Send the command arguments in the HTTP headers. The command is still part
of the URL. If the server does not have the 'httpheader' capability, the
client will send the command arguments in the URL as it did previously.
Web servers typically allow more data to be placed within the headers than
in the URL, so this approach will:
- Avoid HTTP errors due to using a URL that is too large.
- Allow Mercurial to implement a more efficient wire protocol.
An alternate approach is to send the arguments as part of the request body.
This approach has been rejected because it requires the use of POST
requests, so it would break any existing configuration that relies on the
request type for authentication or caching.
Extensibility:
- The header size is provided by the server, which makes it possible to
introduce an hgrc setting for it.
- The client ignores the capability value after the first comma, which
allows more information to be included in the future.
The introduction of the new URL parsing code has created a startup
time regression. This is mainly due to the use of url.hasscheme() in
the ui class. It ends up importing many libraries that the url module
requires.
This fix helps marginally, but if we can get rid of the urllib import
in the URL parser all together, startup time will go back to normal.
perfstartup time before the URL refactoring (707e4b1e8064):
! wall 0.050692 comb 0.000000 user 0.000000 sys 0.000000 (best of 100)
current startup time (9ad1dce9e7f4):
! wall 0.070685 comb 0.000000 user 0.000000 sys 0.000000 (best of 100)
after this change:
! wall 0.064667 comb 0.000000 user 0.000000 sys 0.000000 (best of 100)
By default, hgweb_mod supports caching via the ETag header. This can
cause some confusion with browsers which cache aggressively. This change
preserves existing behavior while giving the administrator a knob to
disable the ETag header.
Clicking on the logo image/text in the hgweb interface brings the
user to the Mercurial project page. The majority of users expect that
this would bring them to the top level index. I have added a new template
variable named `logourl' which allows an administrator to change this
behavior. To stay compatible with existing behavior, `logourl' will
default to http://mercurial.selenic.com/. This change is very useful in
large installations where jumping to the index is common.
Repository() raises a number of IOErrors in addition to RepoErrors.
these are just as uninteresting as RepoErrors and should be ignored
The easiest case of this is a repo whose .hg/ directory is -rx
Invalid requests could give an unhandled ErrorResponse.
Now this ErrorResponse is handled like other ErrorResponses so the client gets
an error message which also is logged on the server.
Other exceptions than StandardExceptions were left to the default error handler
which was muted when running in daemon mode.
Now all Exceptions are handled and logged to the log file.
The archive list generator was holding a reference to each
temporary ui copy passed by rawentries(), so the memory
usage for index generation growed proportionally to the
ui object size and the amount of repositories. By returning a
list instead, the temporary reference is dropped immediately.
Add missing calls to close() to many places where files are
opened. Relying on reference counting to catch them soon-ish is not
portable and fails in environments with a proper GC, such as PyPy.
The only revision information yielded by the annotate view was the revision
number itself. The patch allows the use of per-line revision dates in the
corresponding templates.
8aa8db6deb47 moved the subdirectory match inside the repository match
loop. A virtual path existing/path/invalid/path would then match
existing/path, and generate a wrong index page.
PROTOCOL_SSLv3 on the server side doesn't work everywhere. Sometimes the client
reports "EOF occurred in violation of protocol" (for example on Mac and Solaris).
The more compatible PROTOCOL_SSLv23 is now used instead. It works but is less
"secure" for some OpenSSL versions as it can fall back to weak encryption.
pyOpenSSL apparently doesn't work for Python 2.7 and isn't very actively
maintained.
The built-in ssl module seems like a long-term winner, so we now use that with
Python 2.6 and higher.
The https mode failed in super because BaseRequestHandler is an old-style
class.
This introduces the first test of https client/server functionality - and
"hghave ssl". The test is currently only run on Python 2.6.
Without this fix, mod_wsgi and spawning get in a wedged state after
sending a 304 response. Not sending a body fixed that problem. The
header change was discovered by using wsgiref.validate.validator to
check for other errors.
The behaviour between http and ssh still differ:
- the "unsynced changes" is seen as a remote output in the http cases
- but it is correctly seen as a push error for ssh