hgweb: extract web substitutions table generation to own function

It doesn't use any state in hgweb except for the repo instance.
Move it to a standalone function.

mercurial/hgweb/hgweb_mod.py

@@ -1,483 +1,441
 # hgweb/hgweb_mod.py - Web interface for a repository.
 #
 # Copyright 21 May 2005 - (c) 2005 Jake Edge <[e-mail]>
 # Copyright 2005-2007 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
-import os, re
+import os
 from mercurial import ui, hg, hook, error, encoding, templater, util, repoview
 from mercurial.templatefilters import websub
 from mercurial.i18n import _
 from common import get_stat, ErrorResponse, permhooks, caching
 from common import HTTP_OK, HTTP_NOT_MODIFIED, HTTP_BAD_REQUEST
 from common import HTTP_NOT_FOUND, HTTP_SERVER_ERROR
 from request import wsgirequest
 import webcommands, protocol, webutil
 
 perms = {
     'changegroup': 'pull',
     'changegroupsubset': 'pull',
     'getbundle': 'pull',
     'stream_out': 'pull',
     'listkeys': 'pull',
     'unbundle': 'push',
     'pushkey': 'push',
 }
 
 ## Files of interest
 # Used to check if the repository has changed looking at mtime and size of
 # theses files. This should probably be relocated a bit higher in core.
 foi = [('spath', '00changelog.i'),
        ('spath', 'phaseroots'), # ! phase can change content at the same size
        ('spath', 'obsstore'),
        ('path', 'bookmarks'), # ! bookmark can change content at the same size
       ]
 
 def makebreadcrumb(url, prefix=''):
     '''Return a 'URL breadcrumb' list
 
     A 'URL breadcrumb' is a list of URL-name pairs,
     corresponding to each of the path items on a URL.
     This can be used to create path navigation entries.
     '''
     if url.endswith('/'):
         url = url[:-1]
     if prefix:
         url = '/' + prefix + url
     relpath = url
     if relpath.startswith('/'):
         relpath = relpath[1:]
 
     breadcrumb = []
     urlel = url
     pathitems = [''] + relpath.split('/')
     for pathel in reversed(pathitems):
         if not pathel or not urlel:
             break
         breadcrumb.append({'url': urlel, 'name': pathel})
         urlel = os.path.dirname(urlel)
     return reversed(breadcrumb)
 
-
 class requestcontext(object):
     """Holds state/context for an individual request.
 
     Servers can be multi-threaded. Holding state on the WSGI application
     is prone to race conditions. Instances of this class exist to hold
     mutable and race-free state for requests.
     """
     def __init__(self, app):
         object.__setattr__(self, 'app', app)
         object.__setattr__(self, 'repo', app.repo)
 
         object.__setattr__(self, 'archives', ('zip', 'gz', 'bz2'))
 
         object.__setattr__(self, 'maxchanges',
                            self.configint('web', 'maxchanges', 10))
         object.__setattr__(self, 'stripecount',
                            self.configint('web', 'stripes', 1))
         object.__setattr__(self, 'maxshortchanges',
                            self.configint('web', 'maxshortchanges', 60))
         object.__setattr__(self, 'maxfiles',
                            self.configint('web', 'maxfiles', 10))
         object.__setattr__(self, 'allowpull',
                            self.configbool('web', 'allowpull', True))
 
     # Proxy unknown reads and writes to the application instance
     # until everything is moved to us.
     def __getattr__(self, name):
         return getattr(self.app, name)
 
     def __setattr__(self, name, value):
         return setattr(self.app, name, value)
 
     # Servers are often run by a user different from the repo owner.
     # Trust the settings from the .hg/hgrc files by default.
     def config(self, section, name, default=None, untrusted=True):
         return self.repo.ui.config(section, name, default,
                                    untrusted=untrusted)
 
     def configbool(self, section, name, default=False, untrusted=True):
         return self.repo.ui.configbool(section, name, default,
                                        untrusted=untrusted)
 
     def configint(self, section, name, default=None, untrusted=True):
         return self.repo.ui.configint(section, name, default,
                                       untrusted=untrusted)
 
     def configlist(self, section, name, default=None, untrusted=True):
         return self.repo.ui.configlist(section, name, default,
                                        untrusted=untrusted)
 
     archivespecs = {
         'bz2': ('application/x-bzip2', 'tbz2', '.tar.bz2', None),
         'gz': ('application/x-gzip', 'tgz', '.tar.gz', None),
         'zip': ('application/zip', 'zip', '.zip', None),
     }
 
     def archivelist(self, nodeid):
         allowed = self.configlist('web', 'allow_archive')
         for typ, spec in self.archivespecs.iteritems():
             if typ in allowed or self.configbool('web', 'allow%s' % typ):
                 yield {'type': typ, 'extension': spec[2], 'node': nodeid}
 
 class hgweb(object):
     """HTTP server for individual repositories.
 
     Instances of this class serve HTTP responses for a particular
     repository.
 
     Instances are typically used as WSGI applications.
 
     Some servers are multi-threaded. On these servers, there may
     be multiple active threads inside __call__.
     """
     def __init__(self, repo, name=None, baseui=None):
         if isinstance(repo, str):
             if baseui:
                 u = baseui.copy()
             else:
                 u = ui.ui()
             r = hg.repository(u, repo)
         else:
             # we trust caller to give us a private copy
             r = repo
 
         r = self._getview(r)
         r.ui.setconfig('ui', 'report_untrusted', 'off', 'hgweb')
         r.baseui.setconfig('ui', 'report_untrusted', 'off', 'hgweb')
         r.ui.setconfig('ui', 'nontty', 'true', 'hgweb')
         r.baseui.setconfig('ui', 'nontty', 'true', 'hgweb')
         # displaying bundling progress bar while serving feel wrong and may
         # break some wsgi implementation.
         r.ui.setconfig('progress', 'disable', 'true', 'hgweb')
         r.baseui.setconfig('progress', 'disable', 'true', 'hgweb')
         self.repo = r
         hook.redirect(True)
         self.repostate = None
         self.mtime = -1
         self.reponame = name
         # we use untrusted=False to prevent a repo owner from using
         # web.templates in .hg/hgrc to get access to any file readable
         # by the user running the CGI script
         self.templatepath = self.config('web', 'templates', untrusted=False)
-        self.websubtable = self.loadwebsub()
+        self.websubtable = webutil.getwebsubs(r)
 
     # The CGI scripts are often run by a user different from the repo owner.
     # Trust the settings from the .hg/hgrc files by default.
     def config(self, section, name, default=None, untrusted=True):
         return self.repo.ui.config(section, name, default,
                                    untrusted=untrusted)
 
     def _getview(self, repo):
         """The 'web.view' config controls changeset filter to hgweb. Possible
         values are ``served``, ``visible`` and ``all``. Default is ``served``.
         The ``served`` filter only shows changesets that can be pulled from the
         hgweb instance.  The``visible`` filter includes secret changesets but
         still excludes "hidden" one.
 
         See the repoview module for details.
 
         The option has been around undocumented since Mercurial 2.5, but no
         user ever asked about it. So we better keep it undocumented for now."""
         viewconfig = repo.ui.config('web', 'view', 'served',
                                     untrusted=True)
         if viewconfig == 'all':
             return repo.unfiltered()
         elif viewconfig in repoview.filtertable:
             return repo.filtered(viewconfig)
         else:
             return repo.filtered('served')
 
     def refresh(self):
         repostate = []
         # file of interrests mtime and size
         for meth, fname in foi:
             prefix = getattr(self.repo, meth)
             st = get_stat(prefix, fname)
             repostate.append((st.st_mtime, st.st_size))
         repostate = tuple(repostate)
         # we need to compare file size in addition to mtime to catch
         # changes made less than a second ago
         if repostate != self.repostate:
             r = hg.repository(self.repo.baseui, self.repo.url())
             self.repo = self._getview(r)
             # update these last to avoid threads seeing empty settings
             self.repostate = repostate
             # mtime is needed for ETag
             self.mtime = st.st_mtime
 
     def run(self):
         """Start a server from CGI environment.
 
         Modern servers should be using WSGI and should avoid this
         method, if possible.
         """
         if not os.environ.get('GATEWAY_INTERFACE', '').startswith("CGI/1."):
             raise RuntimeError("This function is only intended to be "
                                "called while running as a CGI script.")
         import mercurial.hgweb.wsgicgi as wsgicgi
         wsgicgi.launch(self)
 
     def __call__(self, env, respond):
         """Run the WSGI application.
 
         This may be called by multiple threads.
         """
         req = wsgirequest(env, respond)
         return self.run_wsgi(req)
 
     def run_wsgi(self, req):
         """Internal method to run the WSGI application.
 
         This is typically only called by Mercurial. External consumers
         should be using instances of this class as the WSGI application.
         """
         self.refresh()
         rctx = requestcontext(self)
 
         # This state is global across all threads.
         encoding.encoding = rctx.config('web', 'encoding', encoding.encoding)
         rctx.repo.ui.environ = req.env
 
         # work with CGI variables to create coherent structure
         # use SCRIPT_NAME, PATH_INFO and QUERY_STRING as well as our REPO_NAME
 
         req.url = req.env['SCRIPT_NAME']
         if not req.url.endswith('/'):
             req.url += '/'
         if 'REPO_NAME' in req.env:
             req.url += req.env['REPO_NAME'] + '/'
 
         if 'PATH_INFO' in req.env:
             parts = req.env['PATH_INFO'].strip('/').split('/')
             repo_parts = req.env.get('REPO_NAME', '').split('/')
             if parts[:len(repo_parts)] == repo_parts:
                 parts = parts[len(repo_parts):]
             query = '/'.join(parts)
         else:
             query = req.env['QUERY_STRING'].split('&', 1)[0]
             query = query.split(';', 1)[0]
 
         # process this if it's a protocol request
         # protocol bits don't need to create any URLs
         # and the clients always use the old URL structure
 
         cmd = req.form.get('cmd', [''])[0]
         if protocol.iscmd(cmd):
             try:
                 if query:
                     raise ErrorResponse(HTTP_NOT_FOUND)
                 if cmd in perms:
                     self.check_perm(rctx, req, perms[cmd])
                 return protocol.call(self.repo, req, cmd)
             except ErrorResponse as inst:
                 # A client that sends unbundle without 100-continue will
                 # break if we respond early.
                 if (cmd == 'unbundle' and
                     (req.env.get('HTTP_EXPECT',
                                  '').lower() != '100-continue') or
                     req.env.get('X-HgHttp2', '')):
                     req.drain()
                 else:
                     req.headers.append(('Connection', 'Close'))
                 req.respond(inst, protocol.HGTYPE,
                             body='0\n%s\n' % inst.message)
                 return ''
 
         # translate user-visible url structure to internal structure
 
         args = query.split('/', 2)
         if 'cmd' not in req.form and args and args[0]:
 
             cmd = args.pop(0)
             style = cmd.rfind('-')
             if style != -1:
                 req.form['style'] = [cmd[:style]]
                 cmd = cmd[style + 1:]
 
             # avoid accepting e.g. style parameter as command
             if util.safehasattr(webcommands, cmd):
                 req.form['cmd'] = [cmd]
 
             if cmd == 'static':
                 req.form['file'] = ['/'.join(args)]
             else:
                 if args and args[0]:
                     node = args.pop(0).replace('%2F', '/')
                     req.form['node'] = [node]
                 if args:
                     req.form['file'] = args
 
             ua = req.env.get('HTTP_USER_AGENT', '')
             if cmd == 'rev' and 'mercurial' in ua:
                 req.form['style'] = ['raw']
 
             if cmd == 'archive':
                 fn = req.form['node'][0]
                 for type_, spec in rctx.archivespecs.iteritems():
                     ext = spec[2]
                     if fn.endswith(ext):
                         req.form['node'] = [fn[:-len(ext)]]
                         req.form['type'] = [type_]
 
         # process the web interface request
 
         try:
             tmpl = self.templater(req)
             ctype = tmpl('mimetype', encoding=encoding.encoding)
             ctype = templater.stringify(ctype)
 
             # check read permissions non-static content
             if cmd != 'static':
                 self.check_perm(rctx, req, None)
 
             if cmd == '':
                 req.form['cmd'] = [tmpl.cache['default']]
                 cmd = req.form['cmd'][0]
 
             if rctx.configbool('web', 'cache', True):
                 caching(self, req) # sets ETag header or raises NOT_MODIFIED
             if cmd not in webcommands.__all__:
                 msg = 'no such method: %s' % cmd
                 raise ErrorResponse(HTTP_BAD_REQUEST, msg)
             elif cmd == 'file' and 'raw' in req.form.get('style', []):
                 self.ctype = ctype
                 content = webcommands.rawfile(rctx, req, tmpl)
             else:
                 content = getattr(webcommands, cmd)(rctx, req, tmpl)
                 req.respond(HTTP_OK, ctype)
 
             return content
 
         except (error.LookupError, error.RepoLookupError) as err:
             req.respond(HTTP_NOT_FOUND, ctype)
             msg = str(err)
             if (util.safehasattr(err, 'name') and
                 not isinstance(err,  error.ManifestLookupError)):
                 msg = 'revision not found: %s' % err.name
             return tmpl('error', error=msg)
         except (error.RepoError, error.RevlogError) as inst:
             req.respond(HTTP_SERVER_ERROR, ctype)
             return tmpl('error', error=str(inst))
         except ErrorResponse as inst:
             req.respond(inst, ctype)
             if inst.code == HTTP_NOT_MODIFIED:
                 # Not allowed to return a body on a 304
                 return ['']
             return tmpl('error', error=inst.message)
 
-    def loadwebsub(self):
-        websubtable = []
-        websubdefs = self.repo.ui.configitems('websub')
-        # we must maintain interhg backwards compatibility
-        websubdefs += self.repo.ui.configitems('interhg')
-        for key, pattern in websubdefs:
-            # grab the delimiter from the character after the "s"
-            unesc = pattern[1]
-            delim = re.escape(unesc)
-
-            # identify portions of the pattern, taking care to avoid escaped
-            # delimiters. the replace format and flags are optional, but
-            # delimiters are required.
-            match = re.match(
-                r'^s%s(.+)(?:(?<=\\\\)|(?<!\\))%s(.*)%s([ilmsux])*$'
-                % (delim, delim, delim), pattern)
-            if not match:
-                self.repo.ui.warn(_("websub: invalid pattern for %s: %s\n")
-                                  % (key, pattern))
-                continue
-
-            # we need to unescape the delimiter for regexp and format
-            delim_re = re.compile(r'(?<!\\)\\%s' % delim)
-            regexp = delim_re.sub(unesc, match.group(1))
-            format = delim_re.sub(unesc, match.group(2))
-
-            # the pattern allows for 6 regexp flags, so set them if necessary
-            flagin = match.group(3)
-            flags = 0
-            if flagin:
-                for flag in flagin.upper():
-                    flags |= re.__dict__[flag]
-
-            try:
-                regexp = re.compile(regexp, flags)
-                websubtable.append((regexp, format))
-            except re.error:
-                self.repo.ui.warn(_("websub: invalid regexp for %s: %s\n")
-                                  % (key, regexp))
-        return websubtable
-
     def templater(self, req):
 
         # determine scheme, port and server name
         # this is needed to create absolute urls
 
         proto = req.env.get('wsgi.url_scheme')
         if proto == 'https':
             proto = 'https'
             default_port = "443"
         else:
             proto = 'http'
             default_port = "80"
 
         port = req.env["SERVER_PORT"]
         port = port != default_port and (":" + port) or ""
         urlbase = '%s://%s%s' % (proto, req.env['SERVER_NAME'], port)
         logourl = self.config("web", "logourl", "http://mercurial.selenic.com/")
         logoimg = self.config("web", "logoimg", "hglogo.png")
         staticurl = self.config("web", "staticurl") or req.url + 'static/'
         if not staticurl.endswith('/'):
             staticurl += '/'
 
         # some functions for the templater
 
         def motd(**map):
             yield self.config("web", "motd", "")
 
         # figure out which style to use
 
         vars = {}
         styles = (
             req.form.get('style', [None])[0],
             self.config('web', 'style'),
             'paper',
         )
         style, mapfile = templater.stylemap(styles, self.templatepath)
         if style == styles[0]:
             vars['style'] = style
 
         start = req.url[-1] == '?' and '&' or '?'
         sessionvars = webutil.sessionvars(vars, start)
 
         if not self.reponame:
             self.reponame = (self.config("web", "name")
                              or req.env.get('REPO_NAME')
                              or req.url.strip('/') or self.repo.root)
 
         def websubfilter(text):
             return websub(text, self.websubtable)
 
         # create the templater
 
         tmpl = templater.templater(mapfile,
                                    filters={"websub": websubfilter},
                                    defaults={"url": req.url,
                                              "logourl": logourl,
                                              "logoimg": logoimg,
                                              "staticurl": staticurl,
                                              "urlbase": urlbase,
                                              "repo": self.reponame,
                                              "encoding": encoding.encoding,
                                              "motd": motd,
                                              "sessionvars": sessionvars,
                                              "pathdef": makebreadcrumb(req.url),
                                              "style": style,
                                             })
         return tmpl
 
     def check_perm(self, rctx, req, op):
         for permhook in permhooks:
             permhook(rctx, req, op)

mercurial/hgweb/webutil.py

@@ -1,542 +1,584
 # hgweb/webutil.py - utility library for the web interface.
 #
 # Copyright 21 May 2005 - (c) 2005 Jake Edge <[e-mail]>
 # Copyright 2005-2007 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 import os, copy
+import re
 from mercurial import match, patch, error, ui, util, pathutil, context
 from mercurial.i18n import _
 from mercurial.node import hex, nullid, short
 from mercurial.templatefilters import revescape
 from common import ErrorResponse, paritygen
 from common import HTTP_NOT_FOUND
 import difflib
 
 def up(p):
     if p[0] != "/":
         p = "/" + p
     if p[-1] == "/":
         p = p[:-1]
     up = os.path.dirname(p)
     if up == "/":
         return "/"
     return up + "/"
 
 def _navseq(step, firststep=None):
     if firststep:
         yield firststep
         if firststep >= 20 and firststep <= 40:
             firststep = 50
             yield firststep
         assert step > 0
         assert firststep > 0
         while step <= firststep:
             step *= 10
     while True:
         yield 1 * step
         yield 3 * step
         step *= 10
 
 class revnav(object):
 
     def __init__(self, repo):
         """Navigation generation object
 
         :repo: repo object we generate nav for
         """
         # used for hex generation
         self._revlog = repo.changelog
 
     def __nonzero__(self):
         """return True if any revision to navigate over"""
         return self._first() is not None
 
     def _first(self):
         """return the minimum non-filtered changeset or None"""
         try:
             return iter(self._revlog).next()
         except StopIteration:
             return None
 
     def hex(self, rev):
         return hex(self._revlog.node(rev))
 
     def gen(self, pos, pagelen, limit):
         """computes label and revision id for navigation link
 
         :pos: is the revision relative to which we generate navigation.
         :pagelen: the size of each navigation page
         :limit: how far shall we link
 
         The return is:
             - a single element tuple
             - containing a dictionary with a `before` and `after` key
             - values are generator functions taking arbitrary number of kwargs
             - yield items are dictionaries with `label` and `node` keys
         """
         if not self:
             # empty repo
             return ({'before': (), 'after': ()},)
 
         targets = []
         for f in _navseq(1, pagelen):
             if f > limit:
                 break
             targets.append(pos + f)
             targets.append(pos - f)
         targets.sort()
 
         first = self._first()
         navbefore = [("(%i)" % first, self.hex(first))]
         navafter = []
         for rev in targets:
             if rev not in self._revlog:
                 continue
             if pos < rev < limit:
                 navafter.append(("+%d" % abs(rev - pos), self.hex(rev)))
             if 0 < rev < pos:
                 navbefore.append(("-%d" % abs(rev - pos), self.hex(rev)))
 
 
         navafter.append(("tip", "tip"))
 
         data = lambda i: {"label": i[0], "node": i[1]}
         return ({'before': lambda **map: (data(i) for i in navbefore),
                  'after':  lambda **map: (data(i) for i in navafter)},)
 
 class filerevnav(revnav):
 
     def __init__(self, repo, path):
         """Navigation generation object
 
         :repo: repo object we generate nav for
         :path: path of the file we generate nav for
         """
         # used for iteration
         self._changelog = repo.unfiltered().changelog
         # used for hex generation
         self._revlog = repo.file(path)
 
     def hex(self, rev):
         return hex(self._changelog.node(self._revlog.linkrev(rev)))
 
 
 def _siblings(siblings=[], hiderev=None):
     siblings = [s for s in siblings if s.node() != nullid]
     if len(siblings) == 1 and siblings[0].rev() == hiderev:
         return
     for s in siblings:
         d = {'node': s.hex(), 'rev': s.rev()}
         d['user'] = s.user()
         d['date'] = s.date()
         d['description'] = s.description()
         d['branch'] = s.branch()
         if util.safehasattr(s, 'path'):
             d['file'] = s.path()
         yield d
 
 def parents(ctx, hide=None):
     if isinstance(ctx, context.basefilectx):
         introrev = ctx.introrev()
         if ctx.changectx().rev() != introrev:
             return _siblings([ctx.repo()[introrev]], hide)
     return _siblings(ctx.parents(), hide)
 
 def children(ctx, hide=None):
     return _siblings(ctx.children(), hide)
 
 def renamelink(fctx):
     r = fctx.renamed()
     if r:
         return [{'file': r[0], 'node': hex(r[1])}]
     return []
 
 def nodetagsdict(repo, node):
     return [{"name": i} for i in repo.nodetags(node)]
 
 def nodebookmarksdict(repo, node):
     return [{"name": i} for i in repo.nodebookmarks(node)]
 
 def nodebranchdict(repo, ctx):
     branches = []
     branch = ctx.branch()
     # If this is an empty repo, ctx.node() == nullid,
     # ctx.branch() == 'default'.
     try:
         branchnode = repo.branchtip(branch)
     except error.RepoLookupError:
         branchnode = None
     if branchnode == ctx.node():
         branches.append({"name": branch})
     return branches
 
 def nodeinbranch(repo, ctx):
     branches = []
     branch = ctx.branch()
     try:
         branchnode = repo.branchtip(branch)
     except error.RepoLookupError:
         branchnode = None
     if branch != 'default' and branchnode != ctx.node():
         branches.append({"name": branch})
     return branches
 
 def nodebranchnodefault(ctx):
     branches = []
     branch = ctx.branch()
     if branch != 'default':
         branches.append({"name": branch})
     return branches
 
 def showtag(repo, tmpl, t1, node=nullid, **args):
     for t in repo.nodetags(node):
         yield tmpl(t1, tag=t, **args)
 
 def showbookmark(repo, tmpl, t1, node=nullid, **args):
     for t in repo.nodebookmarks(node):
         yield tmpl(t1, bookmark=t, **args)
 
 def branchentries(repo, stripecount, limit=0):
     tips = []
     heads = repo.heads()
     parity = paritygen(stripecount)
     sortkey = lambda item: (not item[1], item[0].rev())
 
     def entries(**map):
         count = 0
         if not tips:
             for tag, hs, tip, closed in repo.branchmap().iterbranches():
                 tips.append((repo[tip], closed))
         for ctx, closed in sorted(tips, key=sortkey, reverse=True):
             if limit > 0 and count >= limit:
                 return
             count += 1
             if closed:
                 status = 'closed'
             elif ctx.node() not in heads:
                 status = 'inactive'
             else:
                 status = 'open'
             yield {
                 'parity': parity.next(),
                 'branch': ctx.branch(),
                 'status': status,
                 'node': ctx.hex(),
                 'date': ctx.date()
             }
 
     return entries
 
 def cleanpath(repo, path):
     path = path.lstrip('/')
     return pathutil.canonpath(repo.root, '', path)
 
 def changeidctx(repo, changeid):
     try:
         ctx = repo[changeid]
     except error.RepoError:
         man = repo.manifest
         ctx = repo[man.linkrev(man.rev(man.lookup(changeid)))]
 
     return ctx
 
 def changectx(repo, req):
     changeid = "tip"
     if 'node' in req.form:
         changeid = req.form['node'][0]
         ipos = changeid.find(':')
         if ipos != -1:
             changeid = changeid[(ipos + 1):]
     elif 'manifest' in req.form:
         changeid = req.form['manifest'][0]
 
     return changeidctx(repo, changeid)
 
 def basechangectx(repo, req):
     if 'node' in req.form:
         changeid = req.form['node'][0]
         ipos = changeid.find(':')
         if ipos != -1:
             changeid = changeid[:ipos]
             return changeidctx(repo, changeid)
 
     return None
 
 def filectx(repo, req):
     if 'file' not in req.form:
         raise ErrorResponse(HTTP_NOT_FOUND, 'file not given')
     path = cleanpath(repo, req.form['file'][0])
     if 'node' in req.form:
         changeid = req.form['node'][0]
     elif 'filenode' in req.form:
         changeid = req.form['filenode'][0]
     else:
         raise ErrorResponse(HTTP_NOT_FOUND, 'node or filenode not given')
     try:
         fctx = repo[changeid][path]
     except error.RepoError:
         fctx = repo.filectx(path, fileid=changeid)
 
     return fctx
 
 def changelistentry(web, ctx, tmpl):
     '''Obtain a dictionary to be used for entries in a changelist.
 
     This function is called when producing items for the "entries" list passed
     to the "shortlog" and "changelog" templates.
     '''
     repo = web.repo
     rev = ctx.rev()
     n = ctx.node()
     showtags = showtag(repo, tmpl, 'changelogtag', n)
     files = listfilediffs(tmpl, ctx.files(), n, web.maxfiles)
 
     return {
         "author": ctx.user(),
         "parent": parents(ctx, rev - 1),
         "child": children(ctx, rev + 1),
         "changelogtag": showtags,
         "desc": ctx.description(),
         "extra": ctx.extra(),
         "date": ctx.date(),
         "files": files,
         "rev": rev,
         "node": hex(n),
         "tags": nodetagsdict(repo, n),
         "bookmarks": nodebookmarksdict(repo, n),
         "inbranch": nodeinbranch(repo, ctx),
         "branches": nodebranchdict(repo, ctx)
     }
 
 def symrevorshortnode(req, ctx):
     if 'node' in req.form:
         return revescape(req.form['node'][0])
     else:
         return short(ctx.node())
 
 def changesetentry(web, req, tmpl, ctx):
     '''Obtain a dictionary to be used to render the "changeset" template.'''
 
     showtags = showtag(web.repo, tmpl, 'changesettag', ctx.node())
     showbookmarks = showbookmark(web.repo, tmpl, 'changesetbookmark',
                                  ctx.node())
     showbranch = nodebranchnodefault(ctx)
 
     files = []
     parity = paritygen(web.stripecount)
     for blockno, f in enumerate(ctx.files()):
         template = f in ctx and 'filenodelink' or 'filenolink'
         files.append(tmpl(template,
                           node=ctx.hex(), file=f, blockno=blockno + 1,
                           parity=parity.next()))
 
     basectx = basechangectx(web.repo, req)
     if basectx is None:
         basectx = ctx.p1()
 
     style = web.config('web', 'style', 'paper')
     if 'style' in req.form:
         style = req.form['style'][0]
 
     parity = paritygen(web.stripecount)
     diff = diffs(web.repo, tmpl, ctx, basectx, None, parity, style)
 
     parity = paritygen(web.stripecount)
     diffstatsgen = diffstatgen(ctx, basectx)
     diffstats = diffstat(tmpl, ctx, diffstatsgen, parity)
 
     return dict(
         diff=diff,
         rev=ctx.rev(),
         node=ctx.hex(),
         symrev=symrevorshortnode(req, ctx),
         parent=tuple(parents(ctx)),
         child=children(ctx),
         basenode=basectx.hex(),
         changesettag=showtags,
         changesetbookmark=showbookmarks,
         changesetbranch=showbranch,
         author=ctx.user(),
         desc=ctx.description(),
         extra=ctx.extra(),
         date=ctx.date(),
         phase=ctx.phasestr(),
         files=files,
         diffsummary=lambda **x: diffsummary(diffstatsgen),
         diffstat=diffstats,
         archives=web.archivelist(ctx.hex()),
         tags=nodetagsdict(web.repo, ctx.node()),
         bookmarks=nodebookmarksdict(web.repo, ctx.node()),
         branch=showbranch,
         inbranch=nodeinbranch(web.repo, ctx),
         branches=nodebranchdict(web.repo, ctx))
 
 def listfilediffs(tmpl, files, node, max):
     for f in files[:max]:
         yield tmpl('filedifflink', node=hex(node), file=f)
     if len(files) > max:
         yield tmpl('fileellipses')
 
 def diffs(repo, tmpl, ctx, basectx, files, parity, style):
 
     def countgen():
         start = 1
         while True:
             yield start
             start += 1
 
     blockcount = countgen()
     def prettyprintlines(diff, blockno):
         for lineno, l in enumerate(diff.splitlines(True)):
             difflineno = "%d.%d" % (blockno, lineno + 1)
             if l.startswith('+'):
                 ltype = "difflineplus"
             elif l.startswith('-'):
                 ltype = "difflineminus"
             elif l.startswith('@'):
                 ltype = "difflineat"
             else:
                 ltype = "diffline"
             yield tmpl(ltype,
                        line=l,
                        lineno=lineno + 1,
                        lineid="l%s" % difflineno,
                        linenumber="% 8s" % difflineno)
 
     if files:
         m = match.exact(repo.root, repo.getcwd(), files)
     else:
         m = match.always(repo.root, repo.getcwd())
 
     diffopts = patch.diffopts(repo.ui, untrusted=True)
     if basectx is None:
         parents = ctx.parents()
         if parents:
             node1 = parents[0].node()
         else:
             node1 = nullid
     else:
         node1 = basectx.node()
     node2 = ctx.node()
 
     block = []
     for chunk in patch.diff(repo, node1, node2, m, opts=diffopts):
         if chunk.startswith('diff') and block:
             blockno = blockcount.next()
             yield tmpl('diffblock', parity=parity.next(), blockno=blockno,
                        lines=prettyprintlines(''.join(block), blockno))
             block = []
         if chunk.startswith('diff') and style != 'raw':
             chunk = ''.join(chunk.splitlines(True)[1:])
         block.append(chunk)
     blockno = blockcount.next()
     yield tmpl('diffblock', parity=parity.next(), blockno=blockno,
                lines=prettyprintlines(''.join(block), blockno))
 
 def compare(tmpl, context, leftlines, rightlines):
     '''Generator function that provides side-by-side comparison data.'''
 
     def compline(type, leftlineno, leftline, rightlineno, rightline):
         lineid = leftlineno and ("l%s" % leftlineno) or ''
         lineid += rightlineno and ("r%s" % rightlineno) or ''
         return tmpl('comparisonline',
                     type=type,
                     lineid=lineid,
                     leftlineno=leftlineno,
                     leftlinenumber="% 6s" % (leftlineno or ''),
                     leftline=leftline or '',
                     rightlineno=rightlineno,
                     rightlinenumber="% 6s" % (rightlineno or ''),
                     rightline=rightline or '')
 
     def getblock(opcodes):
         for type, llo, lhi, rlo, rhi in opcodes:
             len1 = lhi - llo
             len2 = rhi - rlo
             count = min(len1, len2)
             for i in xrange(count):
                 yield compline(type=type,
                                leftlineno=llo + i + 1,
                                leftline=leftlines[llo + i],
                                rightlineno=rlo + i + 1,
                                rightline=rightlines[rlo + i])
             if len1 > len2:
                 for i in xrange(llo + count, lhi):
                     yield compline(type=type,
                                    leftlineno=i + 1,
                                    leftline=leftlines[i],
                                    rightlineno=None,
                                    rightline=None)
             elif len2 > len1:
                 for i in xrange(rlo + count, rhi):
                     yield compline(type=type,
                                    leftlineno=None,
                                    leftline=None,
                                    rightlineno=i + 1,
                                    rightline=rightlines[i])
 
     s = difflib.SequenceMatcher(None, leftlines, rightlines)
     if context < 0:
         yield tmpl('comparisonblock', lines=getblock(s.get_opcodes()))
     else:
         for oc in s.get_grouped_opcodes(n=context):
             yield tmpl('comparisonblock', lines=getblock(oc))
 
 def diffstatgen(ctx, basectx):
     '''Generator function that provides the diffstat data.'''
 
     stats = patch.diffstatdata(util.iterlines(ctx.diff(basectx)))
     maxname, maxtotal, addtotal, removetotal, binary = patch.diffstatsum(stats)
     while True:
         yield stats, maxname, maxtotal, addtotal, removetotal, binary
 
 def diffsummary(statgen):
     '''Return a short summary of the diff.'''
 
     stats, maxname, maxtotal, addtotal, removetotal, binary = statgen.next()
     return _(' %d files changed, %d insertions(+), %d deletions(-)\n') % (
              len(stats), addtotal, removetotal)
 
 def diffstat(tmpl, ctx, statgen, parity):
     '''Return a diffstat template for each file in the diff.'''
 
     stats, maxname, maxtotal, addtotal, removetotal, binary = statgen.next()
     files = ctx.files()
 
     def pct(i):
         if maxtotal == 0:
             return 0
         return (float(i) / maxtotal) * 100
 
     fileno = 0
     for filename, adds, removes, isbinary in stats:
         template = filename in files and 'diffstatlink' or 'diffstatnolink'
         total = adds + removes
         fileno += 1
         yield tmpl(template, node=ctx.hex(), file=filename, fileno=fileno,
                    total=total, addpct=pct(adds), removepct=pct(removes),
                    parity=parity.next())
 
 class sessionvars(object):
     def __init__(self, vars, start='?'):
         self.start = start
         self.vars = vars
     def __getitem__(self, key):
         return self.vars[key]
     def __setitem__(self, key, value):
         self.vars[key] = value
     def __copy__(self):
         return sessionvars(copy.copy(self.vars), self.start)
     def __iter__(self):
         separator = self.start
         for key, value in sorted(self.vars.iteritems()):
             yield {'name': key, 'value': str(value), 'separator': separator}
             separator = '&'
 
 class wsgiui(ui.ui):
     # default termwidth breaks under mod_wsgi
     def termwidth(self):
         return 80
+
+def getwebsubs(repo):
+    websubtable = []
+    websubdefs = repo.ui.configitems('websub')
+    # we must maintain interhg backwards compatibility
+    websubdefs += repo.ui.configitems('interhg')
+    for key, pattern in websubdefs:
+        # grab the delimiter from the character after the "s"
+        unesc = pattern[1]
+        delim = re.escape(unesc)
+
+        # identify portions of the pattern, taking care to avoid escaped
+        # delimiters. the replace format and flags are optional, but
+        # delimiters are required.
+        match = re.match(
+            r'^s%s(.+)(?:(?<=\\\\)|(?<!\\))%s(.*)%s([ilmsux])*$'
+            % (delim, delim, delim), pattern)
+        if not match:
+            repo.ui.warn(_("websub: invalid pattern for %s: %s\n")
+                              % (key, pattern))
+            continue
+
+        # we need to unescape the delimiter for regexp and format
+        delim_re = re.compile(r'(?<!\\)\\%s' % delim)
+        regexp = delim_re.sub(unesc, match.group(1))
+        format = delim_re.sub(unesc, match.group(2))
+
+        # the pattern allows for 6 regexp flags, so set them if necessary
+        flagin = match.group(3)
+        flags = 0
+        if flagin:
+            for flag in flagin.upper():
+                flags |= re.__dict__[flag]
+
+        try:
+            regexp = re.compile(regexp, flags)
+            websubtable.append((regexp, format))
+        except re.error:
+            repo.ui.warn(_("websub: invalid regexp for %s: %s\n")
+                         % (key, regexp))
+    return websubtable