirname(
os.path.realpath(__file__)) + "/luceneIndex"
if not os.path.exists(INDEX_DIR):
os.makedirs(INDEX_DIR)
self.__boAppend = False
else:
self.__boAppend = True
# Initialize lucene and JVM
lucene.initVM()
# Get index storage
if debug:
# Store the index in memory
self.__indexDir = RAMDirectory()
self.__boAppend = False
INDEX_DIR = "RAM Memory"
else:
# Store an index on disk
self.__indexDir = SimpleFSDirectory(Paths.get(INDEX_DIR))
# Create Content FieldType
self.__contentType = FieldType()
self.__contentType.setIndexOptions(IndexOptions.DOCS_AND_FREQS)
self.__contentType.setTokenized(True)
self.__contentType.setStored(True)
self.__contentType.setStoreTermVectors(True)
self.__contentType.setStoreTermVectorPositions(True)
self.__contentType.freeze()
# Get the Analyzer
self.__analyzer = StandardAnalyzer(
StandardAnalyzer.ENGLISH_STOP_WORDS_SET)
# Print Indexer Information
print("Lucene version is: ", lucene.VERSION)
print("Index Directory: ", INDEX_DIR)
def indexDocs(self, lines, writer):
ts = []
for i in range(4):
t1 = FieldType()
t1.setStored(True)
t1.setTokenized(True)
t1.setIndexOptions(IndexOptions.DOCS_AND_FREQS)
t1.storeTermVectors()
ts.append(t1)
file_names = ["all", "context", "last", "response"]
for line in lines:
doc = Document()
for i in range(4):
doc.add(Field(file_names[i], line[i], ts[i]))
writer.addDocument(doc)
def __init__(self):
"""Init possible field types"""
# FIELD_ID: stored, indexed, non-tokenized
self.field_id = FieldType()
#self.field_id.setIndexed(True)
self.field_id.setIndexOptions(IndexOptions.DOCS)
self.field_id.setStored(True)
self.field_id.setTokenized(False)
# FIELD_ID_TV: stored, indexed, not tokenized, with term vectors (without positions)
# for storing IDs with term vector info
self.field_id_tv = FieldType()
#self.field_id_tv.setIndexed(True)
self.field_id_tv.setIndexOptions(IndexOptions.DOCS_AND_FREQS)
self.field_id_tv.setStored(True)
self.field_id_tv.setTokenized(False)
self.field_id_tv.setStoreTermVectors(True)
# FIELD_TEXT: stored, indexed, tokenized, with positions
self.field_text = FieldType()
#self.field_text.setIndexed(True)
self.field_text.setIndexOptions(
IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
self.field_text.setStored(True)
self.field_text.setTokenized(True)
# FIELD_TEXT_TV: stored, indexed, tokenized, with term vectors (without positions)
self.field_text_tv = FieldType()
#self.field_text_tv.setIndexed(True)
self.field_text_tv.setIndexOptions(IndexOptions.DOCS_AND_FREQS)
self.field_text_tv.setStored(True)
self.field_text_tv.setTokenized(True)
self.field_text_tv.setStoreTermVectors(True)
# FIELD_TEXT_TVP: stored, indexed, tokenized, with term vectors and positions
# (but no character offsets)
self.field_text_tvp = FieldType()
#self.field_text_tvp.setIndexed(True)
self.field_text_tvp.setIndexOptions(
IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
self.field_text_tvp.setStored(True)
self.field_text_tvp.setTokenized(True)
self.field_text_tvp.setStoreTermVectors(True)
self.field_text_tvp.setStoreTermVectorPositions(True)
def setUp(self):
super(Test_Bug1842, self).setUp()
self.analyzer = StandardAnalyzer()
w1 = self.getWriter(analyzer=self.analyzer)
doc1 = Document()
ftype = FieldType()
ftype.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
ftype.setTokenized(True)
ftype.setStoreTermVectors(True)
ftype.freeze()
doc1.add(Field("all", "blah blah blah Gesundheit", ftype))
doc1.add(Field('id', '1', StringField.TYPE_NOT_STORED))
w1.addDocument(doc1)
w1.close()
def index_docs(self, input_documents):
for document in tqdm(input_documents, total=len(input_documents)):
doc = Document()
doc.add(StringField(".I", document[".I"].lower(), Field.Store.YES))
doc.add(StringField(".U", document[".U"].lower(), Field.Store.YES))
type = FieldType()
type.setIndexOptions(
IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS)
type.setStored(True)
type.setStoreTermVectors(True)
type.setTokenized(True)
if ".W" in document and ".M" in document:
doc.add(
Field(
"text", " ".join(
tokenizer.tokenize(document[".M"].lower() + " " +
document[".T"].lower() +
document[".W"].lower())), type))
elif ".M" in document and ".W" not in document:
doc.add(
Field(
"text", " ".join(
tokenizer.tokenize(document[".M"].lower() + " " +
document[".T"].lower())), type))
elif ".M" not in document and ".W" in document:
doc.add(
Field(
"text", " ".join(
tokenizer.tokenize(document[".T"].lower() +
document[".W"].lower())), type))
elif ".M" not in document and ".W" not in document:
doc.add(
Field("text",
" ".join(tokenizer.tokenize(document[".T"].lower())),
type))
if self.writer.getConfig().getOpenMode(
) == IndexWriterConfig.OpenMode.CREATE:
self.writer.addDocument(doc)
else:
self.writer.updateDocument(Term(".U", document[".U"]), doc)
self.writer.close()
def create_index(self, index_folder):
os.mkdir(index_folder)
self.t1 = FieldType()
self.t1.setStored(True)
self.t1.setIndexOptions(IndexOptions.DOCS)
self.t2 = FieldType()
self.t2.setStored(True)
self.t2.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
self.t3 = FieldType()
self.t3.setStored(True)
self.t3.setIndexOptions(IndexOptions.NONE)
fsDir = MMapDirectory(Paths.get(index_folder))
writerConfig = IndexWriterConfig(
MySimpleAnalyzer(
CharArraySet(collections.JavaSet(utils.STOPWORDS), True)))
writerConfig.setSimilarity(MyTFIDFSimilarity())
writerConfig.setRAMBufferSizeMB(16384.0) # 14g
self.writer = IndexWriter(fsDir, writerConfig)
logger.info(f"{self.writer.numDocs()} docs in index")
logger.info("Indexing documents...")
doc_ids = self.doc_db.get_doc_ids()
for doc_id in tqdm(doc_ids, total=len(doc_ids)):
text = self.doc_db.get_doc_text(doc_id)
tokens = self.doc_db.get_doc_tokens(doc_id)
self.add_doc(doc_id, text, tokens)
logger.info(f"Indexed {self.writer.numDocs()} docs.")
self.writer.forceMerge(1) # to increase search performance
self.writer.close()
def testAdd(self, goodname, salenum, price, shopname, url, picturename, comment, historyprice):
analyzer = WhitespaceAnalyzer(Version.LUCENE_CURRENT)
analyzer = LimitTokenCountAnalyzer(analyzer, 1048576)
config = IndexWriterConfig(Version.LUCENE_CURRENT, analyzer)
config.setOpenMode(IndexWriterConfig.OpenMode.CREATE_OR_APPEND)
writer = IndexWriter(self.dir, config)
# True，建立新索引，False，建立增量索引
noIndexedString = FieldType()
noIndexedString.setTokenized(False)
noIndexedString.setIndexed(False)
noIndexedString.setStored(True)
try:
print "adding", goodname
goodname_s = unicode(goodname, 'utf8')
seg_list_good = jieba.cut(goodname_s, cut_all=False)
goodname_s = " ".join(seg_list_good) # 默认模式
shopname_s = unicode(shopname, 'utf8')
seg_list_shop = jieba.cut(shopname_s, cut_all=False)
shopname_s = " ".join(seg_list_shop) # 默认模式
shopnameField = Field("shopName", shopname, noIndexedString)
shopnameField_s = TextField("shopName_s", shopname_s, Field.Store.NO)
goodnameField = Field("goodName", goodname, noIndexedString)
goodnameField_s = TextField("goodName_s", goodname_s, Field.Store.NO)
salenumField = IntField("saleNum", salenum, Field.Store.YES)
priceField = DoubleField("price", price, Field.Store.YES)
urlField = Field("url", url, noIndexedString)
pictureField = StringField("pictureName", picturename, Field.Store.YES)
commentField = Field("comments", comment, noIndexedString)
historyPriceField = Field("historyPrice", historyprice, noIndexedString)
doc = Document()
doc.add(shopnameField)
doc.add(shopnameField_s)
doc.add(goodnameField)
doc.add(goodnameField_s)
doc.add(salenumField)
doc.add(priceField)
doc.add(urlField)
doc.add(pictureField)
doc.add(commentField)
doc.add(historyPriceField)
writer.addDocument(doc)
except Exception, e:
print "Failed in indexDocs:", e
def create_index(self, index_folder, docs_path, add_terms=False):
os.mkdir(index_folder)
self.t1 = FieldType()
self.t1.setStored(True)
self.t1.setIndexOptions(IndexOptions.DOCS)
self.t2 = FieldType()
self.t2.setStored(False)
self.t2.setIndexOptions(IndexOptions.DOCS_AND_FREQS)
self.t3 = FieldType()
self.t3.setStored(True)
self.t3.setIndexOptions(IndexOptions.NONE)
fsDir = MMapDirectory(Paths.get(index_folder))
writerConfig = IndexWriterConfig(StandardAnalyzer())
self.writer = IndexWriter(fsDir, writerConfig)
print "%d docs in index" % self.writer.numDocs()
print "Indexing documents..."
doc_id = 0
import corpus_hdf5
corpus = corpus_hdf5.CorpusHDF5(docs_path)
for txt in corpus.get_text_iter():
title = corpus.get_article_title(doc_id)
self.add_doc(doc_id, title, txt, add_terms)
if doc_id % 1000 == 0:
print 'indexing doc', doc_id
doc_id += 1
print "Index of %d docs..." % self.writer.numDocs()
self.writer.close()
def testBinaryFieldInIndex(self):
ft = FieldType()
ft.setStored(True)
bytes = JArray('byte')(self.binaryValStored)
binaryFldStored = StoredField("binaryStored", bytes)
stringFldStored = Field("stringStored", self.binaryValStored, ft)
doc = Document()
doc.add(binaryFldStored)
doc.add(stringFldStored)
# test for field count
self.assertEqual(2, doc.fields.size())
# add the doc to a ram index
writer = self.getWriter(
analyzer=StandardAnalyzer(Version.LUCENE_CURRENT))
writer.addDocument(doc)
writer.close()
# open a reader and fetch the document
reader = self.getReader()
docFromReader = reader.document(0)
self.assert_(docFromReader is not None)
# fetch the binary stored field and compare it's content with the
# original one
bytes = docFromReader.getBinaryValue("binaryStored")
binaryFldStoredTest = bytes.bytes.string_
self.assertEqual(binaryFldStoredTest, self.binaryValStored)
# fetch the string field and compare it's content with the original
# one
stringFldStoredTest = docFromReader.get("stringStored")
self.assertEqual(stringFldStoredTest, self.binaryValStored)
reader.close()
def __init__(self):
"""Init possible field types."""
# FIELD_ID: stored, indexed, non-tokenized
self.field_id = FieldType()
self.field_id.setIndexed(True)
self.field_id.setStored(True)
self.field_id.setTokenized(False)
# FIELD_ID_TV: stored, indexed, not tokenized, with term vectors (without positions)
# for storing IDs with term vector info
self.field_id_tv = FieldType()
self.field_id_tv.setIndexed(True)
self.field_id_tv.setStored(True)
self.field_id_tv.setTokenized(False)
self.field_id_tv.setStoreTermVectors(True)
# FIELD_TEXT: stored, indexed, tokenized, with positions
self.field_text = FieldType()
self.field_text.setIndexed(True)
self.field_text.setStored(True)
self.field_text.setTokenized(True)
# FIELD_TEXT_TV: stored, indexed, tokenized, with term vectors (without positions)
self.field_text_tv = FieldType()
self.field_text_tv.setIndexed(True)
self.field_text_tv.setStored(True)
self.field_text_tv.setTokenized(True)
self.field_text_tv.setStoreTermVectors(True)
# FIELD_TEXT_TVP: stored, indexed, tokenized, with term vectors and positions
# (but no character offsets)
self.field_text_tvp = FieldType()
self.field_text_tvp.setIndexed(True)
self.field_text_tvp.setStored(True)
self.field_text_tvp.setTokenized(True)
self.field_text_tvp.setStoreTermVectors(True)
self.field_text_tvp.setStoreTermVectorPositions(True)
# FIELD_TEXT_NTV: not stored, indexed, tokenized, with term vectors (without positions)
self.field_text_ntv = FieldType()
self.field_text_ntv.setIndexed(True)
self.field_text_ntv.setStored(False)
self.field_text_ntv.setTokenized(True)
self.field_text_ntv.setStoreTermVectors(True)
# FIELD_TEXT_TVP: not stored, indexed, tokenized, with term vectors and positions
# (but no character offsets)
self.field_text_ntvp = FieldType()
self.field_text_ntvp.setIndexed(True)
self.field_text_ntvp.setStored(False)
self.field_text_ntvp.setTokenized(True)
self.field_text_ntvp.setStoreTermVectors(True)
self.field_text_ntvp.setStoreTermVectorPositions(True)
def testBinaryFieldInIndex(self):
ft = FieldType()
ft.setStored(True)
bytes = JArray('byte')(self.binaryValStored)
binaryFldStored = StoredField("binaryStored", bytes)
stringFldStored = Field("stringStored", self.binaryValStored, ft)
doc = Document()
doc.add(binaryFldStored)
doc.add(stringFldStored)
# test for field count
self.assertEqual(2, doc.fields.size())
# add the doc to a ram index
writer = self.getWriter(analyzer=StandardAnalyzer())
writer.addDocument(doc)
writer.close()
# open a reader and fetch the document
reader = self.getReader()
docFromReader = reader.document(0)
self.assertTrue(docFromReader is not None)
# fetch the binary stored field and compare it's content with the
# original one
bytes = docFromReader.getBinaryValue("binaryStored")
binaryFldStoredTest = bytes.bytes.bytes_
self.assertEqual(binaryFldStoredTest, self.binaryValStored)
# fetch the string field and compare it's content with the original
# one
stringFldStoredTest = docFromReader.get("stringStored")
self.assertEqual(stringFldStoredTest, self.binaryValStored.decode())
reader.close()
def setUp(self):
super(Test_Bug1842, self).setUp()
self.analyzer = StandardAnalyzer()
w1 = self.getWriter(analyzer=self.analyzer)
doc1 = Document()
ftype = FieldType()
ftype.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
ftype.setTokenized(True)
ftype.setStoreTermVectors(True)
ftype.freeze()
doc1.add(Field("all", "blah blah blah Gesundheit", ftype))
doc1.add(Field('id', '1', StringField.TYPE_NOT_STORED))
w1.addDocument(doc1)
w1.close()
def setUp(self):
super(Test_Bug1842, self).setUp()
self.analyzer = StandardAnalyzer(Version.LUCENE_CURRENT)
w1 = self.getWriter(analyzer=self.analyzer)
doc1 = Document()
ftype = FieldType()
ftype.setStored(False)
ftype.setIndexed(True)
ftype.setStoreTermVectors(True)
doc1.add(Field("all", "blah blah blah Gesundheit", ftype))
doc1.add(Field('id', '1', StringField.TYPE_NOT_STORED))
w1.addDocument(doc1)
w1.close()
def setUp(self):
super(Test_Bug1842, self).setUp()
self.analyzer = StandardAnalyzer(Version.LUCENE_CURRENT)
w1 = self.getWriter(analyzer=self.analyzer)
doc1 = Document()
ftype = FieldType()
ftype.setStored(False)
ftype.setIndexed(True)
ftype.setStoreTermVectors(True)
doc1.add(Field("all", "blah blah blah Gesundheit", ftype))
doc1.add(Field('id', '1', StringField.TYPE_NOT_STORED))
w1.addDocument(doc1)
w1.close()
def lazyImport():
global imported
if imported:
return
from meresco.pylucene import getJVM
getJVM()
from java.nio.file import Paths
from org.apache.lucene.document import Document, StringField, Field, FieldType
from org.apache.lucene.search import IndexSearcher, TermQuery
from org.apache.lucene.index import DirectoryReader, Term, IndexWriter, IndexWriterConfig, IndexOptions
from org.apache.lucene.store import FSDirectory
from org.apache.lucene.util import Version
from org.apache.lucene.analysis.core import WhitespaceAnalyzer
UNINDEXED_TYPE = FieldType()
UNINDEXED_TYPE.setIndexOptions(IndexOptions.NONE)
UNINDEXED_TYPE.setStored(True)
UNINDEXED_TYPE.setTokenized(False)
imported = True
globals().update(locals())
def indexer(self, root, writer):
t1 = FieldType()
t1.setStored(True)
t1.setTokenized(True)
t1.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
for root, dirnames, filenames in os.walk(root):
i = 0
for filename in filenames:
i += 1
with open(os.path.join(root, filename), encoding='utf-8') as f:
for line in f.readlines():
line = line.split(' ', 2)
docname = line[0] + ' ' + line[1]
name = self.repalcer(line[0])
contents = line[2]
name_contents = name + ' ' + contents
doc = Document()
doc.add(Field('name-sid', docname, t1))
doc.add(Field('name', name, t1))
doc.add(Field('contents', contents, t1))
doc.add(Field('name-contents', name_contents, t1))
writer.addDocument(doc)
print('File %d done indexing' % i)
def add_documents(self, input_file, writer_obj):
"""
Read the input file and create the index from the paragraph text
Each paragraph is one document. Store the document ids, frequencies and
positions in the index for each term
:param input_file: the input file used for creating the index
:param writer_obj: the index writer object
:return:
"""
contentType = FieldType()
contentType.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
contentType.setStored(True)
contentType.setTokenized(True)
paras = get_paragraphs(input_file)
for p in paras:
doc = Document()
lines_split = p.split('|__|')
doc.add(StringField("id", lines_split[0], Field.Store.YES))
doc.add(
Field("paragraphs", get_stemmed_string(lines_split[1].lower()),
contentType))
writer_obj.addDocument(doc)
writer_obj.close()
def lazyImport():
global imported
if imported:
return
from meresco.pylucene import getJVM
getJVM()
from java.nio.file import Paths
from org.apache.lucene.document import Document, StringField, Field, FieldType
from org.apache.lucene.search import IndexSearcher, TermQuery
from org.apache.lucene.index import DirectoryReader, Term, IndexWriter, IndexWriterConfig, IndexOptions
from org.apache.lucene.store import FSDirectory
from org.apache.lucene.util import Version
from org.apache.lucene.analysis.core import WhitespaceAnalyzer
UNINDEXED_TYPE = FieldType()
UNINDEXED_TYPE.setIndexOptions(IndexOptions.NONE)
UNINDEXED_TYPE.setStored(True)
UNINDEXED_TYPE.setTokenized(False)
imported = True
globals().update(locals())
def create_index(self, index_folder, docs_path, add_terms=False):
print('Loading Vocab...')
if not self.vocab:
self.vocab = utils.load_vocab(prm.vocab_path, prm.n_words)
os.mkdir(index_folder)
self.t1 = FieldType()
self.t1.setStored(True)
self.t1.setIndexOptions(IndexOptions.DOCS)
self.t2 = FieldType()
self.t2.setStored(False)
self.t2.setIndexOptions(IndexOptions.DOCS_AND_FREQS)
self.t3 = FieldType()
self.t3.setStored(True)
self.t3.setIndexOptions(IndexOptions.NONE)
fsDir = MMapDirectory(Paths.get(index_folder))
writerConfig = IndexWriterConfig(StandardAnalyzer())
self.writer = IndexWriter(fsDir, writerConfig)
print("%d docs in index" % self.writer.numDocs())
print("Indexing documents...")
# import corpus_hdf5
# corpus = corpus_hdf5.MSMARCOCorpusHDF5(docs_path)
import pickle
with open(docs_path, "rb") as read_file:
corpus = pickle.load(read_file)
idx_cnt = 0
# for doc_id, txt in zip(corpus.get_id_iter(), corpus.get_text_iter()):
# for doc_id, txt in corpus.items():
for txt in corpus:
self.add_doc(idx_cnt, txt, add_terms) # not lowered
if idx_cnt % 1000 == 0:
print('indexing doc', idx_cnt)
idx_cnt += 1
print("Index of %d docs..." % self.writer.numDocs())
self.writer.close()
def index_docs(self, train_set, writer):
t1 = FieldType()
t1.setIndexed(True)
t1.setStored(True)
t1.setTokenized(False)
t1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
t2 = FieldType()
t2.setIndexed(True)
t2.setStored(True)
t2.setTokenized(True)
t2.setStoreTermVectorOffsets(True)
t2.setStoreTermVectorPayloads(True)
t2.setStoreTermVectorPositions(True)
t2.setStoreTermVectors(True)
t2.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
for ii in train_set:
doc = Document()
doc.add(Field("answer", ii['Answer'], t1))
doc.add(Field("qid", ii['Question ID'], t1))
doc.add(Field("category", ii['category'], t1))
doc.add(Field("position", ii['Sentence Position'], t1))
doc.add(Field("question", ii['Question Text'], t2))
doc.add(Field("wiki_plain",
self.wiki_reader.get_text(ii['Answer']), t2))
writer.addDocument(doc)
def indexDocs(self, root, writer):
t1 = FieldType()
t1.setIndexed(True)
t1.setStored(True)
t1.setTokenized(False)
t1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
t2 = FieldType()
t2.setIndexed(True)
t2.setStored(False)
t2.setTokenized(True)
t2.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
for root, dirnames, filenames in os.walk(root):
for filename in filenames:
if not filename.endswith('.html'):
continue
print "adding", filename
try:
path = os.path.join(root, filename)
file = open(path)
contents = unicode(file.read(), 'iso-8859-1')
file.close()
doc = Document()
doc.add(Field("name", filename, t1))
doc.add(Field("path", root, t1))
if len(contents) > 0:
doc.add(Field("contents", contents, t2))
else:
print "warning: no content in %s" % filename
writer.addDocument(doc)
except Exception, e:
print "Failed in indexDocs:", e
directory = FSDirectory.open(File(path))
analyzer = StandardAnalyzer(Version.LUCENE_43)
config = IndexWriterConfig(Version.LUCENE_43, analyzer)
writer = IndexWriter(directory, config)
return writer
def open_searcher(writer):
from org.apache.lucene.search import IndexSearcher
reader = writer.getReader()
searcher = IndexSearcher(reader)
return reader, searcher
from org.apache.lucene.document import Document, Field, FieldType, TextField, StringField
from org.apache.lucene.util import BytesRef, BytesRefIterator
from org.apache.lucene.index import Term
vectorFieldType = FieldType(TextField.TYPE_NOT_STORED)
vectorFieldType.setIndexed(True)
vectorFieldType.setTokenized(True)
vectorFieldType.setStoreTermVectors(True)
vectorFieldType.setStoreTermVectorPositions(False)
writer = open_writer('data/index')
def addToIndex(lxmlNode):
uri = xpathFirst(lxmlNode, '//oa:hasTarget/@rdf:resource')
print uri
seen = set()
doc = Document()
for fieldName in FIELD_NAMES:
seen.clear()
for subpath in [
def reindex(self):
''' Re-indexes the entire database into Index file'''
start = time.time()
# get all posts
posts = self._tuples_to_dict(self._fetch_all_questions(), self._posts_fields)
if not posts:
raise Exception("FATAL Error: Could not fetch posts from Database")
# open indexer
# lucene.initVM(vmargs=['-Djava.awt.headless=true'])
# print 'lucene', lucene.VERSION
store = SimpleFSDirectory(File(self.index_dir))
analyzer = StandardAnalyzer(Version.LUCENE_CURRENT)
config = IndexWriterConfig(Version.LUCENE_CURRENT, analyzer)
config.setOpenMode(IndexWriterConfig.OpenMode.CREATE_OR_APPEND)
writer = IndexWriter(store, config)
indexedField = FieldType()
indexedField.setIndexed(True)
indexedField.setStored(True)
indexedField.setTokenized(True)
indexedField.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
storedField = FieldType()
storedField.setIndexed(False)
storedField.setStored(True)
storedField.setTokenized(False)
storedField.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
fieldTypes = {
'type' : storedField,
'id' : storedField,
'title' : indexedField,
'question' : indexedField,
'answer' : indexedField,
# 'comment' : indexedField,
'tag' : indexedField,
'extra' : indexedField,
}
# get their comments
num_docs = 0
for post in posts:
if self.status_mode: print "\r {0:.2f} %complete".format(((num_docs/142627.0)*100)),
if self.debug : print "\n","*"*20,"\nIndexing post: ", post['id'], "from ", post['extra']
if self.debug and self.verbose_values: print post
answers = self._tuples_to_dict(self._fetch_all_answers(post['id'], post['extra']), self._answer_fields)
# add comment field
for answer in answers:
num_docs += 1
if self.debug: print "\n","+"*10, "\nMaking new Document"
doc = Document()
if self.debug: print "Adding doc type"
doc.add(Field("type", self.doctype, fieldTypes['type']))
# make fields
if self.debug: print "Adding post fields"
for i in xrange(len(self._posts_fields)):
f = Field(self._posts_fields[i], self._cleanup_tag(post[self._posts_fields[i]]), fieldTypes[self._posts_fields[i]])
f.setBoost(self._fields_boost[self._posts_fields[i]])
doc.add(f)
if self.status_mode: print "\t Indexing answer: ", answer['answer_id']
if self.debug and self.verbose_values: print answer
# answered_doc = copy.deepcopy(doc)
# make comment field
f = Field("answer", self._cleanup_tag(answer['answer']), fieldTypes['answer'])
f.setBoost(self._fields_boost['answer'])
doc.add(f)
# calculate paths
# commented_doc = copy.deepcopy(answered_doc)
# comments = self._comments_to_comment_string(self._tuples_to_dict(self._fetch_all_comments(answer['id']), self._comment_fields))
# if self.debug: print "\t\tAdding comments: ", comments
# commented_doc.add(Field("comment", self._cleanup_tag(comments), fieldTypes['comment']))
# write index
if self.debug: print "\tAdding document {doc_id} to index".format(doc_id=post['id'])
writer.addDocument(doc)
# del answered_doc
# del commented_doc
if self.debug: print "Commiting document to index"
writer.commit()
# close index
if self.status_mode: print "Closing index write"
writer.close()
end = time.time() - start
if self.status_mode: print "\n","-"*20, \
"\nTotal time spent in indexing: ", end, "seconds" \
"\nIndexed {num_docs} documents".format(num_docs=num_docs)
def tweetIndexer(self, writer):
t1 = FieldType()
t1.setIndexed(True)
t1.setStored(True)
t1.setTokenized(False)
t1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
t2 = FieldType()
t2.setIndexed(True)
t2.setStored(True)
t2.setTokenized(True)
t2.setStoreTermVectorOffsets(True)
t2.setStoreTermVectorPayloads(True)
t2.setStoreTermVectorPositions(True)
t2.setStoreTermVectors(True)
t2.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
x = 0
for i in range(0,500):
if not os.path.isfile("json/tweets-" + str(i) + ".json"):
break
print "adding tweets-" + str(i) + ".json"
tweets = open("json/tweets-" + str(i) + ".json", "r")
for line in tweets.readlines():
tweet = json.loads(line)
if 'limit' in tweet:
continue
try:
doc = Document()
doc.add(Field("file", "json/tweets-" + str(i) + ".json", t1))
sname = tweet['user']['screen_name']
tid = str(tweet['id'])
text = tweet['text']
uname = tweet['user']['name']
created = tweet['created_at']
tstamp = tweet['timestamp_ms']
place = ""
if tweet['place']:
place = tweet['place']['full_name'] + ", " + tweet['place']['country']
lat = ""
lng = ""
titles = ""
urls = ""
exist = "false"
if tweet['coordinates']:
lat = str(tweet['coordinates']['coordinates'][1])
lng = str(tweet['coordinates']['coordinates'][0])
else:
lat = str((tweet['place']['bounding_box']['coordinates'][0][0][1] + tweet['place']['bounding_box']['coordinates'][0][2][1])/2)
lng = str((tweet['place']['bounding_box']['coordinates'][0][0][0] + tweet['place']['bounding_box']['coordinates'][0][2][0])/2)
if len(tweet['entities']['urls']) != 0:
exist = "true"
for index in range(len(tweet['entities']['urls'])):
title = tweet['entities']['urls'][index]['url_title']
if title == None:
titles += ",-"
else:
title = title.encode('ascii','ignore')
titles += "," + str(title)
urls += " " + str(tweet['entities']['urls'][index]['expanded_url'])
searchable = text + " " + urls + " " + uname + " " + sname + " " + place
doc.add(Field("lookup", searchable, t2))
doc.add(Field("text", text, t2))
doc.add(Field("user_name", uname, t2))
doc.add(Field("screen_name", sname, t2))
doc.add(Field("tweet_id", tid, t2))
doc.add(Field("created_at", created, t2))
doc.add(Field("geo_lat", lat, t2))
doc.add(Field("geo_lng", lng, t2))
doc.add(Field("url_exist", exist, t2))
doc.add(Field("url_url", urls, t2))
doc.add(Field("url_title", titles, t2))
doc.add(Field("timestamp", tstamp, t2))
writer.addDocument(doc)
x += 1
except Exception, e:
pass
tweets.close()
def __init__(self):
self.mDocumentDirectory = settings.ADMINS_ENGINE.mDocumentDirectory
self.mIndexDirectory = settings.ADMINS_ENGINE.mIndexDirectory
self.mAnalyzers = settings.ADMINS_ENGINE.getIndexingAnalyzers()
############################# Writer Configurattion #####################################
map = HashMap()
map.put('name', self.mAnalyzers['name'])
map.put('parent', self.mAnalyzers['parent'])
map.put('content', self.mAnalyzers['default'])
map.put('id', self.mAnalyzers['id'])
analyzerWrapper = PerFieldAnalyzerWrapper(self.mAnalyzers['default'], map)
self.mWriterConfig = IndexWriterConfig(Version.LUCENE_CURRENT, analyzerWrapper)
self.mWriterConfig.setOpenMode(settings.ADMINS_ENGINE.mOpenMode)
if settings.ADMINS_ENGINE.mSimilarity != None:
self.mWriterConfig.setSimilarity(settings.ADMINS_ENGINE.mSimilarity)
########################################################################################
directory = SimpleFSDirectory(File(self.mIndexDirectory))
self.mIndexWriter = IndexWriter(directory, self.mWriterConfig)
############################# FieldType Prepration #####################
nameField = FieldType()
nameField.setIndexed(True)
nameField.setStored(True)
nameField.setTokenized(True)
nameField.setIndexOptions(FieldInfo.IndexOptions.DOCS_ONLY)
parentField = FieldType()
parentField.setIndexed(True)
parentField.setStored(True)
parentField.setTokenized(True)
parentField.setIndexOptions(FieldInfo.IndexOptions.DOCS_ONLY)
contentField = FieldType()
contentField.setIndexed(True)
contentField.setStored(True)
contentField.setTokenized(True)
contentField.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS)
idField = FieldType()
idField.setIndexed(True)
idField.setStored(True)
idField.setTokenized(False)
idField.setIndexOptions(FieldInfo.IndexOptions.DOCS_ONLY)
self.mFieldTypes = {
'name' : nameField,
'parent' : parentField,
'content' : contentField,
'id' : idField
}
#######################################################################
self.mLog = ""
def index_article(writer, art_id, art_body):
art_id_field = FieldType()
art_id_field.setIndexed(True)
art_id_field.setStored(True)
art_id_field.setTokenized(False)
art_id_field.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
art_body_field = FieldType()
art_body_field.setIndexed(True)
art_body_field.setStored(True)
art_body_field.setTokenized(True)
art_body_field.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
doc = Document()
doc.add(Field("art_id", str(art_id), art_id_field))
doc.add(Field("art_body", art_body, art_body_field))
writer.addDocument(doc)
from org.apache.lucene.analysis.miscellaneous import LimitTokenCountAnalyzer
from org.apache.lucene.analysis.standard import StandardAnalyzer
from org.apache.lucene.store import RAMDirectory
from org.apache.lucene.document import Document, Field, FieldType
from org.apache.lucene.util import BytesRef, BytesRefIterator
from org.apache.lucene.index import \
IndexWriterConfig, IndexWriter, DirectoryReader, IndexOptions
if __name__ == '__main__':
lucene.initVM(vmargs=['-Djava.awt.headless=true'])
directory = RAMDirectory()
iconfig = IndexWriterConfig(LimitTokenCountAnalyzer(StandardAnalyzer(), 100))
iwriter = IndexWriter(directory, iconfig)
ft = FieldType()
ft.setStored(True)
ft.setTokenized(True)
ft.setStoreTermVectors(True)
ft.setStoreTermVectorOffsets(True)
ft.setStoreTermVectorPositions(True)
ft.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS)
ts = ["this bernhard is the text to be index text",
"this claudia is the text to be indexed"]
for t in ts:
doc = Document()
doc.add(Field("fieldname", t, ft))
iwriter.addDocument(doc)
iwriter.commit()
def _getIndex(self, even, odd):
mergePolicy = LogDocMergePolicy()
mergePolicy.setMergeFactor(1000)
directory = RAMDirectory()
self.dirs.append(directory)
writer = self.getWriter(directory=directory,
analyzer=SimpleAnalyzer(Version.LUCENE_CURRENT),
maxBufferedDocs=2, mergePolicy=mergePolicy)
if self.dvStringSorted:
# Index sorted
stringDVType = FieldInfo.DocValuesType.SORTED
elif self.notSorted:
# Index non-sorted
stringDVType = FieldInfo.DocValuesType.BINARY
else:
# sorted anyway
stringDVType = FieldInfo.DocValuesType.SORTED
ft1 = FieldType()
ft1.setStored(True)
ft2 = FieldType()
ft2.setIndexed(True)
for i in xrange(len(self.data)):
if (i % 2 == 0 and even) or (i % 2 == 1 and odd):
doc = Document()
doc.add(Field("tracer", self.data[i][0], ft1))
doc.add(TextField("contents", self.data[i][1], Field.Store.NO))
if self.data[i][2] is not None:
doc.add(StringField("int", self.data[i][2], Field.Store.NO))
if self.supportsDocValues:
doc.add(NumericDocValuesField("int_dv", Long.parseLong(self.data[i][2])))
if self.data[i][3] is not None:
doc.add(StringField("float", self.data[i][3], Field.Store.NO))
if self.supportsDocValues:
doc.add(FloatDocValuesField("float_dv", Float.parseFloat(self.data[i][3])))
if self.data[i][4] is not None:
doc.add(StringField("string", self.data[i][4], Field.Store.NO))
if self.supportsDocValues:
if stringDVType == FieldInfo.DocValuesType.SORTED:
doc.add(SortedDocValuesField("string_dv", BytesRef(self.data[i][4])))
elif stringDVType == FieldInfo.DocValuesType.BINARY:
doc.add(BinaryDocValuesField("string_dv", BytesRef(self.data[i][4])))
else:
raise ValueError("unknown type " + stringDVType)
if self.data[i][5] is not None:
doc.add(StringField("custom", self.data[i][5], Field.Store.NO))
if self.data[i][6] is not None:
doc.add(StringField("i18n", self.data[i][6], Field.Store.NO))
if self.data[i][7] is not None:
doc.add(StringField("long", self.data[i][7], Field.Store.NO))
if self.data[i][8] is not None:
doc.add(StringField("double", self.data[i][8], Field.Store.NO))
if self.supportsDocValues:
doc.add(NumericDocValuesField("double_dv", Double.doubleToRawLongBits(Double.parseDouble(self.data[i][8]))))
if self.data[i][9] is not None:
doc.add(StringField("short", self.data[i][9], Field.Store.NO))
if self.data[i][10] is not None:
doc.add(StringField("byte", self.data[i][10], Field.Store.NO))
if self.data[i][11] is not None:
doc.add(StringField("parser", self.data[i][11], Field.Store.NO))
for f in doc.getFields():
if f.fieldType().indexed() and not f.fieldType().omitNorms():
Field.cast_(f).setBoost(2.0)
writer.addDocument(doc)
reader = writer.getReader()
writer.close()
return self.getSearcher(reader=reader)
def indexTable(self, writer):
#connection
con = None
#define the index of all the fields
#---------step 2：connect to mysql----------
con = mdb.connect('localhost','root','testgce','douban_movie_v3')
#t_num = FieldType.NumericType it is wrong!!
t_num = FieldType()
t_num.setStored(False)
t1 = FieldType()
t1.setIndexed(True)
t1.setStored(True)
t1.setTokenized(False)
t1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
t2 = FieldType()
t2.setIndexed(True)
t2.setStored(False)
t2.setTokenized(True)
t2.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
t3 = FieldType()
t3.setIndexed(True)
t3.setStored(True)
t3.setTokenized(True)
t3.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
maxDict = utils.maxDict
#加权数值范围
base = DOC_BOOST_RANGE[0]
upper = DOC_BOOST_RANGE[1]
with con:
# Careful with codecs
con.set_character_set('utf8')
cur = con.cursor()
# Aagin the codecs
cur.execute('SET NAMES utf8;')
cur.execute('SET CHARACTER SET utf8;')
cur.execute('SET character_set_connection=utf8;')
#------step 3： choose the right table------
cur.execute("SELECT * FROM movie_items")
numrows = int(cur.rowcount)
print 'numrows:',numrows
for i in range(numrows):
print
row = cur.fetchone()
#------step 4：Index your field------
summary = row[SUMMARY]
subject_id = row[SUBJECT_ID]
print 'id'+subject_id
year = utils.formatYear(row[YEAR])
try:
date = DateTools.stringToDate(year.replace('-',' '))
wtfFile = open('wtf.txt','a')
dateStr = DateTools.dateToString(date,DateTools.Resolution.DAY)
except:
wtfFile.write(year+'\n')
doc = Document()
#boosting
boostProb = utils.calcBoostProb(row,maxDict,dateStr)
boost = base + boostProb*(upper-base)
doc.add(FloatField("boost",boost,Field.Store.YES))
doc.add(StringField("year",dateStr,Field.Store.YES))
print 'dateStr:'+dateStr
#A text field is a sequence of terms that has been tokenized while a string field is a single term (although it can also be multivalued.)
do_count = row[DO_COUNT] if row[DO_COUNT] != None else 0
wish_count = row[COLLECT_COUNT] if row[WISH_COUNT] != None else 0
#fields which should not be analyzed
doc.add(FloatField("rating_average",float(row[RATING_AVERAGE]),Field.Store.YES))
doc.add(FloatField("rating_stars", float(row[RATING_STARS]), Field.Store.YES))
doc.add(IntField("reviews_count", int(row[REVIEWS_COUNT]), Field.Store.YES))
#doc.add(FloatField("year", float(row[YEAR]), Field.Store.YES).setBoost(boost))
doc.add(IntField("collect_count", int(row[COLLECT_COUNT]), Field.Store.YES))
doc.add(IntField("do_count", int(do_count), Field.Store.YES))
doc.add(IntField("wish_count", int(wish_count), Field.Store.YES))
doc.add(IntField("subject_id", int(row[SUBJECT_ID]), Field.Store.YES))
doc.add(IntField("comments_count", int(row[COMMENTS_COUNT]), Field.Store.YES))
doc.add(IntField("ratings_count", int(row[RATINGS_COUNT]), Field.Store.YES))
doc.add(StringField("image_small", row[IMAGE_SMALL], Field.Store.YES))
#fields which should be analyzed with WhitespaceAnalyzer
#attention!!! dont use a long sentence like :
#doc.add(Field("genres", row[GENRES].replace(delim,' '), t3).setBoost(boost))
#or you'll get a null pointer error
f = Field("countries", row[COUNTRIES].replace(delim,' '), t3)
f.setBoost(boost)
doc.add(f)
#process casts
raw_casts = row[CASTS].replace(delim,' ')
f = Field("raw_casts", raw_casts , t1)
f.setBoost(boost)
doc.add(f)
#将英文人名中的 ·
raw_casts = raw_casts.replace('·',' ')
if len(raw_casts.split(' '))<CASTS_LEN:
#平局人名长度是4
casts = raw_casts + ' ￥￥￥￥'*(CASTS_LEN-len(raw_casts.split(' ')))
f = Field("casts", casts , t3)
f.setBoost(boost)
doc.add(f)
#process directors
raw_directors = row[DIRECTORS].replace(delim,' ')
f = Field("raw_directors",raw_directors, t1)
f.setBoost(boost)
doc.add(f)
#将英文人名中的 · 替换
raw_directors = raw_directors.replace('·',' ')
if len(raw_directors.split(' '))<DIRECTORS_LEN:
#平局人名长度是4
directors = raw_directors + ' ￥￥￥￥'*(DIRECTORS_LEN-len(raw_directors.split(' ')))
f = Field("directors", directors, t3)
f.setBoost(boost)
doc.add(f)
Field("genres", row[GENRES].replace(delim,' '), t3)
f.setBoost(boost)
doc.add(f)
Field("subtype", row[SUBTYPE].replace(delim,' '), t3)
f.setBoost(boost)
doc.add(f)
#it is wrong cause indexable field has no method setBoost
# fieldList = doc.getFields() # is not a python 'list' , but a 'List' which is unindexable
# for eachField in fieldList:
# eachField.setBoost(boost)
#user_tags 原始字符串要存，reRank要用：
doc.add(StringField("raw_user_tags",row[USER_TAGS],Field.Store.YES))
doc.add(StringField("raw_others_like",row[OTHERS_LIKE],Field.Store.YES))
user_tags_str = ''
others_like_str = ''
tags_len = 0
if row[USER_TAGS]!='':
user_tags_list = row[USER_TAGS].split(delim)
for tag_pair in user_tags_list:
if tag_pair!='':#字符串的最后一个字符是￥，这样split之后最后一个元素是空字符
#print 'tag_pair'+tag_pair+'hhe'
tag_name = tag_pair.split(delim_uo)[0]+' ' # dont forget this space !!
tag_num = tag_pair.split(delim_uo)[1]
tag_num_processed = int(int(tag_num)/TAG_SPAN)+1 #最小为1
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
user_tags_str = user_tags_str +' '+ tag_name * tag_num_processed
tags_len = tags_len + tag_num_processed #最后得到总共词的个数
if tags_len<TAGS_AVER_LEN:
#填充tags，目测3是平均长度,所以使用 ￥￥￥
user_tags_str = user_tags_str +' ￥￥￥'*(TAGS_AVER_LEN - tags_len)
#
if row[OTHERS_LIKE]!='':
for like_pair in row[OTHERS_LIKE].split(delim):
if like_pair!='':
others_like_str = others_like_str +' '+like_pair.split(delim_uo)[1]
#start process adjs
if row[ADJS] != None:
raw_adjs = row[ADJS][:-1]
adjs_str = ''
adjs_len = 0
if row[ADJS] != '' and row[ADJS] != '\n':
#'重要=4.0,特殊=4.0'
adjs_str = row[ADJS]
adjs_list = adjs_str.split(',')
for adj_pair in adjs_list:
#print 'adj_pair:'+adj_pair+'hhe'
adj_name = adj_pair.split('=')[0]
adj_num = adj_pair.split('=')[1]
#去换行符,转换int
if adj_num[-1] == '\n':
adj_num = adj_num[0:-1]
adj_num = int(float(adj_num))
add_adj=''
# #同义词
# adj_name_bro = searchDictValue(adjMap,adj_name)
# if adj_name_bro == -1: #表示没有结果，即未找到近义词，不添加
# add_adj = ''
# else:
# add_adj = (adj_name_bro+' ')*adj_num
# raw_adjs = raw_adjs + ',' + adj_name_bro+'='+str(adj_num)
adjs_str = adjs_str + ' ' + (adj_name+' ') * adj_num +add_adj
adjs_len = adjs_len + adj_num #最后得到总共tags的个数
#print raw_adjs
doc.add(StringField("raw_adjs",raw_adjs,Field.Store.YES))
if adjs_len<ADJS_AVER_LEN:
#填充 adjs_str，目测2是平均长度,所以使用 "￥￥"
adjs_str = adjs_str +' ￥￥'*(ADJS_AVER_LEN - adjs_len)
f = Field("adjs", adjs_str, t3)
f.setBoost(boost)
doc.add(f)
f = Field("user_tags", user_tags_str, t3)
f.setBoost(boost)
doc.add(f)
f = Field("others_like", others_like_str, t3)
f.setBoost(boost)
doc.add(f)
#fields which should be analyzed with good analyzer
f = Field("title", row[TITLE], t3)
f.setBoost(boost)
doc.add(f)
f = Field("original_title", row[ORIGINAL_TITLE], t3)
f.setBoost(boost)
doc.add(f)
f = Field("summary_segmentation", row[SUMMARY_SEGMENTATION], t3)
f.setBoost(boost)
doc.add(f)
f = Field("aka", row[AKA], t2)
f.setBoost(boost)
doc.add(f)
if len(summary) > 0:
print subject_id +'--->'+':\n '+ row[TITLE]
try:
summary_unicoded = unicode(summary, 'utf-8') #test the encoding
except Exception,e:
print "Decode Failed: ", e
f = Field('summary', summary, t2)
f.setBoost(boost)
doc.add(f)
else:
print "warning:\n" + subject_id +'---> No content!'
print 'boosting:' + str(boost)
#for debug
if boost>upper:
print boostProb
print maxDict
exit(0)
writer.addDocument(doc)
def indexDocs(self, root, writer):
t1 = FieldType()
t1.setIndexed(True)
t1.setStored(True)
t1.setTokenized(True)
t1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
t2 = FieldType()
t2.setIndexed(True)
t2.setStored(False)
t2.setTokenized(True)
t2.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
for root, dirnames, filenames in os.walk(root):
# traverse through the doc directory
for filename in filenames:
# if not filename.endswith('.cdc'):
# continue
try:
# only add the filename and path for indexing
path = os.path.join(root, filename)
print "adding file : ", path
file = open(path)
contents = unicode(file.read(), 'utf-8')
file.close()
doc = Document()
doc.add(Field("name", filename, t1))
doc.add(Field("path", root, t1))
if len(contents) > 0:
doc.add(Field("contents", contents, t2))
else:
print "warning: no content in ", filename
writer.addDocument(doc)
except Exception, e:
print "failed in indexDocs:", e
former_scents = ' '.join(scents_list[0])
doc.add(Field("former_scents", former_scents, t3))
mid_scents = ' '.join(scents_list[1])
doc.add(Field("mid_scents", mid_scents, t3))
last_scents = ' '.join(scents_list[2])
doc.add(Field("last_scents", last_scents, t3))
except Exception, e:
print "Failed in indexDocs:", e
print line
print cnt
cnt += 1
writer.addDocument(doc)
except Exception, e:
print "Failed in indexDocs:", e
t1 = FieldType()
t1.setIndexed(True)
t1.setStored(True)
t1.setTokenized(False)
t1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
t3 = FieldType()
t3.setIndexed(True)
t3.setStored(True)
t3.setTokenized(True)
t3.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
t2 = FieldType()
t2.setIndexed(True)
t2.setStored(False)
t2.setTokenized(False)
def indexDocs(self, root, writer):
# t1 is used for filenames and t2 is used for contents
t1 = FieldType()
t1.setIndexed(True)
t1.setStored(True)
t1.setTokenized(False)
t1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
t2 = FieldType()
t2.setIndexed(True)
t2.setStored(False)
t2.setTokenized(True)
t2.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
for root, dirnames, filenames in os.walk(root):
for filename in filenames:
# We can index only a certain types of files
if not (filename.endswith('.txt') or filename.endswith('.pdf') or filename.endswith('.xml') or filename.endswith('.doc') or filename.endswith('.odt')):
continue
try:
file_path = os.path.join(root, filename)
outfile_path = file_path
# First convert PDF and DOC files to text
if filename.endswith('.pdf'):
outfile = filename.replace('.pdf', '.txt')
outfile_path = os.path.join(root, outfile)
cmd = 'pdftotext ' + '-layout ' + "'"+ file_path + "'" + ' ' + "'" + outfile_path + "'"
subprocess.check_output(cmd, shell=True)
file_path = outfile_path
elif filename.endswith('.doc'):
outfile = filename.replace('.doc', '.txt')
outfile_path = os.path.join(root, outfile)
cmd = 'antiword ' + file_path + ' >> ' + outfile_path
subprocess.check_output(cmd, shell=True)
file_path = outfile_path
elif filename.endswith('.odt'):
outfile = filename.replace('.odt', '.txt')
outfile_path = os.path.join(root, outfile)
cmd = 'odttotext ' + '-layout ' + "'"+ file_path + "'" + ' ' + "'" + outfile_path + "'"
subprocess.check_output(cmd, shell=True)
file_path = outfile_path
file = open(file_path)
contents = unicode(file.read(), 'iso-8859-1')
file.close()
doc = Document()
doc.add(Field("name", filename, t1))
doc.add(Field("path", root, t1))
if len(contents) > 0:
doc.add(Field("contents", contents, t2))
else:
logging.debug('warning: no content in %s', filename)
writer.addDocument(doc)
except Exception, e:
logging.debug('Failed in indexDocs: %s', e)
class Indexer(object):
def __init__(self, **kwargs):
""" Initialize a new instance of the Indexer
:param output: The output directory of the underlying index
:param anaylzer: The overloaded analyzer to work with
"""
self.output = kwargs.get("root", "index")
if not os.path.exists(self.output):
os.mkdir(self.output)
self.analyzer = kwargs.get("analyzer", StandardAnalyzer(Version.LUCENE_CURRENT))
self.analyzer = LimitTokenCountAnalyzer(self.analyzer, 1048576)
self.config = IndexWriterConfig(Version.LUCENE_CURRENT, self.analyzer)
self.config.setOpenMode(IndexWriterConfig.OpenMode.CREATE)
self.store = SimpleFSDirectory(File(self.output))
self.writer = IndexWriter(self.store, self.config)
self.create_field_types()
def index(self, document):
""" Given a new document, add it to the index.
:param document: The document to add to the indexer
"""
try:
self.writer.addDocument(document)
except Exception:
logger.exception("Failed to index the supplied document")
def shutdown(self):
""" Shutdown the currently processing indexer.
"""
try:
# self.writer.optimize()
self.writer.close()
except Exception:
logger.exception("Failed to shutdown the indexer correctly")
def create_field_types(self):
""" Create the field types that will be used to specify
what actions lucene should take on the various fields
supplied to index.
"""
self.field_clean = FieldType()
self.field_clean.setIndexed(True)
self.field_clean.setStored(True)
self.field_clean.setTokenized(False)
self.field_clean.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
self.field_dirty = FieldType()
self.field_dirty.setIndexed(True)
self.field_dirty.setStored(False)
self.field_dirty.setTokenized(True)
self.field_dirty.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
def indexDocs(self, root, writer):
for root, dirnames, filenames in os.walk(root):
for filename in filenames:
print "adding", filename
doc_parser = HTMLDocumentParser()
try:
path = os.path.join(root, filename)
file = open(path)
contents = unicode(file.read(), 'iso-8859-1')
doc_parser.feed(contents)
contents = doc_parser.contents
html_doc = HTMLDocument(contents)
flag = False
if flag:
print '=============='
print 'Title: ' + html_doc.title
print 'Description: ' + html_doc.description
print 'Month: ' + html_doc.month
print 'Year: ' + html_doc.year
print 'Authors: ' + str(html_doc.authors)
print 'Keywords: ' + str(html_doc.keywords)
print 'Timestamp: ' + str(html_doc.timestamp)
print ' '
file.close()
doc = Document()
field_filename = FieldType()
field_filename.setIndexed(True)
field_filename.setStored(True)
field_filename.setTokenized(False)
field_filename.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
doc.add(Field("filename", filename.replace('.html', ''), field_filename))
field_path = FieldType()
field_path.setIndexed(True)
field_path.setStored(True)
field_path.setTokenized(True)
field_path.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
doc.add(Field("path", root, field_path))
field_title = FieldType()
field_title.setIndexed(True)
field_title.setStored(True)
field_title.setTokenized(True)
field_title.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
doc.add(Field("title", html_doc.title, field_title))
field_description = FieldType()
if html_doc.has_description():
field_description.setIndexed(True)
else:
field_description.setIndexed(True)
field_description.setStored(True)
field_description.setTokenized(True)
field_description.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
doc.add(Field("description", html_doc.description, field_description))
field_month = FieldType()
field_month.setIndexed(True)
field_month.setStored(True)
field_month.setTokenized(False)
field_month.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
doc.add(Field("month", html_doc.month, field_month))
field_year = FieldType()
field_year.setIndexed(True)
field_year.setStored(True)
field_year.setTokenized(False)
field_year.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
doc.add(Field("year", html_doc.year, field_year))
if html_doc.has_authors():
field_author = FieldType()
field_author.setIndexed(True)
field_author.setStored(True)
field_author.setTokenized(True)
field_author.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
for author in html_doc.authors:
doc.add(Field("author", author, field_author))
if html_doc.has_keywords():
field_keyword = FieldType()
field_keyword.setIndexed(True)
field_keyword.setStored(True)
field_keyword.setTokenized(True)
field_keyword.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
for keyword in html_doc.keywords:
doc.add(Field("keyword", keyword, field_keyword))
field_timestamp = FieldType()
field_timestamp.setIndexed(False)
field_timestamp.setStored(True)
field_timestamp.setTokenized(False)
field_timestamp.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
doc.add(Field("timestamp", html_doc.timestamp, field_timestamp))
if len(contents) > 0:
field_source = FieldType()
field_source.setIndexed(True)
field_source.setStored(True)
field_source.setTokenized(True)
field_source.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
doc.add(Field("contents", contents, field_source))
else:
print "warning: no content in %s" % filename
writer.addDocument(doc)
except Exception, e:
print "Failed in indexDocs:", e
def indexDocs(root, writer):
"""
indexed: name title content
stored: date name tilte sumary
:param root:
:param writer:
:return:
"""
#index and store
t1 = FieldType()
t1.setIndexed(True)
t1.setStored(True)
t1.setTokenized(False)
t1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
#only index, but not store
t2 = FieldType()
t2.setIndexed(True)
t2.setStored(False)
t2.setTokenized(True)
t2.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
#only store
t3 = FieldType()
t3.setIndexed(False)
t3.setStored(True)
t3.setTokenized(False)
t3.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
for root, dirnames, filenames in os.walk(root):
print filenames
for filename in filenames:
if not filename.endswith('.md'):
continue
print "adding", filename
try:
path = os.path.join(root, filename)
file = open(path)
contents = unicode(file.read(), 'utf-8')
file.close()
date, name = get_date_name(filename)
title, content = get_post_title_content(contents)
summary = content[:200] if content else ''
print date, name, title
doc = Document()
doc.add(Field('date', date, t3))
doc.add(Field('name', name, t1))
doc.add(Field('title', title, t1))
doc.add(Field('content', content, t2))
doc.add(Field('summary', summary, t3))
# doc.add(Field("name", filename, t1))
# doc.add(Field("path", root, t1))
# if len(contents) > 0:
# doc.add(Field("contents", contents, t2))
# else:
# print "warning: no content in %s" % filename
writer.addDocument(doc)
except Exception, e:
print "Failed in indexDocs:", e
def indexDocs(self, url, writer):
type1 = FieldType()
type1.setIndexed(True)
type1.setStored(True)
type1.setTokenized(False)
type1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
type2 = FieldType()
type2.setIndexed(True)
type2.setStored(True)
type2.setTokenized(True)
type2.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS)
# Read Feeds
feeds = feedparser.parse(url)
for item in feeds["entries"]:
print "adding", item["title"]
try:
link = item["link"]
contents = item["description"].encode("utf-8")
contents = re.sub('<[^<]+?>', '', ''.join(contents))
title = item["title"]
doc = Document()
doc.add(Field("url", link, type1))
doc.add(Field("title", title, type1))
if len(contents) > 0:
doc.add(Field("contents", contents, type2))
else:
print "warning: no content in %s" % item["title"]
writer.addDocument(doc)
except Exception, e:
print "Failed in indexDocs:", e
def index_wiki(wiki_xmlfile, index_directory_name):
# Initialize index directory and analyzer.
version = Version.LUCENE_CURRENT
store = FSDirectory.open(File(index_directory_name))
analyzer = StandardAnalyzer(version)
# Creates config file.
config = IndexWriterConfig(version, analyzer)
config.setOpenMode(IndexWriterConfig.OpenMode.CREATE)
writer = IndexWriter(store, config)
# Set document content field type.
content_fieldtype = FieldType()
content_fieldtype.setIndexed(True)
content_fieldtype.setStored(True)
content_fieldtype.setTokenized(True)
content_fieldtype.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
# Set document title field type.
title_fieldtype = FieldType()
title_fieldtype.setIndexed(True)
title_fieldtype.setStored(True)
title_fieldtype.setTokenized(True)
title_fieldtype.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
# Set document url field type.
url_fieldtype = FieldType()
url_fieldtype.setIndexed(True)
url_fieldtype.setStored(True)
url_fieldtype.setTokenized(False)
url_fieldtype.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
for xmldoc in wikicorpusxml((wiki_xmlfile)):
content = xmldoc.partition('>')[2].partition('<')[0].strip()
title = xmldoc.partition(' title="')[2].partition('"')[0].strip()
url = xmldoc.partition(' url="')[2].partition('"')[0].strip()
doc = Document()
doc.add(Field("contents", content, content_fieldtype))
doc.add(Field("title", title, title_fieldtype))
doc.add(Field("url", url, url_fieldtype))
writer.addDocument(doc)
writer.commit()
writer.close()
def indexDocs(self, root, writer):
t1 = FieldType()
t1.setIndexed(True)
t1.setStored(True)
t1.setTokenized(False)
t1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
t2 = FieldType()
t2.setIndexed(True)
t2.setStored(False)
t2.setTokenized(True)
t2.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
for root, dirnames, filenames in os.walk(root):
for filename in filenames:
if not filename.endswith('.txt'):
continue
print "adding", filename
try:
path = os.path.join(root, filename)
file = open(path)
contents = unicode(file.read(), 'gbk')
file.close()
doc = Document()
doc.add(Field("name", filename, t1))
doc.add(Field("path", root, t1))
if len(contents) > 0:
doc.add(Field("contents", contents, t2))
else:
print "warning: no content in %s" % filename
writer.addDocument(doc)
except Exception, e:
print "Failed in indexDocs:", e
class LuceneDocumentField(object):
"""Internal handler class for possible field types"""
def __init__(self):
"""Init possible field types"""
# FIELD_ID: stored, indexed, non-tokenized
self.field_id = FieldType()
self.field_id.setIndexed(True)
self.field_id.setStored(True)
self.field_id.setTokenized(False)
# FIELD_ID_TV: stored, indexed, not tokenized, with term vectors (without positions)
# for storing IDs with term vector info
self.field_id_tv = FieldType()
self.field_id_tv.setIndexed(True)
self.field_id_tv.setStored(True)
self.field_id_tv.setTokenized(False)
self.field_id_tv.setStoreTermVectors(True)
# FIELD_TEXT: stored, indexed, tokenized, with positions
self.field_text = FieldType()
self.field_text.setIndexed(True)
self.field_text.setStored(True)
self.field_text.setTokenized(True)
# FIELD_TEXT_TV: stored, indexed, tokenized, with term vectors (without positions)
self.field_text_tv = FieldType()
self.field_text_tv.setIndexed(True)
self.field_text_tv.setStored(True)
self.field_text_tv.setTokenized(True)
self.field_text_tv.setStoreTermVectors(True)
# FIELD_TEXT_TVP: stored, indexed, tokenized, with term vectors and positions
# (but no character offsets)
self.field_text_tvp = FieldType()
self.field_text_tvp.setIndexed(True)
self.field_text_tvp.setStored(True)
self.field_text_tvp.setTokenized(True)
self.field_text_tvp.setStoreTermVectors(True)
self.field_text_tvp.setStoreTermVectorPositions(True)
def get_field(self, type):
"""Get Lucene FieldType object for the corresponding internal FIELDTYPE_ value"""
if type == Lucene.FIELDTYPE_ID:
return self.field_id
elif type == Lucene.FIELDTYPE_ID_TV:
return self.field_id_tv
elif type == Lucene.FIELDTYPE_TEXT:
return self.field_text
elif type == Lucene.FIELDTYPE_TEXT_TV:
return self.field_text_tv
elif type == Lucene.FIELDTYPE_TEXT_TVP:
return self.field_text_tvp
else:
raise Exception("Unknown field type")
class LuceneSearch():
def __init__(self):
self.env = lucene.initVM(initialheap='28g',
maxheap='28g',
vmargs=['-Djava.awt.headless=true'])
self.vocab = None
BooleanQuery.setMaxClauseCount(2048)
if not os.path.exists(prm.index_folder):
print 'Creating index at', prm.index_folder
if prm.docs_path == prm.docs_path_term:
add_terms = True
else:
add_terms = False
self.create_index(prm.index_folder, prm.docs_path, add_terms)
if prm.local_index_folder:
print 'copying index from', prm.index_folder, 'to', prm.local_index_folder
if os.path.exists(prm.local_index_folder):
print 'Folder', prm.local_index_folder, 'already exists! Doing nothing.'
else:
shutil.copytree(prm.index_folder, prm.local_index_folder)
self.index_folder = prm.local_index_folder
else:
self.index_folder = prm.index_folder
fsDir = MMapDirectory(Paths.get(prm.index_folder))
self.searcher = IndexSearcher(DirectoryReader.open(fsDir))
if prm.docs_path != prm.docs_path_term:
if not os.path.exists(prm.index_folder_term):
print 'Creating index at', prm.index_folder_term
self.create_index(prm.index_folder_term,
prm.docs_path_term,
add_terms=True)
if prm.local_index_folder_term:
print 'copying index from', prm.index_folder_term, 'to', prm.local_index_folder_term
if os.path.exists(prm.local_index_folder_term):
print 'Folder', prm.local_index_folder_term, 'already exists! Doing nothing.'
else:
shutil.copytree(prm.index_folder_term,
prm.local_index_folder_term)
self.index_folder_term = prm.local_index_folder_term
else:
self.index_folder_term = prm.index_folder_term
fsDir_term = MMapDirectory(Paths.get(prm.index_folder_term))
self.searcher_term = IndexSearcher(
DirectoryReader.open(fsDir_term))
self.analyzer = StandardAnalyzer()
self.pool = ThreadPool(processes=prm.n_threads)
self.cache = {}
print 'Loading Title-ID mapping...'
self.title_id_map, self.id_title_map = self.get_title_id_map()
def get_title_id_map(self):
# get number of docs
n_docs = self.searcher.getIndexReader().numDocs()
title_id = {}
id_title = {}
query = MatchAllDocsQuery()
hits = self.searcher.search(query, n_docs)
for hit in hits.scoreDocs:
doc = self.searcher.doc(hit.doc)
idd = int(doc['id'])
title = doc['title']
title_id[title] = idd
id_title[idd] = title
return title_id, id_title
def add_doc(self, doc_id, title, txt, add_terms):
doc = Document()
txt = utils.clean(txt)
if add_terms:
txt_ = txt.lower()
words_idx, words = utils.text2idx2([txt_], self.vocab,
prm.max_terms_per_doc)
words_idx = words_idx[0]
words = words[0]
doc.add(Field("id", str(doc_id), self.t1))
doc.add(Field("title", title, self.t1))
doc.add(Field("text", txt, self.t2))
if add_terms:
doc.add(Field("word_idx", ' '.join(map(str, words_idx)), self.t3))
doc.add(Field("word", '<&>'.join(words), self.t3))
self.writer.addDocument(doc)
def create_index(self, index_folder, docs_path, add_terms=False):
print 'Loading Vocab...'
if not self.vocab:
self.vocab = utils.load_vocab(prm.vocab_path, prm.n_words)
os.mkdir(index_folder)
self.t1 = FieldType()
self.t1.setStored(True)
self.t1.setIndexOptions(IndexOptions.DOCS)
self.t2 = FieldType()
self.t2.setStored(False)
self.t2.setIndexOptions(IndexOptions.DOCS_AND_FREQS)
self.t3 = FieldType()
self.t3.setStored(True)
self.t3.setIndexOptions(IndexOptions.NONE)
fsDir = MMapDirectory(Paths.get(index_folder))
writerConfig = IndexWriterConfig(StandardAnalyzer())
self.writer = IndexWriter(fsDir, writerConfig)
print "%d docs in index" % self.writer.numDocs()
print "Indexing documents..."
doc_id = 0
import corpus_hdf5
corpus = corpus_hdf5.CorpusHDF5(docs_path)
for txt in corpus.get_text_iter():
title = corpus.get_article_title(doc_id)
self.add_doc(doc_id, title, txt, add_terms)
if doc_id % 1000 == 0:
print 'indexing doc', doc_id
doc_id += 1
print "Index of %d docs..." % self.writer.numDocs()
self.writer.close()
def search_multithread(self, qs, max_cand, max_full_cand, searcher):
self.max_cand = max_cand
self.max_full_cand = max_full_cand
self.curr_searcher = searcher
out = self.pool.map(self.search_multithread_part, qs)
return out
def search_multithread_part(self, q):
if not self.env.isCurrentThreadAttached():
self.env.attachCurrentThread()
if q in self.cache:
return self.cache[q]
else:
try:
q = q.replace('AND',
'\\AND').replace('OR',
'\\OR').replace('NOT', '\\NOT')
query = QueryParser("text",
self.analyzer).parse(QueryParser.escape(q))
except:
print 'Unexpected error when processing query:', str(q)
print 'Using query "dummy".'
q = 'dummy'
query = QueryParser("text",
self.analyzer).parse(QueryParser.escape(q))
c = OrderedDict()
hits = self.curr_searcher.search(query, self.max_cand)
for i, hit in enumerate(hits.scoreDocs):
doc = self.curr_searcher.doc(hit.doc)
if i < self.max_full_cand:
word_idx = map(int, doc['word_idx'].split(' '))
word = doc['word'].split('<&>')
else:
word_idx = []
word = []
c[int(doc['id'])] = [word_idx, word]
return c
def search_singlethread(self, qs, max_cand, max_full_cand, curr_searcher):
out = []
for q in qs:
if q in self.cache:
out.append(self.cache[q])
else:
try:
q = q.replace('AND', '\\AND').replace('OR',
'\\OR').replace(
'NOT', '\\NOT')
query = QueryParser("text", self.analyzer).parse(
QueryParser.escape(q))
except:
print 'Unexpected error when processing query:', str(q)
print 'Using query "dummy".'
query = QueryParser("text", self.analyzer).parse(
QueryParser.escape('dummy'))
c = OrderedDict()
hits = curr_searcher.search(query, max_cand)
for i, hit in enumerate(hits.scoreDocs):
doc = curr_searcher.doc(hit.doc)
if i < max_full_cand:
word_idx = map(int, doc['word_idx'].split(' '))
word = doc['word'].split('<&>')
else:
word_idx = []
word = []
c[int(doc['id'])] = [word_idx, word]
out.append(c)
return out
def get_candidates(self,
qs,
max_cand,
max_full_cand=None,
save_cache=False,
extra_terms=True):
if not max_full_cand:
max_full_cand = max_cand
if prm.docs_path != prm.docs_path_term:
max_cand2 = 0
else:
max_cand2 = max_full_cand
if prm.n_threads > 1:
out = self.search_multithread(qs, max_cand, max_cand2,
self.searcher)
if (prm.docs_path != prm.docs_path_term) and extra_terms:
terms = self.search_multithread(qs, max_full_cand,
max_full_cand,
self.searcher_term)
else:
out = self.search_singlethread(qs, max_cand, max_cand2,
self.searcher)
if (prm.docs_path != prm.docs_path_term) and extra_terms:
terms = self.search_singlethread(qs, max_full_cand,
max_full_cand,
self.searcher_term)
if (prm.docs_path != prm.docs_path_term) and extra_terms:
for outt, termss in itertools.izip(out, terms):
for cand_id, term in itertools.izip(
outt.keys()[:max_full_cand], termss.values()):
outt[cand_id] = term
if save_cache:
for q, c in itertools.izip(qs, out):
if q not in self.cache:
self.cache[q] = c
return out
def indexDocs(self,root,writer):
t1 = FieldType()
t1.setIndexed(True)
t1.setStored(True)
t1.setTokenized(True)
t1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
t2 = FieldType()
t2.setIndexed(True)
t2.setStored(False)
t2.setTokenized(True)
t2.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
for root, dirnames,filenames in os.walk(root):
# traverse through the doc directory
for filename in filenames:
# only if this file ends with '.c'
if not filename.endswith('.c'):
continue
try:
# only add the filename and path for indexing
path = os.path.join(root,filename)
print "adding file : ",path
file = open(path)
contents = unicode(file.read(),'utf-8')
file.close()
doc = Document()
doc.add(Field("name",filename,t1))
doc.add(Field("path",root,t1))
# if len(contents) > 0:
# doc.add(Field("contents",contents,t2))
# else:
# print "warning: no content in ",filename
writer.addDocument(doc)
except Exception,e:
print "failed in indexDocs:",e
def indexTable(self, writer):
#connection
con = None
#define the index of all the fields
#---------step 2----------
con = mdb.connect('localhost','root','testgce','moviedata')
#t_num = FieldType.NumericType it is wrong!!
t_num = FieldType()
t_num.setStored(False)
t1 = FieldType()
t1.setIndexed(True)
t1.setStored(True)
t1.setTokenized(False)
t1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
t2 = FieldType()
t2.setIndexed(True)
t2.setStored(False)
t2.setTokenized(True)
t2.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
t3 = FieldType()
t3.setIndexed(True)
t3.setStored(True)
t3.setTokenized(True)
t3.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS)
with con:
# Careful with codecs
con.set_character_set('utf8')
cur = con.cursor()
# Aagin the codecs
cur.execute('SET NAMES utf8;')
cur.execute('SET CHARACTER SET utf8;')
cur.execute('SET character_set_connection=utf8;')
#------step 3------
cur.execute("SELECT * FROM movie_items")
numrows = int(cur.rowcount)
print 'numrows:',numrows
for i in range(numrows):
row = cur.fetchone()
#------step 4------
summary = row[SUMMARY]
subject_id = row[SUBJECT_ID]
print 'id'+subject_id
#print 'summary'+summary+'end'
doc = Document()
#fields which should not be analyzed
doc.add(FloatField("rating_average",float(row[RATING_AVERAGE]),Field.Store.NO))
doc.add(FloatField("rating_stars", float(row[RATING_STARS]), Field.Store.NO))
doc.add(IntField("reviews_count", int(row[REVIEWS_COUNT]), Field.Store.NO))
#doc.add(FloatField("year", float(row[YEAR]), Field.Store.NO))
doc.add(IntField("collect_count", int(row[COLLECT_COUNT]), Field.Store.NO))
doc.add(IntField("subject_id", int(subject_id), Field.Store.YES))
doc.add(IntField("comments_count", int(row[COMMENTS_COUNT]), Field.Store.NO))
doc.add(IntField("ratings_count", int(row[RATINGS_COUNT]), Field.Store.NO))
doc.add(Field("image_small", row[IMAGE_SMALL], t1))
#fields which should be analyzed with WhitespaceAnalyzer
doc.add(Field("countries", row[COUNTRIES].replace(delim,' '), t3))
doc.add(Field("casts", row[CASTS].replace(delim,' '), t3))
doc.add(Field("genres", row[GENRES].replace(delim,' '), t3))
doc.add(Field("subtype", row[SUBTYPE].replace(delim,' '), t2))
doc.add(Field("directors", row[DIRECTORS].replace(delim,' '), t3))
user_tags_str = ''
others_like_str = ''
# print 'user_tags'+row[USER_TAGS]
# print 'others_like'+row[OTHERS_LIKE]
if row[USER_TAGS]!='':
for tag_pair in row[USER_TAGS].split(delim):
if tag_pair!='':#字符串的最后一个字符是:，这样split之后最后一个元素是空字符
user_tags_str = user_tags_str +' '+tag_pair.split(delim_uo)[0]
if row[OTHERS_LIKE]!='':
for like_pair in row[OTHERS_LIKE].split(delim):
if like_pair!='':
others_like_str = others_like_str +' '+like_pair.split(delim_uo)[1]
# print user_tags_str
# print others_like_str
doc.add(Field("user_tags", user_tags_str, t3))
doc.add(Field("others_like", others_like_str, t3))
#fields which should be analyzed with good analyzer
doc.add(Field("title", row[TITLE], t3))
doc.add(Field("original_title", row[ORIGINAL_TITLE], t2))
doc.add(Field("summary_segmentation", row[SUMMARY_SEGMENTATION], t2))
doc.add(Field("aka", row[AKA], t2))
if len(summary) > 0:
print subject_id +'--->'+':\n '+ row[TITLE]
try:
summary_unicoded = unicode(summary, 'utf-8') #test the encoding
except Exception,e:
print "Decode Failed: ", e
doc.add(Field('summary', summary, t2))
else:
print "warning:\n" + subject_id +'---> No content!'
writer.addDocument(doc)
def _getFullStrings(self):
mergePolicy = LogDocMergePolicy()
mergePolicy.setMergeFactor(97)
directory = RAMDirectory()
self.dirs.append(directory)
writer = self.getWriter(directory=directory,
analyzer=SimpleAnalyzer(Version.LUCENE_CURRENT),
maxBufferedDocs=4, mergePolicy=mergePolicy)
onlyStored = FieldType()
onlyStored.setStored(True)
fixedLen = self.getRandomNumber(2, 8)
fixedLen2 = self.getRandomNumber(1, 4)
for i in xrange(NUM_STRINGS):
doc = Document()
num = self.getRandomCharString(self.getRandomNumber(2, 8), 48, 52)
doc.add(Field("tracer", num, onlyStored))
doc.add(StringField("string", num, Field.Store.NO))
if self.supportsDocValues:
if self.dvStringSorted:
doc.add(SortedDocValuesField("string_dv", BytesRef(num)))
else:
doc.add(BinaryDocValuesField("string_dv", BytesRef(num)))
num2 = self.getRandomCharString(self.getRandomNumber(1, 4), 48, 50)
doc.add(StringField("string2", num2, Field.Store.NO))
if self.supportsDocValues:
if self.dvStringSorted:
doc.add(SortedDocValuesField("string2_dv", BytesRef(num2)))
else:
doc.add(BinaryDocValuesField("string2_dv", BytesRef(num2)))
doc.add(Field("tracer2", num2, onlyStored))
for f2 in doc.getFields():
if f2.fieldType().indexed() and not f2.fieldType().omitNorms():
Field.cast_(f2).setBoost(2.0)
numFixed = self.getRandomCharString(fixedLen, 48, 52)
doc.add(Field("fixed_tracer", numFixed, onlyStored))
doc.add(StringField("string_fixed", numFixed, Field.Store.NO))
if self.supportsDocValues:
if self.dvStringSorted:
doc.add(SortedDocValuesField("string_fixed_dv", BytesRef(numFixed)))
else:
doc.add(BinaryDocValuesField("string_fixed_dv", BytesRef(numFixed)))
num2Fixed = self.getRandomCharString(fixedLen2, 48, 52)
doc.add(StringField("string2_fixed", num2Fixed, Field.Store.NO))
if self.supportsDocValues:
if self.dvStringSorted:
doc.add(SortedDocValuesField("string2_fixed_dv", BytesRef(num2Fixed)))
else:
doc.add(BinaryDocValuesField("string2_fixed_dv", BytesRef(num2Fixed)))
doc.add(Field("tracer2_fixed", num2Fixed, onlyStored))
for f2 in doc.getFields():
if f2.fieldType().indexed() and not f2.fieldType().omitNorms():
Field.cast_(f2).setBoost(2.0)
writer.addDocument(doc)
writer.close()
return self.getSearcher(directory=directory)
def indexDocs(self, root, writer):
t1 = FieldType()
t1.setIndexed(False)
t1.setStored(True)
t1.setTokenized(False)
t2 = FieldType()
t2.setIndexed(True)
t2.setStored(True)
t2.setTokenized(True)
t2.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
count = 0
filedic = {}
urlfile = open("sportsinformation.txt", "r")
for line in urlfile.readlines():
urlname, pagename, newsdate = line.split('*my_sep*')
newsdate = newsdate.strip('\n')
if len(newsdate) == 0:
continue
filedic[pagename] = [urlname, newsdate]
'''pageANDurl = line.split('\t')
urlname = pageANDurl[0]
webpage = pageANDurl[1].strip('\n')
filedic[webpage] = urlname'''
urlfile.close()
for root, dirnames, filenames in os.walk(root):
for filename in filenames:
if (filename.endswith('apk') or filename.endswith('pdf') \
or filename.endswith('exe') or filename.endswith('rar') \
or filename.endswith('zip')):
print filename, " SKIP THIS FILE!"
continue
count += 1
print "adding", filename, " ", count
try:
path = os.path.join(root, filename)
file = open(path)
contents = unicode(file.read(), 'utf8', 'ignore')
soup = BeautifulSoup(contents, "html.parser")
title = soup.find("title")
if title:
title = title.text
else:
title = "This page has no title"
contents = ''.join(soup.findAll(text=True))
contents = ' '.join(jieba.cut(contents))
file.close()
url = filedic[filename][0]
newsdate = str(filedic[filename][1])
print type(newsdate)
print type(url)
doc = Document()
doc.add(Field("title", title, t1))
doc.add(Field("url", url, t1))
doc.add(Field("date", newsdate, t1))
sites = []
site = urlparse.urlparse(url).netloc
siteparts = site.split(".")
length = len(siteparts)
while (length > 0):
length -= 1
site = '.'.join(siteparts[length:])
sites.append(site)
site = ' '.join(sites)
doc.add(Field("site", site, t2))
if len(contents) > 0:
print "yes"
doc.add(Field("contents", contents, t2))
else:
print "warning: no content in %s" % filename
writer.addDocument(doc)
except Exception, e:
print "Failed in indexDocs:", e
def index_docs(self, tweets, writer):
t1 = FieldType()
t1.setIndexed(True)
t1.setStored(True)
t1.setTokenized(True)
t1.setIndexOptions(FieldInfo.IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
t1.setStoreTermVectors(True)
t1.setStoreTermVectorOffsets(True)
# add each tweet to the index
for tweet in tweets:
try:
# strip out URLs because they provide false index matches
contents = []
for word in tweet[1].text.split():
if word.startswith("http://") or word.startswith("https://"):
continue
contents.append(word)
contents = " ".join(contents)
if len(contents) == 0: continue
doc = Document()
doc.add(Field("contents", contents, t1))
writer.addDocument(doc)
except Exception, e:
print "Failed in index_docs:", e
def _create_index(self, index_dir: str) -> None:
"""Index documents
Parameters
----------
index_dir : str
The dir to store index
"""
os.mkdir(index_dir)
TITLE_FIELD = FieldType() # pylint: disable=invalid-name
TITLE_FIELD.setStored(True)
TITLE_FIELD.setIndexOptions(IndexOptions.DOCS)
TEXT_FIELD = FieldType() # pylint: disable=invalid-name
TEXT_FIELD.setStored(True)
TEXT_FIELD.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
fs_dir = MMapDirectory(Paths.get(index_dir))
writer_config = IndexWriterConfig(StandardAnalyzer())
writer_config.setRAMBufferSizeMB(16384.0) # 14g
self.writer = IndexWriter(fs_dir, writer_config)
logger.info("%d docs in index", self.writer.numDocs())
logger.info("Indexing documents...")
doc_ids = self.doc_db.get_doc_ids()
for doc_id in tqdm(doc_ids, total=len(doc_ids)):
text = self.doc_db.get_doc_text(doc_id)
doc = Document()
doc.add(Field("title", doc_id, TITLE_FIELD))
doc.add(Field("text", text, TEXT_FIELD))
self.writer.addDocument(doc)
logger.info("Indexed %d docs.", self.writer.numDocs())
self.writer.forceMerge(1) # to increase search performance
self.writer.close()
def _createNoTermsFrequencyFieldType():
f = FieldType()
f.setIndexed(True)
f.setTokenized(True)
f.setOmitNorms(True)
f.setIndexOptions(FieldInfo.IndexOptions.DOCS_ONLY)
f.freeze()
return f
