1 <!DOCTYPE article PUBLIC "-//OASIS//DTD DocBook V3.1//EN"[
2 <!entity % dummy "IGNORE">
3 <!entity supported SYSTEM "supported.sgml">
4 <!entity newfeatures SYSTEM "newfeatures.sgml">
5 <!entity p-intro SYSTEM "privoxy.sgml">
6 <!entity history SYSTEM "history.sgml">
7 <!entity seealso SYSTEM "seealso.sgml">
8 <!entity p-version "3.0.30">
9 <!entity p-status "UNRELEASED">
10 <!entity % p-not-stable "INCLUDE">
11 <!entity % p-stable "IGNORE">
12 <!entity % p-text "IGNORE"> <!-- define we are not a text only doc -->
13 <!entity % p-doc "INCLUDE"> <!-- and we are a formal doc -->
14 <!entity % seealso-extra "INCLUDE"> <!-- extra stuff from seealso.sgml -->
15 <!entity my-copy "©"> <!-- kludge for docbook2man -->
18 File : doc/source/developer-manual.sgml
20 Purpose : developer manual
22 Copyright (C) 2001-2020 Privoxy Developers https://www.privoxy.org/
25 ========================================================================
26 NOTE: Please read developer-manual/documentation.html before touching
27 anything in this, or other Privoxy documentation. You have been warned!
28 Failure to abide by this rule will result in the revocation of your license
29 to live a peaceful existence!
30 ========================================================================
36 <title>Privoxy Developer Manual</title>
39 <!-- Completely the wrong markup, but very little is allowed -->
40 <!-- in this part of an article. FIXME -->
41 <ulink url="https://www.privoxy.org/user-manual/copyright.html">Copyright</ulink>
42 &my-copy; 2001-2020 by
43 <ulink url="https://www.privoxy.org/">Privoxy Developers</ulink>
49 Note: this should generate a separate page, and a live link to it.
50 But it doesn't for some mysterious reason. Please leave commented
51 unless it can be fixed proper. For the time being, the copyright
52 statement will be in copyright.smgl.
56 <legalnotice id="legalnotice">
58 text goes here ........
69 This is here to keep vim syntax file from breaking :/
70 If I knew enough to fix it, I would.
71 PLEASE DO NOT REMOVE! HB: hal@foobox.net
76 The developer manual provides guidance on coding, testing, packaging, documentation
77 and other issues of importance to those involved with
78 <application>Privoxy</application> development. It is mandatory (and helpful!) reading
79 for anyone who wants to join the team. Note that it's currently out of date
80 and may not be entirely correct. As always, patches are welcome.
83 <!-- Include privoxy.sgml boilerplate text: -->
85 <!-- &p-intro; Someone interested enough in the project to contribute
86 will already know at this point what Privoxy is. -->
88 <!-- end boilerplate -->
91 Please note that this document is constantly evolving. This copy represents
92 the state at the release of version &p-version;.
93 You can find the latest version of the this manual at <ulink
94 url="https://www.privoxy.org/developer-manual/">https://www.privoxy.org/developer-manual/</ulink>.
95 Please have a look at the
96 <ulink url="https://www.privoxy.org/user-manual/contact.html">contact section in the user manual</ulink>
97 if you are interested in contacting the developers.
104 <!-- ~~~~~ New section ~~~~~ -->
105 <sect1 id="introduction"><title>Introduction</title>
108 I don't like seeing blank space :) So added *something* here.
112 <application>Privoxy</application>, as an heir to
113 <application>Junkbuster</application>, is a <ulink
114 url="https://www.privoxy.org/user-manual/copyright.html">Free Software</ulink> project.
115 As such, <application>Privoxy</application> development is potentially open
116 to anyone who has the time, knowledge, and desire to contribute
117 in any capacity. Our goals are simply to continue the mission,
118 to improve <application>Privoxy</application>, and
119 to make it available to as wide an audience as possible.
122 One does not have to be a programmer to contribute. Packaging, testing,
123 documenting and porting, are all important jobs as well.
126 <!-- ~~~~~ New section ~~~~~ -->
127 <sect2 id="quickstart"><title>Quickstart to Privoxy Development</title>
129 The first step is to join the <ulink
130 url="https://lists.privoxy.org/mailman/listinfo/privoxy-devel">privoxy-devel mailing list</ulink>.
131 You can submit your ideas or, even better, patches.
132 Patches can also be submitted to the
133 <ulink url="https://sourceforge.net/p/ijbswa/patches/">Sourceforge patch tracker</ulink>.
136 You will also need to have a git package installed,
137 in order to access the git repository.
138 Having the GNU build tools is also going to be important (particularly,
142 For the time being (read, this section is under construction), you can
143 also refer to the extensive comments in the source code. In fact,
144 reading the code is recommended in any case.
149 <!-- ~~~~~ New section ~~~~~ -->
150 <sect1 id="git"><title>The Git Repository</title>
152 If you become part of the active development team, you will eventually
153 need write access to our holy grail, the Git repository. One of the
154 team members will need to set this up for you. Please read
155 this chapter completely before accessing via Git.
158 <sect2 id="gitaccess"><title>Access to Git</title>
160 The project's Git repository is hosted on the
161 <ulink url="https://www.privoxy.org/">Privoxy webserver</ulink>.
162 For Privoxy team members with push privileges the Git repository URL is
163 <literal>ssh://git@git.privoxy.org:23/git/privoxy.git</literal>.
166 Contributors without push privileges can
167 <quote>git clone https://www.privoxy.org/git/privoxy.git</quote>.
170 The central repository is called <literal>privoxy</literal>, and the
171 source branch is called <literal>master</literal>. Subfolders exist
172 within the project for target-dependent build and packaging tools, each
173 including the name of the target operating system in their name (e.g.
174 Windows, OSXPackageBuilder, debian). There is a webview of the Git
176 <ulink url="https://www.privoxy.org/gitweb/?p=privoxy.git;a=tree">
177 https://www.privoxy.org/gitweb/?p=privoxy.git;a=tree</ulink>,
178 which might help with visualizing how these pieces fit together.
182 <sect2 id="gitbranches">
183 <title>Branches</title>
185 Whilst the central repository contains only the master branch, developers
186 are of course free to create branches in their local repositories as they
187 develop features, fixes, or update the target-dependent tools. Only once
188 such changes are fully tested ought they be pushed back to the central
189 repository master branch.
192 Before pushing stuff, please rebase it on a current master so we get
193 an uncomplicated commit history. Avoid merges where possible.
196 Here's an example git sesssion that should result in a merge-free history:
199 fk@t520 ~/git/privoxy $git checkout master
200 Switched to branch 'master'
201 Your branch is up to date with 'origin/master'.
202 # Make sure you have the latest changes
203 fk@t520 ~/git/privoxy $git pull
205 # Create a local banch for changes
206 fk@t520 ~/git/privoxy $git checkout -b local-branch
207 Switched to a new branch 'local-branch'
209 fk@t520 ~/git/privoxy $gmake dok dok-tidy
212 fk@t520 ~/git/privoxy $git diff
214 # Commit your changes if they look goood
215 fk@t520 ~/git/privoxy $git commit -m "developer-manual: Regenerate" doc/webserver/
216 [local-branch 1abb7316] developer-manual: Regenerate
217 1 file changed, 2 insertions(+), 2 deletions(-)
219 fk@t520 ~/git/privoxy $git show
221 # Go to the master branch
222 fk@t520 ~/git/privoxy $git checkout master
223 Switched to branch 'master'
224 Your branch is up to date with 'origin/master'.
225 # Make sure you are still in sync
226 fk@t520 ~/git/privoxy $git pull
229 # Apply the commit you made to the local-branch
230 fk@t520 ~/git/privoxy $git cherry-pick local-branch
231 [master 046e85e2] developer-manual: Regenerate
232 Date: Tue Dec 15 05:10:07 2020 +0100
233 1 file changed, 2 insertions(+), 2 deletions(-)
234 # Make sure the history looks as expected
235 fk@t520 ~/git/privoxy $git log -p
236 # Finally push your change to the Privoxy repository
237 fk@t520 ~/git/privoxy $git push
239 # Go back to the local branch
240 fk@t520 ~/git/privoxy $git checkout local-branch
241 # Rebase on top of master and continue hacking
242 fk@t520 ~/git/privoxy $git rebase master
243 Successfully rebased and updated refs/heads/local-branch.
247 Branches are used to fork a sub-development path from the main trunk.
248 Within the <literal>current</literal> module where the sources are, there
249 is always at least one <quote>branch</quote> from the main trunk
250 devoted to a stable release series. The main trunk is where active
251 development takes place for the next stable series (e.g. 3.2.x).
252 So just prior to each stable series (e.g. 3.0.x), a branch is created
253 just for stable series releases (e.g. 3.0.0 -> 3.0.1 -> 3.0.2, etc).
254 Once the initial stable release of any stable branch has taken place,
255 this branch is <emphasis>only used for bugfixes</emphasis>, which have
256 had prior testing before being committed to Git. (See <link
257 linkend="versionnumbers">Version Numbers</link> below for details on
262 At one time there were two distinct branches: stable and unstable. The
263 more drastic changes were to be in the unstable branch. These branches
264 have now been merged to minimize time and effort of maintaining two
269 This will result in at least two active branches, which means there may
270 be occasions that require the same (or similar) item to be
271 checked into to two different places (assuming its a bugfix and needs
272 fixing in both the stable and unstable trees). This also means that in
273 order to have access to both trees, both will have to be checked out
274 separately. Use the <literal>cvs -r</literal> flag to check out a
275 branch, e.g: <literal>cvs co -r v_3_0_branch current</literal>.
280 <sect2 id="gitcommit"><title>Git Commit Guidelines</title>
282 The source tree is the heart of every software project. Every effort must
283 be made to ensure that it is readable, compilable and consistent at all
284 times. <!-- There are differing guidelines for the stable branch and the
285 main development trunk, and --> We expect anyone with Git access to strictly
286 adhere to the following guidelines:
290 Basic Guidelines, for all branches:
294 Please don't commit even
295 a small change without testing it thoroughly first. When we're
296 close to a public release, ask a fellow developer to review your
300 Your commit message should give a concise overview of <emphasis>what you
301 changed</emphasis> (no big details) and <emphasis>why you changed it</emphasis>
302 Just check previous messages for good examples.
305 Don't use the same message on multiple files, unless it equally applies to
309 If your changes span multiple files, and the code won't recompile unless
310 all changes are committed (e.g. when changing the signature of a function),
311 then commit all files one after another, without long delays in between.
312 If necessary, prepare the commit messages in advance.
315 Before changing things on Git, make sure that your changes are in line
316 with the team's general consensus on what should be done.
320 Note that near a major public release, we get more cautious.
321 There is always the possibility to submit a patch to the <ulink
322 url="https://sourceforge.net/p/ijbswa/patches/">patch
323 tracker</ulink> or the <ulink
324 url="https://lists.privoxy.org/mailman/listinfo/privoxy-devel">privoxy-devel mailing list</ulink>
332 Stable branches are handled with more care, especially after the
333 initial *.*.0 release, and we are just in bugfix mode. In addition to
334 the above, the below applies only to the stable branch (currently the
335 <literal>v_3_0_branch</literal> branch):
342 Do not commit <emphasis>anything</emphasis> unless your proposed
343 changes have been well tested first, preferably by other members of the
344 project, or have prior approval of the project leaders or consensus
350 Where possible, bugfixes and changes should be tested in the main
351 development trunk first. There may be occasions where this is not
357 Alternately, proposed changes can be submitted as patches output by
358 <literal>git format-patch</literal> to the privoxy-devel mailing list
359 or alternatively to the patch tracker on Sourceforge:
360 <ulink url="https://sourceforge.net/p/ijbswa/patches/">
361 https://sourceforge.net/p/ijbswa/patches/</ulink>.
362 Then ask for peer review.
367 Do not even think about anything except bugfixes. No new features!
378 <!-- ~~~~~ New section ~~~~~ -->
379 <sect1 id="documentation"><title>Documentation Guidelines</title>
381 All formal documents are maintained in Docbook SGML and located in the
382 <computeroutput>doc/source/*</computeroutput> directory. You will need
383 <ulink url="https://www.docbook.org/">Docbook</ulink>, the Docbook
384 DTD's and the Docbook modular stylesheets (or comparable alternatives),
385 and either <application>jade</application> or
386 <application>openjade</application> (recommended) installed in order to
387 build docs from source. Currently there is <ulink
388 url="../user-manual/index.html"><citetitle>user-manual</citetitle></ulink>,
389 <ulink url="../faq/index.html"><citetitle>FAQ</citetitle></ulink>, and, of
390 course this, the <citetitle>developer-manual</citetitle> in this format.
391 The <citetitle>README</citetitle>, <citetitle>AUTHORS</citetitle>,
392 <citetitle>INSTALL</citetitle>,
393 <citetitle>privoxy.1</citetitle> (man page), and
394 <citetitle>config</citetitle> files are also now maintained as Docbook
395 SGML. These files, when built, in the top-level source directory are
396 generated files! Also, the <application>Privoxy</application> <filename>index.html</filename> (and a
397 variation on this file, <filename>privoxy-index.html</filename>,
398 meant for inclusion with doc packages), are maintained as SGML as well.
399 <emphasis>DO NOT edit these directly</emphasis>. Edit the SGML source, or
400 contact someone involved in the documentation.
403 <filename>config</filename> requires some special handling. The reason it
404 is maintained this way is so that the extensive comments in the file
405 mirror those in <citetitle>user-manual</citetitle>. But the conversion
406 process requires going from SGML to HTML to text to special formatting
407 required for the embedded comments. Some of this does not survive so
408 well. Especially some of the examples that are longer than 80 characters.
411 Other, less formal documents (e.g. <filename>LICENSE</filename>) are
412 maintained as plain text files in the top-level source directory.
415 Packagers are encouraged to include this documentation. For those without
416 the ability to build the docs locally, text versions of each are kept in
417 Git. HTML versions are also being kept in Git under
418 <filename>doc/webserver/*</filename>.
421 Formal documents are built with the Makefile targets of
422 <computeroutput>make dok</computeroutput>.
423 The build process uses the document SGML sources in
424 <computeroutput>doc/source/*/*</computeroutput> to update all text files in
425 <computeroutput>doc/text/</computeroutput> and to update all HTML
426 documents in <computeroutput>doc/webserver/</computeroutput>.
429 Documentation writers should please make sure documents build
430 successfully before committing to Git, if possible.
433 How do you update the webserver (i.e. the pages on privoxy.org)?
436 <orderedlist numeration="arabic">
438 First, build the docs by running <computeroutput>make
439 dok dok-tidy</computeroutput>.
442 Run <computeroutput>make webserver</computeroutput> which copies all
443 files from <computeroutput>doc/webserver</computeroutput> to the
444 sourceforge webserver via ssh.
449 Finished docs should be occasionally submitted to Git
450 (<filename>doc/webserver/*/*.html</filename>) so that those without
451 the ability to build them locally, have access to them if needed.
452 This is especially important just prior to a new release! Please
453 do this <emphasis>after</emphasis> the <literal>$VERSION</literal> and
454 other release specific data in <filename>configure.in</filename> has been
455 updated (this is done just prior to a new release).
458 <!-- ~~~~~ New section ~~~~~ -->
460 <title>Quickstart to Docbook and SGML</title>
462 If you are not familiar with SGML, it is a markup language similar to HTML.
463 Actually, not a mark up language per se, but a language used to define
464 markup languages. In fact, HTML is an SGML application. Both will use
465 <quote>tags</quote> to format text and other content. SGML tags can be much
466 more varied, and flexible, but do much of the same kinds of things. The tags,
467 or <quote>elements</quote>, are definable in SGML. There is no set
468 <quote>standards</quote>. Since we are using
469 <application>Docbook</application>, our tags are those that are defined by
470 <application>Docbook</application>. Much of how the finish document is
471 rendered is determined by the <quote>stylesheets</quote>.
472 The stylesheets determine how each tag gets translated to HTML, or other
477 Tags in Docbook SGML need to be always <quote>closed</quote>. If not, you
478 will likely generate errors. Example: <literal><title>My
479 Title</title></literal>. They are also case-insensitive, but we
480 strongly suggest using all lower case. This keeps compatibility with
481 [Docbook] <application>XML</application>.
485 Our documents use <quote>sections</quote> for the most part. Sections
486 will be processed into HTML headers (e.g. <literal>h1</literal> for
487 <literal>sect1</literal>). The <application>Docbook</application> stylesheets
488 will use these to also generate the Table of Contents for each doc. Our
489 TOC's are set to a depth of three. Meaning <literal>sect1</literal>,
490 <literal>sect2</literal>, and <literal>sect3</literal> will have TOC
491 entries, but <literal>sect4</literal> will not. Each section requires
492 a <literal><title></literal> element, and at least one
493 <literal><para></literal>. There is a limit of five section
494 levels in Docbook, but generally three should be sufficient for our
499 Some common elements that you likely will use:
504 <emphasis><para></para></emphasis>, paragraph delimiter. Most
505 text needs to be within paragraph elements (there are some exceptions).
508 <emphasis><emphasis></emphasis></emphasis>, the stylesheets
512 <emphasis><filename></filename></emphasis>, files and directories.
515 <emphasis><command></command></emphasis>, command examples.
518 <emphasis><literallayout></literallayout></emphasis>, like
519 <literal><pre></literal>, more or less.
522 <emphasis><itemizedlist></itemizedlist></emphasis>, list with bullets.
525 <emphasis><listitem></listitem></emphasis>, member of the above.
528 <emphasis><screen></screen></emphasis>, screen output, implies
529 <literal><literallayout></literal>.
532 <emphasis><ulink url="example.com"></ulink></emphasis>, like
533 HTML <literal><a></literal> tag.
536 <emphasis><quote></quote></emphasis>, for, doh, quoting text.
541 Look at any of the existing docs for examples of all these and more.
546 <!-- <quote><ulink url="http://opensource.bureau-cornavin.com/crash-course/index.html">
547 domain no longer exists so link to the wayback archive -->
548 <quote><ulink url="https://web.archive.org/web/20160315230758/http://opensource.bureau-cornavin.com/crash-course/index.html">
549 Writing Documentation Using DocBook - A Crash Course</ulink></quote> useful.
553 <!-- ~~~~~ New section ~~~~~ -->
554 <sect2 id="docstyle">
555 <title><application>Privoxy</application> Documentation Style</title>
557 It will be easier if everyone follows a similar writing style. This
558 just makes it easier to read what someone else has written if it
559 is all done in a similar fashion.
567 All tags should be lower case.
572 Tags delimiting a <emphasis>block</emphasis> of text (even small
573 blocks) should be on their own line. Like:
581 Tags marking individual words, or few words, should be in-line:
584 Just to <emphasis>emphasize</emphasis>, some text goes here.
589 Tags should be nested and step indented for block text like: (except
597 Some text goes here in our list example.
600 </itemizedlist>
604 This makes it easier to find the text amongst the tags ;-)
609 Use white space to separate logical divisions within a document,
610 like between sections. Running everything together consistently
611 makes it harder to read and work on.
616 Do not hesitate to make comments. Comments can either use the
617 <comment> element, or the <!-- --> style comment
618 familiar from HTML. (Note in Docbook v4.x <comment> is
619 replaced by <remark>.)
624 We have an international audience. Refrain from slang, or English
625 idiosyncrasies (too many to list :). Humor also does not translate
631 Try to keep overall line lengths in source files to 80 characters or less
632 for obvious reasons. This is not always possible, with lengthy URLs for
638 Our documents are available in differing formats. Right now, they
639 are just plain text and/or HTML, but others are always a
640 future possibility. Be careful with URLs (<ulink>), and avoid
644 My favorite site is <ulink url="http://example.com">here</ulink>.
647 This will render as <quote>My favorite site is here</quote>, which is
648 not real helpful in a text doc. Better like this:
651 My favorite site is <ulink url="http://example.com">example.com</ulink>.
656 All documents should be spell checked occasionally.
657 <application>aspell</application> can check SGML with the
658 <literal>-H</literal> option. (<application>ispell</application> I think
668 <!-- ~~~~~ New section ~~~~~ -->
670 <sect2 id="custom-entities"><title>Privoxy Custom Entities</title>
672 <application>Privoxy</application> documentation is using
673 a number of customized <quote>entities</quote> to facilitate
674 documentation maintenance.
677 We are using a set of <quote>boilerplate</quote> files with generic text,
678 that is used by multiple docs. This way we can write something once, and use
679 it repeatedly without having to re-write the same content over and over again.
680 If editing such a file, keep in mind that it should be
681 <emphasis>generic</emphasis>. That is the purpose; so it can be used in varying
682 contexts without additional modifications.
685 We are also using what <application>Docbook</application> calls
686 <quote>internal entities</quote>. These are like variables in
687 programming. Well, sort of. For instance, we have the
688 <literal>p-version</literal> entity that contains the current
689 <application>Privoxy</application> version string. You are strongly
690 encouraged to use these where possible. Some of these obviously
691 require re-setting with each release (done by the Makefile). A sampling of
692 custom entities are listed below. See any of the main docs for examples.
698 Re- <quote>boilerplate</quote> text entities are defined like:
701 <literal><!entity supported SYSTEM "supported.sgml"></literal>
704 In this example, the contents of the file,
705 <filename>supported.sgml</filename> is available for inclusion anywhere
706 in the doc. To make this happen, just reference the now defined
707 entity: <literal>&supported;</literal> (starts with an ampersand
708 and ends with a semi-colon), and the contents will be dumped into
709 the finished doc at that point.
715 Commonly used <quote>internal entities</quote>:
719 <emphasis>p-version</emphasis>: the <application>Privoxy</application>
720 version string, e.g. <quote>&p-version;</quote>.
723 <emphasis>p-status</emphasis>: the project status, either
724 <quote>alpha</quote>, <quote>beta</quote>, or <quote>stable</quote>.
727 <emphasis>p-not-stable</emphasis>: use to conditionally include
728 text in <quote>not stable</quote> releases (e.g. <quote>beta</quote>).
731 <emphasis>p-stable</emphasis>: just the opposite.
734 <emphasis>p-text</emphasis>: this doc is only generated as text.
740 There are others in various places that are defined for a specific
741 purpose. Read the source!
748 <!-- <listitem><para>be consistent with the redirect script (i.e. the <application>Privoxy</application> program -->
749 <!-- points via the redirect URL at sf to valid end-points in the document)</para></listitem> -->
751 <!-- ~~~~~ New section ~~~~~ -->
752 <sect1 id="coding"><title>Coding Guidelines</title>
754 <sect2 id="s1"><title>Introduction</title>
756 <para>This set of standards is designed to make our lives easier. It is
757 developed with the simple goal of helping us keep the "new and improved
758 <application>Privoxy</application>" consistent and reliable. Thus making
759 maintenance easier and increasing chances of success of the
762 <para>And that of course comes back to us as individuals. If we can
763 increase our development and product efficiencies then we can solve more
764 of the request for changes/improvements and in general feel good about
765 ourselves. ;-></para>
769 <sect2 id="s2"><title>Using Comments</title>
772 <sect3 id="s3"><title>Comment, Comment, Comment</title>
774 <para><emphasis>Explanation:</emphasis></para>
776 <para>Comment as much as possible without commenting the obvious.
777 For example do not comment "variable_a is equal to variable_b".
778 Instead explain why variable_a should be equal to the variable_b.
779 Just because a person can read code does not mean they will
780 understand why or what is being done. A reader may spend a lot
781 more time figuring out what is going on when a simple comment
782 or explanation would have prevented the extra research. Please
783 help your fellow Privoxy developers out!</para>
785 <para>The comments will also help justify the intent of the code.
786 If the comment describes something different than what the code
787 is doing then maybe a programming error is occurring.</para>
789 <para><emphasis>Example:</emphasis></para>
791 /* if page size greater than 1k ... */
792 if (page_length() > 1024)
794 ... "block" the page up ...
797 /* if page size is small, send it in blocks */
798 if (page_length() > 1024)
800 ... "block" the page up ...
803 This demonstrates 2 cases of "what not to do". The first is a
804 "syntax comment". The second is a comment that does not fit what
805 is actually being done.
811 <sect3 id="s4"><title>Use blocks for comments</title>
813 <para><emphasis>Explanation:</emphasis></para>
815 <para>Comments can help or they can clutter. They help when they
816 are differentiated from the code they describe. One line
817 comments do not offer effective separation between the comment
818 and the code. Block identifiers do, by surrounding the code
819 with a clear, definable pattern.</para>
821 <para><emphasis>Example:</emphasis></para>
823 /*********************************************************************
824 * This will stand out clearly in your code!
825 *********************************************************************/
826 if (this_variable == that_variable)
828 do_something_very_important();
832 /* unfortunately, this may not */
833 if (this_variable == that_variable)
835 do_something_very_important();
839 if (this_variable == that_variable) /* this may not either */
841 do_something_very_important();
844 <para><emphasis>Exception:</emphasis></para>
846 <para>If you are trying to add a small logic comment and do not
847 wish to "disrupt" the flow of the code, feel free to use a 1
848 line comment which is NOT on the same line as the code.</para>
854 <sect3 id="s5"><title>Keep Comments on their own line</title>
856 <para><emphasis>Explanation:</emphasis></para>
858 <para>It goes back to the question of readability. If the comment
859 is on the same line as the code it will be harder to read than
860 the comment that is on its own line.</para>
862 <para>There are three exceptions to this rule, which should be
863 violated freely and often: during the definition of variables,
864 at the end of closing braces, when used to comment
867 <para><emphasis>Example:</emphasis></para>
869 /*********************************************************************
870 * This will stand out clearly in your code,
871 * But the second example won't.
872 *********************************************************************/
873 if (this_variable == this_variable)
875 do_something_very_important();
878 if (this_variable == this_variable) /*can you see me?*/
880 do_something_very_important(); /*not easily*/
884 /*********************************************************************
885 * But, the encouraged exceptions:
886 *********************************************************************/
887 int urls_read = 0; /* # of urls read + rejected */
888 int urls_rejected = 0; /* # of urls rejected */
892 do_something_very_important();
896 short do_something_very_important(
897 short firstparam, /* represents something */
898 short nextparam /* represents something else */ )
902 } /* -END- do_something_very_important */
907 <sect3 id="s6"><title>Comment each logical step</title>
909 <para><emphasis>Explanation:</emphasis></para>
911 <para>Logical steps should be commented to help others follow the
912 intent of the written code and comments will make the code more
915 <para>If you have 25 lines of code without a comment, you should
916 probably go back into it to see where you forgot to put
919 <para>Most "for", "while", "do", etc... loops _probably_ need a
920 comment. After all, these are usually major logic
927 <sect3 id="s7"><title>Comment All Functions Thoroughly</title>
929 <para><emphasis>Explanation:</emphasis></para>
931 <para>A reader of the code should be able to look at the comments
932 just prior to the beginning of a function and discern the
933 reason for its existence and the consequences of using it. The
934 reader should not have to read through the code to determine if
935 a given function is safe for a desired use. The proper
936 information thoroughly presented at the introduction of a
937 function not only saves time for subsequent maintenance or
938 debugging, it more importantly aids in code reuse by allowing a
939 user to determine the safety and applicability of any function
940 for the problem at hand. As a result of such benefits, all
941 functions should contain the information presented in the
942 addendum section of this document.</para>
948 <sect3 id="s8"><title>Comment at the end of braces if the
949 content is more than one screen length</title>
951 <para><emphasis>Explanation:</emphasis></para>
953 <para>Each closing brace should be followed on the same line by a
954 comment that describes the origination of the brace if the
955 original brace is off of the screen, or otherwise far away from
956 the closing brace. This will simplify the debugging,
957 maintenance, and readability of the code.</para>
959 <para>As a suggestion , use the following flags to make the
960 comment and its brace more readable:</para>
962 <para>use following a closing brace: } /* -END- if() or while ()
965 <para><emphasis>Example:</emphasis></para>
969 do_something_very_important();
970 ...some long list of commands...
971 } /* -END- if x is 1 */
977 do_something_very_important();
978 ...some long list of commands...
979 } /* -END- if (1 == X) */
985 <sect2 id="s9"><title>Naming Conventions</title>
989 <sect3 id="s10"><title>Variable Names</title>
991 <para><emphasis>Explanation:</emphasis></para>
993 <para>Use all lowercase, and separate words via an underscore
994 ('_'). Do not start an identifier with an underscore. (ANSI C
995 reserves these for use by the compiler and system headers.) Do
996 not use identifiers which are reserved in ANSI C++. (E.g.
997 template, class, true, false, ...). This is in case we ever
998 decide to port Privoxy to C++.</para>
1000 <para><emphasis>Example:</emphasis></para>
1002 int ms_iis5_hack = 0;</programlisting>
1004 <para><emphasis>Instead of:</emphasis></para>
1007 int msiis5hack = 0; int msIis5Hack = 0;
1014 <sect3 id="s11"><title>Function Names</title>
1016 <para><emphasis>Explanation:</emphasis></para>
1018 <para>Use all lowercase, and separate words via an underscore
1019 ('_'). Do not start an identifier with an underscore. (ANSI C
1020 reserves these for use by the compiler and system headers.) Do
1021 not use identifiers which are reserved in ANSI C++. (E.g.
1022 template, class, true, false, ...). This is in case we ever
1023 decide to port Privoxy to C++.</para>
1025 <para><emphasis>Example:</emphasis></para>
1027 int load_some_file(struct client_state *csp)</programlisting>
1029 <para><emphasis>Instead of:</emphasis></para>
1032 int loadsomefile(struct client_state *csp)
1033 int loadSomeFile(struct client_state *csp)
1040 <sect3 id="s12"><title>Header file prototypes</title>
1042 <para><emphasis>Explanation:</emphasis></para>
1044 <para>Use a descriptive parameter name in the function prototype
1045 in header files. Use the same parameter name in the header file
1046 that you use in the c file.</para>
1048 <para><emphasis>Example:</emphasis></para>
1050 (.h) extern int load_aclfile(struct client_state *csp);
1051 (.c) int load_aclfile(struct client_state *csp)</programlisting>
1053 <para><emphasis>Instead of:</emphasis></para>
1055 (.h) extern int load_aclfile(struct client_state *); or
1056 (.h) extern int load_aclfile();
1057 (.c) int load_aclfile(struct client_state *csp)
1064 <sect3 id="s13"><title>Enumerations, and #defines</title>
1066 <para><emphasis>Explanation:</emphasis></para>
1068 <para>Use all capital letters, with underscores between words. Do
1069 not start an identifier with an underscore. (ANSI C reserves
1070 these for use by the compiler and system headers.)</para>
1072 <para><emphasis>Example:</emphasis></para>
1074 (enumeration) : enum Boolean {FALSE, TRUE};
1075 (#define) : #define DEFAULT_SIZE 100;</programlisting>
1077 <para><emphasis>Note:</emphasis> We have a standard naming scheme for #defines
1078 that toggle a feature in the preprocessor: FEATURE_>, where
1079 > is a short (preferably 1 or 2 word) description.</para>
1081 <para><emphasis>Example:</emphasis></para>
1083 #define FEATURE_FORCE 1
1085 #ifdef FEATURE_FORCE
1086 #define FORCE_PREFIX blah
1087 #endif /* def FEATURE_FORCE */
1092 <sect3 id="s14"><title>Constants</title>
1094 <para><emphasis>Explanation:</emphasis></para>
1096 <para>Spell common words out entirely (do not remove vowels).</para>
1098 <para>Use only widely-known domain acronyms and abbreviations.
1099 Capitalize all letters of an acronym.</para>
1101 <para>Use underscore (_) to separate adjacent acronyms and
1102 abbreviations. Never terminate a name with an underscore.</para>
1104 <para><emphasis>Example:</emphasis></para>
1106 #define USE_IMAGE_LIST 1</programlisting>
1108 <para><emphasis>Instead of:</emphasis></para>
1111 #define USE_IMG_LST 1 or
1112 #define _USE_IMAGE_LIST 1 or
1113 #define USE_IMAGE_LIST_ 1 or
1114 #define use_image_list 1 or
1115 #define UseImageList 1
1124 <sect2 id="s15"><title>Using Space</title>
1128 <sect3 id="s16"><title>Put braces on a line by themselves.</title>
1130 <para><emphasis>Explanation:</emphasis></para>
1132 <para>The brace needs to be on a line all by itself, not at the
1133 end of the statement. Curly braces should line up with the
1134 construct that they're associated with. This practice makes it
1135 easier to identify the opening and closing braces for a
1138 <para><emphasis>Example:</emphasis></para>
1145 <para><emphasis>Instead of:</emphasis></para>
1147 <para>if (this == that) { ... }</para>
1151 <para>if (this == that) { ... }</para>
1153 <para><emphasis>Note:</emphasis> In the special case that the if-statement is
1154 inside a loop, and it is trivial, i.e. it tests for a
1155 condition that is obvious from the purpose of the block,
1156 one-liners as above may optically preserve the loop structure
1157 and make it easier to read.</para>
1159 <para><emphasis>Status:</emphasis> developer-discretion.</para>
1161 <para><emphasis>Example exception:</emphasis></para>
1163 while (more lines are read)
1165 /* Please document what is/is not a comment line here */
1166 if (it's a comment) continue;
1174 <sect3 id="s17"><title>ALL control statements should have a
1177 <para><emphasis>Explanation:</emphasis></para>
1179 <para>Using braces to make a block will make your code more
1180 readable and less prone to error. All control statements should
1181 have a block defined.</para>
1183 <para><emphasis>Example:</emphasis></para>
1188 do_something_else();
1191 <para><emphasis>Instead of:</emphasis></para>
1193 <para>if (this == that) do_something(); do_something_else();</para>
1197 <para>if (this == that) do_something();</para>
1199 <para><emphasis>Note:</emphasis> The first example in "Instead of" will execute
1200 in a manner other than that which the developer desired (per
1201 indentation). Using code braces would have prevented this
1202 "feature". The "explanation" and "exception" from the point
1203 above also applies.</para>
1209 <sect3 id="s18"><title>Do not belabor/blow-up boolean
1212 <para><emphasis>Example:</emphasis></para>
1214 structure->flag = (condition);</programlisting>
1216 <para><emphasis>Instead of:</emphasis></para>
1218 <para>if (condition) { structure->flag = 1; } else {
1219 structure->flag = 0; }</para>
1221 <para><emphasis>Note:</emphasis> The former is readable and concise. The later
1222 is wordy and inefficient. Please assume that any developer new
1223 to the project has at least a "good" knowledge of C/C++. (Hope
1224 I do not offend by that last comment ... 8-)</para>
1230 <sect3 id="s19"><title>Use white space freely because it is
1233 <para><emphasis>Explanation:</emphasis></para>
1235 <para>Make it readable. The notable exception to using white space
1236 freely is listed in the next guideline.</para>
1238 <para><emphasis>Example:</emphasis></para>
1240 int first_value = 0;
1242 int another_value = 0;
1243 int this_variable = 0;
1248 <sect3 id="s20"><title>Don't use white space around structure
1251 <para><emphasis>Explanation:</emphasis></para>
1253 <para>- structure pointer operator ( "->" ) - member operator (
1254 "." ) - functions and parentheses</para>
1256 <para>It is a general coding practice to put pointers, references,
1257 and function parentheses next to names. With spaces, the
1258 connection between the object and variable/function name is not
1261 <para><emphasis>Example:</emphasis></para>
1265 function_name();</programlisting>
1267 <para><emphasis>Instead of:</emphasis> a_struct -> a_member; a_struct . a_member;
1268 function_name ();</para>
1274 <sect3 id="s21"><title>Make the last brace of a function stand
1277 <para><emphasis>Example:</emphasis></para>
1279 int function1( ... )
1284 } /* -END- function1 */
1287 int function2( ... )
1289 } /* -END- function2 */
1292 <para><emphasis>Instead of:</emphasis></para>
1294 <para>int function1( ... ) { ...code... return(ret_code); } int
1295 function2( ... ) { }</para>
1297 <para><emphasis>Note:</emphasis> Use 1 blank line before the closing brace and 2
1298 lines afterward. This makes the end of function standout to
1299 the most casual viewer. Although function comments help
1300 separate functions, this is still a good coding practice. In
1301 fact, I follow these rules when using blocks in "for", "while",
1302 "do" loops, and long if {} statements too. After all whitespace
1305 <para><emphasis>Status:</emphasis> developer-discretion on the number of blank
1306 lines. Enforced is the end of function comments.</para>
1312 <sect3 id="s22"><title>Use 3 character indentions</title>
1314 <para><emphasis>Explanation:</emphasis></para>
1316 <para>If some use 8 character TABs and some use 3 character TABs,
1317 the code can look *very* ragged. So use 3 character indentions
1318 only. If you like to use TABs, pass your code through a filter
1319 such as "expand -t3" before checking in your code.</para>
1321 <para><emphasis>Example:</emphasis></para>
1323 static const char * const url_code_map[256] =
1329 int function1( ... )
1337 return HOW_DID_YOU_GET_HERE;
1340 return NEVER_GETS_HERE;
1349 <sect2 id="s23"><title>Initializing</title>
1353 <sect3 id="s24"><title>Initialize all variables</title>
1355 <para><emphasis>Explanation:</emphasis></para>
1357 <para>Do not assume that the variables declared will not be used
1358 until after they have been assigned a value somewhere else in
1359 the code. Remove the chance of accidentally using an unassigned
1362 <para><emphasis>Example:</emphasis></para>
1366 struct *ptr = NULL;</programlisting>
1368 <para><emphasis>Note:</emphasis> It is much easier to debug a SIGSEGV if the
1369 message says you are trying to access memory address 00000000
1370 and not 129FA012; or array_ptr[20] causes a SIGSEV vs.
1371 array_ptr[0].</para>
1373 <para><emphasis>Status:</emphasis> developer-discretion if and only if the
1374 variable is assigned a value "shortly after" declaration.</para>
1380 <sect2 id="s25"><title>Functions</title>
1384 <sect3 id="s26"><title>Name functions that return a boolean as a
1387 <para><emphasis>Explanation:</emphasis></para>
1389 <para>Value should be phrased as a question that would logically
1390 be answered as a true or false statement</para>
1392 <para><emphasis>Example:</emphasis></para>
1394 should_we_block_this();
1395 contains_an_image();
1396 is_web_page_blank();
1401 <sect3 id="s27"><title>Always specify a return type for a
1404 <para><emphasis>Explanation:</emphasis></para>
1406 <para>The default return for a function is an int. To avoid
1407 ambiguity, create a return for a function when the return has a
1408 purpose, and create a void return type if the function does not
1409 need to return anything.</para>
1415 <sect3 id="s28"><title>Minimize function calls when iterating by
1416 using variables</title>
1418 <para><emphasis>Explanation:</emphasis></para>
1420 <para>It is easy to write the following code, and a clear argument
1421 can be made that the code is easy to understand:</para>
1423 <para><emphasis>Example:</emphasis></para>
1425 for (size_t cnt = 0; cnt < block_list_length(); cnt++)
1430 <para><emphasis>Note:</emphasis> Unfortunately, this makes a function call for
1431 each and every iteration. This increases the overhead in the
1432 program, because the compiler has to look up the function each
1433 time, call it, and return a value. Depending on what occurs in
1434 the block_list_length() call, it might even be creating and
1435 destroying structures with each iteration, even though in each
1436 case it is comparing "cnt" to the same value, over and over.
1437 Remember too - even a call to block_list_length() is a function
1438 call, with the same overhead.</para>
1440 <para>Instead of using a function call during the iterations,
1441 assign the value to a variable, and evaluate using the
1444 <para><emphasis>Example:</emphasis></para>
1446 size_t len = block_list_length();
1448 for (size_t cnt = 0; cnt < len; cnt++)
1453 <para><emphasis>Exceptions:</emphasis> if the value of block_list_length()
1454 *may* change or could *potentially* change, then you must code the
1455 function call in the for/while loop.</para>
1461 <sect3 id="s29"><title>Pass and Return by Const Reference</title>
1463 <para><emphasis>Explanation:</emphasis></para>
1465 <para>This allows a developer to define a const pointer and call
1466 your function. If your function does not have the const
1467 keyword, we may not be able to use your function. Consider
1468 strcmp, if it were defined as: extern int strcmp(char *s1,
1471 <para>I could then not use it to compare argv's in main: int
1472 main(int argc, const char *argv[]) { strcmp(argv[0], "privoxy");
1475 <para>Both these pointers are *const*! If the c runtime library
1476 maintainers do it, we should too.</para>
1482 <sect3 id="s30"><title>Pass and Return by Value</title>
1484 <para><emphasis>Explanation:</emphasis></para>
1486 <para>Most structures cannot fit onto a normal stack entry (i.e.
1487 they are not 4 bytes or less). Aka, a function declaration
1488 like: int load_aclfile(struct client_state csp)</para>
1490 <para>would not work. So, to be consistent, we should declare all
1491 prototypes with "pass by value": int load_aclfile(struct
1492 client_state *csp)</para>
1498 <sect3 id="s31"><title>Names of include files</title>
1500 <para><emphasis>Explanation:</emphasis></para>
1502 <para>Your include statements should contain the file name without
1503 a path. The path should be listed in the Makefile, using -I as
1504 processor directive to search the indicated paths. An exception
1505 to this would be for some proprietary software that utilizes a
1506 partial path to distinguish their header files from system or
1507 other header files.</para>
1509 <para><emphasis>Example:</emphasis></para>
1511 #include <iostream.h> /* This is not a local include */
1512 #include "config.h" /* This IS a local include */
1515 <para><emphasis>Exception:</emphasis></para>
1518 /* This is not a local include, but requires a path element. */
1519 #include <sys/fileName.h>
1522 <para><emphasis>Note:</emphasis> Please! do not add "-I." to the Makefile
1523 without a _very_ good reason. This duplicates the #include
1524 "file.h" behavior.</para>
1530 <sect3 id="s32"><title>Provide multiple inclusion
1533 <para><emphasis>Explanation:</emphasis></para>
1535 <para>Prevents compiler and linker errors resulting from
1536 redefinition of items.</para>
1538 <para>Wrap each header file with the following syntax to prevent
1539 multiple inclusions of the file. Of course, replace PROJECT_H
1540 with your file name, with "." Changed to "_", and make it
1543 <para><emphasis>Example:</emphasis></para>
1545 #ifndef PROJECT_H_INCLUDED
1546 #define PROJECT_H_INCLUDED
1548 #endif /* ndef PROJECT_H_INCLUDED */
1553 <sect3 id="s33"><title>Use `extern "C"` when appropriate</title>
1555 <para><emphasis>Explanation:</emphasis></para>
1557 <para>If our headers are included from C++, they must declare our
1558 functions as `extern "C"`. This has no cost in C, but increases
1559 the potential re-usability of our code.</para>
1561 <para><emphasis>Example:</emphasis></para>
1566 #endif /* def __cplusplus */
1568 ... function definitions here ...
1572 #endif /* def __cplusplus */
1577 <sect3 id="s34"><title>Where Possible, Use Forward Struct
1578 Declaration Instead of Includes</title>
1580 <para><emphasis>Explanation:</emphasis></para>
1582 <para>Useful in headers that include pointers to other struct's.
1583 Modifications to excess header files may cause needless
1586 <para><emphasis>Example:</emphasis></para>
1588 /*********************************************************************
1589 * We're avoiding an include statement here!
1590 *********************************************************************/
1592 extern file_list *xyz;</programlisting>
1594 <para><emphasis>Note:</emphasis> If you declare "file_list xyz;" (without the
1595 pointer), then including the proper header file is necessary.
1596 If you only want to prototype a pointer, however, the header
1597 file is unnecessary.</para>
1599 <para><emphasis>Status:</emphasis> Use with discretion.</para>
1605 <sect2 id="s35"><title>General Coding Practices</title>
1609 <sect3 id="s36"><title>Turn on warnings</title>
1611 <para><emphasis>Explanation</emphasis></para>
1613 <para>Compiler warnings are meant to help you find bugs. You
1614 should turn on as many as possible. With GCC, the switch is
1615 "-Wall". Try and fix as many warnings as possible.</para>
1621 <sect3 id="s37"><title>Provide a default case for all switch
1624 <para><emphasis>Explanation:</emphasis></para>
1626 <para>What you think is guaranteed is never really guaranteed. The
1627 value that you don't think you need to check is the one that
1628 someday will be passed. So, to protect yourself from the
1629 unknown, always have a default step in a switch statement.</para>
1631 <para><emphasis>Example:</emphasis></para>
1633 switch (hash_string(cmd))
1635 case hash_actions_file:
1645 ... anomaly code goes here ...
1646 continue; / break; / exit( 1 ); / etc ...
1648 } /* end switch (hash_string(cmd)) */</programlisting>
1650 <para><emphasis>Note:</emphasis> If you already have a default condition, you
1651 are obviously exempt from this point. Of note, most of the
1652 WIN32 code calls `DefWindowProc' after the switch statement.
1653 This API call *should* be included in a default statement.</para>
1655 <para><emphasis>Another Note:</emphasis> This is not so much a readability issue
1656 as a robust programming issue. The "anomaly code goes here" may
1657 be no more than a print to the STDERR stream (as in
1658 load_config). Or it may really be an abort condition.</para>
1660 <para><emphasis>Status:</emphasis> Programmer discretion is advised.</para>
1666 <sect3 id="s38"><title>Try to avoid falling through cases in a
1667 switch statement.</title>
1669 <para><emphasis>Explanation:</emphasis></para>
1671 <para>In general, you will want to have a 'break' statement within
1672 each 'case' of a switch statement. This allows for the code to
1673 be more readable and understandable, and furthermore can
1674 prevent unwanted surprises if someone else later gets creative
1675 and moves the code around.</para>
1677 <para>The language allows you to plan the fall through from one
1678 case statement to another simply by omitting the break
1679 statement within the case statement. This feature does have
1680 benefits, but should only be used in rare cases. In general,
1681 use a break statement for each case statement.</para>
1683 <para>If you choose to allow fall through, you should comment both
1684 the fact of the fall through and reason why you felt it was
1691 <sect3 id="s40"><title>Don't mix size_t and other types</title>
1693 <para><emphasis>Explanation:</emphasis></para>
1695 <para>The type of size_t varies across platforms. Do not make
1696 assumptions about whether it is signed or unsigned, or about
1697 how long it is. Do not compare a size_t against another
1698 variable of a different type (or even against a constant)
1699 without casting one of the values.</para>
1705 <sect3 id="s41"><title>Declare each variable and struct on its
1708 <para><emphasis>Explanation:</emphasis></para>
1710 <para>It can be tempting to declare a series of variables all on
1711 one line. Don't.</para>
1713 <para><emphasis>Example:</emphasis></para>
1717 long c = 0;</programlisting>
1719 <para><emphasis>Instead of:</emphasis></para>
1721 <para>long a, b, c;</para>
1723 <para><emphasis>Explanation:</emphasis> - there is more room for comments on the
1724 individual variables - easier to add new variables without
1725 messing up the original ones - when searching on a variable to
1726 find its type, there is less clutter to "visually"
1729 <para><emphasis>Exceptions:</emphasis> when you want to declare a bunch of loop
1730 variables or other trivial variables; feel free to declare them
1731 on one line. You should, although, provide a good comment on
1732 their functions.</para>
1734 <para><emphasis>Status:</emphasis> developer-discretion.</para>
1740 <sect3 id="s42"><title>Use malloc/zalloc sparingly</title>
1742 <para><emphasis>Explanation:</emphasis></para>
1744 <para>Create a local struct (on the stack) if the variable will
1745 live and die within the context of one function call.</para>
1747 <para>Only "malloc" a struct (on the heap) if the variable's life
1748 will extend beyond the context of one function call.</para>
1750 <para><emphasis>Example:</emphasis></para>
1752 If a function creates a struct and stores a pointer to it in a
1753 list, then it should definitely be allocated via `malloc'.
1758 <sect3 id="s43"><title>The Programmer Who Uses 'malloc' is
1759 Responsible for Ensuring 'free'</title>
1761 <para><emphasis>Explanation:</emphasis></para>
1763 <para>If you have to "malloc" an instance, you are responsible for
1764 insuring that the instance is `free'd, even if the deallocation
1765 event falls within some other programmer's code. You are also
1766 responsible for ensuring that deletion is timely (i.e. not too
1767 soon, not too late). This is known as "low-coupling" and is a
1768 "good thing (tm)". You may need to offer a
1769 free/unload/destructor type function to accommodate this.</para>
1771 <para><emphasis>Example:</emphasis></para>
1773 int load_re_filterfile(struct client_state *csp) { ... }
1774 static void unload_re_filterfile(void *f) { ... }</programlisting>
1776 <para><emphasis>Exceptions:</emphasis></para>
1778 <para>The developer cannot be expected to provide `free'ing
1779 functions for C run-time library functions ... such as
1782 <para><emphasis>Status:</emphasis> developer-discretion. The "main" use of this
1783 standard is for allocating and freeing data structures (complex
1790 <sect3 id="s44"><title>Add loaders to the `file_list' structure
1791 and in order</title>
1793 <para><emphasis>Explanation:</emphasis></para>
1795 <para>I have ordered all of the "blocker" file code to be in alpha
1796 order. It is easier to add/read new blockers when you expect a
1797 certain order.</para>
1799 <para><emphasis>Note:</emphasis> It may appear that the alpha order is broken in
1800 places by POPUP tests coming before PCRS tests. But since
1801 POPUPs can also be referred to as KILLPOPUPs, it is clear that
1802 it should come first.</para>
1808 <sect3 id="s45"><title>"Uncertain" new code and/or changes to
1809 existing code, use XXX</title>
1811 <para><emphasis>Explanation:</emphasis></para>
1813 <para>If you have enough confidence in new code or confidence in
1814 your changes, but are not *quite* sure of the repercussions,
1817 <para>/* XXX: this code has a logic error on platform XYZ, *
1818 attempting to fix */ #ifdef PLATFORM ...changed code here...
1823 <para>/* XXX: I think the original author really meant this...
1824 */ ...changed code here...</para>
1828 <para>/* XXX: new code that *may* break something else... */
1829 ...new code here...</para>
1831 <para><emphasis>Note:</emphasis> If you make it clear that this may or may not
1832 be a "good thing (tm)", it will be easier to identify and
1833 include in the project (or conversely exclude from the
1841 <sect2 id="s46"><title>Addendum: Template for files and function
1842 comment blocks:</title>
1844 <para><emphasis>Example for file comments:</emphasis></para>
1846 /*********************************************************************
1850 * Purpose : (Fill me in with a good description!)
1852 * Copyright : Written by and Copyright (C) 2001-2009
1853 * the Privoxy team. https://www.privoxy.org/
1855 * This program is free software; you can redistribute it
1856 * and/or modify it under the terms of the GNU General
1857 * Public License as published by the Free Software
1858 * Foundation; either version 2 of the License, or (at
1859 * your option) any later version.
1861 * This program is distributed in the hope that it will
1862 * be useful, but WITHOUT ANY WARRANTY; without even the
1863 * implied warranty of MERCHANTABILITY or FITNESS FOR A
1864 * PARTICULAR PURPOSE. See the GNU General Public
1865 * License for more details.
1867 * The GNU General Public License should be included with
1868 * this file. If not, you can view it at
1869 * http://www.gnu.org/licenses/old-licenses/gpl-2.0.html
1870 * or write to the Free Software Foundation, Inc.,
1871 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 ,
1874 *********************************************************************/
1879 ...necessary include files for us to do our work...
1881 const char FILENAME_h_rcs[] = FILENAME_H_VERSION;
1884 <para><emphasis>Note:</emphasis> This declares the rcs variables that should be
1885 added to the "show-version" page. If this is a brand new
1886 creation by you, you are free to change the "Copyright" section
1887 to represent the rights you wish to maintain.</para>
1889 <para><emphasis>Note:</emphasis> The formfeed character that is present right
1890 after the comment flower box is handy for (X|GNU)Emacs users to
1891 skip the verbiage and get to the heart of the code (via
1892 `forward-page' and `backward-page'). Please include it if you
1895 <para><emphasis>Example for file header comments:</emphasis></para>
1899 /*********************************************************************
1903 * Purpose : (Fill me in with a good description!)
1905 * Copyright : Written by and Copyright (C) 2001-2009
1906 * the Privoxy team. https://www.privoxy.org/
1908 * This program is free software; you can redistribute it
1909 * and/or modify it under the terms of the GNU General
1910 * Public License as published by the Free Software
1911 * Foundation; either version 2 of the License, or (at
1912 * your option) any later version.
1914 * This program is distributed in the hope that it will
1915 * be useful, but WITHOUT ANY WARRANTY; without even the
1916 * implied warranty of MERCHANTABILITY or FITNESS FOR A
1917 * PARTICULAR PURPOSE. See the GNU General Public
1918 * License for more details.
1920 * The GNU General Public License should be included with
1921 * this file. If not, you can view it at
1922 * http://www.gnu.org/licenses/old-licenses/gpl-2.0.html
1923 * or write to the Free Software Foundation, Inc.,
1924 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 ,
1927 *********************************************************************/
1930 #include "project.h"
1936 ... function headers here ...
1939 /* Revision control strings from this header and associated .c file */
1940 extern const char FILENAME_rcs[];
1941 extern const char FILENAME_h_rcs[];
1948 #endif /* ndef _FILENAME_H */
1957 <para><emphasis>Example for function comments:</emphasis></para>
1959 /*********************************************************************
1961 * Function : FUNCTION_NAME
1963 * Description : (Fill me in with a good description!)
1966 * 1 : param1 = pointer to an important thing
1967 * 2 : x = pointer to something else
1969 * Returns : 0 => Ok, everything else is an error.
1971 *********************************************************************/
1972 int FUNCTION_NAME(void *param1, const char *x)
1980 <para><emphasis>Note:</emphasis> If we all follow this practice, we should be
1981 able to parse our code to create a "self-documenting" web
1988 <!-- ~~~~~ New section ~~~~~ -->
1989 <sect1 id="testing"><title>Testing Guidelines</title>
1993 <!-- ~~~~~ New section ~~~~~ -->
1994 <sect2 id="testing-plan"><title>Testplan for releases</title>
1996 Explain release numbers. major, minor. developer releases. etc.
1998 <orderedlist numeration="arabic">
2000 Remove any existing rpm with rpm -e
2003 Remove any file that was left over. This includes (but is not limited to)
2006 <listitem><para>/var/log/privoxy</para></listitem>
2007 <listitem><para>/etc/privoxy</para></listitem>
2008 <listitem><para>/usr/sbin/privoxy</para></listitem>
2009 <listitem><para>/etc/init.d/privoxy</para></listitem>
2010 <listitem><para>/usr/doc/privoxy*</para></listitem>
2014 Install the rpm. Any error messages?
2016 <listitem><para>start,stop,status <application>Privoxy</application> with the specific script
2017 (e.g. /etc/rc.d/init/privoxy stop). Reboot your machine. Does
2018 autostart work?</para></listitem>
2019 <listitem><para>Start browsing. Does <application>Privoxy</application> work? Logfile written?</para></listitem>
2020 <listitem><para>Remove the rpm. Any error messages? All files removed?</para></listitem>
2023 <!-- XXX: Document how to write test reports and where to send them -->
2025 <!-- ~~~~~ New section ~~~~~ -->
2026 <sect2 id="privoxy-regression-test"><title>Testing with <application>Privoxy-Regression-Test</application></title>
2028 If you compiled, packaged or merely installed Privoxy, it is recommended to run
2029 <application>Privoxy-Regression-Test</application> to verify that at least
2030 the tested parts of <application>Privoxy</application> are working as expected.
2033 This is actually pretty easy. For details, please see
2034 <command>perldoc privoxy-regression-test.pl</command>.
2037 Here is an example of what <application>Privoxy-Regression-Test</application> can do for you:
2041 fk@t520 ~ $privoxy-regression-test.pl
2042 2020-12-14 12:16:32: Asking Privoxy for the number of action files available ...
2043 2020-12-14 12:16:32: Gathering regression tests from 9 action file(s) delivered by Privoxy 3.0.30.
2044 2020-12-14 12:16:32: Executing regression tests ...
2045 2020-12-14 12:16:41: Ooops. Expected removal but: 'Referer: https://p.p/' is still there.
2046 2020-12-14 12:16:41: Failure for test 785. Header 'Referer: https://p.p/' and tag 'hide-referrer{conditional-block}'
2047 2020-12-14 12:16:41: Ooops. Got: 'Referer: https://p.p/' while expecting: 'Referer: http://p.p/'
2048 2020-12-14 12:16:41: Failure for test 791. Header 'Referer: https://p.p/' and tag 'hide-referrer{conditional-forge}'
2049 2020-12-14 12:16:44: Executed 1087 regression tests. Skipped 115. 1085 successes, 2 failures.
2050 # Repeat one of the failing tests and get a curl command to quickly reproduce the problem
2051 # without causing too much log noise.
2052 fk@t520 ~ $privoxy-regression-test.pl --test-number 785 --verbose --debug 4
2053 2020-12-14 12:17:55: Asking Privoxy for the number of action files available ...
2055 2020-12-14 12:17:56: Executing regression tests ...
2056 2020-12-14 12:17:56: Executing: curl --include -H 'Proxy-Connection:' -H 'Connection: close' -s -S --user-agent 'Privoxy-Regression-Test 0.7.2' --max-time '5' --globoff -H 'X-Privoxy-Control: hide-referrer{conditional-block}' -H 'Referer: https://p.p/' http://p.p/show-request 2>&1
2057 2020-12-14 12:17:56: Ooops. Expected removal but: 'Referer: https://p.p/' is still there.
2058 2020-12-14 12:17:56: Failure for test 785 (0/13/5). Header 'Referer: https://p.p/' and tag 'hide-referrer{conditional-block}'
2059 2020-12-14 12:17:56: Executed 1 regression tests. Skipped 1201. 0 successes, 1 failures.
2062 Use the if the <command>--privoxy-address</command> option if the
2063 http_proxy environment variable isn't configured.
2067 <!-- ~~~~~ New section ~~~~~ -->
2068 <sect2 id="fuzzing"><title>Fuzzing Privoxy</title>
2070 To make fuzzing more convenient, Privoxy can be configured
2071 with --enable-fuzz which will result in the --fuzz option
2075 Example (tested on ElectroBSD):
2078 # Compile Privoxy with instrumentation for afl
2079 $ export CC=afl-clang
2080 $ export CFLAGS="-fsanitize=address -ggdb"
2081 $ export CPPFLAGS=-I/usr/local/include/
2082 $ export LDFLAGS="-fsanitize=address -L/usr/local/lib"
2083 $ export AFL_USE_ASAN=1
2084 $ export AFL_HARDEN=1
2085 $ ./configure --with-debug --enable-extended-host-patterns --enable-accept-filter --enable-no-gifs --enable-compression --enable-strptime-sanity-checks --enable-external-filters --enable-fuzz
2088 Privoxy version 3.0.24 (http://www.privoxy.org/)
2089 Usage: ./privoxy [--config-test] [--chroot] [--help] [--no-daemon] [--pidfile pidfile] [--pre-chroot-nslookup hostname] [--user user[.group]] [--version] [configfile]
2090 ./privoxy --fuzz fuzz-mode ./path/to/fuzzed/input [--stfu]
2092 Supported fuzz modes and the expected input:
2093 action: Text to parse as action file.
2094 client-request: Client request to parse. Currently incomplete
2095 client-header: Client header to parse.
2096 chunked-transfer-encoding: Chunk-encoded data to dechunk.
2097 deflate: deflate-compressed data to decompress.
2098 filter: Text to parse as filter file.
2099 gif: gif to deanimate.
2100 gzip: gzip-compressed data to decompress.
2101 pcrs-substitute: A pcrs-substitute to compile. Not a whole pcrs job! Example: Bla $1 bla C $3 blah.
2102 server-header: Server header to parse.
2103 server-response: Server response to parse.
2105 The following fuzz modes read data from stdin if the 'file' is '-'
2108 chunked-transfer-encoding
2118 $ export ASAN_OPTIONS='abort_on_error=1'
2119 $ mkdir input output
2120 $ echo '$1 bla fasel $2' > input/pcrs
2121 $ afl-fuzz -i input -o output -m none ~/git/privoxy/privoxy --fuzz pcrs-substitute - --stfu
2123 $ cat >input/pcrs.txt
2125 s@(.{1})[432](\d+)@$1$2$hostname@UgisT
2127 $ afl-fuzz -i input/ -o output/ -f bla.filter -m none privoxy --fuzz filter bla.filter --stfu
2132 <!-- ~~~~~ New section ~~~~~ -->
2133 <sect1 id="newrelease"><title>Releasing a New Version</title>
2135 When we release versions of <application>Privoxy</application>,
2136 our work leaves our cozy secret lab and has to work in the cold
2137 RealWorld[tm]. Once it is released, there is no way to call it
2138 back, so it is very important that great care is taken to ensure
2139 that everything runs fine, and not to introduce problems in the
2143 So when releasing a new version, please adhere exactly to the
2144 procedure outlined in this chapter.
2148 The following programs are required to follow this process:
2149 <filename>ssh</filename>,
2150 <filename>gmake</filename> (GNU's version of make), autoconf, git,
2154 <sect2 id="versionnumbers">
2155 <title>Version numbers</title>
2158 First you need to determine which version number the release will have.
2159 <application>Privoxy</application> version numbers consist of three numbers,
2160 separated by dots, like in X.Y.Z (e.g. 3.0.0), where:
2165 X, the version major, is rarely ever changed. It is increased by one if
2166 turning a development branch into stable substantially changes the functionality,
2167 user interface or configuration syntax. Majors 1 and 2 were
2168 <application>Junkbuster</application>, and 3 is the first stable
2169 <application>Privoxy</application> release.
2174 <!-- FIXME this is not the way it works anymore -->
2175 Y, the version minor, represents the branch within the major version.
2176 At any point in time, there are two branches being maintained:
2177 The stable branch, with an even minor, say, 2N, in which no functionality is
2178 being added and only bug-fixes are made, and 2N+1, the development branch, in
2179 which the further development of <application>Privoxy</application> takes
2181 This enables us to turn the code upside down and inside out, while at the same time
2182 providing and maintaining a stable version.
2183 The minor is reset to zero (and one) when the major is incremented. When a development
2184 branch has matured to the point where it can be turned into stable, the old stable branch
2185 2N is given up (i.e. no longer maintained), the former development branch 2N+1 becomes the
2186 new stable branch 2N+2, and a new development branch 2N+3 is opened.
2191 Z, the point or sub version, represents a release of the software within a branch.
2192 It is therefore incremented immediately after each software release.
2193 <!-- FIXME this is not the way it works any more
2194 In development branches, only the even point versions correspond to actual releases,
2195 while the odd ones denote the evolving state of the sources on Git in between.
2196 It follows that Z is odd on Git in development branches most of the time. There, it gets
2197 increased to an even number immediately before a code freeze, and is increased to an odd
2198 number again immediately thereafter.
2199 This ensures that builds from Git snapshots are easily distinguished from released versions.
2200 didn't Fabian get rid of the even=stable, odd=dev convention for release numbering? -->
2201 The point version is reset to zero when the minor changes.
2204 Stable branches work a little differently, since there should be
2205 little to no development happening in such branches. Remember,
2206 only bugfixes, which presumably should have had some testing
2207 before being committed. Stable branches will then have their
2208 version reported as <literal>0.0.0</literal>, during that period
2209 between releases when changes are being added. This is to denote
2210 that this code is <emphasis>not for release</emphasis>. Then
2211 as the release nears, the version is bumped according: e.g.
2212 <literal>3.0.1 -> 0.0.0 -> 3.0.2</literal>.
2217 In summary, the main Git trunk is the development branch where new
2218 features are being worked on for the next stable series. This should
2219 almost always be where the most activity takes place. There is always at
2220 least one stable branch from the trunk, e.g now it is
2221 <literal>3.0</literal>, which is only used to release stable versions.
2222 Once the initial *.0 release of the stable branch has been done, then as a
2223 rule, only bugfixes that have had prior testing should be committed to
2224 the stable branch. Once there are enough bugfixes to justify a new
2225 release, the version of this branch is again incremented Example: 3.0.0
2226 -> 3.0.1 -> 3.0.2, etc are all stable releases from within the stable
2227 branch. 3.1.x is currently the main trunk, and where work on 3.2.x is
2228 taking place. If any questions, please post to the devel list
2229 <emphasis>before</emphasis> committing to a stable branch!
2232 Developers should remember too that if they commit a bugfix to the stable
2233 branch, this will more than likely require a separate submission to the
2234 main trunk, since these are separate development trees within Git. If you
2235 are working on both, then this would require at least two separate check
2236 outs (i.e main trunk, <emphasis>and</emphasis> the stable release branch,
2237 which is <literal>v_3_0_branch</literal> at the moment).
2242 <sect2 id="beforerelease">
2243 <title>Before the Release</title>
2245 The following <emphasis>must be done by one of the
2246 developers</emphasis> prior to each new release.
2251 Make sure that everybody who has worked on the code in the last
2252 couple of days has had a chance to yell <quote>no!</quote> in case
2253 they have pending changes/fixes in their pipelines. Announce the
2254 freeze so that nobody will interfere with last minute changes.
2259 Update the code status (CODE_STATUS="xxx") in configure.in to one of
2260 "alpha", "beta" or "stable".
2265 Rebuild configure and GNUMakefile to make sure the updated values are being used.
2269 $ autoheader && autoconf # rebuild configure
2270 $ ./configure # rebuild GNUmakefile
2275 <command>make dok-release</command> to update the sgml documentation source files.
2280 If action file processing has changed and is not backward-compatible,
2281 make sure the "for-privoxy-version=x.y.z" minimum version number in
2282 default.action.master has been updated:
2286 #############################################################################
2287 #MASTER# COMMENT: The minimum Privoxy version:
2288 for-privoxy-version=3.0.11
2293 Create the change log:
2298 $ git log [last release tag]..master > /tmp/log
2299 # get the commit log since the last release
2300 $ utils/makeChangeLog /tmp/log > /tmp/change.log
2301 # reformat the commit log
2304 Edit <filename>/tmp/change.log</filename> to remove trivial
2305 changes and group the changes under general headings like:
2309 - Action file improvements:
2310 - Filter file improvements:
2311 - General improvements:
2312 - Documentation improvements:
2313 - Build system improvements:
2315 - Privoxy-Log-Parser:
2316 - Privoxy-Regression-Test:
2319 Add the contents of <filename>/tmp/change.log</filename> to the
2320 start of <filename>ChangeLog</filename> and re-create
2321 <filename>doc/source/changelog.sgml</filename>:
2324 $ utils/changelog2doc.pl /tmp/change.log >| doc/source/changelog.sgml
2329 All developers should look at the <filename>ChangeLog</filename> and
2330 make sure noteworthy changes are referenced.
2335 All documentation should be rebuilt:
2343 Finished docs should be then be committed to Git (for those
2344 without the ability to build these). Some docs may require
2345 rather obscure processing tools. <filename>config</filename>,
2346 the man page (and the html version of the man page)
2347 fall in this category. README, the man page, AUTHORS, and config
2348 should all also be committed to Git for other packagers. The
2349 formal docs should be uploaded to the webserver. See the section
2350 <ulink url="webserver-update.html">"Updating the webserver"</ulink>
2351 in this manual for details.
2356 <emphasis>Commit all files that were changed in the above steps!</emphasis>
2361 The <citetitle>User Manual</citetitle> is also used for context
2362 sensitive help for the CGI editor. This is version sensitive, so that
2363 the user will get appropriate help for his/her release. So with
2364 each release a fresh version should be uploaded to the webserver
2365 (this is in addition to the main <citetitle>User Manual</citetitle>
2366 link from the main page since we need to keep manuals for various
2367 versions available). The CGI pages will link to something like
2368 <literal>https://www.privoxy.org/$(VERSION)/user-manual/</literal>. This
2369 needs to be updated for each new release and is done with the
2370 <quote>webserver</quote> target.
2375 Tag all files in Git with the version number with
2376 <quote><command>git tag -s v_X_Y_Z</command></quote>.
2377 Don't use vX_Y_Z, ver_X_Y_Z, v_X.Y.Z (won't work) etc.
2382 On the webserver, copy the user manual to a new top-level directory
2383 called <filename>X.Y.Z</filename>. This ensures that help links from the CGI
2384 pages, which have the version as a prefix, will go into the right version of the manual.
2385 If this is a development branch release, also symlink <filename>X.Y.(Z-1)</filename>
2386 to <filename>X.Y.Z</filename> and <filename>X.Y.(Z+1)</filename> to
2387 <filename>.</filename> (i.e. dot).
2393 <sect2 id="therelease">
2394 <title>Building and Releasing the Packages</title>
2396 Now the individual packages can be built and released. Note that for
2397 GPL reasons the first package to be released is always the source tarball.
2401 For <emphasis>all</emphasis> types of packages, including the source tarball,
2402 <emphasis>you must make sure that you build from clean sources by exporting
2403 the right version from Git into an empty directory</emphasis> (just press return when
2404 asked for a password):
2408 mkdir dist # delete or choose different name if it already exists
2410 git clone https://www.privoxy.org/git/privoxy.git
2412 git checkout v_X_Y_Z
2416 <emphasis>Do NOT change</emphasis> a single bit, including, but not limited to
2417 version information after export from Git. This is to make sure that
2418 all release packages, and with them, all future bug reports, are based
2419 on exactly the same code.
2424 Every significant release of Privoxy has included at least one
2425 package that either had incorrect versions of files, missing files,
2426 or incidental leftovers from a previous build process that gave
2427 unknown numbers of users headaches to try to figure out what was
2428 wrong. PLEASE, make sure you are using pristene sources, and are
2429 following the prescribed process!
2434 Please find additional instructions for the source tarball and the
2435 individual platform dependent binary packages below. And details
2436 on the Sourceforge release process below that.
2439 <sect3 id="pack-guidelines">
2440 <title>Note on Privoxy Packaging</title>
2442 Please keep these general guidelines in mind when putting together
2443 your package. These apply to <emphasis>all</emphasis> platforms!
2448 <application>Privoxy</application> <emphasis>requires</emphasis>
2449 write access to: all <filename>*.action</filename> files, all
2450 logfiles, and the <filename>trust</filename> file. You will
2451 need to determine the best way to do this for your platform.
2456 Please include up to date documentation. At a bare minimum:
2460 <filename>LICENSE</filename> (top-level directory)
2465 <filename>README</filename> (top-level directory)
2470 <filename>AUTHORS</filename> (top-level directory)
2475 <filename>man page</filename> (top-level directory, Unix-like
2481 <filename>The User Manual</filename> (doc/webserver/user-manual/)
2486 <filename>FAQ</filename> (doc/webserver/faq/)
2490 Also suggested: <filename>Developer Manual</filename>
2491 (doc/webserver/developer-manual) and <filename>ChangeLog</filename>
2492 (top-level directory). <filename>FAQ</filename> and the manuals are
2496 The documentation has been designed such that the manuals are linked
2497 to each other from parallel directories, and should be packaged
2498 that way. <filename>privoxy-index.html</filename> can also be
2499 included and can serve as a focal point for docs and other links of
2500 interest (and possibly renamed to <filename>index.html</filename>).
2501 This should be one level up from the manuals. There is a link also
2502 on this page to an HTMLized version of the man page. To avoid 404 for
2503 this, it is in Git as
2504 <filename>doc/webserver/man-page/privoxy-man-page.html</filename>,
2505 and should be included along with the manuals. There is also a
2506 css stylesheets that can be included for better presentation:
2507 <filename>p_doc.css</filename>. This should be in the same directory
2508 with <filename>privoxy-index.html</filename>, (i.e. one level up from
2509 the manual directories).
2514 <filename>user.action</filename> and <filename>user.filter</filename>
2515 are designed for local preferences. Make sure these do not get overwritten!
2516 <filename>config</filename> should not be overwritten either. This
2517 has especially important configuration data in it.
2518 <filename>trust</filename> should be left in tact as well.
2523 Other configuration files (<filename>default.action</filename>,
2524 <filename>regression-tests.action</filename> and
2525 <filename>default.filter</filename>) should be installed as the new
2526 defaults, but all previously installed configuration files should be
2527 preserved as backups. This is just good manners :-) These files are
2528 likely to change between releases and contain important new features
2534 Please check platform specific notes in this doc, if you haven't
2535 done <quote>Privoxy</quote> packaging before for other platform
2536 specific issues. Conversely, please add any notes that you know
2537 are important for your platform (or contact one of the doc
2538 maintainers to do this if you can't).
2543 Packagers should do a <quote>clean</quote> install of their
2544 package after building it. So any previous installs should be
2545 removed first to ensure the integrity of the newly built package.
2546 Then run the package for a while to make sure there are no
2547 obvious problems, before uploading.
2555 <sect3 id="newrelease-tarball"><title>Source Tarball</title>
2557 First, <emphasis>make sure that you have freshly exported the right
2558 version into an empty directory</emphasis>. (See "Building and releasing
2559 packages" above). Then run from that directory:
2562 autoheader && autoconf && ./configure
2572 <sect3 id="NEWRELEASE-WINDOWS"><title>Windows</title>
2573 <!-- so annoying: docbook generated ids are UPPERCASE so
2574 links to "whatever.html#idtag" DO NOT WORK!!
2575 They have to be "whatever.html#IDTAG".
2576 So be consistent and use uppercase on the definition.
2579 Note that the docbook generated files might need some hand editing,
2580 so the Windows build makefile does not rebuild the docs.
2584 First, <emphasis>make sure that you have freshly exported the right
2585 version into an empty directory</emphasis>. (See "Building and releasing
2587 <!-- XXX ??? are we still basing releases off a tarball???
2591 Then you can build the package. This is fully automated, and is
2592 controlled by <filename>windows/GNUmakefile</filename>.
2593 All you need to do is:
2600 Now you can manually rename <filename>privoxy_setup.exe</filename> to
2601 <filename>privoxy_setup_X.Y.Z.exe</filename>, and the <filename>build</filename>
2602 directory to <filename>privoxy_X.Y.Z</filename>.
2603 Create a .zip file of the newly renamed <filename>privoxy_X.Y.Z</filename> directory,
2604 GPG sign the installer and zip file,
2607 $ gpg --armor --detach --sign <filename>privoxy_setup_X.Y.Z.exe</filename>
2608 $ gpg --armor --detach --sign <filename>privoxy_X.Y.Z.zip</filename>
2611 and upload the files to SourceForge.
2615 When releasing the package on SourceForge, use the release notes
2616 and Change Log from the source tarball package.
2620 <sect3 id="newrelease-debian"><title>Debian</title>
2622 First, <emphasis>make sure that you have freshly exported the
2623 right version into an empty directory</emphasis>. (See
2624 "Building and releasing packages" above). Then add a log
2625 entry to <filename>debian/changelog</filename>, if it is not
2626 already there, for example by running:
2629 debchange -v &p-version;-&p-status;-1 "New upstream version"
2635 dpkg-buildpackage -rfakeroot -us -uc -b
2639 <filename>../privoxy_&p-version;-&p-status;-1_i386.deb</filename>
2640 which can be uploaded.
2644 <sect3 id="newrelease-macosx"><title>Mac OS X</title>
2646 First, <emphasis>make sure that you have freshly exported the right
2647 version into an empty directory</emphasis>. (See "Building and releasing
2651 There are three modules available in the CVS repository backups for use on Mac
2652 OS X, though technically only two of them generate a release (the other
2653 can be used to install from source).
2655 <sect4 id="OS-X-OSXPackageBuilder-module">
2656 <title>OSXPackageBuilder module (Documentation out of date)</title>
2658 The OSXPackageBuilder module generates OS X installer packages
2659 supporting all Macs running OS X 10.4 and above. Obtain it from CVS as
2660 follows into a folder parallel to the exported privoxy source:
2663 cvs -z3 -d:pserver:anonymous@ijbswa.cvs.sourceforge.net:/cvsroot/ijbswa co OSXPackageBuilder
2666 The module contains complete instructions on its usage in the file
2667 <filename>OS X Package Builder HOWTO.txt</filename>.
2670 Once the package(s) have been generated, you can then upload them
2671 directly to the Files section of the Sourceforge project in the
2672 Macintosh (OS X) folder. Each new version release of Privoxy should
2673 have a new subfolder created in which to store its files. Please
2674 ensure that the folder contains a readme file that makes it clear
2675 which package is for whichversion of OS X.
2678 <sect4 id="OS-X-osxsetup-module">
2679 <title>osxsetup module (DEPRECATED) (Documentation out of date)</title>
2681 <emphasis>This module is deprecated since the installer it generates
2682 places all Privoxy files in one folder in a non-standard location, and
2683 supports only Intel Macs running OS X 10.6 or higher.</emphasis>
2686 Check out the module from CVS as follows into a folder parallel to the
2687 exported privoxy source:
2690 cvs -z3 -d:pserver:anonymous@ijbswa.cvs.sourceforge.net:/cvsroot/ijbswa co osxsetup
2700 This will run <filename>autoheader</filename>, <filename>autoconf</filename>
2701 and <filename>configure</filename> as well as <filename>make</filename>.
2702 Finally, it will copy over the necessary files to the ./osxsetup/files
2703 directory for further processing by <filename>PackageMaker</filename>.
2706 Bring up PackageMaker with the PrivoxyPackage.pmsp definition file,
2707 modify the package name to match the release, and hit the "Create
2708 package" button. If you specify ./Privoxy.pkg as the output package
2709 name, you can then create the distributable zip file with the command:
2712 zip -r privoxyosx_setup_x.y.z.zip Privoxy.pkg
2715 You can then upload this file directly to the Files section of the
2716 Sourceforge project in the Macintosh (OS X) folder. Each new version
2717 release of Privoxy should have a new subfolder created in which to
2719 Please ensure that the folder contains a readme file that makes it
2720 clear which version(s) of OS X the package supports.
2723 <sect4 id="OS-X-macsetup-module">
2724 <title>macsetup module (Documentation out of date)</title>
2726 The macsetup module is ideal if you wish to build and install Privoxy
2727 from source on a single machine.
2730 Check out the module from CVS as follows into a folder parallel to the
2731 exported privoxy source:
2734 cvs -z3 -d:pserver:anonymous@ijbswa.cvs.sourceforge.net:/cvsroot/ijbswa co macsetup
2737 The module contains complete instructions on its usage in its
2738 <filename>README</filename> file. The end result will be the
2739 exported version of Privoxy installed on the build machine.
2744 <sect3 id="newrelease-freebsd"><title>FreeBSD</title>
2746 Update the www/privoxy port and submit a diff upstream.
2747 For details see the <ulink url="https://www.freebsd.org/doc/en_US.ISO8859-1/books/porters-handbook/">FreeBSD Porter's Handbook</ulink>.
2752 <sect2 id="releasing">
2753 <title>Uploading and Releasing Your Package</title>
2755 After the package is ready, it is time to upload it
2756 and go through the release steps. The upload
2758 <ulink url="https://sourceforge.net/projects/ijbswa/upload/">SourceForge</ulink>
2762 Now just follow the prompts. Be sure to add any appropriate Release
2763 notes. You should see your freshly uploaded packages in
2764 <quote>Step 2. Add Files To This Release</quote>. Check the
2765 appropriate box(es). Remember at each step to hit the
2766 <quote>Refresh/Submit</quote> buttons! You should now see your
2767 file(s) listed in Step 3. Fill out the forms with the appropriate
2768 information for your platform, being sure to hit <quote>Update</quote>
2769 for each file. If anyone is monitoring your platform, check the
2770 <quote>email</quote> box at the very bottom to notify them of
2771 the new package. This should do it!
2774 If you have made errors, or need to make changes, you can go through
2775 essentially the same steps, but select <literal>Edit Release</literal>,
2776 instead of <literal>Add Release</literal>.
2780 <sect2 id="afterrelease">
2781 <title>After the Release</title>
2783 When all (or: most of the) packages have been uploaded and made available,
2784 send an email to the
2785 <ulink url="mailto:privoxy-announce@lists.privoxy.org">announce mailing
2786 list</ulink>, Subject: "Announcing Privoxy X.Y.Z $CODE_STATUS". Be sure to
2788 <ulink url="https://sourceforge.net/projects/ijbswa/files/">
2789 download location</ulink>, the release notes and the Changelog. Also, post an
2790 updated News item on the project page Sourceforge, and update the Home
2791 page and docs linked from the Home page (see below). Other news sites
2792 and release oriented sites, such as Freshmeat, should also be notified.
2795 Then update the source code for the next version to be released:
2800 Increment the version number and change the code status to "UNRELEASED"
2801 in <filename>configure.in</filename>
2806 Rebuild configure (<quote><command>autoheader && autoconf</command></quote>)
2807 and GNUMakefile (<quote><command>./configure</command></quote>)
2812 <quote><command>make dok-release</command></quote> to update the sgml documentation source files.
2817 Commit all your changes.
2826 <!-- ~~~~~ New section ~~~~~ -->
2827 <sect1 id="webserver-update"><title>Update the Webserver</title>
2829 The webserver should be updated at least with each stable release. When
2830 updating, please follow these steps to make sure that no broken links,
2831 inconsistent contents or permission problems will occur (as it has many
2832 times in the past!):
2835 If you have changed anything in the stable-branch documentation source
2842 That will generate <filename>doc/webserver/user-manual</filename>,
2843 <filename>doc/webserver/developer-manual</filename>,
2844 <filename>doc/webserver/faq</filename>,
2845 <filename>doc/webserver/index.html</filename> automatically.
2848 If you changed the manual page sources, generate
2849 <filename>doc/webserver/man-page/privoxy-man-page.html</filename>
2850 by running <quote><command>make man</command></quote>. (This is
2851 a separate target due to dependencies on some obscure perl scripts
2852 [now in Git, but not well tested]. See comments in <filename>GNUmakefile</filename>.)
2855 If you want to add new files to the webserver, create them locally in
2856 the <filename>doc/webserver/*</filename> directory (or
2857 create new directories under <filename>doc/webserver</filename>).
2860 Next, commit any changes from the above steps to Git. All set?
2861 If these are docs in the stable branch, then do:
2867 This will do the upload to the SourceForge webserver (which is manually
2868 syncronized with <ulink url="https://www.privoxy.org/">www.privoxy.org</ulink>)
2869 and ensure all files and directories there are group writable.
2872 Please do <emphasis>NOT</emphasis> use any other means of transferring
2873 files to the webserver to avoid permission problems. Also, please do not
2874 upload docs from development branches or versions. The publicly posted
2875 docs should be in sync with the last official release.
2881 This program is free software; you can redistribute it
2882 and/or modify it under the terms of the GNU General
2883 Public License as published by the Free Software
2884 Foundation; either version 2 of the License, or (at
2885 your option) any later version.
2887 This program is distributed in the hope that it will
2888 be useful, but WITHOUT ANY WARRANTY; without even the
2889 implied warranty of MERCHANTABILITY or FITNESS FOR A
2890 PARTICULAR PURPOSE. See the GNU General Public
2891 License for more details.
2893 The GNU General Public License should be included with
2894 this file. If not, you can view it at
2895 http://www.gnu.org/copyleft/gpl.html
2896 or write to the Free Software Foundation, Inc., 59
2897 Temple Place - Suite 330, Boston, MA 02111-1307, USA.