summaryrefslogtreecommitdiffstats
path: root/Makefile.common
Commit message (Collapse)AuthorAgeFilesLines
* When validating XML, be sure to include the dependent files such asTommy Reynolds2006-03-031-1/+1
| | | | | ${1}/fdp-info.xml and ${1}/${DOC_ENTITIES}.ent so there is actually a hope of the XML validating.
* Replaced ${MSGMERGE} processing to update .PO file when the .XML andTommy Reynolds2006-03-031-0/+2
| | | | thus the .POT file changes.
* Deleted the 'DIR_template' foolishness and removed the "${LANG}" dummyTommy Reynolds2006-03-031-17/+2
| | | | | | target, too. Replaced the explicit "mkdir -p ${LANG}" in the rule to derive the non-${PRI_LANG} XML. This cuts down on the unnecessary XML regeneration and speeds things up a good bit.
* Added "-k" switch to ${XML2PO} usages to try and keep entity referencesTommy Reynolds2006-03-031-12/+10
| | | | intact. This doesn't fix the whole problem, but it's a start.
* Oops, removed too much. Sorry.Tommy Reynolds2006-03-021-0/+1
|
* Removed bogus targets that appeared before the "all::" rule, whichTommy Reynolds2006-03-021-9/+0
| | | | absolutely, positively, must be first.
* When doing "make clean", forgot the ".ent" extention on the ${DOC_ENTITIES}Tommy Reynolds2006-03-021-1/+1
| | | | reference.
* Fixed depandancies so that ${LANG}/fdp-info.xml andTommy Reynolds2006-03-021-4/+11
| | | | ${LANG}/${DOC_ENTITIES}.ent both get built
* Hide our workTommy Reynolds2006-03-011-1/+1
|
* Added support for per-document translatable entities file.Tommy Reynolds2006-03-011-4/+39
|
* Reverting, this is just not good.Paul W. Frields2006-03-011-5/+2
|
* Let's generate new entities for a document at build time, into a realPaul W. Frields2006-03-011-2/+5
| | | | | | | | | | file. This particular fix lets the document author specify a "entities.xml" file in the PRI_LANG directory, which can be PO'd and translated like anything else. I'm thinking this should be one of the XMLFILES in the Makefile as well, right? In any case, there's a little bit of comment duplication but nothing that really matters, since the file shouldn't be entered in CVS anyway and will be rebuilt regularly.
* Set proper locale before validating XML. Do a more thourough jobrelease-notes-FC-5-GOLD-for-ISOTommy Reynolds2006-03-011-2/+14
| | | | of cleaning up after ourselves.
* Include "${1}/fdp-info.xml" as a prerequisite to validation.Tommy Reynolds2006-02-281-1/+1
|
* Added "make validate-xml-${LANG}" and "make validate-xml" targets toTommy Reynolds2006-02-281-0/+24
| | | | validate the document XML without performing any subsequent processing.
* Provided viewing hint to "make help" output: use sort(1)!Tommy Reynolds2006-02-281-1/+1
|
* Added "msgmerge -U" switch to update .PO file in-place. Dropped theTommy Reynolds2006-02-281-2/+1
| | | | "*.po.bank" backup file in favor of the msgmerge default behavior.
* Added "fdpcommon" kind of entity. Automatically relocate filesTommy Reynolds2006-02-281-1/+1
| | | | | | marked as "fdpcommon" when building the ".ent" files. Revert Karsten's "Makefile.common" change as a courtesey.
* The file is a legal XML file now, and it is thus named, so now this ↵Karsten Wade2006-02-271-1/+1
| | | | shouldn't break stuff.
* Use UTF-8 for these targets too, methinks. Also, fix rpm-common to do a ↵Paul W. Frields2006-02-261-4/+8
| | | | more brainful job of automagicking. I love making up words.
* Added "set-locale-${LANG}" target to change the local symlinkTommy Reynolds2006-02-261-35/+63
| | | | | | | | "locale-entities.xml" to point to the "docs-common/common/entities/entities-${LANG}.ent" file. The locale gets reset automatically as needed each time the XML is rendered via xmlto(1).
* Need XML files for src-tarballPaul W. Frields2006-02-241-1/+2
|
* Add another phony alias "help" for better across-the-project adherencePaul W. Frields2006-02-241-2/+2
| | | | (q.v. Fedora Extras), plus it makes a weird kind of sense.
* Use xml2po(1) for all XML=>PO=>XML conversions; drop dependancy onTommy Reynolds2006-02-231-2/+1
| | | | | the buggy po2xml(1) from the kdesdk. This allows translators to free up potentially 82Mbyte or so of disk space as a bonus.
* Every .PO file is precious and make(1) should not delete it unlessTommy Reynolds2006-02-221-0/+3
| | | | explicitly directed to do so.
* Revert prior changes. Document template purpose.Tommy Reynolds2006-02-221-2/+4
|
* Move "-p" arg usage from application path to the actual usage.Tommy Reynolds2006-02-221-2/+2
|
* wrong argument sequence :(Yuan Yijun2006-02-221-2/+2
|
* use xml2po instead of po2xmlYuan Yijun2006-02-221-2/+2
|
* a second make should not fail because of mkdirYuan Yijun2006-02-221-1/+1
|
* Added "make targets" to see a list of all targets for the currentTommy Reynolds2006-02-201-2/+128
| | | | | | | | "Makefile". It looks much better if you sort the output: make targets | sort Needs some entries fleshed out.
* Add "text" alias for "txt" for bad typists and lexihounds. Also stripPaul W. Frields2006-02-171-7/+16
| | | | | | | annoying revhistory and index elements from plain text builds. There are very few instances where we will be providing txt builds -- e.g. release notes -- so this doesn't seem like a loss. If it is, XSLT is cheap.
* Whoa Nelly, let's keep those removals to the right stuff... :-)Paul W. Frields2006-02-171-1/+1
|
* Make text build cleanly without unnecessary link lists, esp. given our ↵Paul W. Frields2006-02-171-1/+4
| | | | standards for ulink element usage
* Make sure HTML nochunks files do not overwrite each otherPaul W. Frields2006-02-161-0/+1
|
* Add txt targetsPaul W. Frields2006-02-161-1/+20
|
* Added "${LANG}::" target to use as a dependant to the "xml-${LANG}" rulesTommy Reynolds2006-02-151-2/+13
| | | | so we don't invoke "mkdir -p ${LANG}" for every translated XML file.
* Reorganize the CVS layout to fully embrace I18N translator support.Tommy Reynolds2006-02-151-71/+226
| | | | | | | | | | | | | | | | | | | | | Many, many changes here. Refer to "example-tutorial/" for a model. 1) Only one rpm-info.xml file, and that is in the top-level dir. 2) XML file names are now language independant (para.xml, not para-en.xml). 3) The "Makefile" in the document directory has been completely rewritten. Now, a ${PRI_LANG} variable identifies the original language for the document. Additional language translations are listed in the ${OTHERS} macro. The ${XMLEXTRAFILES-${LANG}} variable is no more; instead an XMLFILES_template must be filled out to enumerate the necessary XML files. 4) Translated XML files are automatically generated from .POT and .PO files located in a "po/" directory. 5) Translated XML files should NOT be archived in CVS, since they are now generated files. 6) Many new virtual targets have been added to "Makefile.common", such as "make xml-de", "make po-de", and the like. Consult the "Makefile.common".
* Oops, remember Tommy's rule of good behavior in targetsPaul W. Frields2006-02-131-1/+1
|
* Add src-tarball phony target for use when VERSION isn't known externallyPaul W. Frields2006-02-131-2/+3
|
* Push to 0.3.1:Paul W. Frields2006-02-131-0/+1
| | | | | | | - Update Makefile to properly remove supporting source files - Support generation of fdp-info in local build environment - Include additional scripts - Revert spec.xsl to distinguish local builds from CVS builds
* Remove the rest of the files correctlyPaul W. Frields2006-02-131-3/+3
|
* Remember to remove RPM files in top level using correct namePaul W. Frields2006-02-131-1/+1
|
* Must make sure we don't remove non-generated files. For example, inPaul W. Frields2006-02-111-5/+5
| | | | | the release-notes directory there is a non-generated "fedora-release.spec" which we were whacking on a "make distclean".
* SHORT VERSION:Paul W. Frields2006-02-091-6/+24
| | | | | | | | | | | | | | | | | * Separate out rpm-common and package-post * Create additional noarch and srpm targets LONG VERSION: We need a target that builds only the SRPM to go with spec file for FE review, so here it is. (Also an additional disincentive to installing drafts, but that's neither here nor there.) Next up, adding this functionality to build-docs in infrastructure/ module. Note that you can now pass flags to prevent the package-prep and/or package-post steps with any RPM building process, although this might be an edge case. Robust! Simple! Elegant! Hopefully not stupid! Now with 50% more targets. (I added these to .PHONY which I think is the right thing to do. Everything still works at least...
* Do the right thing if there's no RPM revision yet for this docPaul W. Frields2006-02-071-0/+4
|
* Trivial, just tired of setting the modePaul W. Frields2006-02-071-0/+5
|
* Actually check in the ${PRI_LANG} changes I mentioned earlier.Tommy Reynolds2006-02-071-2/+7
|
* This flag no longer needed thanks to XSLT sweetnessPaul W. Frields2006-02-041-1/+0
|
* This seems to make packaging work again... probably needs checking for ↵Paul W. Frields2006-02-041-4/+4
| | | | kludginess
generated by cgit (git 2.34.1) at 2024-06-24 16:24:42 +0000