make dist uses git archive

This changes "make dist" to directly use "git archive", rather than
the custom shell script it currently runs.

This is to make the creation of the distribution tarball more directly
traceable to the git repository.  That is why we removed the "make
distprep" step.

"make dist" continues to produce a .gz and a .bz2 tarball as before.

The archives produced this way are deterministic and reproducible,
meaning for a given commit the result file should always be
bit-for-bit identical.  The exception is that if you use a git version
older than 2.38.0, gzip records the platform in the archive, so you'd
get a different output on Windows vs. macOS vs. "UNIX" (everything
else).  In git 2.38.0, this was changed so that everything is recorded
as "UNIX" now.  This is just something to keep in mind.  This issue is
specific to the gzip format, it does not affect other compression
formats.

Meson has its own distribution building command (meson dist), but we
are not using that at this point.  The main problem is that, the way
they have implemented it, it is not deterministic in the above sense.
Also, we want a "make" version for the time being.  But the target
name "dist" in meson is reserved for that reason, so we call the
custom target "pgdist" (so call something like "meson compile -C build
pgdist").

Reviewed-by: Tristan Partin <tristan@neon.tech>
Discussion: https://www.postgresql.org/message-id/flat/40e80f77-a294-4f29-a16f-e21bc7bc75fc%40eisentraut.org

Author: petere

GNUmakefile.in

@@ -87,29 +87,25 @@ update-unicode: | submake-generated-headers submake-libpgport
 distdir	= postgresql-$(VERSION)
 dummy	= =install=
 
-dist: $(distdir).tar.gz $(distdir).tar.bz2
-	rm -rf $(distdir)
+GIT = git
 
-$(distdir).tar: distdir
-	$(TAR) chf $@ $(distdir)
+dist: $(distdir).tar.gz $(distdir).tar.bz2
 
-.INTERMEDIATE: $(distdir).tar
+.PHONY: $(distdir).tar.gz $(distdir).tar.bz2
 
 distdir-location:
 	@echo $(distdir)
 
-distdir:
-	rm -rf $(distdir)* $(dummy)
-	for x in `cd $(top_srcdir) && find . \( -name CVS -prune \) -o \( -name .git -prune \) -o -print`; do \
-	  file=`expr X$$x : 'X\./\(.*\)'`; \
-	  if test -d "$(top_srcdir)/$$file" ; then \
-	    mkdir "$(distdir)/$$file" && chmod 777 "$(distdir)/$$file";	\
-	  else \
-	    ln "$(top_srcdir)/$$file" "$(distdir)/$$file" >/dev/null 2>&1 \
-	      || cp "$(top_srcdir)/$$file" "$(distdir)/$$file"; \
-	  fi || exit; \
-	done
-	$(MAKE) -C $(distdir) distclean
+# Note: core.autocrlf=false is needed to avoid line-ending conversion
+# in case the environment has a different setting.  Without this, a
+# tarball created on Windows might be different than on, and unusable
+# on, Unix machines.
+
+$(distdir).tar.gz:
+	$(GIT) -C $(srcdir) -c core.autocrlf=false archive --format tar.gz -9 --prefix $(distdir)/ HEAD -o $(abs_top_builddir)/$@
+
+$(distdir).tar.bz2:
+	$(GIT) -C $(srcdir) -c core.autocrlf=false -c tar.tar.bz2.command='$(BZIP2) -c' archive --format tar.bz2 --prefix $(distdir)/ HEAD -o $(abs_top_builddir)/$@
 
 distcheck: dist
 	rm -rf $(dummy)
@@ -135,4 +131,4 @@ headerscheck: submake-generated-headers
 cpluspluscheck: submake-generated-headers
 	$(top_srcdir)/src/tools/pginclude/headerscheck --cplusplus $(top_srcdir) $(abs_top_builddir)
 
-.PHONY: dist distdir distcheck docs install-docs world check-world install-world installcheck-world headerscheck cpluspluscheck
+.PHONY: dist distcheck docs install-docs world check-world install-world installcheck-world headerscheck cpluspluscheck

meson.build

@@ -3359,6 +3359,71 @@ run_target('help',
 
 
 
+###############################################################
+# Distribution archive
+###############################################################
+
+# Meson has its own distribution building command (meson dist), but we
+# are not using that at this point.  The main problem is that, the way
+# they have implemented it, it is not deterministic.  Also, we want it
+# to be equivalent to the "make" version for the time being.  But the
+# target name "dist" in meson is reserved for that reason, so we call
+# the custom target "pgdist".
+
+git = find_program('git', required: false, native: true, disabler: true)
+bzip2 = find_program('bzip2', required: false, native: true)
+
+distdir = meson.project_name() + '-' + meson.project_version()
+
+# Note: core.autocrlf=false is needed to avoid line-ending conversion
+# in case the environment has a different setting.  Without this, a
+# tarball created on Windows might be different than on, and unusable
+# on, Unix machines.
+
+tar_gz = custom_target('tar.gz',
+  build_always_stale: true,
+  command: [git, '-C', '@SOURCE_ROOT@',
+            '-c', 'core.autocrlf=false',
+            'archive',
+            '--format', 'tar.gz',
+            '-9',
+            '--prefix', distdir + '/',
+            '-o', join_paths(meson.build_root(), '@OUTPUT@'),
+            'HEAD', '.'],
+  output: distdir + '.tar.gz',
+)
+
+if bzip2.found()
+  tar_bz2 = custom_target('tar.bz2',
+    build_always_stale: true,
+    command: [git, '-C', '@SOURCE_ROOT@',
+              '-c', 'core.autocrlf=false',
+              '-c', 'tar.tar.bz2.command="@0@" -c'.format(bzip2.path()),
+              'archive',
+              '--format', 'tar.bz2',
+              '--prefix', distdir + '/',
+              '-o', join_paths(meson.build_root(), '@OUTPUT@'),
+              'HEAD', '.'],
+    output: distdir + '.tar.bz2',
+  )
+else
+  tar_bz2 = custom_target('tar.bz2',
+    command: [perl, '-e', 'exit 1'],
+    output: distdir + '.tar.bz2',
+  )
+endif
+
+alias_target('pgdist', [tar_gz, tar_bz2])
+
+# Make the standard "dist" command fail, to prevent accidental use.
+# But not if we are in a subproject, in case the parent project wants to
+# create a dist using the standard Meson command.
+if not meson.is_subproject()
+  meson.add_dist_script(perl, '-e', 'exit 1')
+endif
+
+
+
 ###############################################################
 # The End, The End, My Friend
 ###############################################################