add flag for excluding whitespace and punctuation

[mediawiki_dump_tools.git] / wikiq

diff --git a/wikiq b/wikiq
index 731f59ae27b7dc93035c70734a34ab9d43c0e7d4..334f195017dfe9e3853bedaa614fe46fae4220dd 100755 (executable)
--- a/wikiq
+++ b/wikiq
@@ -130,9 +130,8 @@ class WikiqPage(Page):
     #         2          B               A            True
     #         3          A               B            True
     #         4          A               A           False
     #         2          B               A            True
     #         3          A               B            True
     #         4          A               A           False

-    # Post-loop                          A          Always
+   # Post-loop                          A          Always

     def __find_next_revision(self):
     def __find_next_revision(self):

-

         if self.prev_rev is None:
             prev_rev = WikiqPage._correct_sha(next(self.revisions))
             self.prev_rev = prev_rev
         if self.prev_rev is None:
             prev_rev = WikiqPage._correct_sha(next(self.revisions))
             self.prev_rev = prev_rev


@@ -193,7 +192,7 @@ class WikiqPage(Page):
 
 class WikiqParser():
     
 
 class WikiqParser():
     

-    def __init__(self, input_file, output_file, collapse_user=False, persist=None, urlencode=False, namespaces = None):
+    def __init__(self, input_file, output_file, collapse_user=False, persist=None, urlencode=False, namespaces = None, exclude_punct = False, exclude_ws = False):

         """ 
         Parameters:
            persist : what persistence method to use. Takes a PersistMethod value
         """ 
         Parameters:
            persist : what persistence method to use. Takes a PersistMethod value


@@ -210,11 +209,9 @@ class WikiqParser():
         else:
             self.namespace_filter = None
 
         else:
             self.namespace_filter = None
 

-        # create a regex that creates the output filename
-        # output_filename = re.sub(r'^.*/(enwiki\-\d+)\-.*p(\d+)p.*$',
-        #                         r'output/wikiq-\1-\2.tsv',
-        #                         input_filename)
-
+        self.exclude_punct = exclude_punct
+        self.exclude_ws = exclude_ws
+        

         # Construct dump file iterator
         self.dump = WikiqIterator.from_file(self.input_file, self.collapse_user)
       
         # Construct dump file iterator
         self.dump = WikiqIterator.from_file(self.input_file, self.collapse_user)
       


@@ -226,29 +223,6 @@ class WikiqParser():
         if self.persist == PersistMethod.segment:
             self.diff_engine = SegmentMatcher(tokenizer = wikitext_split)
 
         if self.persist == PersistMethod.segment:
             self.diff_engine = SegmentMatcher(tokenizer = wikitext_split)
 

-    # def __get_namespace_from_title(self, title):
-    #     default_ns = None
-
-    #     for ns in self.namespaces:
-    #         # skip if the namespace is not defined
-    #         if ns == None:
-    #             default_ns = self.namespaces[ns]
-    #             continue
-
-    #         if title.startswith(ns + ":"):
-    #             return self.namespaces[ns]
-
-    #     # if we've made it this far with no matches, we return the default namespace
-    #     return default_ns
-
-    # def _set_namespace(self, rev_docs):
-        
-    #     for rev_data in rev_docs:
-    #         if 'namespace' not in rev_data['page']:
-    #             namespace = self.__get_namespace_from_title(page['title'])
-    #             rev_data['page']['namespace'] = namespace
-    #         yield rev_data
-

     def process(self):
         page_count = 0
         rev_count = 0
     def process(self):
         page_count = 0
         rev_count = 0


@@ -332,16 +306,11 @@ class WikiqParser():
                     rev_data['anon'] = ""
                     rev_data['editor'] = ""
 
                     rev_data['anon'] = ""
                     rev_data['editor'] = ""
 

-                #if re.match(r'^#redirect \[\[.*\]\]', rev.text, re.I):
-                #    redirect = True
-                #else:
-                #    redirect = False
-                
-                #TODO missing: additions_size deletions_size
+                # we can easily add redirect info
+                # rev_data['redirect'] = rev.page.redirect

                 
                 

-                # if collapse user was on, lets run that
-                # if self.collapse_user:
-                #     rev_data.collapsed_revs = rev.collapsed_revs
+                if self.collapse_user:
+                    rev_data['collapsed_revs'] = rev.collapsed_revs

 
                 if self.persist != PersistMethod.none:
                     if rev.deleted.text:
 
                 if self.persist != PersistMethod.none:
                     if rev.deleted.text:


@@ -360,8 +329,16 @@ class WikiqParser():
                         if len(window) == PERSISTENCE_RADIUS:
                             old_rev_id, old_rev_data, old_tokens_added, old_tokens_removed = window[0]
                             
                         if len(window) == PERSISTENCE_RADIUS:
                             old_rev_id, old_rev_data, old_tokens_added, old_tokens_removed = window[0]
                             

-                            num_token_revs, num_tokens_added, num_tokens_removed  = calculate_persistence(old_tokens_added, old_tokens_removed, legacy = self.persist == PersistMethod.legacy)
-
+                            num_token_revs, \
+                                num_tokens_added, \
+                                num_tokens_removed = \
+                                                     calculate_persistence(
+                                                         old_tokens_added,
+                                                         old_tokens_removed,
+                                                         exclude_ws = self.exclude_ws,
+                                                         exclude_punct = self.exclude_punct,
+                                                         legacy = self.persist == PersistMethod.legacy)
+                            

                             old_rev_data["token_revs"] = num_token_revs
                             old_rev_data["tokens_added"] = num_tokens_added
                             old_rev_data["tokens_removed"] = num_tokens_removed
                             old_rev_data["token_revs"] = num_token_revs
                             old_rev_data["tokens_added"] = num_tokens_added
                             old_rev_data["tokens_removed"] = num_tokens_removed


@@ -383,7 +360,15 @@ class WikiqParser():
 
                     rev_id, rev_data, tokens_added, tokens_removed = item
 
 
                     rev_id, rev_data, tokens_added, tokens_removed = item
 

-                    num_token_revs, num_tokens_added, num_tokens_removed = calculate_persistence(tokens_added, tokens_removed, legacy = self.persist == PersistMethod.legacy)
+                    num_token_revs, \
+                        num_tokens_added, \
+                        num_tokens_removed = calculate_persistence(
+                            tokens_added,
+                            tokens_removed,
+                            exclude_ws = self.exclude_ws,
+                            exclude_punct = self.exclude_punct,
+                            legacy = self.persist == PersistMethod.legacy)
+

 
                     rev_data["token_revs"] = num_token_revs
                     rev_data["tokens_added"] = num_tokens_added
 
                     rev_data["token_revs"] = num_token_revs
                     rev_data["tokens_added"] = num_tokens_added


@@ -458,7 +443,11 @@ parser.add_argument('-u', '--url-encode', dest="urlencode", action="store_true",
 parser.add_argument('-n', '--namespace-include', dest="namespace_filter", type=int, action='append',
                     help="Id number of namspace to include. Can be specified more than once.")
 
 parser.add_argument('-n', '--namespace-include', dest="namespace_filter", type=int, action='append',
                     help="Id number of namspace to include. Can be specified more than once.")
 

+parser.add_argument('--exclude-whitespace', dest="exclude_ws", action="store_true",
+                    help="Flag to remove whitespace from persistence measures.")

 
 

+parser.add_argument('--exclude-punctuation', dest="exclude_punct", action="store_true",
+                    help="Flag to remove punctuation from persistence measures.")

 
 args = parser.parse_args()
 
 
 args = parser.parse_args()
 


@@ -500,7 +489,9 @@ if len(args.dumpfiles) > 0:
                             collapse_user=args.collapse_user,
                             persist=persist,
                             urlencode=args.urlencode,
                             collapse_user=args.collapse_user,
                             persist=persist,
                             urlencode=args.urlencode,

-                            namespaces = namespaces)
+                            namespaces = namespaces,
+                            exclude_punct = args.exclude_punct,
+                            exclude_ws = args.exclude_ws)

 
         wikiq.process()
 
 
         wikiq.process()