Fixes a bug, a performance issue, and use 'M' instead of 'U' to be consistent with...

[mw] / src / mw / metadir.py

diff --git a/src/mw/metadir.py b/src/mw/metadir.py
index 295481901682441c55276068db16d605f7f53f30..6c3dc8b0e4c299868d52864ab15f0d77dea5f97e 100644 (file)
--- a/src/mw/metadir.py
+++ b/src/mw/metadir.py
@@ -20,7 +20,6 @@ import bzrlib.diff
 import codecs
 import ConfigParser
 import json
-import mw.api
 import os
 from StringIO import StringIO
 import sys
@@ -62,7 +61,7 @@ class Metadir(object):
             os.mkdir(self.location, 0755)
         # metadir versioning
         fd = file(os.path.join(self.location, 'version'), 'w')
-        fd.write('1') # XXX THIS API VERSION NOT LOCKED IN YET
+        fd.write('1')  # XXX THIS API VERSION NOT LOCKED IN YET
         fd.close()
         # create config
         self.config = ConfigParser.RawConfigParser()
@@ -78,6 +77,15 @@ class Metadir(object):
         # create cache/pages/
         os.mkdir(os.path.join(self.location, 'cache', 'pages'), 0755)
 
+    def clean_page(self, pagename):
+        filename = pagename_to_filename(pagename) + '.wiki'
+        cur_content = codecs.open(filename, 'r', 'utf-8').read()
+        if len(cur_content) != 0 and cur_content[-1] == '\n':
+            cur_content = cur_content[:-1]
+        fd = file(filename, 'w')
+        fd.write(cur_content.encode('utf-8'))
+        fd.close()
+
     def pagedict_add(self, pagename, pageid, currentrv):
         fd = file(os.path.join(self.location, 'cache', 'pagedict'), 'r+')
         pagedict = json.loads(fd.read())
@@ -90,6 +98,7 @@ class Metadir(object):
     def get_pageid_from_pagename(self, pagename):
         fd = file(os.path.join(self.location, 'cache', 'pagedict'), 'r')
         pagedict = json.loads(fd.read())
+        pagename = pagename.decode('utf-8')
         if pagename in pagedict.keys():
             return pagedict[pagename]
         else:
@@ -147,7 +156,7 @@ class Metadir(object):
         for full in check:
             name = os.path.split(full)[1]
             if name[-5:] == '.wiki':
-                pagename = mw.api.filename_to_pagename(name[:-5])
+                pagename = filename_to_pagename(name[:-5])
                 pageid = self.get_pageid_from_pagename(pagename)
                 if not pageid:
                     status[os.path.relpath(full, self.root)] = '?'
@@ -155,16 +164,17 @@ class Metadir(object):
                     rvid = self.pages_get_rv_list(pageid)[-1]
                     rv = self.pages_get_rv(pageid, rvid)
                     cur_content = codecs.open(full, 'r', 'utf-8').read()
-                    if cur_content[-1] == '\n':
+                    if (len(cur_content) != 0) and (cur_content[-1] == '\n'):
                         cur_content = cur_content[:-1]
                     if cur_content != rv['content']:
-                        status[os.path.relpath(full, self.root)] = 'U'
+                        status[os.path.relpath(full, self.root)] = 'M'
         return status
 
     def diff_rv_to_working(self, pagename, oldrvid=0, newrvid=0):
         # oldrvid=0 means latest fetched revision
         # newrvid=0 means working copy
-        filename = mw.api.pagename_to_filename(pagename) + '.wiki'
+        filename = pagename_to_filename(pagename) + '.wiki'
+        filename = filename.decode('utf-8')
         pageid = self.get_pageid_from_pagename(pagename)
         if not pageid:
             raise ValueError('page named %s has not been fetched' % pagename)
@@ -173,10 +183,12 @@ class Metadir(object):
                 oldrvid = self.pages_get_rv_list(pageid)[-1]
             oldrv = self.pages_get_rv(pageid, oldrvid)
             oldname = 'a/%s (revision %i)' % (filename, oldrvid)
-            old = [i + '\n' for i in oldrv['content'].split('\n')]
+            old = [i + '\n' for i in \
+                   oldrv['content'].encode('utf-8').split('\n')]
             if newrvid == 0:
                 cur_content = codecs.open(filename, 'r', 'utf-8').read()
-                if cur_content[-1] == '\n':
+                cur_content = cur_content.encode('utf-8')
+                if (len(cur_content) != 0) and (cur_content[-1] == '\n'):
                     cur_content = cur_content[:-1]
                 newname = 'b/%s (working copy)' % filename
                 new = [i + '\n' for i in cur_content.split('\n')]
@@ -186,8 +198,19 @@ class Metadir(object):
                 new = [i + '\n' for i in newrv['content'].split('\n')]
             diff_fd = StringIO()
             bzrlib.diff.internal_diff(oldname, old, newname, new, diff_fd)
-            diff_fd.seek(0)
-            diff = diff_fd.read()
+            diff = diff_fd.getvalue()
             if diff[-1] == '\n':
                 diff = diff[:-1]
             return diff
+
+
+def pagename_to_filename(name):
+    name = name.replace(' ', '_')
+    name = name.replace('/', '!')
+    return name
+
+
+def filename_to_pagename(name):
+    name = name.replace('!', '/')
+    name = name.replace('_', ' ')
+    return name