]> git.saurik.com Git - bison.git/blobdiff - cfg.mk
c++: revamp the support for variants
[bison.git] / cfg.mk
diff --git a/cfg.mk b/cfg.mk
index 392d2112bebbbff8e4df7dde320cc862dfd52708..a79566d3fda8f11879cc1d98fe2a7f303c49c601 100644 (file)
--- a/cfg.mk
+++ b/cfg.mk
@@ -1,5 +1,5 @@
 # Customize maint.mk                           -*- makefile -*-
-# Copyright (C) 2008-2012 Free Software Foundation, Inc.
+# Copyright (C) 2008-2013 Free Software Foundation, Inc.
 
 # This program is free software: you can redistribute it and/or modify
 # it under the terms of the GNU General Public License as published by
@@ -24,6 +24,7 @@ regen: _version
 
 # Used in maint.mk's web-manual rule
 manual_title = The Yacc-compatible Parser Generator
+gendocs_options_ = -I $(abs_top_srcdir)/doc -I $(abs_top_builddir)/doc
 
 # It's useful to run maintainer-*check* targets during development, but we
 # don't want to wait on a recompile because of an update to $(VERSION).  Thus,
@@ -54,17 +55,126 @@ update-copyright: update-b4-copyright update-package-copyright-year
 update-copyright-env = \
   UPDATE_COPYRIGHT_FORCE=1 UPDATE_COPYRIGHT_USE_INTERVALS=1
 
-exclude = $(foreach a,$(1),$(eval exclude_file_name_regexp--sc_$(a)))
+
+## -------------------- ##
+## More syntax-checks.  ##
+## -------------------- ##
+
+# At least for Mac OS X's grep, the order between . and [ in "[^.[]"
+# matters:
+# $ LC_ALL=fr_FR grep -nE '[^[.]' /dev/null
+# $ LC_ALL=C grep -nE '[^[.]' /dev/null
+# grep: invalid collating element or class
+# $ LC_ALL=fr_FR grep -nE '[^.[]' /dev/null
+# $ LC_ALL=C grep -nE '[^.[]' /dev/null
+sc_at_parser_check:
+       @prohibit='AT_PARSER_CHECK\(\[+[^.[]|AT_CHECK\(\[+\./'          \
+       halt='use AT_PARSER_CHECK for and only for generated parsers'   \
+         $(_sc_search_regexp)
+
+# Indent only with spaces.
+# Taken from Coreutils.
+sc_prohibit_tab_based_indentation:
+       @prohibit='^ *  '                                               \
+       halt='TAB in indentation; use only spaces'                      \
+         $(_sc_search_regexp)
+
+# Prohibit the use of `...` in tests/.  Use $(...) instead.
+# Taken from Coreutils.
+# Not ready for Bison yet.
+#sc_prohibit_test_backticks:
+#      @prohibit='`' in_vc_files='^tests/'             \
+#      halt='use $$(...), not `...` in tests/'         \
+#        $(_sc_search_regexp)
+
+# Enforce recommended preprocessor indentation style.
+# Taken from Coreutils.
+sc_preprocessor_indentation:
+       @if cppi --version >/dev/null 2>&1; then                        \
+         $(VC_LIST_EXCEPT) | grep '\.[ch]$$' | xargs cppi -a -c        \
+           || { echo '$(ME): incorrect preprocessor indentation' 1>&2; \
+               exit 1; };                                              \
+       else                                                            \
+         echo '$(ME): skipping test $@: cppi not installed' 1>&2;      \
+       fi
+
+###########################################################
+# Taken from Coreutils.
+_p0 = \([^"'/]\|"\([^\"]\|[\].\)*"\|'\([^\']\|[\].\)*'
+_pre = $(_p0)\|[/][^"'/*]\|[/]"\([^\"]\|[\].\)*"\|[/]'\([^\']\|[\].\)*'\)*
+_pre_anchored = ^\($(_pre)\)
+_comment_and_close = [^*]\|[*][^/*]\)*[*][*]*/
+# help font-lock mode: '
+
+# A sed expression that removes ANSI C and ISO C99 comments.
+# Derived from the one in GNU gettext's 'moopp' preprocessor.
+_sed_remove_comments =                                 \
+/[/][/*]/{                                             \
+  ta;                                                  \
+  :a;                                                  \
+  s,$(_pre_anchored)//.*,\1,;                          \
+  te;                                                  \
+  s,$(_pre_anchored)/[*]\($(_comment_and_close),\1 ,;  \
+  ta;                                                  \
+  /^$(_pre)[/][*]/{                                    \
+    s,$(_pre_anchored)/[*].*,\1 ,;                     \
+    tu;                                                        \
+    :u;                                                        \
+    n;                                                 \
+    s,^\($(_comment_and_close),,;                      \
+    tv;                                                        \
+    s,^.*$$,,;                                         \
+    bu;                                                        \
+    :v;                                                        \
+  };                                                   \
+  :e;                                                  \
+}
+# Quote all single quotes.
+_sed_rm_comments_q = $(subst ','\'',$(_sed_remove_comments))
+# help font-lock mode: '
+
+_space_before_paren_exempt =? \\n\\$$
+_space_before_paren_exempt = \
+  (^ *\#|\\n\\$$|%s\(to %s|(date|group|character)\(s\))
+# Ensure that there is a space before each open parenthesis in C code.
+sc_space_before_open_paren:
+       @if $(VC_LIST_EXCEPT) | grep -l '\.[ch]$$' > /dev/null; then    \
+         fail=0;                                                       \
+         for c in $$($(VC_LIST_EXCEPT) | grep '\.[ch]$$'); do          \
+           sed '$(_sed_rm_comments_q)' $$c 2>/dev/null                 \
+             | grep -i '[[:alnum:]]('                                  \
+             | grep -vE '$(_space_before_paren_exempt)'                \
+             | grep . && { fail=1; echo "*** $$c"; };                  \
+         done;                                                         \
+         test $$fail = 1 &&                                            \
+           { echo '$(ME): the above files lack a space-before-open-paren' \
+               1>&2; exit 1; } || :;                                   \
+       else :;                                                         \
+       fi
+
+## -------------------------- ##
+## syntax-checks exceptions.  ##
+## -------------------------- ##
+
+exclude = \
+  $(foreach a,$(1),$(eval $(subst $$,$$$$,exclude_file_name_regexp--sc_$(a))))
 $(call exclude,                                                                \
   bindtextdomain=^lib/main.c$$                                         \
+  preprocessor_indentation=^data/|^lib/|^src/parse-gram.[ch]$$         \
   program_name=^lib/main.c$$                                           \
-  prohibit_always-defined_macros=^data/yacc.c|^djgpp/                  \
+  prohibit_always-defined_macros=^data/yacc.c$$|^djgpp/                        \
   prohibit_always-defined_macros+=?|^lib/timevar.c$$                   \
   prohibit_always-defined_macros+=?|^src/(parse-gram.c|system.h)$$     \
   prohibit_always-defined_macros+=?|^tests/regression.at$$             \
-  prohibit_empty_lines_at_EOF=^src/parse-gram.h$$                      \
-  prohibit_strcmp=^doc/bison\.texinfo$$                                        \
+  prohibit_always_true_header_tests=^djgpp/subpipe.h$$|^lib/timevar.c$$        \
+  prohibit_always_true_header_tests+=?|^m4/timevar.m4$$                        \
+  prohibit_defined_have_decl_tests=?|^lib/timevar.c$$                  \
+  prohibit_doubled_word=^tests/named-refs.at$$                          \
+  prohibit_magic_number_exit=^doc/bison.texi$$                         \
+  prohibit_magic_number_exit+=?|^tests/(conflicts|regression).at$$     \
+  prohibit_strcmp=^doc/bison\.texi$$                                   \
+  prohibit_tab_based_indentation=\.(am|mk)$$|^djgpp/|^\.git            \
   require_config_h_first=^(lib/yyerror|data/(glr|yacc))\.c$$           \
   space_tab=^tests/(input|c\+\+)\.at$$                                 \
-  unmarked_diagnostics=^djgpp/                                         \
+  unmarked_diagnostics=^(djgpp/|doc/bison.texi$$|tests/c\+\+\.at$$)    \
 )