]> projects.mako.cc - mw/blobdiff - src/mw/metadir.py
implementing edit conflict & collision detection
[mw] / src / mw / metadir.py
index 22b9c5cf0049144092ecc6c0b2d1aee4e7365020..3f8d96782551c01bb2f4cf66c34b6b72bebb67ea 100644 (file)
@@ -1,6 +1,6 @@
 ###
 # mw - VCS-like nonsense for MediaWiki websites
-# Copyright (C) 2009  Ian Weller <ian@ianweller.org>
+# Copyright (C) 2010  Ian Weller <ian@ianweller.org>
 #
 # This program is free software; you can redistribute it and/or modify
 # it under the terms of the GNU General Public License as published by
 # with this program.  If not, see <http://www.gnu.org/licenses/>.
 ###
 
+import bzrlib.diff
+import codecs
 import ConfigParser
 import json
+import mw.api
 import os
+from StringIO import StringIO
 import sys
-import time
+
 
 class Metadir(object):
+
     def __init__(self):
         self.me = os.path.basename(sys.argv[0])
         root = os.getcwd()
@@ -30,7 +35,7 @@ class Metadir(object):
             if '.mw' in os.listdir(root):
                 self.root = root
                 break
-            (head, tail) = os.path.split(root)
+            head = os.path.split(root)[0]
             if head == root:
                 self.root = os.getcwd()
                 break
@@ -44,47 +49,157 @@ class Metadir(object):
         else:
             self.config = None
 
+    def save_config(self):
+        with open(self.config_loc, 'wb') as config_file:
+            self.config.write(config_file)
+
     def create(self, api_url):
         # create the directory
-        try:
-            os.mkdir(self.location, 0755)
-        except OSError, e:
+        if os.path.isdir(self.location):
             print '%s: you are already in a mw repo' % self.me
             sys.exit(1)
+        else:
+            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.close()
         # create config
         self.config = ConfigParser.RawConfigParser()
         self.config.add_section('remote')
         self.config.set('remote', 'api_url', api_url)
-        with open(self.config_loc, 'wb') as config_file:
-            self.config.write(config_file)
-        # create cache
+        self.save_config()
+        # create cache/
         os.mkdir(os.path.join(self.location, 'cache'))
-        # create cache/page
-        fd = file(os.path.join(self.location, 'cache', 'page'), 'w')
-        fd.write(json.dumps({}))
-        # create cache/rv
-        fd = file(os.path.join(self.location, 'cache', 'rv'), 'w')
+        # create cache/pagedict
+        fd = file(os.path.join(self.location, 'cache', 'pagedict'), 'w')
         fd.write(json.dumps({}))
+        fd.close()
+        # create cache/pages/
+        os.mkdir(os.path.join(self.location, 'cache', 'pages'), 0755)
+
 
-    def add_page_info(self, pageid, pagename, rvids):
-        lulz = file(os.path.join(self.location, 'cache', 'page'), 'r')
-        conf = json.loads(lulz.read())
-        conf[pageid] = {'name': pagename, 'rv': rvids}
-        fd = file(os.path.join(self.location, 'cache', 'page'), 'w')
-        fd.write(json.dumps(conf))
 
-    def add_rv_info(self, rv):
-        lulz = file(os.path.join(self.location, 'cache', 'rv'), 'r')
-        conf = json.loads(lulz.read())
+    def clean_page(self, pagename):
+        filename = mw.api.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())
+        pagedict[pagename] = {'id': int(pageid), 'currentrv': int(currentrv)}
+        fd.seek(0)
+        fd.write(json.dumps(pagedict))
+        fd.truncate()
+        fd.close()
+
+    def get_pageid_from_pagename(self, pagename):
+        fd = file(os.path.join(self.location, 'cache', 'pagedict'), 'r')
+        pagedict = json.loads(fd.read())
+        if pagename in pagedict.keys():
+            return pagedict[pagename]
+        else:
+            return None
+
+    def pages_add_rv(self, pageid, rv):
+        pagefile = os.path.join(self.location, 'cache', 'pages', str(pageid))
+        fd = file(pagefile, 'w+')
+        pagedata_raw = fd.read()
+        if pagedata_raw == '':
+            pagedata = {}
+        else:
+            pagedata = json.loads(pagedata_raw)
         rvid = int(rv['revid'])
-        conf[rvid] = {
-                'user': rv['user'], 'timestamp': rv['timestamp'],
-                'content': rv['*']
+        pagedata[rvid] = {
+                'user': rv['user'],
+                'timestamp': rv['timestamp'],
         }
-        conf[rvid]['minor'] = 'minor' in rv
-        if 'comment' in rv:
-            conf[rvid]['comment'] = rv['comment']
+        if '*' in rv.keys():
+            pagedata[rvid]['content'] = rv['*']
+        fd.seek(0)
+        fd.write(json.dumps(pagedata))
+        fd.truncate()
+        fd.close()
+
+    def pages_get_rv_list(self, pageid):
+        pagefile = os.path.join(self.location, 'cache', 'pages',
+                                str(pageid['id']))
+        fd = file(pagefile, 'r')
+        pagedata = json.loads(fd.read())
+        rvs = [int(x) for x in pagedata.keys()]
+        rvs.sort()
+        return rvs
+
+    def pages_get_rv(self, pageid, rvid):
+        pagefile = os.path.join(self.location, 'cache', 'pages',
+                                str(pageid['id']))
+        fd = file(pagefile, 'r')
+        pagedata = json.loads(fd.read())
+        return pagedata[str(rvid)]
+
+    def working_dir_status(self, files=None):
+        status = {}
+        check = []
+        if files == None or files == []:
+            for root, dirs, files in os.walk(self.root):
+                if root == self.root:
+                    dirs.remove('.mw')
+                for name in files:
+                    check.append(os.path.join(root, name))
+        else:
+            for file in files:
+                check.append(os.path.join(os.getcwd(), file))
+        check.sort()
+        for full in check:
+            name = os.path.split(full)[1]
+            if name[-5:] == '.wiki':
+                pagename = mw.api.filename_to_pagename(name[:-5])
+                pageid = self.get_pageid_from_pagename(pagename)
+                if not pageid:
+                    status[os.path.relpath(full, self.root)] = '?'
+                else:
+                    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 (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'
+        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'
+        pageid = self.get_pageid_from_pagename(pagename)
+        if not pageid:
+            raise ValueError('page named %s has not been fetched' % pagename)
         else:
-            conf[rvid]['comment'] = None
-        fd = file(os.path.join(self.location, 'cache', 'rv'), 'w')
-        fd.write(json.dumps(conf))
+            if oldrvid == 0:
+                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')]
+            if newrvid == 0:
+                cur_content = codecs.open(filename, 'r', 'utf-8').read()
+                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')]
+            else:
+                newrv = self.pages_get_rv(pageid, newrvid)
+                newname = 'b/%s (revision %i)' % (filename, newrvid)
+                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()
+            if diff[-1] == '\n':
+                diff = diff[:-1]
+            return diff

Benjamin Mako Hill || Want to submit a patch?