last-svn-commit-82-gba6a3f8 Generate the list of split outputs in the documentation.

18 Mar 2010

* doc/Makefile.am ($(srcdir)/split-outputs.mk): New rule.
	(regen-split-outputs-mk): New (phony) target.
	($(srcdir)/split-outputs.stamp): Remove rule.
	(refman_dependencies)
	(tutorial_dependencies)
	(ref_guide_dependencies)
	(EXTRA_DIST, MAINTAINERCLEANFILES):
	Remove $(srcdir)/split-outputs.stamp
	Add $(SPLIT_OUTPUTS).
	(regen-am): Remake regen-split-outputs-mk.
	* doc/gen-split-outputs-mk: New script.
	Use it to generate...
	* doc/split-outputs.mk: ...this (new) Make helper.
---
 milena/ChangeLog                                   |   18 ++++++
 milena/doc/Makefile.am                             |   38 ++++++------
 milena/doc/gen-split-examples-mk                   |    2 +
 ...{gen-split-examples-mk => gen-split-outputs-mk} |   24 +++++---
 milena/doc/split-outputs.mk                        |   60 ++++++++++++++++++++
 5 files changed, 114 insertions(+), 28 deletions(-)
 copy milena/doc/{gen-split-examples-mk => gen-split-outputs-mk} (82%)
 create mode 100644 milena/doc/split-outputs.mk

diff --git a/milena/ChangeLog b/milena/ChangeLog
index f52482d..7b7982d 100644
--- a/milena/ChangeLog
+++ b/milena/ChangeLog
@@ -1,5 +1,23 @@
 2010-03-18  Roland Levillain  <roland@lrde.epita.fr>
 
+	Generate the list of split outputs in the documentation.
+
+	* doc/Makefile.am ($(srcdir)/split-outputs.mk): New rule.
+	(regen-split-outputs-mk): New (phony) target.
+	($(srcdir)/split-outputs.stamp): Remove rule.
+	(refman_dependencies)
+	(tutorial_dependencies)
+	(ref_guide_dependencies)
+	(EXTRA_DIST, MAINTAINERCLEANFILES):
+	Remove $(srcdir)/split-outputs.stamp
+	Add $(SPLIT_OUTPUTS).
+	(regen-am): Remake regen-split-outputs-mk.
+	* doc/gen-split-outputs-mk: New script.
+	Use it to generate...
+	* doc/split-outputs.mk: ...this (new) Make helper.
+
+2010-03-18  Roland Levillain  <roland@lrde.epita.fr>
+
 	Prepare doc source files for the automation of output splitting.
 
 	* doc/examples/tuto3/println.cc,
diff --git a/milena/doc/Makefile.am b/milena/doc/Makefile.am
index f40cd4b..d2b6300 100644
--- a/milena/doc/Makefile.am
+++ b/milena/doc/Makefile.am
@@ -244,7 +244,7 @@ regen-split-examples-mk:
 #    corresponding to a bunch of files.
 refman_dependencies =				\
   $(SPLIT_EXAMPLES)				\
-  $(srcdir)/split-outputs.stamp			\
+  $(SPLIT_OUTPUTS)				\
   $(PNG_FIGURES)				\
   $(srcdir)/tutorial/tutorial.hh		\
   $(srcdir)/ref_guide/ref_guide.hh
@@ -534,7 +534,7 @@ tutorial_dir = $(doc_dir)/tutorial
 #    corresponding to a bunch of files.
 tutorial_dependencies =				\
   $(SPLIT_EXAMPLES)				\
-  $(doc_dir)/split-outputs.stamp		\
+  $(SPLIT_OUTPUTS)				\
   $(PNG_FIGURES)				\
   $(DOC_PACKAGES)
 
@@ -732,7 +732,7 @@ ref_guide_dir = $(doc_dir)/ref_guide
 #    corresponding to a bunch of files.
 ref_guide_dependencies =			\
   $(SPLIT_EXAMPLES)				\
-  $(doc_dir)/split-outputs.stamp		\
+  $(SPLIT_OUTPUTS)				\
   $(PNG_FIGURES)				\
   $(DOC_PACKAGES)
 
@@ -909,22 +909,21 @@ include $(srcdir)/figures.mk
 ## Split outputs.  ##
 ## --------------- ##
 
-EXTRA_DIST += $(srcdir)/split-outputs.stamp
-MAINTAINERCLEANFILES += $(srcdir)/split-outputs.stamp
-# FIXME: Loops are bad, as they prevent parallelism.  And we prefer
-# data-driven actions anyway.  This would imply that we know in
-# advance the list out outputs (here, split outputs).  Maybe we can
-# infer it from the LaTeX document.
-$(srcdir)/split-outputs.stamp: $(OUTPUTS)
-	@rm -f $@.tmp
-	@touch $@.tmp
-	failcom='exit 1';				\
-	for file in $(OUTPUTS); do			\
-	  $(DOC_SRCDIR)/tools/split_sample.sh		\
-	      $$file txt "" $(SPLIT_OUTPUTS_SRCDIR)	\
-	    || eval $$failcom;				\
-	done
-	@mv -f $@.tmp $@
+$(srcdir)/split-outputs.mk: $(srcdir)/gen-split-outputs-mk
+	rm -f $@.tmp
+	cd $(srcdir) && ./gen-split-outputs-mk >split-outputs.mk.tmp
+	mv -f $@.tmp $@
+	chmod -w $@
+include $(srcdir)/split-outputs.mk
+EXTRA_DIST += $(SPLIT_OUTPUTS)
+MAINTAINERCLEANFILES += $(SPLIT_OUTPUTS)
+
+.PHONY: regen-split-outputs-mk
+regen-split-outputs-mk:
+	rm -f $(srcdir)/split-outputs.mk.tmp
+	cd $(srcdir) && ./gen-split-outputs-mk >split-outputs.mk.tmp
+	mv -f $(srcdir)/split-outputs.mk.tmp $(srcdir)/split-outputs.mk
+	chmod -w $(srcdir)/split-outputs.mk
 
 
 ## -------------- ##
@@ -942,6 +941,7 @@ regen-am:
 	$(MAKE) $(AM_MAKEFLAGS) regen-split-examples-mk
 	$(MAKE) $(AM_MAKEFLAGS) regen-outputs-mk
 	$(MAKE) $(AM_MAKEFLAGS) regen-examples-outputs-mk
+	$(MAKE) $(AM_MAKEFLAGS) regen-split-outputs-mk
 
 
 ## -------------- ##
diff --git a/milena/doc/gen-split-examples-mk b/milena/doc/gen-split-examples-mk
index fb29a02..a049c46 100755
--- a/milena/doc/gen-split-examples-mk
+++ b/milena/doc/gen-split-examples-mk
@@ -18,6 +18,8 @@
 
 # To be used from milena/doc/'s source directory.
 
+# FIXME: Factor gen-split-examples-mk and gen-split-outputs-mk.
+
 me=`basename $0`
 
 die ()
diff --git a/milena/doc/gen-split-examples-mk b/milena/doc/gen-split-outputs-mk
similarity index 82%
copy from milena/doc/gen-split-examples-mk
copy to milena/doc/gen-split-outputs-mk
index fb29a02..22d0a71 100755
--- a/milena/doc/gen-split-examples-mk
+++ b/milena/doc/gen-split-outputs-mk
@@ -18,6 +18,8 @@
 
 # To be used from milena/doc/'s source directory.
 
+# FIXME: Factor gen-split-examples-mk and gen-split-outputs-mk.
+
 me=`basename $0`
 
 die ()
@@ -44,10 +46,11 @@ backslashify()
 # Get this list as argument?
 inputs()
 {
+  # FIXME: Or « find examples -name \*.cc -a \! -path examples/trash/\* » ?
   find examples -name \*.cc
 }
 
-output_dir=examples/split
+output_dir=outputs/split
 
 # Use a temporary file to dump rules at the end of the output.
 tempfile=`mktemp`
@@ -63,11 +66,12 @@ EOF
 
 # Output target names in a variable.
 {
-  echo "SPLIT_EXAMPLES ="
+  echo "SPLIT_OUTPUTS ="
   inputs | while read file; do
+
     # Count the number of opening and closing marks.
-    nopening=`grep -c '// *\\\{' "$file"`
-    nclosing=`grep -c '// *\\\}' "$file"`
+    nopening=`grep -c 'doc::begin_output' "$file"`
+    nclosing=`grep -c 'doc::end_output'   "$file"`
     test "$nopening" -eq "$nclosing" \
       || die "Unbalanced marks in $file ($nopening vs $nclosing)"
 
@@ -81,15 +85,17 @@ EOF
                       | sed -e 's,examples/,,'	\
                             -e 'y,/,_,'		\
                             -e 's,\.cc$,,'`
-    # Timestamp split examples (outputs) depend on.
-    timestamp="\$(srcdir)/$output_dir/$canonical_name.cc.stamp"
+    # Output file to be split.
+    output="\$(srcdir)/outputs/$canonical_name.txt"
+    # Timestamp split outputs depend on.
+    timestamp="\$(srcdir)/$output_dir/$canonical_name.txt.stamp"
     # List of targets associated to FILE.
     targets=
     {
       i=1
       while test "$i" -le "$nopening"; do
         # i-th part of the split file.
-        part="\$(srcdir)/$output_dir/$canonical_name-$i.cc.raw"
+        part="\$(srcdir)/$output_dir/$canonical_name-$i.txt"
 
         if test -z "$targets"; then
           targets="$part"
@@ -104,10 +110,10 @@ EOF
     cat >>"$tempfile" <<EOF
 
 
-$timestamp: \$(srcdir)/$file
+$timestamp: $output
 	@rm -f \$@.tmp
 	@touch \$@.tmp
-	\$(srcdir)/tools/split_sample.sh \$< cc raw \$(srcdir)/$output_dir
+	\$(srcdir)/tools/split_sample.sh \$< txt "" \$(srcdir)/$output_dir
 	@mv -f \$@.tmp \$@
 EXTRA_DIST += $timestamp
 MAINTAINERCLEANFILES += $timestamp
diff --git a/milena/doc/split-outputs.mk b/milena/doc/split-outputs.mk
new file mode 100644
index 0000000..417c25d
--- /dev/null
+++ b/milena/doc/split-outputs.mk
@@ -0,0 +1,60 @@
+## Generated by gen-split-outputs-mk.  Do not edit by hand.
+
+SPLIT_OUTPUTS = \
+  $(srcdir)/outputs/split/graph-iter-1.txt \
+  $(srcdir)/outputs/split/graph-iter-2.txt \
+  $(srcdir)/outputs/split/graph-iter-3.txt \
+  $(srcdir)/outputs/split/graph-iter-4.txt \
+  $(srcdir)/outputs/split/tuto3_println-1.txt \
+  $(srcdir)/outputs/split/tuto3_println-2.txt \
+  $(srcdir)/outputs/split/tuto4_site_set_create-1.txt \
+  $(srcdir)/outputs/split/tuto4_site_set_create-2.txt \
+  $(srcdir)/outputs/split/tuto4_site_set_create-3.txt
+
+
+$(srcdir)/outputs/split/graph-iter.txt.stamp: $(srcdir)/outputs/graph-iter.txt
+	@rm -f $@.tmp
+	@touch $@.tmp
+	$(srcdir)/tools/split_sample.sh $< txt "" $(srcdir)/outputs/split
+	@mv -f $@.tmp $@
+EXTRA_DIST += $(srcdir)/outputs/split/graph-iter.txt.stamp
+MAINTAINERCLEANFILES += $(srcdir)/outputs/split/graph-iter.txt.stamp
+
+$(srcdir)/outputs/split/graph-iter-1.txt $(srcdir)/outputs/split/graph-iter-2.txt $(srcdir)/outputs/split/graph-iter-3.txt $(srcdir)/outputs/split/graph-iter-4.txt: $(srcdir)/outputs/split/graph-iter.txt.stamp
+## Recover from the removal of $@
+	@if test -f $@; then :; else \
+	  rm -f $<; \
+	  $(MAKE) $(AM_MAKEFLAGS) $<; \
+	fi
+
+
+$(srcdir)/outputs/split/tuto3_println.txt.stamp: $(srcdir)/outputs/tuto3_println.txt
+	@rm -f $@.tmp
+	@touch $@.tmp
+	$(srcdir)/tools/split_sample.sh $< txt "" $(srcdir)/outputs/split
+	@mv -f $@.tmp $@
+EXTRA_DIST += $(srcdir)/outputs/split/tuto3_println.txt.stamp
+MAINTAINERCLEANFILES += $(srcdir)/outputs/split/tuto3_println.txt.stamp
+
+$(srcdir)/outputs/split/tuto3_println-1.txt $(srcdir)/outputs/split/tuto3_println-2.txt: $(srcdir)/outputs/split/tuto3_println.txt.stamp
+## Recover from the removal of $@
+	@if test -f $@; then :; else \
+	  rm -f $<; \
+	  $(MAKE) $(AM_MAKEFLAGS) $<; \
+	fi
+
+
+$(srcdir)/outputs/split/tuto4_site_set_create.txt.stamp: $(srcdir)/outputs/tuto4_site_set_create.txt
+	@rm -f $@.tmp
+	@touch $@.tmp
+	$(srcdir)/tools/split_sample.sh $< txt "" $(srcdir)/outputs/split
+	@mv -f $@.tmp $@
+EXTRA_DIST += $(srcdir)/outputs/split/tuto4_site_set_create.txt.stamp
+MAINTAINERCLEANFILES += $(srcdir)/outputs/split/tuto4_site_set_create.txt.stamp
+
+$(srcdir)/outputs/split/tuto4_site_set_create-1.txt $(srcdir)/outputs/split/tuto4_site_set_create-2.txt $(srcdir)/outputs/split/tuto4_site_set_create-3.txt: $(srcdir)/outputs/split/tuto4_site_set_create.txt.stamp
+## Recover from the removal of $@
+	@if test -f $@; then :; else \
+	  rm -f $<; \
+	  $(MAKE) $(AM_MAKEFLAGS) $<; \
+	fi
-- 
1.5.6.5

    

Roland Levillain

tags

participants (1)