summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2003-06-06(connect_to_upstream): Fixed an off-by-one error in the snprintf()Robert James Kaes1-5/+6
call used to build the URL for the upstream proxy. [Patch suggested by David T. Pierso]
2003-06-02Remove the -O2 optimization if debugging is enabled. Also, remove theRobert James Kaes1-5/+9
-g option if the code does _not_ have debugging enabled.
2003-06-02Included additional examples of how to configure the upstream proxyRobert James Kaes1-10/+15
support for more selective proxying.
2003-06-02(upstream_get):Robert James Kaes2-38/+67
(upstream_add): Added support to allow ip addresses and networks to be used when matching an upstream proxy directive. [Code by Peter da Silva]
2003-05-31Changed the safefree() macro to make it safe to use a conditionalRobert James Kaes1-3/+11
statement, and also safe to use with a rvalue that has a side effect. [Bug fix recommended by Peter da Silva]
2003-05-31# Changed all the for calls to use the != test rather than < test.Robert James Kaes7-31/+31
The change was recommended in the C/C++ User Journal magazine.
2003-05-30# Updated ChangeLogRobert James Kaes1-0/+40
2003-05-30# Bumped up the version numberRobert James Kaes1-2/+2
2003-05-30# Changed the calls to vector_getentry() to use the new callingRobert James Kaes2-4/+5
convention.
2003-05-30(vector_getentry): Changed the API to return the data pointer and haveRobert James Kaes2-18/+19
the length returned in a argument variable pointer. This should be a more natural way of using the function.
2003-05-29# Added a vector_prepend() function and recoded the old vector_insert()Robert James Kaes2-15/+41
to be a general "insert" for both vector_append() and vector_prepend()
2003-05-29# Renamed the vector_insert() calls to vector_append()Robert James Kaes2-4/+4
2003-05-29(vector_append): Renamed the vector_insert() function to moreRobert James Kaes2-9/+9
accurately indicate that entries are appended to the end of the vector.
2003-05-29Improved the upstream proxy support by making the upstream proxyRobert James Kaes5-23/+210
server configurable based on the destination host. [Code written by Peter da Silva]
2003-05-10Added support for installing the HTML error files used by tinyproxy.Robert James Kaes1-2/+17
2003-05-05(establish_http_connection): If the port being requested is not aRobert James Kaes1-7/+23
standard HTTP port (80 or 443) append the port string to the host header; otherwise, leave the host string with only the host's domain name. Replaced all occurrences of constant 80 and 443 with defines HTTP_PORT and HTTP_PORT_SSL.
2003-05-04Fixed a bug that would kill a child process because of an invalidRobert James Kaes2-5/+4
safefree() call. Basically, destroy_conn() was trying to free memory not allocated by malloc. [Fix by David T. Pierson]
2003-04-16# Updated change logRobert James Kaes1-0/+23
2003-04-16# Updated version numberRobert James Kaes1-2/+2
2003-04-16(build_url): Rebuild the URL from the component pieces. This functionRobert James Kaes1-13/+53
is used by the transparent proxy code. [Anatole Shaw] (process_request): Fixed up the transparent proxy code so that filtering can be done on the whole URL. [Anatole Shaw] (pull_client_data): Added a bug fix for Internet Explorer (IE). IE will leave an extra CR and LF after the data in an HTTP POST. The new code will eat the extra bytes if they're present. Thanks to Yannick Koehler for finding the bug and offering an explanation as to why it was happening. Changed all calls of connptr->remote_content_length to connptr->content_length.server
2003-04-16# Changed it again to this time use the TINYPROXY_DEBUG environmentRobert James Kaes1-7/+13
variable to determine whether to wait for a connection from GDB.
2003-04-16# The programmer is now made to _explicitly_ enable the GDB support inRobert James Kaes1-2/+3
the child handling function.
2003-04-16Removed the remote_content_length field in the "conn" structure andRobert James Kaes2-4/+8
replaced it with a smaller structure containing both the remote/server and the local/client content-length fields if they're present in the HTTP response headers.
2003-04-01# Updated change log (should have done this _before_ the tagging of course.)Robert James Kaes1-0/+64
2003-04-01Removed duplicate header tests.Robert James Kaes1-6/+4
2003-04-01Reorganized some of the code, and remove some function tests for whichRobert James Kaes1-19/+13
results were never used.
2003-04-01(add_error_variable): Test whether connptr->error_variable is NULL,Robert James Kaes1-2/+6
and if so call safemalloc(). This is needed since saferealloc() will assert() if the first argument is a NULL pointer.
2003-03-26(establish_http_connection): Always include the port number for theRobert James Kaes1-3/+4
requested server. This fixes a problem when the server is not listening on the default port, 80. [Fix suggested by duncan@sapio.co.uk]
2003-03-17# Moved the location of the "initializing" log message to below theRobert James Kaes1-3/+3
processing of the command line options.
2003-03-17Create the doc/report.sh script from the new template file.Robert James Kaes1-3/+15
2003-03-17Moved the doc/report.sh script to doc/report.sh.tmpl file. The scriptRobert James Kaes1-2/+4
is then updated by the make procedure to reflect the correct location of the tinyproxy executable.
2003-03-14Made get_html_file() and lookup_variable() static functions since theyRobert James Kaes1-3/+3
are only used with this file.
2003-03-14# Added copyright notices for Steve. Reformatted the source code toRobert James Kaes2-24/+62
better match the existing tinyproxy practise. Included a few bug fixes from Steve.
2003-03-14(send_http_message): Changed the function to use the new http_messageRobert James Kaes1-31/+20
API.
2003-03-14# Updated all the calls to indicate_http_error() to include aRobert James Kaes1-18/+35
terminating NULL. The va_arg() function requires it to work properly.
2003-03-13# Include the HTML variable substitution documentation files with theRobert James Kaes1-1/+5
tarball.
2003-03-13# Updated ChangeLog with Steven Young's log of changes to include theRobert James Kaes1-0/+26
variable substitution code.
2003-03-13Allow the URL for the statistic page to be controlled from theRobert James Kaes2-2/+9
configuration file, rather than being hard-coded in the program. [John M Wright]
2003-03-13Include example information for the new "file" directives (ErrorFile,Robert James Kaes1-0/+24
DefaultErrorFile, StatFile.) [Steven Young]
2003-03-13# Added default HTML files for HTTP error messages and the statsRobert James Kaes4-0/+80
page. [Steven Young]
2003-03-13Moved the send_http_error_message() and indicate_http_error()Robert James Kaes2-74/+2
functions into the htmlerror.c file, and recoded them to use the new variable substitution system. [Steven Young]
2003-03-13# Added variables to config structure to keep track of the files to beRobert James Kaes2-10/+37
displayed for various HTTP errors and the stats page. [Steven Young]
2003-03-13# Changed showstats() to use the HTML variable functions whenRobert James Kaes1-12/+35
possible. It still retains the hard-coded page for when an HTML file is not available. [Steven Young]
2003-03-13# Changed calls to indicate_http_error() to use the new HTML variableRobert James Kaes1-34/+53
substitution mechanism. [Steven Young]
2003-03-13# Added parser support for the error file configuration keywordsRobert James Kaes2-3/+19
(ErrorFile, DefaultErrorFile, StatFile) [Steven Young]
2003-03-13# Added variables to keep track of the variables to be substituted inRobert James Kaes2-4/+29
.html files displayed to the client [Steven Young]
2003-03-13# Steve Young's code to implement file based HTML error messages,Robert James Kaes2-0/+262
rather than the hard coded string.
2003-03-13# Added htmlerror.c and htmlerror.hRobert James Kaes1-3/+2
2003-03-13# Changed the wording of the header comment to remove the reference toRobert James Kaes1-3/+3
"dnsserver"
2003-03-13# (child_main): If this is a debugging build output the child processRobert James Kaes1-2/+7
ID and wait for 10 seconds so we have time to connect gdb to the child. This is needed if we want to use gdb against the child process.