X-Git-Url: https://git.stderr.nl/gitweb?p=matthijs%2Fupstream%2Fblosxom-plugins.git;a=blobdiff_plain;f=general%2Fautotrack;h=7152e6ada3e5b5e1676f975c802225d37fa87bb1;hp=031e5f603e01ef9bb07e40b46f840101723a225f;hb=93245008113e831d1db6fa8a856c0e177315adaa;hpb=f728c3fb2f7140671e4d7a2d70c843c608d6d057 diff --git a/general/autotrack b/general/autotrack index 031e5f6..7152e6a 100644 --- a/general/autotrack +++ b/general/autotrack @@ -1,7 +1,7 @@ # Blosxom Plugin: autotrack -*- cperl -*- # Author: Todd Larason (jtl@molehill.org) # Version: 0+2i -# Blosxom Home/Docs/Licensing: http://www.raelity.org/blosxom +# Blosxom Home/Docs/Licensing: http://blosxom.sourceforge.net/ # AutoTrack plugin Home/Docs/Licensing: # http://molelog.molehill.org/blox/Computers/Internet/Web/Blosxom/AutoTrack/ package autotrack; @@ -275,114 +275,119 @@ sub story { 1; =head1 NAME - Blosxom Plug-in: autotrack +Blosxom Plug-in: autotrack =head1 SYNOPSIS - Automatically or semi-automatically sends trackback pings for new stories. +Automatically or semi-automatically sends trackback pings for new stories. + =head1 VERSION - 0+2i +0+2i + +2nd test release - 2nd test release =head1 AUTHOR - Todd Larason http://molelog.molehill.org/ +Todd Larason http://molelog.molehill.org/ + +This plugin is now maintained by the Blosxom Sourceforge Team, +. =head1 BUGS - None known; address bug reports and comments to me or to the Blosxom - mailing list [http://www.yahoogroups.com/groups.blosxom]. +None known; please send bug reports and feedback to the Blosxom +development mailing list . =head1 Trackback Ping URL Discovery - Trackback Ping URLs are discovered two different ways. +Trackback Ping URLs are discovered two different ways. =head2 Manual Ping URLs - If you have the meta plugin installed, and have it set to run prior to the - autotrack plugin, you can give a trackback url with the "meta-tb_ping" - header; the value of the header should be the ping URL to ping. +If you have the meta plugin installed, and have it set to run prior to the +autotrack plugin, you can give a trackback url with the "meta-tb_ping" +header; the value of the header should be the ping URL to ping. =head2 Automatic Ping URL detection - Subject to some exceptions explained below, every URL given in an 'href' in - the story is fetched, and the resulting content is searched for embedded RDF - sections giving trackback URLs for the given URL. This is the preferred way - for all tools to be given trackback URLs, as it requires no human - intervention, but unfortunately not everyone which has a trackback server - includes the appropriate RDF. Even more unfortunately, there's no easy - way to know whether it's included or not, other than examining the source - of the page. +Subject to some exceptions explained below, every URL given in an 'href' in +the story is fetched, and the resulting content is searched for embedded RDF +sections giving trackback URLs for the given URL. This is the preferred way +for all tools to be given trackback URLs, as it requires no human +intervention, but unfortunately not everyone which has a trackback server +includes the appropriate RDF. Even more unfortunately, there's no easy +way to know whether it's included or not, other than examining the source +of the page. - It's always safe to give a meta-tb_ping header; if you give one, and the - same ping URL is found by autodiscovery, it's only pinged once. +It's always safe to give a meta-tb_ping header; if you give one, and the +same ping URL is found by autodiscovery, it's only pinged once. - If you don't want autodiscovery to be used for a given story, you can set - the meta header 'meta-autotrack' to 'no'. If "meta-autotrack: no" is given, - the meta-tb_ping URL is still pinged if it's specified. +If you don't want autodiscovery to be used for a given story, you can set +the meta header 'meta-autotrack' to 'no'. If "meta-autotrack: no" is given, +the meta-tb_ping URL is still pinged if it's specified. =head1 Customization =head2 Configuration Variables - C<$dont_tb_re> is a regular expression agains which URLs are matched; - if it matches, the URL isn't fetched for autodiscovery; this is useful - for classes of URLs that you link to frequently that you know don't - include the autodiscovery RDF, or that you don't wish to be pinged. The - default value matches Amazon and Google URLs, as well as references to - the current weblog. - - C<$start_from_now> is a boolean that controls the behavior if the timestamp - file doesn't exist; if it's true, then it's treated as if it does exist, - with the current time -- no old articles are pinged. If it's false, then - every story seen is treated as new. Defaults to true. - - C<$semi_auto> is a boolean controlling how automatic the pinging is. If - it's false, then the plugin acts in fully automatic mode -- it's always - enabled, and any new story is examined. If it's true, then the plugin - acts in semi-automatic mode -- it's only enabled if the URL being browsed - includes the paramater "autotrack" (ie, ends with "?autotrack=yes"). By - default, this is true. - - C<$networking> controls which networking implementation to use. If set to - "LWP", an implementation which uses the common LWP (libwww-for-perl) perl - module set is used; if set to a string that includes the word 'curl', an - implementation which uses the external 'curl' utility is used, and the value - of $networking is used as the beginning of the command line (this can be used - to specify a full path to curl or to pass additional arguments); if set - to a string which includes the word 'wget', an implementation which uses the - external 'wget' utility is used with $networking used at the beginning of - the command line as with curl. The wget executable must be new enough to - include the --post-data option; currently, that means a recent 1.9 beta. - Defaults to "LWP". - - C<$debug_level> is an int from 0 to 5 controlling how much debugging output - is logged; 0 logs only errors. Defaults to 1. +C<$dont_tb_re> is a regular expression agains which URLs are matched; +if it matches, the URL isn't fetched for autodiscovery; this is useful +for classes of URLs that you link to frequently that you know don't +include the autodiscovery RDF, or that you don't wish to be pinged. The +default value matches Amazon and Google URLs, as well as references to +the current weblog. + +C<$start_from_now> is a boolean that controls the behavior if the timestamp +file doesn't exist; if it's true, then it's treated as if it does exist, +with the current time -- no old articles are pinged. If it's false, then +every story seen is treated as new. Defaults to true. + +C<$semi_auto> is a boolean controlling how automatic the pinging is. If +it's false, then the plugin acts in fully automatic mode -- it's always +enabled, and any new story is examined. If it's true, then the plugin +acts in semi-automatic mode -- it's only enabled if the URL being browsed +includes the paramater "autotrack" (ie, ends with "?autotrack=yes"). By +default, this is true. + +C<$networking> controls which networking implementation to use. If set to +"LWP", an implementation which uses the common LWP (libwww-for-perl) perl +module set is used; if set to a string that includes the word 'curl', an +implementation which uses the external 'curl' utility is used, and the value +of $networking is used as the beginning of the command line (this can be used +to specify a full path to curl or to pass additional arguments); if set +to a string which includes the word 'wget', an implementation which uses the +external 'wget' utility is used with $networking used at the beginning of +the command line as with curl. The wget executable must be new enough to +include the --post-data option; currently, that means a recent 1.9 beta. +Defaults to "LWP". + +C<$debug_level> is an int from 0 to 5 controlling how much debugging output +is logged; 0 logs only errors. Defaults to 1. =head2 CSS and Flavour Files - There is no output, so no customization through these methods. +There is no output, so no customization through these methods. =head1 Timestamp - A timestamp file is kept as $plugin_state_dir/.autotrack.timestamp; stories - are considered 'new' if their timestamp is later than the timestamp file - (see the C<$start_from_now> variable for the behavior if the file doesn't - exist). There is a small race condition between reading the timestamp - file and updating it when the plugin is enabled; one advantage of semi- - automatic mode is that this is rarely a problem, since the plugin is only - enabled when you want it to be. +A timestamp file is kept as $plugin_state_dir/.autotrack.timestamp; stories +are considered 'new' if their timestamp is later than the timestamp file +(see the C<$start_from_now> variable for the behavior if the file doesn't +exist). There is a small race condition between reading the timestamp +file and updating it when the plugin is enabled; one advantage of semi- +automatic mode is that this is rarely a problem, since the plugin is only +enabled when you want it to be. - If trackback pings are attempted but they all fail, the timestamp file is - reverted to its previous value, so the pings will be tried again later. if - some pings succeed and others fail, however, the timestamp is left with the - updated values, and the failed pings won't be retried. +If trackback pings are attempted but they all fail, the timestamp file is +reverted to its previous value, so the pings will be tried again later. if +some pings succeed and others fail, however, the timestamp is left with the +updated values, and the failed pings won't be retried. =head1 THANKS - * Rael Dornfest -- blosxom (of course) and suggesting $start_from_now option - * Taper Wickel -- pointing out wget 1.9's post support +* Rael Dornfest -- blosxom (of course) and suggesting $start_from_now option +* Taper Wickel -- pointing out wget 1.9's post support =head1 LICENSE