hgwebdir: export collections of repos

now you can use [collections] section in hgweb.config to export entire
tree of repos.

hgwebdir.cgi

@@ -1,20 +1,31
 #!/usr/bin/env python
 #
 # An example CGI script to export multiple hgweb repos, edit as necessary
 
 import cgitb, sys
 cgitb.enable()
 
 # sys.path.insert(0, "/path/to/python/lib") # if not a system-wide install
 from mercurial import hgweb
 
-# The config file looks like this:
+# The config file looks like this.  You can have paths to individual
+# repos, collections of repos in a directory tree, or both.
+#
 # [paths]
 # virtual/path = /real/path
 # virtual/path = /real/path
+#
+# [collections]
+# /prefix/to/strip/off = /root/of/tree/full/of/repos
+#
+# collections example: say directory tree /foo contains repos /foo/bar,
+# /foo/quux/baz.  Give this config section:
+#   [collections]
+#   /foo = /foo
+# Then repos will list as bar and quux/baz.
 
 # Alternatively you can pass a list of ('virtual/path', '/real/path') tuples
 # or use a dictionary with entries like 'virtual/path': '/real/path'
 
 h = hgweb.hgwebdir("hgweb.config")
 h.run()

mercurial/hgweb.py

@@ -1,1145 +1,1155
 # hgweb.py - web interface to a mercurial repository
 #
 # Copyright 21 May 2005 - (c) 2005 Jake Edge <[e-mail]>
 # Copyright 2005 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms
 # of the GNU General Public License, incorporated herein by reference.
 
 import os, cgi, sys, urllib
 from demandload import demandload
 demandload(globals(), "mdiff time re socket zlib errno ui hg ConfigParser")
 demandload(globals(), "zipfile tempfile StringIO tarfile BaseHTTPServer util")
 demandload(globals(), "mimetypes")
 from node import *
 from i18n import gettext as _
 
 def templatepath():
     for f in "templates", "../templates":
         p = os.path.join(os.path.dirname(__file__), f)
         if os.path.isdir(p):
             return p
 
 def age(x):
     def plural(t, c):
         if c == 1:
             return t
         return t + "s"
     def fmt(t, c):
         return "%d %s" % (c, plural(t, c))
 
     now = time.time()
     then = x[0]
     delta = max(1, int(now - then))
 
     scales = [["second", 1],
               ["minute", 60],
               ["hour", 3600],
               ["day", 3600 * 24],
               ["week", 3600 * 24 * 7],
               ["month", 3600 * 24 * 30],
               ["year", 3600 * 24 * 365]]
 
     scales.reverse()
 
     for t, s in scales:
         n = delta / s
         if n >= 2 or s == 1:
             return fmt(t, n)
 
 def nl2br(text):
     return text.replace('\n', '<br/>\n')
 
 def obfuscate(text):
     return ''.join(['&#%d;' % ord(c) for c in text])
 
 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 get_mtime(repo_path):
     hg_path = os.path.join(repo_path, ".hg")
     cl_path = os.path.join(hg_path, "00changelog.i")
     if os.path.exists(os.path.join(cl_path)):
         return os.stat(cl_path).st_mtime
     else:
         return os.stat(hg_path).st_mtime
 
 class hgrequest(object):
     def __init__(self, inp=None, out=None, env=None):
         self.inp = inp or sys.stdin
         self.out = out or sys.stdout
         self.env = env or os.environ
         self.form = cgi.parse(self.inp, self.env, keep_blank_values=1)
 
     def write(self, *things):
         for thing in things:
             if hasattr(thing, "__iter__"):
                 for part in thing:
                     self.write(part)
             else:
                 try:
                     self.out.write(str(thing))
                 except socket.error, inst:
                     if inst[0] != errno.ECONNRESET:
                         raise
 
     def header(self, headers=[('Content-type','text/html')]):
         for header in headers:
             self.out.write("%s: %s\r\n" % header)
         self.out.write("\r\n")
 
     def httphdr(self, type, file="", size=0):
 
         headers = [('Content-type', type)]
         if file:
             headers.append(('Content-disposition', 'attachment; filename=%s' % file))
         if size > 0:
             headers.append(('Content-length', str(size)))
         self.header(headers)
 
 class templater(object):
     def __init__(self, mapfile, filters={}, defaults={}):
         self.cache = {}
         self.map = {}
         self.base = os.path.dirname(mapfile)
         self.filters = filters
         self.defaults = defaults
 
         for l in file(mapfile):
             m = re.match(r'(\S+)\s*=\s*"(.*)"$', l)
             if m:
                 self.cache[m.group(1)] = m.group(2)
             else:
                 m = re.match(r'(\S+)\s*=\s*(\S+)', l)
                 if m:
                     self.map[m.group(1)] = os.path.join(self.base, m.group(2))
                 else:
                     raise LookupError(_("unknown map entry '%s'") % l)
 
     def __call__(self, t, **map):
         m = self.defaults.copy()
         m.update(map)
         try:
             tmpl = self.cache[t]
         except KeyError:
             tmpl = self.cache[t] = file(self.map[t]).read()
         return self.template(tmpl, self.filters, **m)
 
     def template(self, tmpl, filters={}, **map):
         while tmpl:
             m = re.search(r"#([a-zA-Z0-9]+)((%[a-zA-Z0-9]+)*)((\|[a-zA-Z0-9]+)*)#", tmpl)
             if m:
                 yield tmpl[:m.start(0)]
                 v = map.get(m.group(1), "")
                 v = callable(v) and v(**map) or v
 
                 format = m.group(2)
                 fl = m.group(4)
 
                 if format:
                     q = v.__iter__
                     for i in q():
                         lm = map.copy()
                         lm.update(i)
                         yield self(format[1:], **lm)
 
                     v = ""
 
                 elif fl:
                     for f in fl.split("|")[1:]:
                         v = filters[f](v)
 
                 yield v
                 tmpl = tmpl[m.end(0):]
             else:
                 yield tmpl
                 return
 
 common_filters = {
     "escape": lambda x: cgi.escape(x, True),
     "urlescape": urllib.quote,
     "strip": lambda x: x.strip(),
     "age": age,
     "date": lambda x: util.datestr(x),
     "addbreaks": nl2br,
     "obfuscate": obfuscate,
     "short": (lambda x: x[:12]),
     "firstline": (lambda x: x.splitlines(1)[0]),
     "permissions": (lambda x: x and "-rwxr-xr-x" or "-rw-r--r--"),
     "rfc822date": lambda x: util.datestr(x, "%a, %d %b %Y %H:%M:%S"),
     }
 
 class hgweb(object):
     def __init__(self, repo, name=None):
         if type(repo) == type(""):
             self.repo = hg.repository(ui.ui(), repo)
         else:
             self.repo = repo
 
         self.mtime = -1
         self.reponame = name
         self.archives = 'zip', 'gz', 'bz2'
 
     def refresh(self):
         mtime = get_mtime(self.repo.root)
         if mtime != self.mtime:
             self.mtime = mtime
             self.repo = hg.repository(self.repo.ui, self.repo.root)
             self.maxchanges = int(self.repo.ui.config("web", "maxchanges", 10))
             self.maxfiles = int(self.repo.ui.config("web", "maxfiles", 10))
             self.allowpull = self.repo.ui.configbool("web", "allowpull", True)
 
     def archivelist(self, nodeid):
         for i in self.archives:
             if self.repo.ui.configbool("web", "allow" + i, False):
                 yield {"type" : i, "node" : nodeid}
 
     def listfiles(self, files, mf):
         for f in files[:self.maxfiles]:
             yield self.t("filenodelink", node=hex(mf[f]), file=f)
         if len(files) > self.maxfiles:
             yield self.t("fileellipses")
 
     def listfilediffs(self, files, changeset):
         for f in files[:self.maxfiles]:
             yield self.t("filedifflink", node=hex(changeset), file=f)
         if len(files) > self.maxfiles:
             yield self.t("fileellipses")
 
     def siblings(self, siblings=[], rev=None, hiderev=None, **args):
         if not rev:
             rev = lambda x: ""
         siblings = [s for s in siblings if s != nullid]
         if len(siblings) == 1 and rev(siblings[0]) == hiderev:
             return
         for s in siblings:
             yield dict(node=hex(s), rev=rev(s), **args)
 
     def renamelink(self, fl, node):
         r = fl.renamed(node)
         if r:
             return [dict(file=r[0], node=hex(r[1]))]
         return []
 
     def showtag(self, t1, node=nullid, **args):
         for t in self.repo.nodetags(node):
              yield self.t(t1, tag=t, **args)
 
     def diff(self, node1, node2, files):
         def filterfiles(filters, files):
             l = [x for x in files if x in filters]
 
             for t in filters:
                 if t and t[-1] != os.sep:
                     t += os.sep
                 l += [x for x in files if x.startswith(t)]
             return l
 
         parity = [0]
         def diffblock(diff, f, fn):
             yield self.t("diffblock",
                          lines=prettyprintlines(diff),
                          parity=parity[0],
                          file=f,
                          filenode=hex(fn or nullid))
             parity[0] = 1 - parity[0]
 
         def prettyprintlines(diff):
             for l in diff.splitlines(1):
                 if l.startswith('+'):
                     yield self.t("difflineplus", line=l)
                 elif l.startswith('-'):
                     yield self.t("difflineminus", line=l)
                 elif l.startswith('@'):
                     yield self.t("difflineat", line=l)
                 else:
                     yield self.t("diffline", line=l)
 
         r = self.repo
         cl = r.changelog
         mf = r.manifest
         change1 = cl.read(node1)
         change2 = cl.read(node2)
         mmap1 = mf.read(change1[0])
         mmap2 = mf.read(change2[0])
         date1 = util.datestr(change1[2])
         date2 = util.datestr(change2[2])
 
         modified, added, removed, deleted, unknown = r.changes(node1, node2)
         if files:
             modified, added, removed = map(lambda x: filterfiles(files, x),
                                            (modified, added, removed))
 
         diffopts = self.repo.ui.diffopts()
         showfunc = diffopts['showfunc']
         ignorews = diffopts['ignorews']
         for f in modified:
             to = r.file(f).read(mmap1[f])
             tn = r.file(f).read(mmap2[f])
             yield diffblock(mdiff.unidiff(to, date1, tn, date2, f,
                             showfunc=showfunc, ignorews=ignorews), f, tn)
         for f in added:
             to = None
             tn = r.file(f).read(mmap2[f])
             yield diffblock(mdiff.unidiff(to, date1, tn, date2, f,
                             showfunc=showfunc, ignorews=ignorews), f, tn)
         for f in removed:
             to = r.file(f).read(mmap1[f])
             tn = None
             yield diffblock(mdiff.unidiff(to, date1, tn, date2, f,
                             showfunc=showfunc, ignorews=ignorews), f, tn)
 
     def changelog(self, pos):
         def changenav(**map):
             def seq(factor, maxchanges=None):
                 if maxchanges:
                     yield maxchanges
                     if maxchanges >= 20 and maxchanges <= 40:
                         yield 50
                 else:
                     yield 1 * factor
                     yield 3 * factor
                 for f in seq(factor * 10):
                     yield f
 
             l = []
             last = 0
             for f in seq(1, self.maxchanges):
                 if f < self.maxchanges or f <= last:
                     continue
                 if f > count:
                     break
                 last = f
                 r = "%d" % f
                 if pos + f < count:
                     l.append(("+" + r, pos + f))
                 if pos - f >= 0:
                     l.insert(0, ("-" + r, pos - f))
 
             yield {"rev": 0, "label": "(0)"}
 
             for label, rev in l:
                 yield {"label": label, "rev": rev}
 
             yield {"label": "tip", "rev": "tip"}
 
         def changelist(**map):
             parity = (start - end) & 1
             cl = self.repo.changelog
             l = [] # build a list in forward order for efficiency
             for i in range(start, end):
                 n = cl.node(i)
                 changes = cl.read(n)
                 hn = hex(n)
 
                 l.insert(0, {"parity": parity,
                              "author": changes[1],
                              "parent": self.siblings(cl.parents(n), cl.rev,
                                                      cl.rev(n) - 1),
                              "child": self.siblings(cl.children(n), cl.rev,
                                                     cl.rev(n) + 1),
                              "changelogtag": self.showtag("changelogtag",n),
                              "manifest": hex(changes[0]),
                              "desc": changes[4],
                              "date": changes[2],
                              "files": self.listfilediffs(changes[3], n),
                              "rev": i,
                              "node": hn})
                 parity = 1 - parity
 
             for e in l:
                 yield e
 
         cl = self.repo.changelog
         mf = cl.read(cl.tip())[0]
         count = cl.count()
         start = max(0, pos - self.maxchanges + 1)
         end = min(count, start + self.maxchanges)
         pos = end - 1
 
         yield self.t('changelog',
                      changenav=changenav,
                      manifest=hex(mf),
                      rev=pos, changesets=count, entries=changelist)
 
     def search(self, query):
 
         def changelist(**map):
             cl = self.repo.changelog
             count = 0
             qw = query.lower().split()
 
             def revgen():
                 for i in range(cl.count() - 1, 0, -100):
                     l = []
                     for j in range(max(0, i - 100), i):
                         n = cl.node(j)
                         changes = cl.read(n)
                         l.append((n, j, changes))
                     l.reverse()
                     for e in l:
                         yield e
 
             for n, i, changes in revgen():
                 miss = 0
                 for q in qw:
                     if not (q in changes[1].lower() or
                             q in changes[4].lower() or
                             q in " ".join(changes[3][:20]).lower()):
                         miss = 1
                         break
                 if miss:
                     continue
 
                 count += 1
                 hn = hex(n)
 
                 yield self.t('searchentry',
                              parity=count & 1,
                              author=changes[1],
                              parent=self.siblings(cl.parents(n), cl.rev),
                              child=self.siblings(cl.children(n), cl.rev),
                              changelogtag=self.showtag("changelogtag",n),
                              manifest=hex(changes[0]),
                              desc=changes[4],
                              date=changes[2],
                              files=self.listfilediffs(changes[3], n),
                              rev=i,
                              node=hn)
 
                 if count >= self.maxchanges:
                     break
 
         cl = self.repo.changelog
         mf = cl.read(cl.tip())[0]
 
         yield self.t('search',
                      query=query,
                      manifest=hex(mf),
                      entries=changelist)
 
     def changeset(self, nodeid):
         cl = self.repo.changelog
         n = self.repo.lookup(nodeid)
         nodeid = hex(n)
         changes = cl.read(n)
         p1 = cl.parents(n)[0]
 
         files = []
         mf = self.repo.manifest.read(changes[0])
         for f in changes[3]:
             files.append(self.t("filenodelink",
                                 filenode=hex(mf.get(f, nullid)), file=f))
 
         def diff(**map):
             yield self.diff(p1, n, None)
 
         yield self.t('changeset',
                      diff=diff,
                      rev=cl.rev(n),
                      node=nodeid,
                      parent=self.siblings(cl.parents(n), cl.rev),
                      child=self.siblings(cl.children(n), cl.rev),
                      changesettag=self.showtag("changesettag",n),
                      manifest=hex(changes[0]),
                      author=changes[1],
                      desc=changes[4],
                      date=changes[2],
                      files=files,
                      archives=self.archivelist(nodeid))
 
     def filelog(self, f, filenode):
         cl = self.repo.changelog
         fl = self.repo.file(f)
         filenode = hex(fl.lookup(filenode))
         count = fl.count()
 
         def entries(**map):
             l = []
             parity = (count - 1) & 1
 
             for i in range(count):
                 n = fl.node(i)
                 lr = fl.linkrev(n)
                 cn = cl.node(lr)
                 cs = cl.read(cl.node(lr))
 
                 l.insert(0, {"parity": parity,
                              "filenode": hex(n),
                              "filerev": i,
                              "file": f,
                              "node": hex(cn),
                              "author": cs[1],
                              "date": cs[2],
                              "rename": self.renamelink(fl, n),
                              "parent": self.siblings(fl.parents(n),
                                                      fl.rev, file=f),
                              "child": self.siblings(fl.children(n),
                                                     fl.rev, file=f),
                              "desc": cs[4]})
                 parity = 1 - parity
 
             for e in l:
                 yield e
 
         yield self.t("filelog", file=f, filenode=filenode, entries=entries)
 
     def filerevision(self, f, node):
         fl = self.repo.file(f)
         n = fl.lookup(node)
         node = hex(n)
         text = fl.read(n)
         changerev = fl.linkrev(n)
         cl = self.repo.changelog
         cn = cl.node(changerev)
         cs = cl.read(cn)
         mfn = cs[0]
 
         mt = mimetypes.guess_type(f)[0]
         rawtext = text
         if util.binary(text):
             text = "(binary:%s)" % mt
 
         def lines():
             for l, t in enumerate(text.splitlines(1)):
                 yield {"line": t,
                        "linenumber": "% 6d" % (l + 1),
                        "parity": l & 1}
 
         yield self.t("filerevision",
                      file=f,
                      filenode=node,
                      path=up(f),
                      text=lines(),
                      raw=rawtext,
                      mimetype=mt,
                      rev=changerev,
                      node=hex(cn),
                      manifest=hex(mfn),
                      author=cs[1],
                      date=cs[2],
                      parent=self.siblings(fl.parents(n), fl.rev, file=f),
                      child=self.siblings(fl.children(n), fl.rev, file=f),
                      rename=self.renamelink(fl, n),
                      permissions=self.repo.manifest.readflags(mfn)[f])
 
     def fileannotate(self, f, node):
         bcache = {}
         ncache = {}
         fl = self.repo.file(f)
         n = fl.lookup(node)
         node = hex(n)
         changerev = fl.linkrev(n)
 
         cl = self.repo.changelog
         cn = cl.node(changerev)
         cs = cl.read(cn)
         mfn = cs[0]
 
         def annotate(**map):
             parity = 1
             last = None
             for r, l in fl.annotate(n):
                 try:
                     cnode = ncache[r]
                 except KeyError:
                     cnode = ncache[r] = self.repo.changelog.node(r)
 
                 try:
                     name = bcache[r]
                 except KeyError:
                     cl = self.repo.changelog.read(cnode)
                     bcache[r] = name = self.repo.ui.shortuser(cl[1])
 
                 if last != cnode:
                     parity = 1 - parity
                     last = cnode
 
                 yield {"parity": parity,
                        "node": hex(cnode),
                        "rev": r,
                        "author": name,
                        "file": f,
                        "line": l}
 
         yield self.t("fileannotate",
                      file=f,
                      filenode=node,
                      annotate=annotate,
                      path=up(f),
                      rev=changerev,
                      node=hex(cn),
                      manifest=hex(mfn),
                      author=cs[1],
                      date=cs[2],
                      rename=self.renamelink(fl, n),
                      parent=self.siblings(fl.parents(n), fl.rev, file=f),
                      child=self.siblings(fl.children(n), fl.rev, file=f),
                      permissions=self.repo.manifest.readflags(mfn)[f])
 
     def manifest(self, mnode, path):
         man = self.repo.manifest
         mn = man.lookup(mnode)
         mnode = hex(mn)
         mf = man.read(mn)
         rev = man.rev(mn)
         node = self.repo.changelog.node(rev)
         mff = man.readflags(mn)
 
         files = {}
 
         p = path[1:]
         if p and p[-1] != "/":
             p += "/"
         l = len(p)
 
         for f,n in mf.items():
             if f[:l] != p:
                 continue
             remain = f[l:]
             if "/" in remain:
                 short = remain[:remain.find("/") + 1] # bleah
                 files[short] = (f, None)
             else:
                 short = os.path.basename(remain)
                 files[short] = (f, n)
 
         def filelist(**map):
             parity = 0
             fl = files.keys()
             fl.sort()
             for f in fl:
                 full, fnode = files[f]
                 if not fnode:
                     continue
 
                 yield {"file": full,
                        "manifest": mnode,
                        "filenode": hex(fnode),
                        "parity": parity,
                        "basename": f,
                        "permissions": mff[full]}
                 parity = 1 - parity
 
         def dirlist(**map):
             parity = 0
             fl = files.keys()
             fl.sort()
             for f in fl:
                 full, fnode = files[f]
                 if fnode:
                     continue
 
                 yield {"parity": parity,
                        "path": os.path.join(path, f),
                        "manifest": mnode,
                        "basename": f[:-1]}
                 parity = 1 - parity
 
         yield self.t("manifest",
                      manifest=mnode,
                      rev=rev,
                      node=hex(node),
                      path=path,
                      up=up(path),
                      fentries=filelist,
                      dentries=dirlist,
                      archives=self.archivelist(hex(node)))
 
     def tags(self):
         cl = self.repo.changelog
         mf = cl.read(cl.tip())[0]
 
         i = self.repo.tagslist()
         i.reverse()
 
         def entries(**map):
             parity = 0
             for k,n in i:
                 yield {"parity": parity,
                        "tag": k,
                        "tagmanifest": hex(cl.read(n)[0]),
                        "date": cl.read(n)[2],
                        "node": hex(n)}
                 parity = 1 - parity
 
         yield self.t("tags",
                      manifest=hex(mf),
                      entries=entries)
 
     def summary(self):
         cl = self.repo.changelog
         mf = cl.read(cl.tip())[0]
 
         i = self.repo.tagslist()
         i.reverse()
 
         def tagentries(**map):
             parity = 0
             count = 0
             for k,n in i:
                 if k == "tip": # skip tip
                     continue;
 
                 count += 1
                 if count > 10: # limit to 10 tags
                     break;
 
                 c = cl.read(n)
                 m = c[0]
                 t = c[2]
 
                 yield self.t("tagentry",
                              parity = parity,
                              tag = k,
                              node = hex(n),
                              date = t,
                              tagmanifest = hex(m))
                 parity = 1 - parity
 
         def changelist(**map):
             parity = 0
             cl = self.repo.changelog
             l = [] # build a list in forward order for efficiency
             for i in range(start, end):
                 n = cl.node(i)
                 changes = cl.read(n)
                 hn = hex(n)
                 t = changes[2]
 
                 l.insert(0, self.t(
                     'shortlogentry',
                     parity = parity,
                     author = changes[1],
                     manifest = hex(changes[0]),
                     desc = changes[4],
                     date = t,
                     rev = i,
                     node = hn))
                 parity = 1 - parity
 
             yield l
 
         cl = self.repo.changelog
         mf = cl.read(cl.tip())[0]
         count = cl.count()
         start = max(0, count - self.maxchanges)
         end = min(count, start + self.maxchanges)
         pos = end - 1
 
         yield self.t("summary",
                  desc = self.repo.ui.config("web", "description", "unknown"),
                  owner = (self.repo.ui.config("ui", "username") or # preferred
                           self.repo.ui.config("web", "contact") or # deprecated
                           self.repo.ui.config("web", "author", "unknown")), # also
                  lastchange = (0, 0), # FIXME
                  manifest = hex(mf),
                  tags = tagentries,
                  shortlog = changelist)
 
     def filediff(self, file, changeset):
         cl = self.repo.changelog
         n = self.repo.lookup(changeset)
         changeset = hex(n)
         p1 = cl.parents(n)[0]
         cs = cl.read(n)
         mf = self.repo.manifest.read(cs[0])
 
         def diff(**map):
             yield self.diff(p1, n, file)
 
         yield self.t("filediff",
                      file=file,
                      filenode=hex(mf.get(file, nullid)),
                      node=changeset,
                      rev=self.repo.changelog.rev(n),
                      parent=self.siblings(cl.parents(n), cl.rev),
                      child=self.siblings(cl.children(n), cl.rev),
                      diff=diff)
 
     def archive(self, req, cnode, type):
         cs = self.repo.changelog.read(cnode)
         mnode = cs[0]
         mf = self.repo.manifest.read(mnode)
         rev = self.repo.manifest.rev(mnode)
         reponame = re.sub(r"\W+", "-", self.reponame)
         name = "%s-%s/" % (reponame, short(cnode))
 
         files = mf.keys()
         files.sort()
 
         if type == 'zip':
             tmp = tempfile.mkstemp()[1]
             try:
                 zf = zipfile.ZipFile(tmp, "w", zipfile.ZIP_DEFLATED)
 
                 for f in files:
                     zf.writestr(name + f, self.repo.file(f).read(mf[f]))
                 zf.close()
 
                 f = open(tmp, 'r')
                 req.httphdr('application/zip', name[:-1] + '.zip',
                         os.path.getsize(tmp))
                 req.write(f.read())
                 f.close()
             finally:
                 os.unlink(tmp)
 
         else:
             tf = tarfile.TarFile.open(mode='w|' + type, fileobj=req.out)
             mff = self.repo.manifest.readflags(mnode)
             mtime = int(time.time())
 
             if type == "gz":
                 encoding = "gzip"
             else:
                 encoding = "x-bzip2"
             req.header([('Content-type', 'application/x-tar'),
                     ('Content-disposition', 'attachment; filename=%s%s%s' %
                         (name[:-1], '.tar.', type)),
                     ('Content-encoding', encoding)])
             for fname in files:
                 rcont = self.repo.file(fname).read(mf[fname])
                 finfo = tarfile.TarInfo(name + fname)
                 finfo.mtime = mtime
                 finfo.size = len(rcont)
                 finfo.mode = mff[fname] and 0755 or 0644
                 tf.addfile(finfo, StringIO.StringIO(rcont))
             tf.close()
 
     # add tags to things
     # tags -> list of changesets corresponding to tags
     # find tag, changeset, file
 
     def run(self, req=hgrequest()):
         def clean(path):
             p = os.path.normpath(path)
             if p[:2] == "..":
                 raise "suspicious path"
             return p
 
         def header(**map):
             yield self.t("header", **map)
 
         def footer(**map):
             yield self.t("footer", **map)
 
         def expand_form(form):
             shortcuts = {
                 'cl': [('cmd', ['changelog']), ('rev', None)],
                 'cs': [('cmd', ['changeset']), ('node', None)],
                 'f': [('cmd', ['file']), ('filenode', None)],
                 'fl': [('cmd', ['filelog']), ('filenode', None)],
                 'fd': [('cmd', ['filediff']), ('node', None)],
                 'fa': [('cmd', ['annotate']), ('filenode', None)],
                 'mf': [('cmd', ['manifest']), ('manifest', None)],
                 'ca': [('cmd', ['archive']), ('node', None)],
                 'tags': [('cmd', ['tags'])],
                 'tip': [('cmd', ['changeset']), ('node', ['tip'])],
             }
 
             for k in shortcuts.iterkeys():
                 if form.has_key(k):
                     for name, value in shortcuts[k]:
                         if value is None:
                             value = form[k]
                         form[name] = value
                     del form[k]
 
         self.refresh()
 
         expand_form(req.form)
 
         t = self.repo.ui.config("web", "templates", templatepath())
         m = os.path.join(t, "map")
         style = self.repo.ui.config("web", "style", "")
         if req.form.has_key('style'):
             style = req.form['style'][0]
         if style:
             b = os.path.basename("map-" + style)
             p = os.path.join(t, b)
             if os.path.isfile(p):
                 m = p
 
         port = req.env["SERVER_PORT"]
         port = port != "80" and (":" + port) or ""
         uri = req.env["REQUEST_URI"]
         if "?" in uri:
             uri = uri.split("?")[0]
         url = "http://%s%s%s" % (req.env["SERVER_NAME"], port, uri)
         if not self.reponame:
             self.reponame = (self.repo.ui.config("web", "name")
                              or uri.strip('/') or self.repo.root)
 
         self.t = templater(m, common_filters,
                            {"url": url,
                             "repo": self.reponame,
                             "header": header,
                             "footer": footer,
                            })
 
         if not req.form.has_key('cmd'):
             req.form['cmd'] = [self.t.cache['default'],]
 
         if req.form['cmd'][0] == 'changelog':
             c = self.repo.changelog.count() - 1
             hi = c
             if req.form.has_key('rev'):
                 hi = req.form['rev'][0]
                 try:
                     hi = self.repo.changelog.rev(self.repo.lookup(hi))
                 except hg.RepoError:
                     req.write(self.search(hi))
                     return
 
             req.write(self.changelog(hi))
 
         elif req.form['cmd'][0] == 'changeset':
             req.write(self.changeset(req.form['node'][0]))
 
         elif req.form['cmd'][0] == 'manifest':
             req.write(self.manifest(req.form['manifest'][0],
                                     clean(req.form['path'][0])))
 
         elif req.form['cmd'][0] == 'tags':
             req.write(self.tags())
 
         elif req.form['cmd'][0] == 'summary':
             req.write(self.summary())
 
         elif req.form['cmd'][0] == 'filediff':
             req.write(self.filediff(clean(req.form['file'][0]),
                                     req.form['node'][0]))
 
         elif req.form['cmd'][0] == 'file':
             req.write(self.filerevision(clean(req.form['file'][0]),
                                         req.form['filenode'][0]))
 
         elif req.form['cmd'][0] == 'annotate':
             req.write(self.fileannotate(clean(req.form['file'][0]),
                                         req.form['filenode'][0]))
 
         elif req.form['cmd'][0] == 'filelog':
             req.write(self.filelog(clean(req.form['file'][0]),
                                    req.form['filenode'][0]))
 
         elif req.form['cmd'][0] == 'heads':
             req.httphdr("application/mercurial-0.1")
             h = self.repo.heads()
             req.write(" ".join(map(hex, h)) + "\n")
 
         elif req.form['cmd'][0] == 'branches':
             req.httphdr("application/mercurial-0.1")
             nodes = []
             if req.form.has_key('nodes'):
                 nodes = map(bin, req.form['nodes'][0].split(" "))
             for b in self.repo.branches(nodes):
                 req.write(" ".join(map(hex, b)) + "\n")
 
         elif req.form['cmd'][0] == 'between':
             req.httphdr("application/mercurial-0.1")
             nodes = []
             if req.form.has_key('pairs'):
                 pairs = [map(bin, p.split("-"))
                          for p in req.form['pairs'][0].split(" ")]
             for b in self.repo.between(pairs):
                 req.write(" ".join(map(hex, b)) + "\n")
 
         elif req.form['cmd'][0] == 'changegroup':
             req.httphdr("application/mercurial-0.1")
             nodes = []
             if not self.allowpull:
                 return
 
             if req.form.has_key('roots'):
                 nodes = map(bin, req.form['roots'][0].split(" "))
 
             z = zlib.compressobj()
             f = self.repo.changegroup(nodes)
             while 1:
                 chunk = f.read(4096)
                 if not chunk:
                     break
                 req.write(z.compress(chunk))
 
             req.write(z.flush())
 
         elif req.form['cmd'][0] == 'archive':
             changeset = self.repo.lookup(req.form['node'][0])
             type = req.form['type'][0]
             if (type in self.archives and
                 self.repo.ui.configbool("web", "allow" + type, False)):
                 self.archive(req, changeset, type)
                 return
 
             req.write(self.t("error"))
 
         else:
             req.write(self.t("error"))
 
 def create_server(repo):
 
     def openlog(opt, default):
         if opt and opt != '-':
             return open(opt, 'w')
         return default
 
     address = repo.ui.config("web", "address", "")
     port = int(repo.ui.config("web", "port", 8000))
     use_ipv6 = repo.ui.configbool("web", "ipv6")
     accesslog = openlog(repo.ui.config("web", "accesslog", "-"), sys.stdout)
     errorlog = openlog(repo.ui.config("web", "errorlog", "-"), sys.stderr)
 
     class IPv6HTTPServer(BaseHTTPServer.HTTPServer):
         address_family = getattr(socket, 'AF_INET6', None)
 
         def __init__(self, *args, **kwargs):
             if self.address_family is None:
                 raise hg.RepoError(_('IPv6 not available on this system'))
             BaseHTTPServer.HTTPServer.__init__(self, *args, **kwargs)
 
     class hgwebhandler(BaseHTTPServer.BaseHTTPRequestHandler):
         def log_error(self, format, *args):
             errorlog.write("%s - - [%s] %s\n" % (self.address_string(),
                                                  self.log_date_time_string(),
                                                  format % args))
 
         def log_message(self, format, *args):
             accesslog.write("%s - - [%s] %s\n" % (self.address_string(),
                                                   self.log_date_time_string(),
                                                   format % args))
 
         def do_POST(self):
             try:
                 self.do_hgweb()
             except socket.error, inst:
                 if inst[0] != errno.EPIPE:
                     raise
 
         def do_GET(self):
             self.do_POST()
 
         def do_hgweb(self):
             query = ""
             p = self.path.find("?")
             if p:
                 query = self.path[p + 1:]
                 query = query.replace('+', ' ')
 
             env = {}
             env['GATEWAY_INTERFACE'] = 'CGI/1.1'
             env['REQUEST_METHOD'] = self.command
             env['SERVER_NAME'] = self.server.server_name
             env['SERVER_PORT'] = str(self.server.server_port)
             env['REQUEST_URI'] = "/"
             if query:
                 env['QUERY_STRING'] = query
             host = self.address_string()
             if host != self.client_address[0]:
                 env['REMOTE_HOST'] = host
                 env['REMOTE_ADDR'] = self.client_address[0]
 
             if self.headers.typeheader is None:
                 env['CONTENT_TYPE'] = self.headers.type
             else:
                 env['CONTENT_TYPE'] = self.headers.typeheader
             length = self.headers.getheader('content-length')
             if length:
                 env['CONTENT_LENGTH'] = length
             accept = []
             for line in self.headers.getallmatchingheaders('accept'):
                 if line[:1] in "\t\n\r ":
                     accept.append(line.strip())
                 else:
                     accept = accept + line[7:].split(',')
             env['HTTP_ACCEPT'] = ','.join(accept)
 
             req = hgrequest(self.rfile, self.wfile, env)
             self.send_response(200, "Script output follows")
             hg.run(req)
 
     hg = hgweb(repo)
     if use_ipv6:
         return IPv6HTTPServer((address, port), hgwebhandler)
     else:
         return BaseHTTPServer.HTTPServer((address, port), hgwebhandler)
 
 # This is a stopgap
 class hgwebdir(object):
     def __init__(self, config):
         def cleannames(items):
-            return [(name.strip('/'), path) for name, path in items]
+            return [(name.strip(os.sep), path) for name, path in items]
 
-        if type(config) == type([]):
+        if isinstance(config, (list, tuple)):
             self.repos = cleannames(config)
-        elif type(config) == type({}):
+        elif isinstance(config, dict):
             self.repos = cleannames(config.items())
             self.repos.sort()
         else:
             cp = ConfigParser.SafeConfigParser()
             cp.read(config)
-            self.repos = cleannames(cp.items("paths"))
+            self.repos = []
+            if cp.has_section('paths'):
+                self.repos.extend(cleannames(cp.items('paths')))
+            if cp.has_section('collections'):
+                for prefix, root in cp.items('collections'):
+                    for path in util.walkrepos(root):
+                        repo = os.path.normpath(path)
+                        name = repo
+                        if name.startswith(prefix):
+                            name = name[len(prefix):]
+                        self.repos.append((name.lstrip(os.sep), repo))
             self.repos.sort()
 
     def run(self, req=hgrequest()):
         def header(**map):
             yield tmpl("header", **map)
 
         def footer(**map):
             yield tmpl("footer", **map)
 
         m = os.path.join(templatepath(), "map")
         tmpl = templater(m, common_filters,
                          {"header": header, "footer": footer})
 
         def entries(**map):
             parity = 0
             for name, path in self.repos:
                 u = ui.ui()
                 try:
                     u.readconfig(os.path.join(path, '.hg', 'hgrc'))
                 except IOError:
                     pass
                 get = u.config
 
                 url = ('/'.join([req.env["REQUEST_URI"].split('?')[0], name])
                        .replace("//", "/"))
 
                 # update time with local timezone
                 try:
                     d = (get_mtime(path), util.makedate()[1])
                 except OSError:
                     continue
 
                 yield dict(contact=(get("ui", "username") or # preferred
                                     get("web", "contact") or # deprecated
                                     get("web", "author", "unknown")), # also
                            name=get("web", "name", name),
                            url=url,
                            parity=parity,
                            shortdesc=get("web", "description", "unknown"),
                            lastupdate=d)
 
                 parity = 1 - parity
 
         virtual = req.env.get("PATH_INFO", "").strip('/')
         if virtual:
             real = dict(self.repos).get(virtual)
             if real:
                 try:
                     hgweb(real).run(req)
                 except IOError, inst:
                     req.write(tmpl("error", error=inst.strerror))
                 except hg.RepoError, inst:
                     req.write(tmpl("error", error=str(inst)))
             else:
                 req.write(tmpl("notfound", repo=virtual))
         else:
             req.write(tmpl("index", entries=entries))

mercurial/util.py

@@ -1,692 +1,705
 """
 util.py - Mercurial utility functions and platform specfic implementations
 
  Copyright 2005 K. Thananchayan <[e-mail]>
 
 This software may be used and distributed according to the terms
 of the GNU General Public License, incorporated herein by reference.
 
 This contains helper routines that are independent of the SCM core and hide
 platform-specific details from the core.
 """
 
 import os, errno
 from i18n import gettext as _
 from demandload import *
 demandload(globals(), "cStringIO errno popen2 re shutil sys tempfile")
 demandload(globals(), "threading time")
 
 def pipefilter(s, cmd):
     '''filter string S through command CMD, returning its output'''
     (pout, pin) = popen2.popen2(cmd, -1, 'b')
     def writer():
         pin.write(s)
         pin.close()
 
     # we should use select instead on UNIX, but this will work on most
     # systems, including Windows
     w = threading.Thread(target=writer)
     w.start()
     f = pout.read()
     pout.close()
     w.join()
     return f
 
 def tempfilter(s, cmd):
     '''filter string S through a pair of temporary files with CMD.
     CMD is used as a template to create the real command to be run,
     with the strings INFILE and OUTFILE replaced by the real names of
     the temporary files generated.'''
     inname, outname = None, None
     try:
         infd, inname = tempfile.mkstemp(prefix='hgfin')
         fp = os.fdopen(infd, 'wb')
         fp.write(s)
         fp.close()
         outfd, outname = tempfile.mkstemp(prefix='hgfout')
         os.close(outfd)
         cmd = cmd.replace('INFILE', inname)
         cmd = cmd.replace('OUTFILE', outname)
         code = os.system(cmd)
         if code: raise Abort(_("command '%s' failed: %s") %
                              (cmd, explain_exit(code)))
         return open(outname, 'rb').read()
     finally:
         try:
             if inname: os.unlink(inname)
         except: pass
         try:
             if outname: os.unlink(outname)
         except: pass
 
 filtertable = {
     'tempfile:': tempfilter,
     'pipe:': pipefilter,
     }
 
 def filter(s, cmd):
     "filter a string through a command that transforms its input to its output"
     for name, fn in filtertable.iteritems():
         if cmd.startswith(name):
             return fn(s, cmd[len(name):].lstrip())
     return pipefilter(s, cmd)
 
 def patch(strip, patchname, ui):
     """apply the patch <patchname> to the working directory.
     a list of patched files is returned"""
     fp = os.popen('patch -p%d < "%s"' % (strip, patchname))
     files = {}
     for line in fp:
         line = line.rstrip()
         ui.status("%s\n" % line)
         if line.startswith('patching file '):
             pf = parse_patch_output(line)
             files.setdefault(pf, 1)
     code = fp.close()
     if code:
         raise Abort(_("patch command failed: %s") % explain_exit(code)[0])
     return files.keys()
 
 def binary(s):
     """return true if a string is binary data using diff's heuristic"""
     if s and '\0' in s[:4096]:
         return True
     return False
 
 def unique(g):
     """return the uniq elements of iterable g"""
     seen = {}
     for f in g:
         if f not in seen:
             seen[f] = 1
             yield f
 
 class Abort(Exception):
     """Raised if a command needs to print an error and exit."""
 
 def always(fn): return True
 def never(fn): return False
 
 def patkind(name, dflt_pat='glob'):
     """Split a string into an optional pattern kind prefix and the
     actual pattern."""
     for prefix in 're', 'glob', 'path', 'relglob', 'relpath', 'relre':
         if name.startswith(prefix + ':'): return name.split(':', 1)
     return dflt_pat, name
 
 def globre(pat, head='^', tail='$'):
     "convert a glob pattern into a regexp"
     i, n = 0, len(pat)
     res = ''
     group = False
     def peek(): return i < n and pat[i]
     while i < n:
         c = pat[i]
         i = i+1
         if c == '*':
             if peek() == '*':
                 i += 1
                 res += '.*'
             else:
                 res += '[^/]*'
         elif c == '?':
             res += '.'
         elif c == '[':
             j = i
             if j < n and pat[j] in '!]':
                 j += 1
             while j < n and pat[j] != ']':
                 j += 1
             if j >= n:
                 res += '\\['
             else:
                 stuff = pat[i:j].replace('\\','\\\\')
                 i = j + 1
                 if stuff[0] == '!':
                     stuff = '^' + stuff[1:]
                 elif stuff[0] == '^':
                     stuff = '\\' + stuff
                 res = '%s[%s]' % (res, stuff)
         elif c == '{':
             group = True
             res += '(?:'
         elif c == '}' and group:
             res += ')'
             group = False
         elif c == ',' and group:
             res += '|'
         else:
             res += re.escape(c)
     return head + res + tail
 
 _globchars = {'[': 1, '{': 1, '*': 1, '?': 1}
 
 def pathto(n1, n2):
     '''return the relative path from one place to another.
     this returns a path in the form used by the local filesystem, not hg.'''
     if not n1: return localpath(n2)
     a, b = n1.split('/'), n2.split('/')
     a.reverse()
     b.reverse()
     while a and b and a[-1] == b[-1]:
         a.pop()
         b.pop()
     b.reverse()
     return os.sep.join((['..'] * len(a)) + b)
 
 def canonpath(root, cwd, myname):
     """return the canonical path of myname, given cwd and root"""
     if root == os.sep:
         rootsep = os.sep
     else:
     	rootsep = root + os.sep
     name = myname
     if not name.startswith(os.sep):
         name = os.path.join(root, cwd, name)
     name = os.path.normpath(name)
     if name.startswith(rootsep):
         return pconvert(name[len(rootsep):])
     elif name == root:
         return ''
     else:
         raise Abort('%s not under root' % myname)
 
 def matcher(canonroot, cwd='', names=['.'], inc=[], exc=[], head='', src=None):
     return _matcher(canonroot, cwd, names, inc, exc, head, 'glob', src)
 
 def cmdmatcher(canonroot, cwd='', names=['.'], inc=[], exc=[], head='', src=None):
     if os.name == 'nt':
         dflt_pat = 'glob'
     else:
         dflt_pat = 'relpath'
     return _matcher(canonroot, cwd, names, inc, exc, head, dflt_pat, src)
 
 def _matcher(canonroot, cwd, names, inc, exc, head, dflt_pat, src):
     """build a function to match a set of file patterns
 
     arguments:
     canonroot - the canonical root of the tree you're matching against
     cwd - the current working directory, if relevant
     names - patterns to find
     inc - patterns to include
     exc - patterns to exclude
     head - a regex to prepend to patterns to control whether a match is rooted
 
     a pattern is one of:
     'glob:<rooted glob>'
     're:<rooted regexp>'
     'path:<rooted path>'
     'relglob:<relative glob>'
     'relpath:<relative path>'
     'relre:<relative regexp>'
     '<rooted path or regexp>'
 
     returns:
     a 3-tuple containing
     - list of explicit non-pattern names passed in
     - a bool match(filename) function
     - a bool indicating if any patterns were passed in
 
     todo:
     make head regex a rooted bool
     """
 
     def contains_glob(name):
         for c in name:
             if c in _globchars: return True
         return False
 
     def regex(kind, name, tail):
         '''convert a pattern into a regular expression'''
         if kind == 're':
             return name
         elif kind == 'path':
             return '^' + re.escape(name) + '(?:/|$)'
         elif kind == 'relglob':
             return head + globre(name, '(?:|.*/)', tail)
         elif kind == 'relpath':
             return head + re.escape(name) + tail
         elif kind == 'relre':
             if name.startswith('^'):
                 return name
             return '.*' + name
         return head + globre(name, '', tail)
 
     def matchfn(pats, tail):
         """build a matching function from a set of patterns"""
         if not pats:
             return
         matches = []
         for k, p in pats:
             try:
                 pat = '(?:%s)' % regex(k, p, tail)
                 matches.append(re.compile(pat).match)
             except re.error:
                 if src: raise Abort("%s: invalid pattern (%s): %s" % (src, k, p))
                 else: raise Abort("invalid pattern (%s): %s" % (k, p))
 
         def buildfn(text):
             for m in matches:
                 r = m(text)
                 if r:
                     return r
 
         return buildfn
 
     def globprefix(pat):
         '''return the non-glob prefix of a path, e.g. foo/* -> foo'''
         root = []
         for p in pat.split(os.sep):
             if contains_glob(p): break
             root.append(p)
         return '/'.join(root)
 
     pats = []
     files = []
     roots = []
     for kind, name in [patkind(p, dflt_pat) for p in names]:
         if kind in ('glob', 'relpath'):
             name = canonpath(canonroot, cwd, name)
             if name == '':
                 kind, name = 'glob', '**'
         if kind in ('glob', 'path', 're'):
             pats.append((kind, name))
         if kind == 'glob':
             root = globprefix(name)
             if root: roots.append(root)
         elif kind == 'relpath':
             files.append((kind, name))
             roots.append(name)
 
     patmatch = matchfn(pats, '$') or always
     filematch = matchfn(files, '(?:/|$)') or always
     incmatch = always
     if inc:
         incmatch = matchfn(map(patkind, inc), '(?:/|$)')
     excmatch = lambda fn: False
     if exc:
         excmatch = matchfn(map(patkind, exc), '(?:/|$)')
 
     return (roots,
             lambda fn: (incmatch(fn) and not excmatch(fn) and
                         (fn.endswith('/') or
                          (not pats and not files) or
                          (pats and patmatch(fn)) or
                          (files and filematch(fn)))),
             (inc or exc or (pats and pats != [('glob', '**')])) and True)
 
 def system(cmd, errprefix=None):
     """execute a shell command that must succeed"""
     rc = os.system(cmd)
     if rc:
         errmsg = "%s %s" % (os.path.basename(cmd.split(None, 1)[0]),
                             explain_exit(rc)[0])
         if errprefix:
             errmsg = "%s: %s" % (errprefix, errmsg)
         raise Abort(errmsg)
 
 def rename(src, dst):
     """forcibly rename a file"""
     try:
         os.rename(src, dst)
     except:
         os.unlink(dst)
         os.rename(src, dst)
 
 def unlink(f):
     """unlink and remove the directory if it is empty"""
     os.unlink(f)
     # try removing directories that might now be empty
     try: os.removedirs(os.path.dirname(f))
     except: pass
 
 def copyfiles(src, dst, hardlink=None):
     """Copy a directory tree using hardlinks if possible"""
 
     if hardlink is None:
         hardlink = (os.stat(src).st_dev ==
                     os.stat(os.path.dirname(dst)).st_dev)
 
     if os.path.isdir(src):
         os.mkdir(dst)
         for name in os.listdir(src):
             srcname = os.path.join(src, name)
             dstname = os.path.join(dst, name)
             copyfiles(srcname, dstname, hardlink)
     else:
         if hardlink:
             try:
                 os_link(src, dst)
             except:
                 hardlink = False
                 shutil.copy(src, dst)
         else:
             shutil.copy(src, dst)
 
 def opener(base):
     """
     return a function that opens files relative to base
 
     this function is used to hide the details of COW semantics and
     remote file access from higher level code.
     """
     p = base
 
     def mktempcopy(name):
         d, fn = os.path.split(name)
         fd, temp = tempfile.mkstemp(prefix=fn, dir=d)
         fp = os.fdopen(fd, "wb")
         try:
             fp.write(file(name, "rb").read())
         except:
             try: os.unlink(temp)
             except: pass
             raise
         fp.close()
         st = os.lstat(name)
         os.chmod(temp, st.st_mode)
         return temp
 
     class atomicfile(file):
         """the file will only be copied on close"""
         def __init__(self, name, mode, atomic=False):
             self.__name = name
             self.temp = mktempcopy(name)
             file.__init__(self, self.temp, mode)
         def close(self):
             if not self.closed:
                 file.close(self)
                 rename(self.temp, self.__name)
         def __del__(self):
             self.close()
 
     def o(path, mode="r", text=False, atomic=False):
         f = os.path.join(p, path)
 
         if not text:
             mode += "b" # for that other OS
 
         if mode[0] != "r":
             try:
                 nlink = nlinks(f)
             except OSError:
                 d = os.path.dirname(f)
                 if not os.path.isdir(d):
                     os.makedirs(d)
             else:
                 if atomic:
                     return atomicfile(f, mode)
                 if nlink > 1:
                     rename(mktempcopy(f), f)
         return file(f, mode)
 
     return o
 
 def _makelock_file(info, pathname):
     ld = os.open(pathname, os.O_CREAT | os.O_WRONLY | os.O_EXCL)
     os.write(ld, info)
     os.close(ld)
 
 def _readlock_file(pathname):
     return file(pathname).read()
 
 def nlinks(pathname):
     """Return number of hardlinks for the given file."""
     return os.stat(pathname).st_nlink
 
 if hasattr(os, 'link'):
     os_link = os.link
 else:
     def os_link(src, dst):
         raise OSError(0, _("Hardlinks not supported"))
 
 # Platform specific variants
 if os.name == 'nt':
     demandload(globals(), "msvcrt")
     nulldev = 'NUL:'
 
     class winstdout:
         '''stdout on windows misbehaves if sent through a pipe'''
 
         def __init__(self, fp):
             self.fp = fp
 
         def __getattr__(self, key):
             return getattr(self.fp, key)
 
         def close(self):
             try:
                 self.fp.close()
             except: pass
 
         def write(self, s):
             try:
                 return self.fp.write(s)
             except IOError, inst:
                 if inst.errno != 0: raise
                 self.close()
                 raise IOError(errno.EPIPE, 'Broken pipe')
 
     sys.stdout = winstdout(sys.stdout)
 
     try:
         import win32api, win32process
         filename = win32process.GetModuleFileNameEx(win32api.GetCurrentProcess(), 0)
         systemrc = os.path.join(os.path.dirname(filename), 'mercurial.ini')
 
     except ImportError:
         systemrc = r'c:\mercurial\mercurial.ini'
         pass
 
     rcpath = (systemrc,
               os.path.join(os.path.expanduser('~'), 'mercurial.ini'))
 
     def parse_patch_output(output_line):
         """parses the output produced by patch and returns the file name"""
         pf = output_line[14:]
         if pf[0] == '`':
             pf = pf[1:-1] # Remove the quotes
         return pf
 
     try: # ActivePython can create hard links using win32file module
         import win32file
 
         def os_link(src, dst): # NB will only succeed on NTFS
             win32file.CreateHardLink(dst, src)
 
         def nlinks(pathname):
             """Return number of hardlinks for the given file."""
             try:
                 fh = win32file.CreateFile(pathname,
                     win32file.GENERIC_READ, win32file.FILE_SHARE_READ,
                     None, win32file.OPEN_EXISTING, 0, None)
                 res = win32file.GetFileInformationByHandle(fh)
                 fh.Close()
                 return res[7]
             except:
                 return os.stat(pathname).st_nlink
 
     except ImportError:
         pass
 
     def is_exec(f, last):
         return last
 
     def set_exec(f, mode):
         pass
 
     def set_binary(fd):
         msvcrt.setmode(fd.fileno(), os.O_BINARY)
 
     def pconvert(path):
         return path.replace("\\", "/")
 
     def localpath(path):
         return path.replace('/', '\\')
 
     def normpath(path):
         return pconvert(os.path.normpath(path))
 
     makelock = _makelock_file
     readlock = _readlock_file
 
     def explain_exit(code):
         return _("exited with status %d") % code, code
 
 else:
     nulldev = '/dev/null'
 
     def rcfiles(path):
         rcs = [os.path.join(path, 'hgrc')]
         rcdir = os.path.join(path, 'hgrc.d')
         try:
             rcs.extend([os.path.join(rcdir, f) for f in os.listdir(rcdir)
                         if f.endswith(".rc")])
         except OSError, inst: pass
         return rcs
     rcpath = []
     if len(sys.argv) > 0:
         rcpath.extend(rcfiles(os.path.dirname(sys.argv[0]) + '/../etc/mercurial'))
     rcpath.extend(rcfiles('/etc/mercurial'))
     rcpath.append(os.path.expanduser('~/.hgrc'))
     rcpath = [os.path.normpath(f) for f in rcpath]
 
     def parse_patch_output(output_line):
         """parses the output produced by patch and returns the file name"""
         pf = output_line[14:]
         if pf.startswith("'") and pf.endswith("'") and pf.find(" ") >= 0:
             pf = pf[1:-1] # Remove the quotes
         return pf
 
     def is_exec(f, last):
         """check whether a file is executable"""
         return (os.stat(f).st_mode & 0100 != 0)
 
     def set_exec(f, mode):
         s = os.stat(f).st_mode
         if (s & 0100 != 0) == mode:
             return
         if mode:
             # Turn on +x for every +r bit when making a file executable
             # and obey umask.
             umask = os.umask(0)
             os.umask(umask)
             os.chmod(f, s | (s & 0444) >> 2 & ~umask)
         else:
             os.chmod(f, s & 0666)
 
     def set_binary(fd):
         pass
 
     def pconvert(path):
         return path
 
     def localpath(path):
         return path
 
     normpath = os.path.normpath
 
     def makelock(info, pathname):
         try:
             os.symlink(info, pathname)
         except OSError, why:
             if why.errno == errno.EEXIST:
                 raise
             else:
                 _makelock_file(info, pathname)
 
     def readlock(pathname):
         try:
             return os.readlink(pathname)
         except OSError, why:
             if why.errno == errno.EINVAL:
                 return _readlock_file(pathname)
             else:
                 raise
 
     def explain_exit(code):
         """return a 2-tuple (desc, code) describing a process's status"""
         if os.WIFEXITED(code):
             val = os.WEXITSTATUS(code)
             return _("exited with status %d") % val, val
         elif os.WIFSIGNALED(code):
             val = os.WTERMSIG(code)
             return _("killed by signal %d") % val, val
         elif os.WIFSTOPPED(code):
             val = os.WSTOPSIG(code)
             return _("stopped by signal %d") % val, val
         raise ValueError(_("invalid exit code"))
 
 class chunkbuffer(object):
     """Allow arbitrary sized chunks of data to be efficiently read from an
     iterator over chunks of arbitrary size."""
 
     def __init__(self, in_iter, targetsize = 2**16):
         """in_iter is the iterator that's iterating over the input chunks.
         targetsize is how big a buffer to try to maintain."""
         self.in_iter = iter(in_iter)
         self.buf = ''
         self.targetsize = int(targetsize)
         if self.targetsize <= 0:
             raise ValueError(_("targetsize must be greater than 0, was %d") %
                              targetsize)
         self.iterempty = False
 
     def fillbuf(self):
         """Ignore target size; read every chunk from iterator until empty."""
         if not self.iterempty:
             collector = cStringIO.StringIO()
             collector.write(self.buf)
             for ch in self.in_iter:
                 collector.write(ch)
             self.buf = collector.getvalue()
             self.iterempty = True
 
     def read(self, l):
         """Read L bytes of data from the iterator of chunks of data.
         Returns less than L bytes if the iterator runs dry."""
         if l > len(self.buf) and not self.iterempty:
             # Clamp to a multiple of self.targetsize
             targetsize = self.targetsize * ((l // self.targetsize) + 1)
             collector = cStringIO.StringIO()
             collector.write(self.buf)
             collected = len(self.buf)
             for chunk in self.in_iter:
                 collector.write(chunk)
                 collected += len(chunk)
                 if collected >= targetsize:
                     break
             if collected < targetsize:
                 self.iterempty = True
             self.buf = collector.getvalue()
         s, self.buf = self.buf[:l], buffer(self.buf, l)
         return s
 
 def filechunkiter(f, size = 65536):
     """Create a generator that produces all the data in the file size
     (default 65536) bytes at a time.  Chunks may be less than size
     bytes if the chunk is the last chunk in the file, or the file is a
     socket or some other type of file that sometimes reads less data
     than is requested."""
     s = f.read(size)
     while len(s) > 0:
         yield s
         s = f.read(size)
 
 def makedate():
     lt = time.localtime()
     if lt[8] == 1 and time.daylight:
         tz = time.altzone
     else:
         tz = time.timezone
     return time.mktime(lt), tz
 
 def datestr(date=None, format='%c'):
     """represent a (unixtime, offset) tuple as a localized time.
     unixtime is seconds since the epoch, and offset is the time zone's
     number of seconds away from UTC."""
     t, tz = date or makedate()
     return ("%s %+03d%02d" %
             (time.strftime(format, time.gmtime(float(t) - tz)),
              -tz / 3600,
              ((-tz % 3600) / 60)))
+
+def walkrepos(path):
+    '''yield every hg repository under path, recursively.'''
+    def errhandler(err):
+        if err.filename == path:
+            raise err
+
+    for root, dirs, files in os.walk(path, onerror=errhandler):
+        for d in dirs:
+            if d == '.hg':
+                yield root
+                dirs[:] = []
+                break