Mercurial > hg
view mercurial/hgweb.py @ 226:1536ccac47e9
merge: don't bail on outstanding changes
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
merge: don't bail on outstanding changes
With multiple heads, we don't need to worry about the working dir's
uncommitted changes at pull time
manifest hash: 5b4e024f220fa616732310ce5f48e71abfa910e0
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)
iD8DBQFCoMFQywK+sNU5EO8RApLyAKCoNDF84wFzgnpS+WLuXdkGxeHFPwCdFsMy
CysB458dNcFuB/vDFhgJr58=
=gG+u
-----END PGP SIGNATURE-----
author | mpm@selenic.com |
---|---|
date | Fri, 03 Jun 2005 12:45:04 -0800 |
parents | 87484f627422 |
children | 4f802588cdfb 4f802588cdfb 3b92f8fe47ae |
line wrap: on
line source
#!/usr/bin/env python # # hgweb.py - 0.2 - 21 May 2005 - (c) 2005 Jake Edge <jake@edge2.net> # - web interface to a mercurial repository # # This software may be used and distributed according to the terms # of the GNU General Public License, incorporated herein by reference. # useful for debugging import cgitb cgitb.enable() import os, cgi, time, re, difflib, sys, zlib from mercurial.hg import * from mercurial.ui import * 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(t): 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() delta = max(1, int(now - t)) 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 httphdr(type): print 'Content-type: %s\n' % type def write(*things): for thing in things: if hasattr(thing, "__iter__"): for part in thing: write(part) else: sys.stdout.write(str(thing)) def template(tmpl, filters = {}, **map): while tmpl: m = re.search(r"#([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() or v fl = m.group(2) if fl: for f in fl.split("|")[1:]: v = filters[f](v) yield v tmpl = tmpl[m.end(0):] else: yield tmpl return class templater: def __init__(self, mapfile, filters = {}): self.cache = {} self.map = {} self.base = os.path.dirname(mapfile) self.filters = filters 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 "unknown map entry '%s'" % l def __call__(self, t, **map): try: tmpl = self.cache[t] except KeyError: tmpl = self.cache[t] = file(self.map[t]).read() return template(tmpl, self.filters, **map) class hgweb: maxchanges = 20 maxfiles = 10 def __init__(self, path, name, templates = ""): self.templates = templates or templatepath() self.reponame = name self.repo = repository(ui(), path) self.viewonly = 0 self.filters = { "escape": cgi.escape, "age": age, "date": (lambda x: time.asctime(time.gmtime(x))), "addbreaks": nl2br, "obfuscate": obfuscate, "firstline": (lambda x: x.splitlines(1)[0]), } def date(self, cs): return time.asctime(time.gmtime(float(cs[2].split(' ')[0]))) 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 parent(self, t1, node=nullid, rev=-1, **args): if node != hex(nullid): yield self.t(t1, node = node, rev = rev, **args) def diff(self, node1, node2, files): def filterfiles(list, files): l = [ x for x in list if x in files ] for f in files: if f[-1] != os.sep: f += os.sep l += [ x for x in list if x.startswith(f) ] return l parity = [0] def diffblock(diff, f, fn): yield self.t("diffblock", lines = prettyprintlines(diff), parity = parity[0], file = f, filenode = hex(fn)) 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 = self.date(change1) date2 = self.date(change2) c, a, d = r.diffrevs(node1, node2) c, a, d = map(lambda x: filterfiles(x, files), (c, a, d)) for f in c: to = r.file(f).read(mmap1[f]) tn = r.file(f).read(mmap2[f]) yield diffblock(mdiff.unidiff(to, date1, tn, date2, f), f, tn) for f in a: to = "" tn = r.file(f).read(mmap2[f]) yield diffblock(mdiff.unidiff(to, date1, tn, date2, f), f, tn) for f in d: to = r.file(f).read(mmap1[f]) tn = "" yield diffblock(mdiff.unidiff(to, date1, tn, date2, f), f, tn) def header(self): yield self.t("header", repo = self.reponame) def footer(self): yield self.t("footer", repo = self.reponame) def changelog(self, pos): def changenav(): def seq(factor = 1): yield 1 * factor yield 3 * factor #yield 5 * factor for f in seq(factor * 10): yield f l = [] for f in seq(): if f < self.maxchanges / 2: continue if f > count: break r = "%d" % f if pos + f < count - (f/2): l.append(("+" + r, pos + f)) if pos - f >= 0 + (f/2): l.insert(0, ("-" + r, pos - f)) yield self.t("naventry", rev = 0, label="(0)") for label, rev in l: yield self.t("naventry", label = label, rev = rev) yield self.t("naventry", label="tip") def changelist(): parity = (start - end) & 1 cl = self.repo.changelog l = [] # build a list in forward order for efficiency for i in range(start, end + 1): n = cl.node(i) changes = cl.read(n) hn = hex(n) p1, p2 = cl.parents(n) t = float(changes[2].split(' ')[0]) l.insert(0, self.t( 'changelogentry', parity = parity, author = changes[1], parent1 = self.parent("changelogparent", hex(p1), cl.rev(p1)), parent2 = self.parent("changelogparent", hex(p2), cl.rev(p2)), p1 = hex(p1), p2 = hex(p2), p1rev = cl.rev(p1), p2rev = cl.rev(p2), manifest = hex(changes[0]), desc = changes[4], date = t, files = self.listfilediffs(changes[3], n), rev = i, node = hn)) parity = 1 - parity yield l cl = self.repo.changelog mf = cl.read(cl.tip())[0] count = cl.count() end = min(pos, count - 1) start = max(0, pos - self.maxchanges) end = min(count - 1, start + self.maxchanges) yield self.t('changelog', header = self.header(), footer = self.footer(), repo = self.reponame, changenav = changenav, manifest = hex(mf), rev = pos, changesets = count, entries = changelist) def changeset(self, nodeid): n = bin(nodeid) cl = self.repo.changelog changes = cl.read(n) p1, p2 = cl.parents(n) p1rev, p2rev = cl.rev(p1), cl.rev(p2) t = float(changes[2].split(' ')[0]) files = [] mf = self.repo.manifest.read(changes[0]) for f in changes[3]: files.append(self.t("filenodelink", filenode = hex(mf[f]), file = f)) def diff(): yield self.diff(p1, n, changes[3]) yield self.t('changeset', header = self.header(), footer = self.footer(), repo = self.reponame, diff = diff, rev = cl.rev(n), node = nodeid, parent1 = self.parent("changesetparent", hex(p1), cl.rev(p1)), parent2 = self.parent("changesetparent", hex(p2), cl.rev(p2)), p1 = hex(p1), p2 = hex(p2), p1rev = cl.rev(p1), p2rev = cl.rev(p2), manifest = hex(changes[0]), author = changes[1], desc = changes[4], date = t, files = files) def filelog(self, f, filenode): cl = self.repo.changelog fl = self.repo.file(f) count = fl.count() def entries(): 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)) p1, p2 = fl.parents(n) t = float(cs[2].split(' ')[0]) l.insert(0, self.t("filelogentry", parity = parity, filenode = hex(n), filerev = i, file = f, node = hex(cn), author = cs[1], date = t, desc = cs[4], p1 = hex(p1), p2 = hex(p2), p1rev = fl.rev(p1), p2rev = fl.rev(p2))) parity = 1 - parity yield l yield self.t("filelog", header = self.header(), footer = self.footer(), repo = self.reponame, file = f, filenode = filenode, entries = entries) def filerevision(self, f, node): fl = self.repo.file(f) n = bin(node) text = fl.read(n) changerev = fl.linkrev(n) cl = self.repo.changelog cn = cl.node(changerev) cs = cl.read(cn) p1, p2 = fl.parents(n) t = float(cs[2].split(' ')[0]) mfn = cs[0] def lines(): for l, t in enumerate(text.splitlines(1)): yield self.t("fileline", line = t, linenumber = "% 6d" % (l + 1), parity = l & 1) yield self.t("filerevision", file = f, header = self.header(), footer = self.footer(), repo = self.reponame, filenode = node, path = up(f), text = lines(), rev = changerev, node = hex(cn), manifest = hex(mfn), author = cs[1], date = t, parent1 = self.parent("filerevparent", hex(p1), fl.rev(p1), file=f), parent2 = self.parent("filerevparent", hex(p2), fl.rev(p2), file=f), p1 = hex(p1), p2 = hex(p2), p1rev = fl.rev(p1), p2rev = fl.rev(p2)) def fileannotate(self, f, node): bcache = {} ncache = {} fl = self.repo.file(f) n = bin(node) changerev = fl.linkrev(n) cl = self.repo.changelog cn = cl.node(changerev) cs = cl.read(cn) p1, p2 = fl.parents(n) t = float(cs[2].split(' ')[0]) mfn = cs[0] def annotate(): 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) name = cl[1] f = name.find('@') if f >= 0: name = name[:f] bcache[r] = name if last != cnode: parity = 1 - parity last = cnode yield self.t("annotateline", parity = parity, node = hex(cnode), rev = r, author = name, file = f, line = l) yield self.t("fileannotate", header = self.header(), footer = self.footer(), repo = self.reponame, file = f, filenode = node, annotate = annotate, path = up(f), rev = changerev, node = hex(cn), manifest = hex(mfn), author = cs[1], date = t, parent1 = self.parent("fileannotateparent", hex(p1), fl.rev(p1), file=f), parent2 = self.parent("fileannotateparent", hex(p2), fl.rev(p2), file=f), p1 = hex(p1), p2 = hex(p2), p1rev = fl.rev(p1), p2rev = fl.rev(p2)) def manifest(self, mnode, path): mf = self.repo.manifest.read(bin(mnode)) rev = self.repo.manifest.rev(bin(mnode)) node = self.repo.changelog.node(rev) files = {} p = path[1:] 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(): parity = 0 fl = files.keys() fl.sort() for f in fl: full, fnode = files[f] if fnode: yield self.t("manifestfileentry", file = full, manifest = mnode, filenode = hex(fnode), parity = parity, basename = f) else: yield self.t("manifestdirentry", parity = parity, path = os.path.join(path, f), manifest = mnode, basename = f[:-1]) parity = 1 - parity yield self.t("manifest", header = self.header(), footer = self.footer(), repo = self.reponame, manifest = mnode, rev = rev, node = hex(node), path = path, up = up(path), entries = filelist) def tags(self): cl = self.repo.changelog mf = cl.read(cl.tip())[0] self.repo.lookup(0) # prime the cache i = self.repo.tags.items() n = [ (cl.rev(e[1]), e) for e in i ] # sort by revision n.sort() n.reverse() i = [ e[1] for e in n ] def entries(): parity = 0 for k,n in i: yield self.t("tagentry", parity = parity, tag = k, node = hex(n)) parity = 1 - parity yield self.t("tags", header = self.header(), footer = self.footer(), repo = self.reponame, manifest = hex(mf), entries = entries) def filediff(self, file, changeset): n = bin(changeset) cl = self.repo.changelog p1 = cl.parents(n)[0] cs = cl.read(n) mf = self.repo.manifest.read(cs[0]) def diff(): yield self.diff(p1, n, file) yield self.t("filediff", header = self.header(), footer = self.footer(), repo = self.reponame, file = file, filenode = hex(mf[file]), node = changeset, rev = self.repo.changelog.rev(n), p1 = hex(p1), p1rev = self.repo.changelog.rev(p1), diff = diff) # add tags to things # tags -> list of changesets corresponding to tags # find tag, changeset, file def run(self): args = cgi.parse() m = os.path.join(self.templates, "map") if args.has_key('style'): b = os.path.basename("map-" + args['style'][0]) p = os.path.join(self.templates, b) if os.path.isfile(p): m = p self.t = templater(m, self.filters) if not args.has_key('cmd') or args['cmd'][0] == 'changelog': hi = self.repo.changelog.count() if args.has_key('rev'): hi = args['rev'][0] try: hi = self.repo.changelog.rev(self.repo.lookup(hi)) except KeyError: hi = self.repo.changelog.count() write(self.changelog(hi)) elif args['cmd'][0] == 'changeset': write(self.changeset(args['node'][0])) elif args['cmd'][0] == 'manifest': write(self.manifest(args['manifest'][0], args['path'][0])) elif args['cmd'][0] == 'tags': write(self.tags()) elif args['cmd'][0] == 'filediff': write(self.filediff(args['file'][0], args['node'][0])) elif args['cmd'][0] == 'file': write(self.filerevision(args['file'][0], args['filenode'][0])) elif args['cmd'][0] == 'annotate': write(self.fileannotate(args['file'][0], args['filenode'][0])) elif args['cmd'][0] == 'filelog': write(self.filelog(args['file'][0], args['filenode'][0])) elif args['cmd'][0] == 'heads': httphdr("text/plain") h = self.repo.heads() sys.stdout.write(" ".join(map(hex, h)) + "\n") elif args['cmd'][0] == 'branches': httphdr("text/plain") nodes = [] if args.has_key('nodes'): nodes = map(bin, args['nodes'][0].split(" ")) for b in self.repo.branches(nodes): sys.stdout.write(" ".join(map(hex, b)) + "\n") elif args['cmd'][0] == 'between': httphdr("text/plain") nodes = [] if args.has_key('pairs'): pairs = [ map(bin, p.split("-")) for p in args['pairs'][0].split(" ") ] for b in self.repo.between(pairs): sys.stdout.write(" ".join(map(hex, b)) + "\n") elif args['cmd'][0] == 'changegroup': httphdr("application/hg-changegroup") nodes = [] if self.viewonly: return if args.has_key('roots'): nodes = map(bin, args['roots'][0].split(" ")) z = zlib.compressobj() for chunk in self.repo.changegroup(nodes): sys.stdout.write(z.compress(chunk)) sys.stdout.write(z.flush()) else: write(self.t("error")) def server(path, name, templates, address, port): import BaseHTTPServer import sys, os class hgwebhandler(BaseHTTPServer.BaseHTTPRequestHandler): def do_POST(self): self.do_hgweb() def do_GET(self): self.do_hgweb() 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 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) os.environ.update(env) save = sys.argv, sys.stdin, sys.stdout, sys.stderr try: sys.stdin = self.rfile sys.stdout = self.wfile sys.argv = ["hgweb.py"] if '=' not in query: sys.argv.append(query) self.send_response(200, "Script output follows") hg.run() finally: sys.argv, sys.stdin, sys.stdout, sys.stderr = save hg = hgweb(path, name, templates) httpd = BaseHTTPServer.HTTPServer((address, port), hgwebhandler) httpd.serve_forever()