[Spambayes-checkins] spambayes dbExpImp.py,1.2,1.3
Tim Stone
timstone4 at users.sourceforge.net
Sat Feb 1 11:07:12 EST 2003
Update of /cvsroot/spambayes/spambayes
In directory sc8-pr-cvs1:/tmp/cvs-serv18578
Modified Files:
dbExpImp.py
Log Message:
Cleaned up and added some informational messages.
Index: dbExpImp.py
===================================================================
RCS file: /cvsroot/spambayes/spambayes/dbExpImp.py,v
retrieving revision 1.2
retrieving revision 1.3
diff -C2 -d -r1.2 -r1.3
*** dbExpImp.py 1 Feb 2003 17:57:04 -0000 1.2
--- dbExpImp.py 1 Feb 2003 19:07:10 -0000 1.3
***************
*** 92,96 ****
from __future__ import generators
! import storage
from spambayes.Options import options
import sys, os, getopt, errno, re
--- 92,96 ----
from __future__ import generators
! import spambayes.storage
from spambayes.Options import options
import sys, os, getopt, errno, re
***************
*** 99,109 ****
def runExport(dbFN, useDBM, outFN):
- print "running export on %s" % (dbFN)
if useDBM:
! bayes = storage.DBDictClassifier(dbFN)
words = bayes.db.keys()
words.remove(bayes.statekey)
else:
! bayes = storage.PickledClassifier(dbFN)
words = bayes.wordinfo.keys()
--- 99,108 ----
def runExport(dbFN, useDBM, outFN):
if useDBM:
! bayes = spambayes.storage.DBDictClassifier(dbFN)
words = bayes.db.keys()
words.remove(bayes.statekey)
else:
! bayes = spambayes.storage.PickledClassifier(dbFN)
words = bayes.wordinfo.keys()
***************
*** 116,120 ****
nham = bayes.nham;
nspam = bayes.nspam;
! print "nham %s, nspam %s" % (nham, nspam)
fp.write("%s,%s,\n" % (nham, nspam))
--- 115,122 ----
nham = bayes.nham;
nspam = bayes.nspam;
!
! print "Exporting database %s to file %s" % (dbFN, outFN)
! print "Database has %s ham, %s spam, and %s words" \
! % (nham, nspam, len(words))
fp.write("%s,%s,\n" % (nham, nspam))
***************
*** 138,145 ****
raise
if useDBM:
! bayes = storage.DBDictClassifier(dbFN)
else:
! bayes = storage.PickledClassifier(dbFN)
try:
--- 140,159 ----
raise
+ try:
+ os.unlink(dbFN+".dat")
+ except OSError, e:
+ if e.errno != 2: # errno.<WHAT>
+ raise
+
+ try:
+ os.unlink(dbFN+".dir")
+ except OSError, e:
+ if e.errno != 2: # errno.<WHAT>
+ raise
+
if useDBM:
! bayes = spambayes.storage.DBDictClassifier(dbFN)
else:
! bayes = spambayes.storage.PickledClassifier(dbFN)
try:
***************
*** 150,154 ****
nline = fp.readline()
- print nline
(nham, nspam, junk) = re.split(',', nline)
--- 164,167 ----
***************
*** 160,163 ****
--- 173,183 ----
bayes.nspam += nspam
+ if newDBM:
+ impType = "Importing"
+ else:
+ impType = "Merging"
+
+ print "%s database %s using file %s" % (impType, dbFN, inFN)
+
lines = fp.readlines()
***************
*** 177,181 ****
--- 197,216 ----
fp.close()
+
+ print "Storing database, please be patient. Even moderately large"
+ print "databases may take a very long time to store."
bayes.store()
+ print "Finished storing database"
+
+ if useDBM:
+ words = bayes.db.keys()
+ words.remove(bayes.statekey)
+ else:
+ words = bayes.wordinfo.keys()
+
+ print "Database has %s ham, %s spam, and %s words" \
+ % (bayes.nham, bayes.nspam, len(words))
+
+
More information about the Spambayes-checkins
mailing list