Fabian Keil [Wed, 31 Aug 2011 13:36:15 +0000 (13:36 +0000)]
Add a dok-tidy target to clean up the messy HTML generated by the other dok targets
Not yet connected to any other target,
but probably should be after some testing.
Fabian Keil [Wed, 31 Aug 2011 13:35:21 +0000 (13:35 +0000)]
Ditch a duplicated dot in a comment
Fabian Keil [Sat, 27 Aug 2011 10:36:45 +0000 (10:36 +0000)]
Change the HTTP version in the snipplets to 1.1
The snipplets are compliant, so there's no reason not to.
Fabian Keil [Fri, 26 Aug 2011 16:15:11 +0000 (16:15 +0000)]
Rebuild with updated copyright
Fabian Keil [Fri, 26 Aug 2011 16:14:24 +0000 (16:14 +0000)]
Rebuild with updated copyright and minor changes
Fabian Keil [Fri, 26 Aug 2011 16:13:40 +0000 (16:13 +0000)]
Rebuild after ditching GNUisms in GNUMakefile.in
Fabian Keil [Fri, 26 Aug 2011 16:12:34 +0000 (16:12 +0000)]
Deduplicate a line in the dok-release target
Fabian Keil [Fri, 26 Aug 2011 16:12:14 +0000 (16:12 +0000)]
Add #97: Add support for Tor's 'optimistic data' socks extension.
Fabian Keil [Fri, 26 Aug 2011 16:11:59 +0000 (16:11 +0000)]
Once more, with feeling: GNUisms be gone
This time the dok target has actually been tested on
a GNU/Linux system using GNU make.
Fabian Keil [Mon, 22 Aug 2011 18:58:29 +0000 (18:58 +0000)]
When sending out the "no-server-data" response, do not log two LOG_LEVEL_CLF messages
send_crunch_response() already takes care of emitting the
message, so there's no need for chat() to emit another one.
Noticed while looking into #
3394462 reported by Andy.
Fabian Keil [Thu, 18 Aug 2011 11:45:02 +0000 (11:45 +0000)]
Don't use unspecified MSN sites as examples for User-Agent-based descrimination
Without knowing the URLs, nobody can easily verify it and it could
be mistaken as FUD. I also assume that it's no longer an issue anyway.
Fabian Keil [Thu, 18 Aug 2011 11:42:50 +0000 (11:42 +0000)]
Bump some more documentation copyright ranges.
Fabian Keil [Thu, 18 Aug 2011 11:41:18 +0000 (11:41 +0000)]
Also highlight the compression level, now that it is logged, too
Fabian Keil [Wed, 17 Aug 2011 10:41:21 +0000 (10:41 +0000)]
Rebuild README
Fabian Keil [Wed, 17 Aug 2011 10:41:03 +0000 (10:41 +0000)]
Index updates.
Fabian Keil [Wed, 17 Aug 2011 10:40:07 +0000 (10:40 +0000)]
Update the entities.
This commit is chronological out of order.
Fabian Keil [Wed, 17 Aug 2011 10:37:48 +0000 (10:37 +0000)]
Rebuild user-manual, developer-manual and faq.
Looks like the code status UNRELEASED isn't properly dealt with everywhere,
but as this build is not going to get released, I think we can live with
it for now.
Also ampersands in URLs weren't properly escaped, but hopefully
none of the broken links slipped into this commit. This needs to
be addressed before the next release anyway.
Fabian Keil [Wed, 17 Aug 2011 10:32:44 +0000 (10:32 +0000)]
Formattting fixes missed in the previous commit and forgotten to squash
Fabian Keil [Wed, 17 Aug 2011 10:31:40 +0000 (10:31 +0000)]
Rebuild config with the recently added compression directives
Fabian Keil [Wed, 17 Aug 2011 10:30:36 +0000 (10:30 +0000)]
Bumpy copyright year
Fabian Keil [Wed, 17 Aug 2011 10:30:11 +0000 (10:30 +0000)]
Revert the previous commit for now.
Turns out that it broke a the dok targets. Ooops.
Fabian Keil [Wed, 17 Aug 2011 10:26:47 +0000 (10:26 +0000)]
In compress_buffer(), let zlib figure out the proper length for the new buffer
Simply adding 30 bytes is somewhat "suboptimal".
Fabian Keil [Wed, 17 Aug 2011 10:25:43 +0000 (10:25 +0000)]
Fix a compiler warning on GNU/Linux i386
Fabian Keil [Wed, 17 Aug 2011 10:24:55 +0000 (10:24 +0000)]
When loading action sections, verify that the referenced filters exist
Currently missing filters only result in an error message,
but eventually the severity should be upgraded to fatal.
Fabian Keil [Wed, 17 Aug 2011 10:23:42 +0000 (10:23 +0000)]
Add #96: Enabled filters should be easier to look up
Fabian Keil [Wed, 17 Aug 2011 10:23:23 +0000 (10:23 +0000)]
Get the language for #42 right.
While at it, remove the bible reference as it arrived a while ago.
Fabian Keil [Tue, 2 Aug 2011 18:21:21 +0000 (18:21 +0000)]
Fabian Keil [Sat, 30 Jul 2011 15:22:39 +0000 (15:22 +0000)]
Add tests for missing socks4 and socks4a forwarders
The code paths are different than for socks5 and the recent
memory leak in socks5_connect() was only detected due to the
already existing test.
Currently no such leaks exist in socks4_connect() but I may
add some in the future and would prefer to know about it.
Fabian Keil [Sat, 30 Jul 2011 15:17:35 +0000 (15:17 +0000)]
Deduplicate the INADDR_NONE definition for Solaris by moving it to jbsockets.h
While at it, add a comment about whether or not the workaround is still needed.
Fabian Keil [Sat, 30 Jul 2011 15:15:25 +0000 (15:15 +0000)]
In block_url(), ditch the obsolete workaround for ancient Netscape versions that supposedly couldn't properly deal with status code 403.
Fabian Keil [Sat, 30 Jul 2011 15:13:32 +0000 (15:13 +0000)]
White-space fixes for add_loader() and run_loader()
Fabian Keil [Sat, 30 Jul 2011 15:12:55 +0000 (15:12 +0000)]
Remove a useless NULL pointer check in load_trustfile()
Fabian Keil [Sat, 30 Jul 2011 15:12:42 +0000 (15:12 +0000)]
Remove two useless NULL pointer checks in load_one_re_filterfile().
csp has already been dereferenced earlier and there's
no reason why it should be NULL anyway.
Fabian Keil [Sat, 30 Jul 2011 15:12:02 +0000 (15:12 +0000)]
Set socket_error to errno if connecting fails in rfc2553_connect_to()
Previously rejected direct connections could be incorrectly reported as DNS issues.
Fabian Keil [Sat, 30 Jul 2011 15:05:49 +0000 (15:05 +0000)]
Slightly restructure the error handling in socks5_connect().
No (intended) change of behaviour.
Fabian Keil [Sat, 30 Jul 2011 15:05:23 +0000 (15:05 +0000)]
Add the filter-file loaders before the action-file loaders
This way all filters are loaded by the time the action files
are parsed and filter{} actions specifying non-existent filters
can be detected and rejected (with code not yet committed).
Fabian Keil [Sat, 30 Jul 2011 15:02:57 +0000 (15:02 +0000)]
Disable banners-by-size filters for '.thinkgeek.com/'
The filter only seems to catch pictures of the inventory.
Fabian Keil [Sat, 30 Jul 2011 15:00:20 +0000 (15:00 +0000)]
Block requests for 'go.idmnet.bbelements.com/please/showit/'
Reported by kacperdominik in #
3372959.
Fabian Keil [Tue, 19 Jul 2011 11:01:14 +0000 (11:01 +0000)]
Add #95: Support a non-standard client header in CONNECT requests that contains the URL of the requested ressource
Fabian Keil [Tue, 19 Jul 2011 11:00:38 +0000 (11:00 +0000)]
Remove #71 as it has been implemented a few days ago
Fabian Keil [Tue, 19 Jul 2011 11:00:20 +0000 (11:00 +0000)]
Minor style fix for load_config()
Fabian Keil [Sun, 17 Jul 2011 13:52:11 +0000 (13:52 +0000)]
Make sure --privoxy-address works with IPv6 addresses containing brackets, too
Fabian Keil [Sun, 17 Jul 2011 13:38:05 +0000 (13:38 +0000)]
Minor style fixes for bind_port_helper()
Fabian Keil [Sun, 17 Jul 2011 13:37:32 +0000 (13:37 +0000)]
Reduce code duplication in bind_port_helper()
Fabian Keil [Sun, 17 Jul 2011 13:37:14 +0000 (13:37 +0000)]
Minor style fix for unload_configfile()
Fabian Keil [Sun, 17 Jul 2011 13:36:48 +0000 (13:36 +0000)]
Add a comment about possibly questionable behaviour in accept_connection()
Fabian Keil [Sun, 17 Jul 2011 13:36:27 +0000 (13:36 +0000)]
Style and comment fixes for accept_connection()
Fabian Keil [Sun, 17 Jul 2011 13:36:06 +0000 (13:36 +0000)]
Restructure bind_ports_helper() a bit
Setting config->need_bind isn't necessary.
Fabian Keil [Sun, 17 Jul 2011 13:35:32 +0000 (13:35 +0000)]
Do not claim to listen on a socket until we actually do.
Patch submitted by Petr PÃsaÅ™ in #
3354485
Fabian Keil [Sun, 17 Jul 2011 13:34:36 +0000 (13:34 +0000)]
Allow to bind to multiple separate addresses.
Patch set submitted by Petr PÃsaÅ™ in #
3354485.
Fabian Keil [Sun, 17 Jul 2011 13:31:35 +0000 (13:31 +0000)]
Fix a memory leak if the socks5 server is unreachable.
I introduced the leak after the last release so this
fix is irrelevant as far as the ChangeLog is concerned.
Fabian Keil [Sun, 17 Jul 2011 13:31:02 +0000 (13:31 +0000)]
Minor style fix for listen_loop().
Fabian Keil [Sun, 17 Jul 2011 13:30:43 +0000 (13:30 +0000)]
Remove #13 which is already implemented.
Fabian Keil [Sun, 17 Jul 2011 13:30:24 +0000 (13:30 +0000)]
Also log the client socket when dropping a connection.
Fabian Keil [Fri, 8 Jul 2011 13:31:40 +0000 (13:31 +0000)]
Document compression-level
Fabian Keil [Fri, 8 Jul 2011 13:31:17 +0000 (13:31 +0000)]
Document enable-compression
Fabian Keil [Fri, 8 Jul 2011 13:30:59 +0000 (13:30 +0000)]
Add a 'to' that was missing in the FEATURE_COMPRESSION description
Fabian Keil [Fri, 8 Jul 2011 13:30:08 +0000 (13:30 +0000)]
Add an enable-compression directive and disable compression by default
Fabian Keil [Fri, 8 Jul 2011 13:29:39 +0000 (13:29 +0000)]
parse_toggle_state() only returns 0 or 1, so check for 1 directly.
Fabian Keil [Fri, 8 Jul 2011 13:29:22 +0000 (13:29 +0000)]
Also let parse_toggle_state() reject values like '1[insert garbage here]'
Fabian Keil [Fri, 8 Jul 2011 13:29:06 +0000 (13:29 +0000)]
Add a dedicated function to parse the values of toggles
Reduces duplicated code in load_config() and provides
better error handling. Invalid or missing toggle values
are now a fatal error instead of being silently ignored.
Fabian Keil [Fri, 8 Jul 2011 13:28:11 +0000 (13:28 +0000)]
If compressing the buffer failed, include buffer size and compression level in the error message
Fabian Keil [Fri, 8 Jul 2011 13:27:56 +0000 (13:27 +0000)]
In compress_buffer(), add some room for the zlib overhead
If the entropy is high enough, the data may not be compressable,
or at least not compressable enough to make room for the header.
If the compression level is 0, the "compressed" is guaranteed
to take more room than the "uncompressed" data.
Fabian Keil [Fri, 8 Jul 2011 13:27:31 +0000 (13:27 +0000)]
Add the config directive compression-level
There isn't a single best compression level for
all situations, so hard-coding one makes no sense.
Fabian Keil [Mon, 4 Jul 2011 17:48:11 +0000 (17:48 +0000)]
#94: Add a config directive to let Privoxy prefer either IPv4 (or IPv6) addresses
Fabian Keil [Mon, 4 Jul 2011 17:47:50 +0000 (17:47 +0000)]
#93: Add a config directive to let Privoxy explicitly request either IPv4 (or IPv6) addresses
Fabian Keil [Mon, 4 Jul 2011 17:47:29 +0000 (17:47 +0000)]
In bind_port(), stop hinting AI_ADDRCONFIG to getaddrinfo()
It's not strictly needed there and causes GNU libc to fail to
return loopback addresses if no non-loopback IP address is
configured on the system. Apparently this is the result of a
somewhat strict but valid interpretation of RFC 3493 and
considered a feature. Sad but true.
This is mainly an issue if the system is using DHCP and Privoxy
is started before the network is completely configured.
Reported by Raphael Marichez in #
3349356.
Additional insight from Petr Pisar.
Fabian Keil [Sun, 3 Jul 2011 17:55:23 +0000 (17:55 +0000)]
Move LOWER_LENGTH_LIMIT_FOR_COMPRESSION into cgi.[ch] where it belongs
While at it, turn it into a proper variable so it can be changed from the debugger.
Fabian Keil [Sun, 3 Jul 2011 17:54:29 +0000 (17:54 +0000)]
Fix spelling of LOWER_LENGTH_LIMIT_FOR_COMPRESSION
Fabian Keil [Sun, 3 Jul 2011 17:53:00 +0000 (17:53 +0000)]
Fix the FEATURE_ZLIB description. It's mainly used for decompression.
Fabian Keil [Wed, 29 Jun 2011 20:31:05 +0000 (20:31 +0000)]
Unbreak execute_dumb_fetch_test().
Apparently 1.73 wasn't regression-tested. How ironic.
Fabian Keil [Wed, 29 Jun 2011 20:30:18 +0000 (20:30 +0000)]
Remove a stray exit(0)
Fabian Keil [Wed, 29 Jun 2011 20:29:58 +0000 (20:29 +0000)]
Cosmetics for generate_creation_time()
Fabian Keil [Wed, 29 Jun 2011 20:29:29 +0000 (20:29 +0000)]
Remove a couple of unnecessary variables in generate_creation_time()
Fabian Keil [Wed, 29 Jun 2011 20:29:12 +0000 (20:29 +0000)]
Let Getopt::Long perform some sanity checks for parameters that are supposed to have numerical values.
While it doesn't prevent all invalid input, it's better than nothing.
Fabian Keil [Wed, 29 Jun 2011 18:36:49 +0000 (18:36 +0000)]
In finish_http_response(), don't add an Content-Encoding header if compression failed.
Fabian Keil [Wed, 29 Jun 2011 18:36:24 +0000 (18:36 +0000)]
Let Getopt::Long perform some sanity checks for parameters that are supposed to have numerical values.
While it doesn't prevent all invalid input, it's better than nothing.
Fabian Keil [Wed, 29 Jun 2011 18:35:53 +0000 (18:35 +0000)]
Bump version and copyright date
Fabian Keil [Wed, 29 Jun 2011 18:35:38 +0000 (18:35 +0000)]
Bump generated Firefox version to 5.0
Fabian Keil [Wed, 29 Jun 2011 18:35:23 +0000 (18:35 +0000)]
Only randomize the release date if the new --randomize-release-date option is enabled.
Firefox versions after 4 use a fixed date string without meaning.
Fabian Keil [Wed, 29 Jun 2011 18:35:01 +0000 (18:35 +0000)]
Implement a --sleep-time option
Specifies a number of second to sleep between tests, defaults to 0.
Fabian Keil [Wed, 29 Jun 2011 18:34:38 +0000 (18:34 +0000)]
Consistently use quote() in execute_dumb_fetch_test().
Fabian Keil [Wed, 29 Jun 2011 18:34:23 +0000 (18:34 +0000)]
In load_action_files(), use quote() instead of quoting manually.
Fabian Keil [Sat, 25 Jun 2011 12:45:47 +0000 (12:45 +0000)]
Rephrase the sentence describing the default configuration file name so it doesn't mix kernels and operating systems.
Fabian Keil [Sat, 25 Jun 2011 12:41:17 +0000 (12:41 +0000)]
Unblock adainitiative.org/
Fabian Keil [Sat, 25 Jun 2011 12:40:55 +0000 (12:40 +0000)]
Remove a double space in an error message
Fabian Keil [Sat, 25 Jun 2011 12:40:27 +0000 (12:40 +0000)]
Terminate HTML lines in static error messages with \n instead of \r\n.
Fabian Keil [Sat, 25 Jun 2011 12:39:41 +0000 (12:39 +0000)]
Simplify cgi_error_unknown() a bit.
Fabian Keil [Sat, 25 Jun 2011 12:39:18 +0000 (12:39 +0000)]
In get_log_timestamp(), use the ISO 8601 date format %Y-%m-%d.
It's only slightly longer than the old format, but contains
the full date including the year and allows sorting by date
(when grepping in multiple log files) without hassle.
Fabian Keil [Thu, 23 Jun 2011 14:04:29 +0000 (14:04 +0000)]
Let the msn filter hide elements with the id 's_notf_div'.
They only seem to be used to advertise site 'enhancements'.
Fabian Keil [Thu, 23 Jun 2011 14:03:38 +0000 (14:03 +0000)]
Disable the range-requests tagger for tests that break if it's enabled
Fabian Keil [Thu, 23 Jun 2011 14:02:53 +0000 (14:02 +0000)]
Delimit the date in log messages with dashes instead of slashes
Fabian Keil [Thu, 23 Jun 2011 14:02:28 +0000 (14:02 +0000)]
Accept and highlight: Added header: Content-Encoding: deflate
Fabian Keil [Thu, 23 Jun 2011 14:02:02 +0000 (14:02 +0000)]
Accept and highlight: Compressed content from 29258 to 8630 bytes.
Fabian Keil [Thu, 23 Jun 2011 14:01:39 +0000 (14:01 +0000)]
Show FEATURE_COMPRESSION on the show-status page
Fabian Keil [Thu, 23 Jun 2011 14:01:01 +0000 (14:01 +0000)]
Add support for (re-)compression of buffered content and CGI responses.
For now it has to be explicitly enabled with the configure option --enable-compression.
Fabian Keil [Thu, 23 Jun 2011 13:59:56 +0000 (13:59 +0000)]
Add a fast-redirects exception for '.googleusercontent.com/.*=cache'
Fabian Keil [Thu, 23 Jun 2011 13:58:53 +0000 (13:58 +0000)]
Let the yahoo filter hide '.ads'
Fabian Keil [Thu, 23 Jun 2011 13:58:22 +0000 (13:58 +0000)]
Remove a useless check in write_socket()
Fabian Keil [Thu, 23 Jun 2011 13:57:47 +0000 (13:57 +0000)]
In decompress_iob() change a buffer limit check from '==' to the '>=' one would expect at first glance.
At the same time explain why we don't really have to.