Fabian Keil [Fri, 27 May 2011 11:36:36 +0000 (11:36 +0000)]
Ditch two checks that have been commented out for years already, without causing any issues
Fabian Keil [Fri, 27 May 2011 11:36:15 +0000 (11:36 +0000)]
Remove AC_FUNC_SETPGRP as it can cause issues when cross-compiling
I considered only skipping it when cross-compiling, but given
that Privoxy doesn't use setpgrp(2), why should we care about
its prototype?
There's no justification for the check in the revision log
either, except that it was part of "Al's patch".
Fabian Keil [Fri, 27 May 2011 11:34:39 +0000 (11:34 +0000)]
Remove a useless '+' in main().
Fabian Keil [Fri, 27 May 2011 11:34:16 +0000 (11:34 +0000)]
In LogPutString(), don't bother looking at pszText when not actually logging anything
Fabian Keil [Sun, 22 May 2011 10:31:29 +0000 (10:31 +0000)]
Remove -prevent-compression from the fragile alias
We no longer use +prevent-compresssion anywhere by
default and it isn't known to break stuff anyway.
Fabian Keil [Sun, 22 May 2011 10:30:55 +0000 (10:30 +0000)]
Garbage collect the strdup() replacement for mingw32
Supposedly it was added ten years ago to paper over a mingw32
bug where the returned pointer couldn't be free'd, but I doubt
that the bug still exists assuming the problem was analyzed
correctly in the first place.
If it does, we should notice soon enough as I'd consider
free'ing pointers returned by strdup() a rather common
operation.
Fabian Keil [Sun, 22 May 2011 10:26:45 +0000 (10:26 +0000)]
Garbage collect safe_strerror() which isn't used anywhere
Fabian Keil [Sun, 22 May 2011 10:25:26 +0000 (10:25 +0000)]
If only one of the strncmpic() arguments has a length the compiler can figure out, move it to the second place
This will allow to use a macro to figure out the length later on.
Fabian Keil [Sun, 22 May 2011 10:21:54 +0000 (10:21 +0000)]
Change a ssplit() call to use SZ(array) as the third parameter instead of the hardcoded 3
Silences a clang complaint now that the ssplit() prototype has been fixed.
Fabian Keil [Sun, 22 May 2011 10:20:05 +0000 (10:20 +0000)]
Change ssplit()'s fourth parameter from int to the proper size_t
As pointed out by clang, size_t's don't necessarily have the
same width as int's and while I'm not aware of any actual
issues caused by this (here), it doesn't hurt to use the right
type anyway.
Fabian Keil [Sun, 15 May 2011 10:49:33 +0000 (10:49 +0000)]
Mention some of the fun one can experience when not explicitly configuring a IP addresses to listen to.
Fabian Keil [Sun, 15 May 2011 10:48:58 +0000 (10:48 +0000)]
Improve the listen-address documentation
Explicitly mention that hostnames can be used instead of
IP addresses, that only the first address returned will
be used and what happens if the address is invalid.
Requested by Cálestyo in #
3302213.
Fabian Keil [Sun, 8 May 2011 13:31:05 +0000 (13:31 +0000)]
Add a -fast-redirects exception for webcache.googleusercontent.com/
Fabian Keil [Sun, 8 May 2011 13:30:17 +0000 (13:30 +0000)]
Fix white-space.
Fabian Keil [Sun, 8 May 2011 13:29:59 +0000 (13:29 +0000)]
Use a more traditional for loop to iterate over the tests in execute_regression_tests()
While it doesn't really matter for the number of tests we have,
there's no reason to keep all the test numbers in memory.
Fabian Keil [Sun, 8 May 2011 13:28:13 +0000 (13:28 +0000)]
Reduce empty lines in init_our_variables().
Fabian Keil [Sun, 8 May 2011 13:27:53 +0000 (13:27 +0000)]
Simplify check_header_result() a bit.
Fabian Keil [Sun, 8 May 2011 13:27:10 +0000 (13:27 +0000)]
Fix spelling in two error messages.
Fabian Keil [Sun, 8 May 2011 13:26:24 +0000 (13:26 +0000)]
Add a warning that the statistics currently can't be trusted.
Mention Privoxy-Log-Parser's --statistics option as
an alternative for the time being.
Fabian Keil [Sun, 8 May 2011 13:25:14 +0000 (13:25 +0000)]
Add: 92) The statistics currently aren't calculated correctly by Privoxy
Fabian Keil [Tue, 3 May 2011 10:33:53 +0000 (10:33 +0000)]
Do not expect a trailing dot in: Request from 10.0.0.1 marked for blocking. limit-connect{0} doesn't allow CONNECT requests to example.org:443
I intended to add one, but decided to go with the submitted version.
Fabian Keil [Tue, 3 May 2011 10:19:20 +0000 (10:19 +0000)]
Highlight the hostport in: Request from 10.0.0.1 marked for blocking. limit-connect{0} doesn't allow CONNECT requests to example.org:443.
Fabian Keil [Tue, 3 May 2011 10:15:54 +0000 (10:15 +0000)]
Include the destination host in the 'Request ... marked for blocking. limit-connect{...} doesn't allow CONNECT ...' message
Patch submitted by Saperski in #
3296250.
Fabian Keil [Tue, 3 May 2011 10:13:21 +0000 (10:13 +0000)]
Point out that the SourceForge messaging system works like a blackhole and should thus not be used
Fabian Keil [Tue, 3 May 2011 10:12:23 +0000 (10:12 +0000)]
In unbreak_lines_only_loop(), accept ISO 8601 time stamps, too
Fabian Keil [Tue, 3 May 2011 10:11:24 +0000 (10:11 +0000)]
In rfc2553_connect_to(), prevent a duplicated log message if none of the resolved IP addresses were connectable
Fabian Keil [Tue, 3 May 2011 09:55:35 +0000 (09:55 +0000)]
Prevent a NULL-pointer dereference in connect_to()
It happened in case of address resolution failures when
there was no forwarding proxy enabled for the request in
the config file and forward-override{} wasn't used at all.
Regression introduced in v1.98 by yours truly.
Patch submitted by saperski in #
3296527.
Fabian Keil [Wed, 27 Apr 2011 18:11:05 +0000 (18:11 +0000)]
In connect_to(), do not pretent to retry if forwarded-connect-retries is zero or unset.
Fabian Keil [Tue, 26 Apr 2011 16:55:08 +0000 (16:55 +0000)]
Split the --user delimiter skipping across two lines and add comments.
Fabian Keil [Tue, 26 Apr 2011 16:53:21 +0000 (16:53 +0000)]
Make a copy of the --user value and only mess with that when splitting user and group.
On some operating systems modifying the value directly
is reflected in the output of ps and friends and can
be misleading.
While at it, use a pointer name that gives a hint
what the pointer is actually supposed to point to.
Fixes #
3292710 reported by zepard.
Fabian Keil [Tue, 26 Apr 2011 16:50:11 +0000 (16:50 +0000)]
Remove a harmless but pointless line in the code to handle the --user argument.
If p is not NULL, --p already has been set to '\0' previously
and it hadn't, it would be too late now as we are done with p.
Fabian Keil [Tue, 26 Apr 2011 16:48:56 +0000 (16:48 +0000)]
When a specified user or group can't be found, put the name in single-quotes when logging it.
Fabian Keil [Tue, 26 Apr 2011 16:48:04 +0000 (16:48 +0000)]
If the --user argument user[.group] contains a dot, always bail out if no group has been specified.
Previously the intended, but undocumented, behaviour was to
try interpreting the whole argument as user name, but the
detection was flawed and checked for '0' isntead of '\0',
thus merely preventing group names beginning with a zero.
Remove the incorrect check instead of fixing it to get
the documented behaviour.
Fabian Keil [Tue, 19 Apr 2011 13:18:46 +0000 (13:18 +0000)]
Merge a typo fix.
Fabian Keil [Tue, 19 Apr 2011 13:14:10 +0000 (13:14 +0000)]
Fix spelling errors in the documentation. Found with codespell.
Fabian Keil [Tue, 19 Apr 2011 13:10:11 +0000 (13:10 +0000)]
Fix spelling errors found with codespell.
Fabian Keil [Tue, 19 Apr 2011 13:08:51 +0000 (13:08 +0000)]
Fix a perldoc syntax error.
Fabian Keil [Tue, 19 Apr 2011 13:00:47 +0000 (13:00 +0000)]
Fix various spelling errors in the comments.
Found with codespell.
Fabian Keil [Tue, 19 Apr 2011 11:42:51 +0000 (11:42 +0000)]
Add a --disable-ipv6-support switch for platforms where support is detected but doesn't actually work.
Fabian Keil [Tue, 19 Apr 2011 11:42:12 +0000 (11:42 +0000)]
Rename the --disable-acl-files switch to --disable-acl-support
Since about 2001, ACL directives are specified in the standard config file.
Fabian Keil [Sat, 16 Apr 2011 17:31:35 +0000 (17:31 +0000)]
Clarify what kind of directory layout #67 is talking about.
Fabian Keil [Sat, 16 Apr 2011 17:31:24 +0000 (17:31 +0000)]
Mention #41 in #65
Fabian Keil [Sat, 16 Apr 2011 17:31:11 +0000 (17:31 +0000)]
In #42, note that the DTrace bible is preordered.
Fabian Keil [Sat, 16 Apr 2011 17:30:36 +0000 (17:30 +0000)]
Rephrase #40 which could be easily misunderstood.
Fabian Keil [Sat, 16 Apr 2011 17:30:23 +0000 (17:30 +0000)]
Remove #39. jarfile support got garbage collected before the 3.0.12 release.
Fabian Keil [Sat, 16 Apr 2011 17:29:38 +0000 (17:29 +0000)]
Remove #35. The accept_connection() thread does little enough now, I think.
Fabian Keil [Sat, 16 Apr 2011 17:29:24 +0000 (17:29 +0000)]
Shorten #14 by removing the logging part.
Fabian Keil [Sat, 16 Apr 2011 17:29:12 +0000 (17:29 +0000)]
Remove #81. The unlimited-config-lines branch has been merged a while ago.
Fabian Keil [Sat, 16 Apr 2011 17:28:58 +0000 (17:28 +0000)]
Add #91: Add an optional limit for internal redirects.
Fabian Keil [Sat, 16 Apr 2011 17:28:15 +0000 (17:28 +0000)]
Mention why the static PCRE code is actually deprecated.
The URL to the mailing list posting is likely
to get messed up again in the future.
Fabian Keil [Sat, 16 Apr 2011 17:27:40 +0000 (17:27 +0000)]
Update the URL of the 'Removing outdated PCRE version after the next stable release' posting.
The old URL stopped working after one of SF's recent layout pessimizations.
Reported by Han Liu.
Fabian Keil [Sun, 27 Mar 2011 14:14:13 +0000 (14:14 +0000)]
Remove an obsolete paragraph in the forwarded-connect-retries section
The bug that forwarded-connect-retries covered direct connections, too,
has just been fixed (I hope), so there's no need to document it any longer.
Fabian Keil [Sun, 27 Mar 2011 14:08:35 +0000 (14:08 +0000)]
If forwarded-connect-retries is set, only retry if the we are actually forwarding the request.
Fabian Keil [Sun, 27 Mar 2011 14:04:10 +0000 (14:04 +0000)]
Fix a theoretical memory leak in a cannot-realistically-happen situation in rfc2553_connect_to().
Fabian Keil [Sun, 27 Mar 2011 14:03:43 +0000 (14:03 +0000)]
In rfc2553_connect_to(), initialize socket_error to 0
There are a couple of fringe situations where it isn't properly set yet.
Fabian Keil [Sun, 27 Mar 2011 14:03:25 +0000 (14:03 +0000)]
Mark the errno overloading in connect_to() as questionable, too
Fabian Keil [Sun, 27 Mar 2011 14:02:53 +0000 (14:02 +0000)]
Zero out the timeval struct in rfc2553_connect_to() completely and rename it to timeout
It's not used for anything else so there's not need for a generic name.
While at it, use a struct instead of an array of a single struct.
Fabian Keil [Sun, 27 Mar 2011 14:01:46 +0000 (14:01 +0000)]
Remove useless parentheses in get_destination_from_headers()
Fabian Keil [Sun, 27 Mar 2011 14:01:18 +0000 (14:01 +0000)]
Factor the Content-Encoding header ditching out of server_content_encoding() into server_adjust_content_encoding()
Fabian Keil [Sun, 27 Mar 2011 14:00:36 +0000 (14:00 +0000)]
In socks4_connect(), don't comment on connect_to() failures.
The failures are dealt with when they happen now.
Fabian Keil [Sun, 27 Mar 2011 13:58:33 +0000 (13:58 +0000)]
In rfc2553_connect_to(), explain getnameinfo() errors differently.
Fabian Keil [Sun, 27 Mar 2011 13:58:09 +0000 (13:58 +0000)]
Remove a useless log message in chat()
The fact that the connection attempt failed already has been logged
and a reason given earlier on. Also %E can't be used here anyway,
as errno will no longer contain the cause of the problem.
Fabian Keil [Sun, 27 Mar 2011 13:57:41 +0000 (13:57 +0000)]
Bump copyright year
Fabian Keil [Sun, 27 Mar 2011 13:57:28 +0000 (13:57 +0000)]
In rfc2553_connect_to() and no_rfc2553_connect_to(), factor the socket() call out of the #ifdef _WIN32 branch.
Fabian Keil [Sun, 27 Mar 2011 13:57:08 +0000 (13:57 +0000)]
In rfc2553_connect_to(), simplify the OS/2 code path and pray that it still works
Fabian Keil [Sun, 27 Mar 2011 13:56:48 +0000 (13:56 +0000)]
Don't leak csp->http->host_ip_addr_str when calling rfc2553_connect_to() multiple times.
Fabian Keil [Sun, 27 Mar 2011 13:56:27 +0000 (13:56 +0000)]
Don't leak csp->http->host_ip_addr_str when calling no_rfc2553_connect_to() multiple times.
Fabian Keil [Sun, 27 Mar 2011 13:55:55 +0000 (13:55 +0000)]
In rfc2553_connect_to(), start setting cgi->error_message on error
Fabian Keil [Sun, 27 Mar 2011 13:55:09 +0000 (13:55 +0000)]
In rfc2553_connect_to(), rename socket_in_error to socket_error
Fabian Keil [Sun, 27 Mar 2011 13:54:52 +0000 (13:54 +0000)]
In the IPv6-aware connect_to(), allocate http->host_ip_addr_str outside the loop and reuse the buffer in case of errors
I think this also fixes an unlikely memory leak, but I haven't actually verified it.
Fabian Keil [Sun, 27 Mar 2011 13:53:49 +0000 (13:53 +0000)]
In rfc2553_connect_to(), remove a useless check for NULL
Fabian Keil [Sun, 27 Mar 2011 13:53:25 +0000 (13:53 +0000)]
Remove a useless comment in rfc2553_connect_to().
connect(2) is neither Linux nor GNU libc specific
and we don't advertise other man pages either.
Fabian Keil [Sun, 27 Mar 2011 13:53:04 +0000 (13:53 +0000)]
When retrying to connect, also log the maximum number of connection attempts
Fabian Keil [Sun, 27 Mar 2011 13:52:23 +0000 (13:52 +0000)]
Move the forwarded-connect-retries handling into a shiny new connect_to()
In the process, give the two former connect_to() versions different names.
Fabian Keil [Sun, 27 Mar 2011 13:51:04 +0000 (13:51 +0000)]
Remove an incorrect comment about csp not being modified by connect_to(). It's been wrong since years ago.
Fabian Keil [Sun, 27 Mar 2011 13:50:38 +0000 (13:50 +0000)]
Bump Copyright range.
Fabian Keil [Sun, 27 Mar 2011 13:50:15 +0000 (13:50 +0000)]
Polish syntax to drop the GNU make dependency for the build target.
Fabian Keil [Sun, 27 Mar 2011 13:49:16 +0000 (13:49 +0000)]
Add 90) Implement NO-TAG: patterns
Fabian Keil [Sun, 20 Mar 2011 11:51:12 +0000 (11:51 +0000)]
Accept and ignore: SDCH-compressed content detected, content filtering disabled. Consider suppressing SDCH offers made by the client.
Fabian Keil [Sun, 20 Mar 2011 11:50:28 +0000 (11:50 +0000)]
In server_content_encoding(), mark the content as taboo for filtering if SDCH compression is used
If SDCH was combined with a supported compression algorithm,
we'd previously try to decompress it, when successful apply
the enabled filters and ditch the Content-Encoding header
even though the SDCH compression wasn't removed.
Reported by zebul666 in #
3225863.
I get the impression that filtering SDCH-compressed content
actually isn't guaranteed to cause problems as long as the
client still gets a "Content-Encoding: sdch" header, so we
should probably eventually allow users who want to risk it
to do so by explicitly enforcing filtering.
While at it, mark a bunch of other, partly-related, problems
in server_content_encoding() that should be addressed someday.
Fabian Keil [Tue, 8 Mar 2011 18:43:46 +0000 (18:43 +0000)]
Bump generated Firefox version to 3.6.15.
Fabian Keil [Tue, 8 Mar 2011 18:33:28 +0000 (18:33 +0000)]
Change the expected status code returned for p.p/die depending on whether or not FEATURE_GRACEFUL_TERMINATION is available
Fabian Keil [Tue, 8 Mar 2011 18:32:48 +0000 (18:32 +0000)]
Unblock adassier.wordpress.com/ and http://adassier.files.wordpress.com/
Reported on ijbswa-users@, no name exlicitly mentioned.
Message-ID: <ea-mime-
4d758a87-4088-
57d52b28@w-4.netcmail.com>
Fabian Keil [Tue, 8 Mar 2011 18:32:11 +0000 (18:32 +0000)]
Rephrase a log message in compile_dynamic_pcrs_job_list()
Devide the error code and its meaning with a colon.
Call the pcrs job dynamic and not the filter. Filters may
contain dynamic and non-dynamic pcrs jobs at the same time.
Only mention the name of the filter or tagger, but don't
claim it's a filter when it could be a tagger.
Fabian Keil [Tue, 8 Mar 2011 18:30:46 +0000 (18:30 +0000)]
In a fatal error message in load_one_actions_file(), cover both URL and TAG patterns
Fabian Keil [Thu, 3 Mar 2011 14:49:08 +0000 (14:49 +0000)]
In pcrs_strerror(), properly report unknown positive error code values as unknown.
Previously they were handled like 0 (no error).
Fabian Keil [Thu, 3 Mar 2011 14:47:28 +0000 (14:47 +0000)]
In compile_dynamic_pcrs_job_list(), also log the actual error code as pcrs_strerror() doesn't handle all errors reported by pcre
Fabian Keil [Thu, 3 Mar 2011 14:46:37 +0000 (14:46 +0000)]
Remove an incorrect assertion in compile_dynamic_pcrs_job_list()
It could be triggered by a pcrs job with an invalid pcre
pattern (for example one that contains a lone quantifier).
Fabian Keil [Thu, 3 Mar 2011 14:44:00 +0000 (14:44 +0000)]
Don't bother trying to continue chatting if the client didn't ask for it.
Reduces log noise a bit.
Fabian Keil [Thu, 3 Mar 2011 14:42:55 +0000 (14:42 +0000)]
In cgi_die(), mark the client connection for closing.
If the client will fetch the style sheet through another connection
it gets the main thread out of the accept() state and should thus
trigger the actual shutdown.
Fabian Keil [Thu, 3 Mar 2011 14:42:18 +0000 (14:42 +0000)]
Add a proper CGI message for cgi_die().
Fabian Keil [Thu, 3 Mar 2011 14:41:29 +0000 (14:41 +0000)]
Fix an invalid free when compiled with FEATURE_GRACEFUL_TERMINATION and shut down through config.privoxy.org/die
Fabian Keil [Thu, 3 Mar 2011 14:39:57 +0000 (14:39 +0000)]
Improve a comment in load_one_actions_file()
Fabian Keil [Thu, 3 Mar 2011 14:39:34 +0000 (14:39 +0000)]
Make two fatal error message in load_one_actions_file() more descriptive
Fabian Keil [Thu, 3 Mar 2011 14:39:13 +0000 (14:39 +0000)]
In load_one_actions_file(), don't limit the actions_buf to BUFFER_SIZE bytes.
Fabian Keil [Thu, 3 Mar 2011 14:38:36 +0000 (14:38 +0000)]
Don't enforce a logical line length limit in read_config_line()
It means the caller has to free the buffer, but we can live with that.
Fabian Keil [Thu, 3 Mar 2011 14:22:39 +0000 (14:22 +0000)]
Remove duplicated 'the' in a comment
Fabian Keil [Thu, 3 Mar 2011 14:22:12 +0000 (14:22 +0000)]
Remove duplicated 'the'
Fabian Keil [Sun, 20 Feb 2011 13:17:16 +0000 (13:17 +0000)]
Additionally accept ISO 8601 time stamps.
Fabian Keil [Sat, 19 Feb 2011 13:59:59 +0000 (13:59 +0000)]
Bump copyright.