summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2010-04-30Make sure the list of archives generated by archives_text() is unique and ↵Daniel Burrows1-4/+13
sorted. Apparently some users managed to create situations where a package was available twice from the same archive, and got confused by this.
2010-04-29Merge with desktop.Daniel Burrows1-1/+2
2010-04-29Tidy up the area code and factor out generic stuff into generic/util.Daniel Burrows8-311/+449
Several changes rolled into this: * Provide an interface for an append-only list that can emit signals when stuff is appended or removed. (TODO: index-based operations?) * Pull out code with no GTK+ dependencies for possible reuse. Includes the dynamic_list and dynamic_list_impl (above), as well as progress_info and enumerators.
2010-04-29Add some stub files sketching out the classes that I want to exist.Daniel Burrows4-1/+94
2010-04-29Create a new directory under "gtk" to hold the toplevel UI code.Daniel Burrows8-6/+33
2010-04-29Fix how the main source list is generated.Daniel Burrows1-1/+2
Since I want sources (for use with the .po rules), I should call srcnode() on each file. This also means the sanity-check I put in to catch this sort of thing doesn't bail.
2010-04-28Stop using chdir in "make dist".Daniel Burrows1-3/+5
2010-04-28Stop using chdir when building the aptitude manpage.Daniel Burrows1-2/+2
chdir isn't safe with -j (see the scons manual).
2010-04-28Remove the po4a.cfg scanner, which was a development avenue I decided not to ↵Daniel Burrows1-9/+1
take for now.
2010-04-28Add an appropriate rule to cause the po4a pofiles to be rebuilt when the ↵Daniel Burrows1-0/+9
master XMLs change.
2010-04-28Add code to build the po4a po files from the master documentation files, ↵Daniel Burrows1-0/+202
using the po4a.cfg file as the "source". This code is slightly hacky. For instance, it expects the config file to exist before the build starts, rather than allowing it to be an intermediate target. It also has a lot of specific beliefs about the po4a configuration file that may not hold in general. But for aptitude's purposes, it will be enough. This also modifies the po4a call (relative to the Makefiles) to not update translations. It's a pain in scons to deal with files that are generated by two commands, and those already have individual rules.
2010-04-27Don't rebuild pofiles just because the .pot's date changed.Daniel Burrows1-2/+22
I tried a couple implementations, but the trick of writing a signature file that's calculated excluding the date line seemed like the cleanest option. Also, it was the only one that worked.
2010-04-27Sort the output from xgettext so it can be reliably checked for changes.Daniel Burrows1-0/+1
Wasn't necessary previously because make does timestamp-based checks, not content-based checks.
2010-04-27Be more careful not to blow away old files if the build is interrupted.Daniel Burrows2-6/+8
2010-04-27Ensure that the master source list in the top-level file is always in the ↵Daniel Burrows1-0/+3
same order. Otherwise scons will rebuild stuff that depends on this (like the .pot file) when the order changes.
2010-04-26Add an SConscript file for the po/ directory.Daniel Burrows3-0/+105
2010-04-26Flatten the list of aptitude sources before making a set from it, since only ↵Daniel Burrows1-1/+1
the leaf nodes should be stored. Otherwise we try to put lists into a set, which is just wrong.
2010-04-26Add SCons custom tools for msgfmt, msgmerge, and xgettext.Daniel Burrows3-0/+116
2010-04-26Use """ instead of ''' in a docstring with lots of apostrophes.Daniel Burrows1-2/+2
2010-04-26Make my custom Install() suport arbitrary keyword arguments (e.g., for more ↵Daniel Burrows1-2/+3
readable variable substitution).
2010-04-26Return File nodes from aptitude's top-level source directory, not filename ↵Daniel Burrows1-2/+2
strings.
2010-04-26Return the list of all aptitude's sources from the top-level SConscript ↵Daniel Burrows2-4/+18
file, so that it can be used to generate pofiles. This is a bit ugly IMO, and I seriously considered the alternative of writing a list of *all* the source, putting it in a module, and importing that into both src/SConscript and SConstruct. However, there are lots of source files in aptitude, and splitting the build files up would be moderately annoying.
2010-04-26Whoops, use the right variable name.Daniel Burrows1-1/+1
2010-04-26Add a variable and command-line option that can set which rsvg-convert to use.Daniel Burrows2-1/+2
2010-04-26Support explicitly setting the variable name in MakeCheckForExecutable.Daniel Burrows1-2/+3
2010-04-26Use the xsltproc environment variable to build the docs.Daniel Burrows1-1/+1
2010-04-26Split the tests for gettext utilities from the tests for libraries.Daniel Burrows2-3/+13
We should be able to build a gettext-enabled aptitude without the utilities for processing pofiles, and we should be able to process pofiles without being able to build gettext support.
2010-04-26Don't try to require xsltproc.Daniel Burrows1-1/+1
2010-04-26Check for xsltproc.Daniel Burrows2-0/+3
2010-04-26Allow the name of the po4a program to be overridden.Daniel Burrows1-1/+1
2010-04-26Add configure checks for the rest of the tools needed by gettext.Daniel Burrows1-3/+24
2010-04-26Make CheckForExecutable() automatically add an option to override the ↵Daniel Burrows1-2/+10
executable location.
2010-04-26Use all_build_envs as the published "base" environment.Daniel Burrows1-1/+1
This is necessary so that the docs can pick up the gettext/po4a settings. I could instead add a "docs" environment, but that doesn't seem necessary at the moment.
2010-04-26Add a convenience routine to define a check for an executable with a ↵Daniel Burrows1-0/+15
standard form.
2010-04-26Support passing the method name as a parameter to ConfigureCheck.Daniel Burrows1-5/+7
2010-04-26Support passing the method name as a parameter to RegisterCheck.Daniel Burrows1-2/+4
Since @RegisterCheck has to work, this can't be used with the decorator syntax (well, it could be, but it would take some silly Python hackery).
2010-04-26Add the Docbook-XSL images into the generated images directory.Daniel Burrows1-1/+4
2010-04-26Add build files for the Japanese docs.Daniel Burrows2-0/+8
2010-04-26Add build files for the French documentation.Daniel Burrows2-0/+9
2010-04-26Switch to the new calling convention for manpage_postprocess.Daniel Burrows1-1/+1
2010-04-26Don't require fixman to be provided by a function, and get it working.Daniel Burrows1-12/+27
2010-04-26Make the fixman scripts executable.Daniel Burrows2-0/+0
2010-04-25Undo the changes to docbook.py.Daniel Burrows1-13/+7
I don't quite understand what's going wrong, but not having the directory in my dependency tree seems to cause more trouble than it solves.
2010-04-25Fix the Docbook builder to really output to the output directory.Daniel Burrows1-1/+1
2010-04-25Add a build file for the Finnish documentation.Daniel Burrows2-0/+9
Haven't tested the manpage postprocessing yet.
2010-04-25Build the documentation in a variant directory again.Daniel Burrows1-4/+1
Hopefully this won't break anything...
2010-04-25Don't try to generate a build rule for a directory when generating Docbook ↵Daniel Burrows1-5/+14
output. This caused all sorts of confusing behavior due to the way that scons treats directories. Now I just generate a stamp file for the HTML output.
2010-04-25Don't try to include manpage.xml in the Czech docs, since it doesn't exist ↵Daniel Burrows1-0/+1
in that language.
2010-04-25Disable the use of an alternate tree for the documentation, to reduce the ↵Daniel Burrows1-1/+4
number of variables involved in its breakage.
2010-04-25Add an explicit dependency from the built documentation to the manpage XML, ↵Daniel Burrows1-0/+9
to work around a bug in the scons implicit dependency stuff.