[perl5.git] / genpacksizetables.pl

#!/usr/bin/perl -w
# I'm assuming that you're running this on some kind of ASCII system, but
# it will generate EDCDIC too. (TODO)
use strict;
use Encode;

sub make_text {
    my ($chrmap, $letter, $unpredictable, $nocsum, $size, $condition) = @_;
    my $text = "    /* $letter */ $size";
    $text .= " | PACK_SIZE_UNPREDICTABLE" if $unpredictable;
    $text .= " | PACK_SIZE_CANNOT_CSUM"   if $nocsum;
    $text .= ",";

    if ($condition) {
        $condition = join " && ", map {"defined($_)"} split ' ', $condition;
        $text = "#if $condition
$text
#else
    0,
#endif";
    }
    return $text;
}

sub make_tables {
    my %arrays;

    my $chrmap = shift;
    foreach (@_) {
        my ($letter, $shriek, $unpredictable, $nocsum, $size, $condition) =
            /^([A-Za-z])(!?)\t(\S*)\t(\S*)\t([^\t\n]+)(?:\t+(.*))?$/ or
            die "Can't parse '$_'";

        $size = "sizeof($size)" unless $size =~ s/^=//;

        $arrays{$shriek ? 'shrieking' : 'normal'}{ord $chrmap->{$letter}} =
            make_text($chrmap, $letter,
                      $unpredictable, $nocsum, $size, $condition);
    }

    my $text = "STATIC const packprops_t packprops[512] = {\n";
    foreach my $arrayname (qw(normal shrieking)) {
        my $array = $arrays{$arrayname} ||
            die "No defined entries in $arrayname";
        $text .= "    /* $arrayname */\n";
        for my $ch (0..255) {
            $text .= $array->{$ch} || "    0,";
            $text .= "\n";
        }
    }
    # Join "0," entries together
    1 while $text =~ s/\b0,\s*\n\s*0,/0, 0,/g;
    # But split them up again if the sequence gets too long
    $text =~ s/((?:\b0, ){15}0,) /$1\n    /g;
    # Clean up final ,
    $text =~ s/,$//;
    $text .= "};";
    return $text;
}

my @lines = grep {
    s/#.*//;
    /\S/;
} <DATA>;

my %asciimap  = map {chr $_, chr $_} 0..255;
my %ebcdicmap = map {chr $_, Encode::encode("posix-bc", chr $_)} 0..255;

print <<"EOC";
/* These tables are regenerated by genpacksizetables.pl (and then hand pasted
   in).  You're unlikely ever to need to regenerate them.  */

#if TYPE_IS_SHRIEKING != 0x100
   ++++shriek offset should be 256
#endif

typedef U8 packprops_t;
#if 'J'-'I' == 1
/* ASCII */
@{[make_tables (\%asciimap, @lines)]}
#else
/* EBCDIC (or bust) */
@{[make_tables (\%ebcdicmap, @lines)]}
#endif
EOC

__DATA__
#Symbol	unpredictable
#		nocsum	size
c			char
C	*		unsigned char
W	*		unsigned char
U	*		char
s!			short
s			=SIZE16
S!			unsigned short
v			=SIZE16
n			=SIZE16
S			=SIZE16
v!			=SIZE16	PERL_PACK_CAN_SHRIEKSIGN
n!			=SIZE16	PERL_PACK_CAN_SHRIEKSIGN
i			int
i!			int
I			unsigned int
I!			unsigned int
j			=IVSIZE
J			=UVSIZE
l!			long
l			=SIZE32
L!			unsigned long
V			=SIZE32
N			=SIZE32
V!			=SIZE32	PERL_PACK_CAN_SHRIEKSIGN
N!			=SIZE32	PERL_PACK_CAN_SHRIEKSIGN
L			=SIZE32
p		*	char *
w	*	*	char
q			Quad_t	HAS_QUAD
Q			Uquad_t	HAS_QUAD
f			float
d			double
F			=NVSIZE
D			=LONG_DOUBLESIZE	HAS_LONG_DOUBLE USE_LONG_DOUBLE
Commit	Line	Data
78d46eaa NC	1	#!/usr/bin/perl -w
	2	# I'm assuming that you're running this on some kind of ASCII system, but
	3	# it will generate EDCDIC too. (TODO)
	4	use strict;
	5	use Encode;
	6
a7a3cfaa TH	7	sub make_text {
	8	my ($chrmap, $letter, $unpredictable, $nocsum, $size, $condition) = @_;
	9	my $text = " /* $letter */ $size";
	10	$text .= " \| PACK_SIZE_UNPREDICTABLE" if $unpredictable;
	11	$text .= " \| PACK_SIZE_CANNOT_CSUM" if $nocsum;
	12	$text .= ",";
78d46eaa	13
a7a3cfaa TH	14	if ($condition) {
	15	$condition = join " && ", map {"defined($_)"} split ' ', $condition;
	16	$text = "#if $condition
	17	$text
	18	#else
	19	0,
	20	#endif";
	21	}
	22	return $text;
78d46eaa NC	23	}
78d46eaa NC	24
a7a3cfaa TH	25	sub make_tables {
a7a3cfaa TH	26	my %arrays;
78d46eaa	27
a7a3cfaa TH	28	my $chrmap = shift;
	29	foreach (@_) {
	30	my ($letter, $shriek, $unpredictable, $nocsum, $size, $condition) =
	31	/^([A-Za-z])(!?)\t(\S)\t(\S)\t([^\t\n]+)(?:\t+(.*))?$/ or
	32	die "Can't parse '$_'";
78d46eaa	33
a7a3cfaa	34	$size = "sizeof($size)" unless $size =~ s/^=//;
78d46eaa	35
a7a3cfaa TH	36	$arrays{$shriek ? 'shrieking' : 'normal'}{ord $chrmap->{$letter}} =
	37	make_text($chrmap, $letter,
	38	$unpredictable, $nocsum, $size, $condition);
80a13697	39	}
a7a3cfaa	40
f43a9a31	41	my $text = "STATIC const packprops_t packprops[512] = {\n";
a7a3cfaa TH	42	foreach my $arrayname (qw(normal shrieking)) {
	43	my $array = $arrays{$arrayname} \|\|
	44	die "No defined entries in $arrayname";
	45	$text .= " /* $arrayname */\n";
	46	for my $ch (0..255) {
	47	$text .= $array->{$ch} \|\| " 0,";
	48	$text .= "\n";
	49	}
7212898e	50	}
a7a3cfaa TH	51	# Join "0," entries together
	52	1 while $text =~ s/\b0,\s\n\s0,/0, 0,/g;
	53	# But split them up again if the sequence gets too long
	54	$text =~ s/((?:\b0, ){15}0,) /$1\n /g;
	55	# Clean up final ,
	56	$text =~ s/,$//;
	57	$text .= "};";
	58	return $text;
	59	}
78d46eaa	60
a7a3cfaa TH	61	my @lines = grep {
	62	s/#.*//;
	63	/\S/;
	64	} <DATA>;
78d46eaa	65
a7a3cfaa TH	66	my %asciimap = map {chr $_, chr $_} 0..255;
a7a3cfaa TH	67	my %ebcdicmap = map {chr $_, Encode::encode("posix-bc", chr $_)} 0..255;
78d46eaa	68
a7a3cfaa TH	69	print <<"EOC";
	70	/* These tables are regenerated by genpacksizetables.pl (and then hand pasted
	71	in). You're unlikely ever to need to regenerate them. */
78d46eaa	72
a7a3cfaa TH	73	#if TYPE_IS_SHRIEKING != 0x100
	74	++++shriek offset should be 256
	75	#endif
	76
	77	typedef U8 packprops_t;
78d46eaa NC	78	#if 'J'-'I' == 1
78d46eaa NC	79	/* ASCII */
a7a3cfaa	80	@{[make_tables (\%asciimap, @lines)]}
78d46eaa NC	81	#else
78d46eaa NC	82	/* EBCDIC (or bust) */
a7a3cfaa TH	83	@{[make_tables (\%ebcdicmap, @lines)]}
a7a3cfaa TH	84	#endif
78d46eaa	85	EOC
78d46eaa NC	86
78d46eaa NC	87	__DATA__
f337b084 TH	88	#Symbol unpredictable
f337b084 TH	89	# nocsum size
78d46eaa	90	c char
f337b084 TH	91	C * unsigned char
	92	W * unsigned char
	93	U * char
78d46eaa NC	94	s! short
	95	s =SIZE16
	96	S! unsigned short
	97	v =SIZE16
	98	n =SIZE16
	99	S =SIZE16
7212898e NC	100	v! =SIZE16 PERL_PACK_CAN_SHRIEKSIGN
7212898e NC	101	n! =SIZE16 PERL_PACK_CAN_SHRIEKSIGN
78d46eaa NC	102	i int
	103	i! int
	104	I unsigned int
	105	I! unsigned int
	106	j =IVSIZE
	107	J =UVSIZE
	108	l! long
	109	l =SIZE32
	110	L! unsigned long
	111	V =SIZE32
	112	N =SIZE32
7212898e NC	113	V! =SIZE32 PERL_PACK_CAN_SHRIEKSIGN
7212898e NC	114	N! =SIZE32 PERL_PACK_CAN_SHRIEKSIGN
78d46eaa	115	L =SIZE32
a72c271b	116	p * char *
f337b084	117	w * * char
80a13697 NC	118	q Quad_t HAS_QUAD
80a13697 NC	119	Q Uquad_t HAS_QUAD
78d46eaa NC	120	f float
	121	d double
	122	F =NVSIZE
80a13697	123	D =LONG_DOUBLESIZE HAS_LONG_DOUBLE USE_LONG_DOUBLE