# original wikiq headers are: title articleid revid date_time anon
# editor editor_id minor text_size text_entropy text_md5 reversion
# additions_size deletions_size
-import pdb
import argparse
import sys
import os, os.path
def __init__(self, page, namespace_map, collapse_user=False):
self.id = page.id
self.namespace = page.namespace
+ # following mwxml, we assume namespace 0 in cases where
+ # page.namespace is inconsistent with namespace_map
+ if page.namespace not in namespace_map:
+ self.title = page.title
+ page.namespace = 0
if page.namespace != 0:
self.title = ':'.join([namespace_map[page.namespace], page.title])
else:
class WikiqParser():
- def __init__(self, input_file, output_file, collapse_user=False, persist=None, urlencode=False):
+ def __init__(self, input_file, output_file, collapse_user=False, persist=None, urlencode=False, namespaces = None):
"""
Parameters:
persist : what persistence method to use. Takes a PersistMethod value
self.printed_header = False
self.namespaces = []
self.urlencode = urlencode
-
+ if namespaces is not None:
+ self.namespace_filter = set(namespaces)
+ else:
+ self.namespace_filter = None
+
def __get_namespace_from_title(self, title):
default_ns = None
# Iterate through pages
for page in dump:
+ namespace = page.namespace if page.namespace is not None else self.__get_namespace_from_title(page.title)
+
+ # skip namespaces not in the filter
+ if self.namespace_filter is not None:
+ if namespace not in self.namespace_filter:
+ continue
+
rev_detector = mwreverts.Detector()
if self.persist != PersistMethod.none:
from mw.lib import persistence
state = persistence.State()
+
+
# Iterate through a page's revisions
for rev in page:
'articleid' : page.id,
'editor_id' : "" if rev.deleted.user == True or rev.user.id is None else rev.user.id,
'title' : '"' + page.title + '"',
- 'namespace' : page.namespace if page.namespace is not None else self.__get_namespace_from_title(page.title),
+ 'namespace' : namespace,
'deleted' : "TRUE" if rev.deleted.text else "FALSE" }
# if revisions are deleted, /many/ things will be missing
parser.add_argument('-u', '--url-encode', dest="urlencode", action="store_true",
help="Output url encoded text strings. This works around some data issues like newlines in editor names. In the future it may be used to output other text data.")
+parser.add_argument('-n', '--namespace-include', dest="namespace_include", type=int, action='append',
+ help="Id number of namspace to include.")
+
+
args = parser.parse_args()
# set persistence method
else:
persist = PersistMethod.sequence
+if args.namespace_include is not None:
+ namespaces = args.namespace_include
+else:
+ namespaces = None
+
if len(args.dumpfiles) > 0:
for filename in args.dumpfiles:
input_file = open_input_file(filename)
filename = os.path.join(output_dir, os.path.basename(filename))
output_file = open_output_file(filename)
- wikiq = WikiqParser(input_file, output_file,
+ wikiq = WikiqParser(input_file, output_file,
collapse_user=args.collapse_user,
- persist=persist,
- urlencode=args.urlencode)
-
+ persist=persist,
+ urlencode=args.urlencode,
+ namespaces = namespaces)
wikiq.process()