add missing Update line for wheezy
etch is out, wheezy is in
fix static path to maintenance cat
Explicitly close a popen3(patch) fd to avoid lockups
Thanks to Bernhard R. Link for the suggestion.
Thanks to Bernhard R. Link for the suggestion.
remove Update stanza for etch, since it is no longer updated.
Allow blank lines and comments in quilt series files
Thanks to Colin Watson for the suggestion.
Thanks to Colin Watson for the suggestion.
allow a "3.0 (quilt)" format packages without patches.
change some absolute urls to be relative in 500 error handler page
use a more generic title for the diff/tarball section header
now include a 500 error page
allow for compression formats other than gz in the debian.tar files
use "Debian-diff" as "type" for both 1.0 and 3.0 (quilt) formats
there doesn't really seem to be a reason to have it apart from
identifying packages that are native and non-native.
there doesn't really seem to be a reason to have it apart from
identifying packages that are native and non-native.
store the patches in Quilt30Series without the dir name
initial support for source package format 3.0 (quilt)
the changes aren't as pretty as i'd like them to be, but i don't want
to hold things up on mere aesthetic grounds and it seems to work anyway :)
the changes aren't as pretty as i'd like them to be, but i don't want
to hold things up on mere aesthetic grounds and it seems to work anyway :)
add explicit ./ to reprepro confdir
inject some zero-whitespace break characters in long version strings
remove another unused option: static_dir
remove unused/obsolete config output_dir
temporarily disable the body border attribute
the large table is overflowing past the body element, and it looks like
crap (at least in iceweasel). until the proper css/html foo is put in
place the border is disabled.
the large table is overflowing past the body element, and it looks like
crap (at least in iceweasel). until the proper css/html foo is put in
place the border is disabled.
add squeeze to the list of tracked releases
shuffle the logic in PackageCmd to allow exact matches for < 3 letters
Revert "oops, the length should be two not one"
This reverts commit 7bacd4e7b2f521be3d2bf6494547c60ddf2426b3.
actually, it should be one.
This reverts commit 7bacd4e7b2f521be3d2bf6494547c60ddf2426b3.
actually, it should be one.
oops, the length should be two not one
when no version is specified, do not cache PackageCmd
some small notes in TODO
oookay, really fix the latin-1/utf-8 problem
the proper changes were reverted, the wrongly reverted
had the revert reverted, and all is back to normal.
move along, nothing to see here...
the proper changes were reverted, the wrongly reverted
had the revert reverted, and all is back to normal.
move along, nothing to see here...
Revert "try both utf8 and latin-1 for contents in Patch constructor"
This reverts commit ee7e7301434b1cf8b3baa4d839fdbf2f05617e91.
This reverts commit ee7e7301434b1cf8b3baa4d839fdbf2f05617e91.
Revert "Revert "remove some extra print statements that only cause problems""
This reverts commit a9ffd60188742f144d1a8d49f93b05f8e34c39d7.
d'oh, i reverted the wrong commit. this re-adds the wrongly reverted commit.
This reverts commit a9ffd60188742f144d1a8d49f93b05f8e34c39d7.
d'oh, i reverted the wrong commit. this re-adds the wrongly reverted commit.
proper fix for encoding/decoding errors in patch_view contents
this is the proper fix for the previous commit/revert.
this is the proper fix for the previous commit/revert.
Revert "remove some extra print statements that only cause problems"
This reverts commit c2ef848129e03d5486ff3d084d903a7ef7ea658b.
This reverts commit c2ef848129e03d5486ff3d084d903a7ef7ea658b.
try both utf8 and latin-1 for contents in Patch constructor
thanks to Raphael Geissert for catching this
thanks to Raphael Geissert for catching this
remove some extra print statements that only cause problems
while some of them are printed only in the case of error, mod_wsgi
doesn't allow sys.stdout access and therefore they're useless anyway.
while some of them are printed only in the case of error, mod_wsgi
doesn't allow sys.stdout access and therefore they're useless anyway.
switch <code> and <pre> order in patch view template
add mention of diffstat in required packages
add extra indicator config variable for compressed lists file.
also mention python-pysqlite2
add mention of python-profiler in README
move the profile/pstats import into run_profile
this way if the app is being run on a machine that does not have
python-profiler installed it will still function unless specifically
told to profile.
this way if the app is being run on a machine that does not have
python-profiler installed it will still function unless specifically
told to profile.
update apache config to include WSGIDaemonProcess and friends
sample apache2 config now includes the various WSGI options needed
to run as a dedicated user.
sample apache2 config now includes the various WSGI options needed
to run as a dedicated user.
remove superflous import of DiffGzHandler from SourceArchive
track the size of the diff.gz file in DiffGzHandler
this can be used later to do things like pre-emptive
caching of large files, etc.
this can be used later to do things like pre-emptive
caching of large files, etc.
consolidate profiling in pagehandler, removing duplicate code
initial implementation of selective output caching
the PackageCmd and PatchCmd classes are now cached on their
first request.
note that this is incomplete, as the PackageCmd is also currently
used to print non-static information as well (when a package name
doesn't find an exact match and it does a query).
the PackageCmd and PatchCmd classes are now cached on their
first request.
note that this is incomplete, as the PackageCmd is also currently
used to print non-static information as well (when a package name
doesn't find an exact match and it does a query).
small fixups to the pts index feature
add new .gitignore masks for cache, profile and index files
even better profiling, new -o output cmdline option
use cProfile instead of profile, it's much faster and better as
long as it doesn't need to be extended.
the -o option redirects the output to the specified file, instead
of printing it to stdout (or not printing it, both of which aren't
very good for general testing/profiling purposes).
use cProfile instead of profile, it's much faster and better as
long as it doesn't need to be extended.
the -o option redirects the output to the specified file, instead
of printing it to stdout (or not printing it, both of which aren't
very good for general testing/profiling purposes).
allow for pagehandler args to be passed on cmdline
initial support for profiling
add simple getopt support to pagehandler.py
while this isn't intended for mod_wsgi deployments, it opens up some new
oppurtunities for cmdline testing (specifying a URL to avoid invoking
the wsgiref server, enabling profiling, etc).
while this isn't intended for mod_wsgi deployments, it opens up some new
oppurtunities for cmdline testing (specifying a URL to avoid invoking
the wsgiref server, enabling profiling, etc).
initial support for PTS index file generation
a new option (-i) to ./gen-patch-info.py allows for generating a global
"index" file with any and all information that the PTS might want from
this system. The format is:
{
"index-version": N
"packages":
{
"<pkg1>":
{
"url": url-for-all-versions
"versions":
{
"<vers1>":
{
<various attributes>
}
}
}
}
}
index-version is a monotonically increasing version number. when the
"schema" changes, this version is incremented. if we care enough, we
can do a major-minor versioning scheme where BC breaks are differentiated
from "new attribute" type changes by incrementing the major or minor
versions, respectively.
a new option (-i) to ./gen-patch-info.py allows for generating a global
"index" file with any and all information that the PTS might want from
this system. The format is:
{
"index-version": N
"packages":
{
"<pkg1>":
{
"url": url-for-all-versions
"versions":
{
"<vers1>":
{
<various attributes>
}
}
}
}
}
index-version is a monotonically increasing version number. when the
"schema" changes, this version is incremented. if we care enough, we
can do a major-minor versioning scheme where BC breaks are differentiated
from "new attribute" type changes by incrementing the major or minor
versions, respectively.
changes to work with >> lenny versions of reprepro and python-debian
reprepro now seems to default to passching the lists files uncompressed
to the hook, so we deal with that gracefully either way.
python-debian seems to have slightly changed how it handles some multiline
fields due to starting to use the built-in apt_pkg module, see the comments
in Conf.py which should point to the bug.
reprepro now seems to default to passching the lists files uncompressed
to the hook, so we deal with that gracefully either way.
python-debian seems to have slightly changed how it handles some multiline
fields due to starting to use the built-in apt_pkg module, see the comments
in Conf.py which should point to the bug.
pass the patch tracker root via PYTHONPATH to reprepro
the diffsonly.py script called via reprepro needs to know where
the patch tracker root is so that it can include configuration
options from the same central location.
the diffsonly.py script called via reprepro needs to know where
the patch tracker root is so that it can include configuration
options from the same central location.
minor changes to required software in README
get README up to date with reality
fail more gracefully if a diff.gz is missing
this should never happen in production as the database contains
only a cache of what's on disk, but during testing on a system with
the database and no archive, or in unexpected circumstances, it
might be helpful to give a more informative error page.
this should never happen in production as the database contains
only a cache of what's on disk, but during testing on a system with
the database and no archive, or in unexpected circumstances, it
might be helpful to give a more informative error page.
add a base href to skeleton template
updated apache conf for wsgi
more gracefully find what should be the installation root
mod_python -> wsgi
i hear it's the new black. code change requirements weren't
very big at all, and apparently this will work better/stabler
on lenny systems, so...
i hear it's the new black. code change requirements weren't
very big at all, and apparently this will work better/stabler
on lenny systems, so...
d'oh. html validation fix for anchor name tags
add background-color attribute for body{}
include anchor name tags in package page
use min-height instead of height, it's prettier
add documentation about "email search"
add a sample apache configuration
missed one module rename from the initial ptcgi -> ReqHandler
migrate from cgi to mod_python
previously this system was a cgi-based execution. however, this prevented
us from doing a number of useful things, like throwing a 404/5xx error
when an invalid page was requested. so, no we use mod_python instead...
previously this system was a cgi-based execution. however, this prevented
us from doing a number of useful things, like throwing a 404/5xx error
when an invalid page was requested. so, no we use mod_python instead...
fix absolute link in quick jump form
add a local configuration file ignored by git
more TODO cut/paste from email
incorporate some suggestions/updates into TODO
two maintainer email fixes
Revert previous database representation changes
This reverts commit 26a4f31bc27b021dbb332c536e7d14c65fdf54c0.
This reverts commit 5578d0fad24d68ecbd12a735c444910754db50ca.
This reverts commit 26a4f31bc27b021dbb332c536e7d14c65fdf54c0.
This reverts commit 5578d0fad24d68ecbd12a735c444910754db50ca.
remove unneeded where condition
unicode fixes
better output for searches, email query fix
don't allow package searches < 3 chars
add utf-8 encoding comment
better database maintainer representation
change root_url to '' by default
consolidate the FK triggers
maintainer support in frontend
initial, slightly stupid, maintainer support in database
add slash to solo root_url to prevent empty href
new ReleaseList class to order project releases
add experimental to reprepro config
add automatic archive updating support
add contrib/non-free components to reprepro cfg
bugfix: don't automatically prune in destructor
otherwise its gets pruned every page visit, whoops.
otherwise its gets pruned every page visit, whoops.
remove superfluous print from Archive class
remove update trigger for package_rel_maps
since now we "mark" new package relationss (which is necessary for removing
old relations), remove the update trigger, which would otherwise end up
deleting everything in the database...
since now we "mark" new package relationss (which is necessary for removing
old relations), remove the update trigger, which would otherwise end up
deleting everything in the database...
use a "marked" field to catch stale/removed pkgs
no such thing as varchar in sqlite
slightly better breadcrumbs in header
left-align diffstat info
shrink the diffstat info a bit
force -p1 for diffstat generation
remove superfluous diffgz in summary
it shows up in the next table below if it exists, anyway.
it shows up in the next table below if it exists, anyway.
fix a long overdue XXX wrt diffgz's in pagehandler
minor template touchups / fixes
