Hello,
I tried Christian's python-lucene package.
It seems the lucene API changed from 1.9.1 and 2.0.0.
Some interfaces were tagged as deprecated in 1.9.1 and were removed in
2.0.0.
The attached patchs permits to use the jToolkit with lucene 2.0.0.
I also think it should work with 1.9.1 (not tested). I did not test
whether it breaks lupy support.
With this patch applied, I found some files in /tmp (lock files for
lucene).
It would be nice (for security reasons and to avoid cluttering /tmp) to put
these lock files in the corresponding index directory.
A second patch is attached.
I've still not checked whether Pootle is faster or not.
Kind Regards,
--
Nekral
Index: jToolkit/data/indexer.py
===================================================================
--- jToolkit/data/indexer.py (révision 4)
+++ jToolkit/data/indexer.py (copie de travail)
@@ -25,6 +25,18 @@
except:
return None, False
+def Occur(required, prohibited):
+ if required == True and prohibited == False:
+ return PyLucene.BooleanClause.Occur.MUST
+ elif required == False and prohibited == False:
+ return PyLucene.BooleanClause.Occur.SHOULD
+ elif required == False and prohibited == True:
+ return PyLucene.BooleanClause.Occur.MUST_NOT
+ else:
+ # It is an error to specify a clause as both required
+ # and prohibited
+ return None
+
class Wrapper:
def __init__(self, **kwargs):
for key, value in kwargs.iteritems():
@@ -331,11 +343,11 @@
contents = unicode(fp.read(), self.encoding)
fp.close()
doc = indexer.Document()
- doc.add(indexer.Field("file_name",os.path.basename(file),True,True,True))
+ doc.add(indexer.Field("file_name",os.path.basename(file),PyLucene.Field.Store.YES, PyLucene.Field.Index.TOKENIZED))
if len(contents) > 0:
- doc.add(indexer.Field("file_contents", contents, True, True, True))
+ doc.add(indexer.Field("file_contents", contents, PyLucene.Field.Store.YES, PyLucene.Field.Index.TOKENIZED))
if ID is not None:
- doc.add(indexer.Field("recordID",ID,True,True,True))
+ doc.add(indexer.Field("recordID",ID,PyLucene.Field.Store.YES, PyLucene.Field.Index.TOKENIZED))
self.writer.addDocument(doc)
self.errorhandler.logtrace("indexer.py: Indexing file %s" % file)
@@ -359,7 +371,7 @@
value = value.decode("charmap")
if not isinstance(value, (str, unicode)):
value = str(value)
- doc.add(indexer.Field(str(field), value, True, True, True))
+ doc.add(indexer.Field(str(field), value, PyLucene.Field.Store.YES, PyLucene.Field.Index.TOKENIZED))
self.writer.addDocument(doc)
def startIndex(self):
@@ -376,7 +388,7 @@
self.dirLock.acquire()
try:
self.writer = indexer.IndexWriter(self.storeDir, self.analyzer, create)
- self.writer.maxFieldLength = 1048576
+ self.writer.setMaxFieldLength(1048576)
success = True
except Exception,e:
self.errorhandler.logerror("Failed to create index. %s" % self.errorhandler.traceback_str())
@@ -484,15 +496,15 @@
combinedquery = indexer.BooleanQuery()
for fieldSearch in fieldSearches:
if isinstance(fieldSearch, indexer.BooleanQuery):
- clause = indexer.BooleanClause(fieldSearch, requireall, False)
+ clause = indexer.BooleanClause(fieldSearch, Occur(requireall, False))
combinedquery.add(clause)
elif isinstance(fieldSearch, tuple):
fieldName, search = fieldSearch
analyzer = self.analyzer
if isinstance(analyzer, PerFieldAnalyzer):
analyzer = analyzer.getAnalyzer(fieldName)
- query = indexer.QueryParser.parse(search, fieldName, analyzer)
- combinedquery.add(query, requireall, False)
+ query = indexer.QueryParser(fieldName,analyzer).parse(search)
+ combinedquery.add(query, Occur(requireall, False))
else:
raise ValueError("unexpected value in fieldSearch: %r" % fieldSearch)
return combinedquery
@@ -500,7 +512,7 @@
def notQuery(self, query):
"""returns a query that matches everything but the query"""
notquery = indexer.BooleanQuery()
- clause = indexer.BooleanClause(query, False, True)
+ clause = indexer.BooleanClause(query, Occur(False, True))
notquery.add(clause)
return notquery
@@ -645,7 +657,7 @@
query = indexer.BooleanQuery()
analyzer = indexer.StandardAnalyzer()
for keyfield in IDFields.keys():
- query.add(indexer.QueryParser.parse(IDFields[keyfield], keyfield, analyzer), True, False)
+ query.add(indexer.QueryParser.parse(IDFields[keyfield], keyfield, analyzer), Occur(True, False))
hits = self.search(query)
modifiedFields.update(IDFields)
for hit, doc in hits:
Index: jToolkit/data/indexer.py
===================================================================
--- jToolkit/data/indexer.py (révision 4)
+++ jToolkit/data/indexer.py (copie de travail)
@@ -197,7 +209,7 @@
self.encoding = encoding
self.errorhandler = errorhandler
self.storeDir = config.indexdir
- lockname = os.path.join(tempfile.gettempdir(),self.storeDir.replace('/','_').replace('\\','_').replace(':','_'))
+ lockname = self.storeDir+".lock"
self.dirLock = glock.GlobalLock(lockname)
if not os.path.exists(self.storeDir):
os.mkdir(self.storeDir)
@@ -437,7 +449,7 @@
self.errorhandler = errorhandler
self.storeDir = storeDir
self.indexReader = self.indexVersion = self.indexSearcher = None
- lockname = os.path.join(tempfile.gettempdir(),self.storeDir.replace('/','_').replace('\\','_').replace(':','_'))
+ lockname = self.storeDir+".lock"
self.dirLock = glock.GlobalLock(lockname)
# if we can't acquire the lock, someone is busy writing, and we should wait for them
self.dirLock.acquire(blocking=True)
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Translate-pootle mailing list
Translate-pootle@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/translate-pootle
