distribute configure, aclocal.m4#1492
Conversation
081b563 to
31c94d5
Compare
|
fyi, |
|
That's also what I thought, but I didn't bother to check... |
31c94d5 to
8dfface
Compare
|
Actually we could even do something like diff --git a/Makefile.in b/Makefile.in
index 146c6dcce..d107957f5 100644
--- a/Makefile.in
+++ b/Makefile.in
@@ -790,11 +790,12 @@ endif
configure aclocal.m4 src/config.h.in &: bootstrap.sh configure.ac acinclude.m4
./$<
-DIST_PREFIX := flint-$(FLINT_MAJOR).$(FLINT_MINOR).$(FLINT_PATCH)/
+DIST_NAME := flint-$(FLINT_MAJOR).$(FLINT_MINOR).$(FLINT_PATCH)
-dist: configure aclocal.m4 src/config.h.in
- git archive --format tar.gz --prefix ${DIST_PREFIX}src/ $(patsubst %,--add-file %,$(filter src/%,$^)) --prefix ${DIST_PREFIX} $(patsubst %,--add-file %,$(filter-out src/%,$^)) origin/flint-$(FLINT_MAJOR).$(FLINT_MINOR) > ../flint-$(FLINT_MAJOR).$(FLINT_MINOR).$(FLINT_PATCH).tar.gz
- git archive --format zip --prefix ${DIST_PREFIX}src/ $(patsubst %,--add-file %,$(filter src/%,$^)) --prefix ${DIST_PREFIX} $(patsubst %,--add-file %,$(filter-out src/%,$^)) origin/flint-$(FLINT_MAJOR).$(FLINT_MINOR) > ../flint-$(FLINT_MAJOR).$(FLINT_MINOR).$(FLINT_PATCH).zip
+../$(DIST_NAME).%: FORCE
+ git archive --format $* --prefix ${DIST_NAME}/src/ --add-file src/config.h.in --prefix ${DIST_NAME}/ --add-file configure --add-file aclocal.m4 origin/flint-$(FLINT_MAJOR).$(FLINT_MINOR) > $@
+
+dist: ../$(DIST_NAME).tar.gz ../$(DIST_NAME).zip
################################################################################
# debugging
@@ -803,4 +804,6 @@ dist: configure aclocal.m4 src/config.h.in
print-%:
@echo "$*=$($*)"
+FORCE:
+
.PHONY: all library shared static examples profile tests check tune valgrind clean distclean install uninstall dist %_TEST_RUN %_VALGRIND_RUN print-% coverageNot sure if this is better of worse. Any opinions? |
|
It is better, as I only need to parse one line instead of two identical lines. What is the point of |
|
Since the output is based on the contents of the git repo itself and not just on the worktree, I think we want |
This comment was marked as outdated.
This comment was marked as outdated.
8dfface to
e91c6b2
Compare
|
ok, here is a version that works for me... |
|
Thanks for the explanation |
|
I tried it out, and the code creates a .tar.gz that I can unpack and that allows me to do "./configure" and "make". As someone used to autotools, the different behaviour is a bit surprising; I am used to trying "make dist" (actually, "make distcheck") just in my source code checkout, whereas here a branch "origin/flint-x.y" is expected. For the maintainer, I suppose this patch does the desired thing. |
|
This seems to duplicate the work I already did on #1449 ? |
|
Indeed! Sorry for the noise. |
sagemathgh-35848: upgrade to flint3 Upgrade to flint3. Current Sage versions are not compatible with flint ≥ 3, and, though the diff is not huge, there are enough changes that versions including this PR will be incompatible with flint < 3. Fixes sagemath#20003. Closes sagemath#35993 as no longer relevant. Related PRs in upstream projects: * Singular/Singular#1177 * flintlib/flint#1408 * flintlib/flint#1489 * flintlib/flint#1492 * flintlib/flint#1611 * algebraic-solving/msolve#76 * flatsurf/e-antic#264 Additional changes still needed for optional packages to work: * sagemath#36677 * upgrade e-antic * possibly more Planned follow-ups: * sagemath#36449 * sagemath#36433 URL: sagemath#35848 Reported by: Marc Mezzarobba Reviewer(s): Vincent Delecroix
sagemathgh-35848: upgrade to flint3 Upgrade to flint3. Current Sage versions are not compatible with flint ≥ 3, and, though the diff is not huge, there are enough changes that versions including this PR will be incompatible with flint < 3. Fixes sagemath#20003. Closes sagemath#35993 as no longer relevant. Related PRs in upstream projects: * Singular/Singular#1177 * flintlib/flint#1408 * flintlib/flint#1489 * flintlib/flint#1492 * flintlib/flint#1611 * algebraic-solving/msolve#76 * flatsurf/e-antic#264 Additional changes still needed for optional packages to work: * sagemath#36677 * upgrade e-antic * possibly more Planned follow-ups: * sagemath#36449 * sagemath#36433 URL: sagemath#35848 Reported by: Marc Mezzarobba Reviewer(s): Vincent Delecroix
No description provided.