[m-rev.] for review: remove compiler Unix dependencies

Fergus Henderson fjh at cs.mu.OZ.AU
Tue Aug 5 16:41:29 AEST 2003
Previous message: [m-rev.] for review: remove compiler Unix dependencies
Next message: [m-rev.] for review: remove compiler Unix dependencies
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
On 02-Aug-2003, Simon Taylor <stayl at cs.mu.OZ.AU> wrote:
> 
> Index: compiler/make.program_target.m
...
> +++ compiler/make.util.m	31 Jul 2003 15:02:50 -0000
> @@ -330,16 +330,24 @@
>  		{ Succeeded = no }
>  	;
>  		{ OptionsResult = yes(ModuleOptionArgs) }, 
> +		globals__io_lookup_bool_option(make_implies_use_subdirs,
> +			MakeImpliesUseSubdirs),
>  		globals__io_get_globals(Globals),
>  
> -		% --invoked-by-mmc-make disables reading DEFAULT_MCFLAGS
> -		% from the environment (DEFAULT_MCFLAGS is included in
> -		% OptionArgs) and generation of `.d' files.
> +		% --invoked-by-mmc-make disables reading the options file.
> +		% (DEFAULT_MCFLAGS is included in OptionArgs) and generation
> +		% of `.d' files.

s/file./file/

> Index: compiler/options.m
...
> @@ -2364,17 +2385,6 @@
>  	->
>  		Arg = """"""
>  	;
> -		list__member(Char, ArgList),
> -		\+ ( char__is_alnum_or_underscore(Char)
> -		; Char = ('-')
> -		; Char = ('/')
> -		; Char = ('.')
> -		; Char = (',')
> -		; Char = (':')
> -		)
> -	->
> -		Arg = """" ++ string__from_char_list(ArgList) ++ """"
> -	;
>  		Arg = string__from_char_list(ArgList)
>  	).
>  
> @@ -2384,15 +2394,24 @@
>  quote_arg_2([Char | Chars0]) = Chars :-
>  	Chars1 = quote_arg_2(Chars0),
>  	( quote_char(Char) ->
> -		Chars = [('\\'), Char | Chars1]
> +		% We want whitespace characters within an argument to not be
> +		% treated as whitespace when splitting the command line
> +		% into words. \newline is still treated as whitespace,
> +		% and not all shells will understand "\n".
> +		% Newlines and tabs within a word don't really make
> +		% sense, so just convert them to spaces.
> +		QuoteChar = ( char__is_whitespace(Char) -> ' ' ; Char ),
> +		Chars = [('\\'), QuoteChar | Chars1]
>  	;
> -		Chars = [Char | Chars1]	
> -	).	
> -
> +		Chars = [Char | Chars1]
> +	).
>  
>  :- pred quote_char(char::in) is semidet.
>  
> -quote_char('\\').
> +quote_char(' ').
> +quote_char('\n').
> +quote_char('\t').
> +quote_char('\\') :- \+ dir__use_windows_paths.
>  quote_char('"').
>  quote_char('`').
>  quote_char('$').

Ouch.  Converting newlines and tabs to spaces doesn't seem
like a good idea -- it might not preserve the semantics.

If the problem is that quoting for a Unix shell doesn't work
when using a Windows shell, then I think it would be better
to use different quoting algorithms for the different shells.

> @@ -3983,7 +4010,13 @@
>  		"\tExecutables and libraries will be symlinked or copied into",
>  		"\tthe current directory.",
>  		"\t`--use-grade-subdirs' does not work with Mmake (it does",
> -		"\twork with `mmc --make')."
> +		"\twork with `mmc --make').",
> +
> +		"--no-make-implies-use-subdirs",
> +		"\tNormally, `--make' implies `--use-subdirs'.",
> +		"\tThis option disables that behaviour.",
> +		"\tLibrary installation does not work with",
> +		"\t--no-make-implies-use-subdirs."

Hmm... would it be better for this option to be the default?

Or alternatively, perhaps we could make --use-subdirs the default?

(Incidentally, isn't this a completely separate issue from removing
the Unix dependencies?  Should this be a separate patch?)

> Index: library/Mmakefile
> ===================================================================
> RCS file: /home/mercury1/repository/mercury/library/Mmakefile,v
> retrieving revision 1.117
> diff -u -u -r1.117 Mmakefile
> --- library/Mmakefile	21 Jul 2003 14:08:40 -0000	1.117
> +++ library/Mmakefile	30 Jul 2003 14:27:55 -0000
> @@ -322,19 +322,12 @@
>  
>  ifeq ($(MMAKE_USE_MMC_MAKE),no)
>  
> -EXTRA_INIT_COMMAND = \
> -	for file in $($(STD_LIB_NAME).ms); do \
> -		grep '^INIT ' $$file; \
> -		true; \
> -	done
> +EXTRA_INIT_COMMAND = print_extra_inits $($(STD_LIB_NAME).ms)
>  
>  else
>  
> -MCFLAGS += --extra-init-command \
> -		"for module in %; do \
> -			grep '^INIT ' $$$${module}.m; \
> -			true; \
> -		done"
> +MCFLAGS += --extra-init-command print_extra_inits
> +
>  endif	# 

Is this relying on "." being in $PATH?

> Index: library/print_extra_inits
> ===================================================================
> RCS file: library/print_extra_inits
> diff -N library/print_extra_inits
> --- /dev/null	1 Jan 1970 00:00:00 -0000
> +++ library/print_extra_inits	30 Jul 2003 14:27:55 -0000
> @@ -0,0 +1,24 @@
> +#!/bin/sh
> +#-----------------------------------------------------------------------------#
> +# Copyright (C) 2003 The University of Melbourne.
> +# This file may only be copied under the terms of the GNU General
> +# Public License - see the file COPYING in the Mercury distribution.
> +#-----------------------------------------------------------------------------#
> +# library/print_extra_inits - print extra .init file lines to stdout.
> +#
> +# Invocation:
> +#	print_extra_inits <mer_std.ms>
> +# 		where <mer_std.ms> is the names of all of the source files
> +#		for the modules	in libmer_std.
> +#
> +#-----------------------------------------------------------------------------#
> +
> +for file; do
> +	if [ -f $file ]; then
> +		grep '^INIT ' $file
> +	else
> +		echo "source file $file not found" 1>&2
> +		exit 1

s/source/$0: source/

> Index: scripts/mercury.bat.in
...
> +rem There is no equivalent of "$@" available on all Windows platforms.
> +rem (Windows XP has `%*').
> +
> +set v1=%1
> +set v2=%2
> +set v3=%3
> +set v4=%4
> +set v5=%5
> +set v6=%6
> +set v7=%7
> +set v8=%8
> +set v9=%9
> +shift
> +shift
> +shift
> +shift
> +shift
> +shift
> +shift
> +shift
> +shift
> +shift
> +set v10=%0
> +set v11=%1
> +set v12=%2
> +set v13=%3
> +set v14=%4
> +set v15=%5
> +set v16=%6
> +set v17=%7
> +set v18=%8
> +set v19=%9
> +shift
> +shift
> +shift
> +shift
> +shift
> +shift
> +shift
> +shift
> +shift
> +shift
> +%MERCURY_COMPILER% %v1% %v2% %v3% %v4% %v5% %v6% %v7% %v8% %v9% %v10% %v11% %v12% %v13% %v14% %v15% %v16% %v17% %v18% %v19% %0 %1 %2 %3 %4 %5 %6 %7 %8 %9

It would be a good idea to check if there are any remaining arguments,
and if so, to issue an error message.

> Index: scripts/mmc.in
> ===================================================================
> RCS file: /home/mercury1/repository/mercury/scripts/mmc.in,v
> retrieving revision 1.35
> diff -u -u -r1.35 mmc.in
> --- scripts/mmc.in	14 Apr 2003 06:50:13 -0000	1.35
> +++ scripts/mmc.in	30 Jul 2003 14:27:55 -0000
> @@ -14,12 +14,11 @@
>  # MERCURY_COMPILER, MERCURY_C_COMPILER, MERCURY_DEFAULT_GRADE,
>  # MERCURY_DEFAULT_OPT_LEVEL.
>  
> -MC=${MERCURY_COMPILER="@LIBDIR@/bin/@FULLARCH@/mercury_compile"}
> -MERCURY_CONFIG_DIR=${MERCURY_CONFIG_DIR-${MERCURY_STDLIB_DIR- at LIBDIR@}}
> -export MERCURY_CONFIG_DIR
> +MERCURY_COMPILER=${MERCURY_COMPILER-'@LIBDIR@/bin/@FULLARCH@/mercury_compile'}
> +MERCURY_CONFIG_DIR=${MERCURY_CONFIG_DIR-${MERCURY_STDLIB_DIR-'@CONFIG_LIBDIR@'}}
> +export MERCURY_COMPILER MERCURY_CONFIG_DIR
>  
>  case $# in
> -	0) exec $MC ;;
> -	*) exec $MC "$@" ;;
> +	0) exec $MERCURY_COMPILER ;;
> +	*) exec $MERCURY_COMPILER "$@" ;;
>  esac
> -

That change wasn't mentioned in the log message.

> Index: util/mkinit.c
> ===================================================================
> RCS file: /home/mercury1/repository/mercury/util/mkinit.c,v
> retrieving revision 1.89
> diff -u -u -r1.89 mkinit.c
> --- util/mkinit.c	25 Sep 2002 07:53:53 -0000	1.89
> +++ util/mkinit.c	30 Jul 2003 14:27:55 -0000
> @@ -876,6 +876,9 @@
>  	if ((position = strrchr(filename, '/')) != NULL) {
>  		filename = position + 1;
>  	}
> +	if ((position = strrchr(filename, '\\')) != NULL) {
> +		filename = position + 1;
> +	}

Missing "else"?

What happens with file names that contain both separators, e.g.
	C:\foo/bar
or
	C:/foo\bar
?

Otherwise that patch looks fine.

-- 
Fergus Henderson <fjh at cs.mu.oz.au>  |  "I have always known that the pursuit
The University of Melbourne         |  of excellence is a lethal habit"
WWW: <http://www.cs.mu.oz.au/~fjh>  |     -- the last words of T. S. Garp.
--------------------------------------------------------------------------
mercury-reviews mailing list
post:  mercury-reviews at cs.mu.oz.au
administrative address: owner-mercury-reviews at cs.mu.oz.au
unsubscribe: Address: mercury-reviews-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-reviews-request at cs.mu.oz.au Message: subscribe
--------------------------------------------------------------------------
Previous message: [m-rev.] for review: remove compiler Unix dependencies
Next message: [m-rev.] for review: remove compiler Unix dependencies
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the reviews mailing list