This is an archive of the discontinued Mercurial Phabricator instance.

Differential D1079

hgweb: rewrite most obviously-native-strings to be native strings
ClosedPublic

Authored by durin42 on Oct 14 2017, 4:16 PM.

Details

Reviewers
indygreg
Group Reviewers
hg-reviewers
Commits
rHGc51380879054: hgweb: rewrite most obviously-native-strings to be native strings
Summary

This clearly won't be everything, but it unblocks a fair amount of
stuff here.

Diff Detail

Repository
rHG Mercurial
Lint
Automatic diff as part of commit; lint not applicable.
Unit
Automatic diff as part of commit; unit tests not applicable.

Event Timeline

durin42 created this revision.Oct 14 2017, 4:16 PM
indygreg added subscribers: yuja, indygreg.Oct 14 2017, 4:56 PM
indygreg accepted this revision.

There might be some subtle issues with encoding for e.g. form values (if we don't have a test for e.g. a Unicode value in a revset search, we should add one). But I'm sure @yuja will flag those if they exist. I'm happy to queue until then.

This revision is now accepted and ready to land.Oct 14 2017, 4:56 PM
Closed by commit rHGc51380879054: hgweb: rewrite most obviously-native-strings to be native strings (authored by durin42). · Explain WhyOct 14 2017, 5:03 PM
This revision was automatically updated to reflect the committed changes.
yuja added a comment.Oct 14 2017, 11:16 PM

Please don't count on me that I can catch unicode issues around
Python URL/HTTP libraries. I hate Python 3 because of the unicode
mess, so I don't write it for any projects other than Mercurial.

Maybe we'll need a clear boundary between Mercurial and Python
types to determine which value should be bytes/unicode.

Revision Contents
Changeset List

Diff 2755

mercurial/hgweb/__init__.py

# hgweb/__init__.py - web interface to a mercurial repository # hgweb/__init__.py - web interface to a mercurial repository
# #
# Copyright 21 May 2005 - (c) 2005 Jake Edge <jake@edge2.net> # Copyright 21 May 2005 - (c) 2005 Jake Edge <jake@edge2.net>
# Copyright 2005 Matt Mackall <mpm@selenic.com> # Copyright 2005 Matt Mackall <mpm@selenic.com>
# #
# This software may be used and distributed according to the terms of the # This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version. # GNU General Public License version 2 or any later version.
from __future__ import absolute_import from __future__ import absolute_import
import os import os
from ..i18n import _ from ..i18n import _
from .. import ( from .. import (
error, error,
pycompat,
util, util,
) )
from . import ( from . import (
hgweb_mod, hgweb_mod,
hgwebdir_mod, hgwebdir_mod,
server, server,
) )
if self.opts['port'] and not self.ui.verbose: if self.opts['port'] and not self.ui.verbose:
return return
if self.httpd.prefix: if self.httpd.prefix:
prefix = self.httpd.prefix.strip('/') + '/' prefix = self.httpd.prefix.strip('/') + '/'
else: else:
prefix = '' prefix = ''
port = ':%d' % self.httpd.port port = r':%d' % self.httpd.port
if port == ':80': if port == r':80':
port = '' port = r''
bindaddr = self.httpd.addr bindaddr = self.httpd.addr
if bindaddr == '0.0.0.0': if bindaddr == r'0.0.0.0':
bindaddr = '*' bindaddr = r'*'
elif ':' in bindaddr: # IPv6 elif r':' in bindaddr: # IPv6
bindaddr = '[%s]' % bindaddr bindaddr = r'[%s]' % bindaddr
fqaddr = self.httpd.fqaddr fqaddr = self.httpd.fqaddr
if ':' in fqaddr: if r':' in fqaddr:
fqaddr = '[%s]' % fqaddr fqaddr = r'[%s]' % fqaddr
if self.opts['port']: if self.opts['port']:
write = self.ui.status write = self.ui.status
else: else:
write = self.ui.write write = self.ui.write
write(_('listening at http://%s%s/%s (bound to %s:%d)\n') % write(_('listening at http://%s%s/%s (bound to %s:%d)\n') %
(fqaddr, port, prefix, bindaddr, self.httpd.port)) (pycompat.sysbytes(fqaddr), pycompat.sysbytes(port),
prefix, pycompat.sysbytes(bindaddr), self.httpd.port))
self.ui.flush() # avoid buffering of status message self.ui.flush() # avoid buffering of status message
def run(self): def run(self):
self.httpd.serve_forever() self.httpd.serve_forever()
def createapp(baseui, repo, webconf): def createapp(baseui, repo, webconf):
if webconf: if webconf:
return hgwebdir_mod.hgwebdir(webconf, baseui=baseui) return hgwebdir_mod.hgwebdir(webconf, baseui=baseui)
else: else:
if not repo: if not repo:
raise error.RepoError(_("there is no Mercurial repository" raise error.RepoError(_("there is no Mercurial repository"
" here (.hg not found)")) " here (.hg not found)"))
return hgweb_mod.hgweb(repo, baseui=baseui) return hgweb_mod.hgweb(repo, baseui=baseui)

mercurial/hgweb/hgweb_mod.py

proto = req.env.get('wsgi.url_scheme') proto = req.env.get('wsgi.url_scheme')
if proto == 'https': if proto == 'https':
proto = 'https' proto = 'https'
default_port = '443' default_port = '443'
else: else:
proto = 'http' proto = 'http'
default_port = '80' default_port = '80'
port = req.env['SERVER_PORT'] port = req.env[r'SERVER_PORT']
port = port != default_port and (':' + port) or '' port = port != default_port and (r':' + port) or r''
urlbase = '%s://%s%s' % (proto, req.env['SERVER_NAME'], port) urlbase = r'%s://%s%s' % (proto, req.env[r'SERVER_NAME'], port)
logourl = self.config('web', 'logourl') logourl = self.config('web', 'logourl')
logoimg = self.config('web', 'logoimg') logoimg = self.config('web', 'logoimg')
staticurl = self.config('web', 'staticurl') or req.url + 'static/' staticurl = self.config('web', 'staticurl') or req.url + 'static/'
if not staticurl.endswith('/'): if not staticurl.endswith('/'):
staticurl += '/' staticurl += '/'
# some functions for the templater # some functions for the templater
# replace it. # replace it.
req.headers = [h for h in req.headers req.headers = [h for h in req.headers
if h[0] != 'Content-Security-Policy'] if h[0] != 'Content-Security-Policy']
req.headers.append(('Content-Security-Policy', rctx.csp)) req.headers.append(('Content-Security-Policy', rctx.csp))
# work with CGI variables to create coherent structure # work with CGI variables to create coherent structure
# use SCRIPT_NAME, PATH_INFO and QUERY_STRING as well as our REPO_NAME # use SCRIPT_NAME, PATH_INFO and QUERY_STRING as well as our REPO_NAME
req.url = req.env['SCRIPT_NAME'] req.url = req.env[r'SCRIPT_NAME']
if not req.url.endswith('/'): if not req.url.endswith('/'):
req.url += '/' req.url += '/'
if req.env.get('REPO_NAME'): if req.env.get('REPO_NAME'):
req.url += req.env['REPO_NAME'] + '/' req.url += req.env[r'REPO_NAME'] + r'/'
if 'PATH_INFO' in req.env: if r'PATH_INFO' in req.env:
parts = req.env['PATH_INFO'].strip('/').split('/') parts = req.env[r'PATH_INFO'].strip('/').split('/')
repo_parts = req.env.get('REPO_NAME', '').split('/') repo_parts = req.env.get(r'REPO_NAME', r'').split(r'/')
if parts[:len(repo_parts)] == repo_parts: if parts[:len(repo_parts)] == repo_parts:
parts = parts[len(repo_parts):] parts = parts[len(repo_parts):]
query = '/'.join(parts) query = '/'.join(parts)
else: else:
query = req.env['QUERY_STRING'].partition('&')[0] query = req.env[r'QUERY_STRING'].partition(r'&')[0]
query = query.partition(';')[0] query = query.partition(r';')[0]
# process this if it's a protocol request # process this if it's a protocol request
# protocol bits don't need to create any URLs # protocol bits don't need to create any URLs
# and the clients always use the old URL structure # and the clients always use the old URL structure
cmd = req.form.get('cmd', [''])[0] cmd = pycompat.sysbytes(req.form.get(r'cmd', [r''])[0])
if protocol.iscmd(cmd): if protocol.iscmd(cmd):
try: try:
if query: if query:
raise ErrorResponse(HTTP_NOT_FOUND) raise ErrorResponse(HTTP_NOT_FOUND)
if cmd in perms: if cmd in perms:
self.check_perm(rctx, req, perms[cmd]) self.check_perm(rctx, req, perms[cmd])
return protocol.call(rctx.repo, req, cmd) return protocol.call(rctx.repo, req, cmd)
except ErrorResponse as inst: except ErrorResponse as inst:
# A client that sends unbundle without 100-continue will # A client that sends unbundle without 100-continue will
# break if we respond early. # break if we respond early.
if (cmd == 'unbundle' and if (cmd == 'unbundle' and
(req.env.get('HTTP_EXPECT', (req.env.get('HTTP_EXPECT',
'').lower() != '100-continue') or '').lower() != '100-continue') or
req.env.get('X-HgHttp2', '')): req.env.get('X-HgHttp2', '')):
req.drain() req.drain()
else: else:
req.headers.append(('Connection', 'Close')) req.headers.append(('Connection', 'Close'))
req.respond(inst, protocol.HGTYPE, req.respond(inst, protocol.HGTYPE,
body='0\n%s\n' % inst) body='0\n%s\n' % inst)
return '' return ''
# translate user-visible url structure to internal structure # translate user-visible url structure to internal structure
args = query.split('/', 2) args = query.split('/', 2)
if 'cmd' not in req.form and args and args[0]: if r'cmd' not in req.form and args and args[0]:
cmd = args.pop(0) cmd = args.pop(0)
style = cmd.rfind('-') style = cmd.rfind('-')
if style != -1: if style != -1:
req.form['style'] = [cmd[:style]] req.form['style'] = [cmd[:style]]
cmd = cmd[style + 1:] cmd = cmd[style + 1:]
# avoid accepting e.g. style parameter as command # avoid accepting e.g. style parameter as command
if util.safehasattr(webcommands, cmd): if util.safehasattr(webcommands, cmd):
req.form['cmd'] = [cmd] req.form[r'cmd'] = [cmd]
if cmd == 'static': if cmd == 'static':
req.form['file'] = ['/'.join(args)] req.form['file'] = ['/'.join(args)]
else: else:
if args and args[0]: if args and args[0]:
node = args.pop(0).replace('%2F', '/') node = args.pop(0).replace('%2F', '/')
req.form['node'] = [node] req.form['node'] = [node]
if args: if args:
ctype = tmpl('mimetype', encoding=encoding.encoding) ctype = tmpl('mimetype', encoding=encoding.encoding)
ctype = templater.stringify(ctype) ctype = templater.stringify(ctype)
# check read permissions non-static content # check read permissions non-static content
if cmd != 'static': if cmd != 'static':
self.check_perm(rctx, req, None) self.check_perm(rctx, req, None)
if cmd == '': if cmd == '':
req.form['cmd'] = [tmpl.cache['default']] req.form[r'cmd'] = [tmpl.cache['default']]
cmd = req.form['cmd'][0] cmd = req.form[r'cmd'][0]
# Don't enable caching if using a CSP nonce because then it wouldn't # Don't enable caching if using a CSP nonce because then it wouldn't
# be a nonce. # be a nonce.
if rctx.configbool('web', 'cache') and not rctx.nonce: if rctx.configbool('web', 'cache') and not rctx.nonce:
caching(self, req) # sets ETag header or raises NOT_MODIFIED caching(self, req) # sets ETag header or raises NOT_MODIFIED
if cmd not in webcommands.__all__: if cmd not in webcommands.__all__:
msg = 'no such method: %s' % cmd msg = 'no such method: %s' % cmd
raise ErrorResponse(HTTP_BAD_REQUEST, msg) raise ErrorResponse(HTTP_BAD_REQUEST, msg)
elif cmd == 'file' and 'raw' in req.form.get('style', []): elif cmd == 'file' and r'raw' in req.form.get(r'style', []):
rctx.ctype = ctype rctx.ctype = ctype
content = webcommands.rawfile(rctx, req, tmpl) content = webcommands.rawfile(rctx, req, tmpl)
else: else:
content = getattr(webcommands, cmd)(rctx, req, tmpl) content = getattr(webcommands, cmd)(rctx, req, tmpl)
req.respond(HTTP_OK, ctype) req.respond(HTTP_OK, ctype)
return content return content