/* Copyright (C) 1995, 2000 artofcode LLC.  All rights reserved.
  
  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 the
  Free Software Foundation; either version 2 of the License, or (at your
  option) any later version.

  This program is distributed in the hope that it will be useful, but
  WITHOUT ANY WARRANTY; without even the implied warranty of
  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  General Public License for more details.

  You should have received a copy of the GNU General Public License along
  with this program; if not, write to the Free Software Foundation, Inc.,
  59 Temple Place, Suite 330, Boston, MA, 02111-1307.

*/

/*$Id: geninit.c,v 1.3.6.1.2.1 2003/01/17 00:49:02 giles Exp $ */
/*
 * Utility for merging all the Ghostscript initialization files (gs_*.ps)
 * into a single file, optionally converting them to C data.  Usage:
 *    geninit [-(I|i) <prefix>] <init-file.ps> <gconfig.h> <merged-init-file.ps>
 *    geninit [-(I|i) <prefix>] <init-file.ps> <gconfig.h> -c <merged-init-file.c>
 *
 * The following special constructs are recognized in the input files:
 *	%% Replace <#lines> (<psfile>)
 *	%% Replace <#lines> INITFILES
 * If the first non-comment, non-blank line in a file ends with the
 * LanguageLevel 2 constructs << or <~, its section of the merged file
 * will begin with
 *	currentobjectformat 1 setobjectformat
 * and end with
 *	setobjectformat
 * and if any line then ends with with <, the following ASCIIHex string
 * will be converted to a binary token.
 */
#include "stdpre.h"
#include <ctype.h>
#include <stdio.h>
#include <stdlib.h>		/* for exit() */
#include <string.h>

/* Forward references */
private FILE *prefix_open(P2(const char *prefix, const char *inname));
private void merge_to_c(P5(const char *prefix, const char *inname, FILE * in,
			   FILE * config, FILE * out));
private void merge_to_ps(P5(const char *prefix, const char *inname, FILE * in,
			    FILE * config, FILE * out));

#define LINE_SIZE 128

int
main(int argc, char *argv[])
{
    int arg_c = argc;
    char **arg_v = argv;
    const char *fin;
    FILE *in;
    const char *fconfig;
    FILE *config;
    const char *fout;
    FILE *out;
    const char *prefix = "";
    bool to_c = false;

    if (arg_c >= 2 && (!strcmp(arg_v[1], "-I") || !strcmp(arg_v[1], "-i"))) {
	prefix = arg_v[2];
	arg_c -= 2;
	arg_v += 2;
    }
    if (arg_c == 4)
	fin = arg_v[1], fconfig = arg_v[2], fout = arg_v[3];
    else if (arg_c == 5 && !strcmp(arg_v[3], "-c"))
	fin = arg_v[1], fconfig = arg_v[2], fout = arg_v[4], to_c = true;
    else {
	fprintf(stderr, "\
Usage: geninit [-(I|i) lib/] gs_init.ps gconfig.h gs_xinit.ps\n\
or     geninit [-(I|i) lib/] gs_init.ps gconfig.h -c gs_init.c\n");
	exit(1);
    }
    in = prefix_open(prefix, fin);
    if (in == 0) {
	fprintf(stderr, "Cannot open %s for reading.\n", fin);
	exit(1);
    }
    config = fopen(fconfig, "r");
    if (config == 0) {
	fprintf(stderr, "Cannot open %s for reading.\n", fconfig);
	fclose(in);
	exit(1);
    }
    out = fopen(fout, "w");
    if (out == 0) {
	fprintf(stderr, "Cannot open %s for writing.\n", fout);
	fclose(config);
	fclose(in);
	exit(1);
    }
    if (to_c)
	merge_to_c(prefix, fin, in, config, out);
    else
	merge_to_ps(prefix, fin, in, config, out);
    fclose(out);
    return 0;
}

/* Open a file with a name prefix. */
private FILE *
prefix_open(const char *prefix, const char *inname)
{
    char fname[LINE_SIZE + 1];

    if (strlen(prefix) + strlen(inname) > LINE_SIZE) {
	fprintf(stderr, "File name > %d characters, too long.\n",
		LINE_SIZE);
	exit(1);
    }
    strcpy(fname, prefix);
    strcat(fname, inname);
    return fopen(fname, "r");
}

/* Read a line from the input. */
private bool
rl(FILE * in, char *str, int len)
{
    /*
     * Unfortunately, we can't use fgets here, because the typical
     * implementation only recognizes the EOL convention of the current
     * platform.
     */
    int i = 0, c = getc(in);

    if (c < 0)
	return false;
    while (i < len - 1) {
	if (c < 0 || c == 10)		/* ^J, Unix EOL */
	    break;
	if (c == 13) {		/* ^M, Mac EOL */
	    c = getc(in);
	    if (c != 10 && c >= 0)	/* ^M^J, PC EOL */
		ungetc(c, in);
	    break;
	}
	str[i++] = c;
	c = getc(in);
    }
    str[i] = 0;
    return true;
}

/* Write a string or a line on the output. */
private void
wsc(FILE * out, const byte *str, int len)
{
    int n = 0;
    const byte *p = str;
    int i;

    for (i = 0; i < len; ++i) {
	int c = str[i];

	fprintf(out,
		(c < 32 || c >= 127 ? "%d," :
		 c == '\'' || c == '\\' ? "'\\%c'," : "'%c',"),
		c);
	if (++n == 15) {
	    fputs("\n", out);
	    n = 0;
	}
    }
    if (n != 0)
	fputc('\n', out);
}
private void
ws(FILE * out, const byte *str, int len, bool to_c)
{
    if (to_c)
	wsc(out, str, len);
    else
	fwrite(str, 1, len, out);
}
private void
wl(FILE * out, const char *str, bool to_c)
{
    ws(out, (const byte *)str, strlen(str), to_c);
    ws(out, (const byte *)"\n", 1, to_c);
}

/*
 * Strip whitespace and comments from a line of PostScript code as possible.
 * Return a pointer to any string that remains, or NULL if none.
 * Note that this may store into the string.
 */
private char *
doit(char *line)
{
    char *str = line;
    char *from;
    char *to;
    int in_string = 0;

    while (*str == ' ' || *str == '\t')		/* strip leading whitespace */
	++str;
    if (*str == 0)		/* all whitespace */
	return NULL;
    if (!strncmp(str, "%END", 4))	/* keep these for .skipeof */
	return str;
    if (str[0] == '%')		/* comment line */
	return NULL;
    /*
     * Copy the string over itself removing:
     *  - All comments not within string literals;
     *  - Whitespace adjacent to []{};
     *  - Whitespace before /(<;
     *  - Whitespace after )>.
     */
    for (to = from = str; (*to = *from) != 0; ++from, ++to) {
	switch (*from) {
	    case '%':
		if (!in_string)
		    break;
		continue;
	    case ' ':
	    case '\t':
		if (to > str && !in_string && strchr(" \t>[]{})", to[-1]))
		    --to;
		continue;
	    case '(':
		++in_string;
	    case '<':
	    case '/':
	    case '[':
	    case ']':
	    case '{':
	    case '}':
		if (to > str && !in_string && strchr(" \t", to[-1]))
		    *--to = *from;
		continue;
	    case ')':
		--in_string;
		continue;
	    case '\\':
		if (from[1] == '\\' || from[1] == '(' || from[1] == ')')
		    *++to = *++from;
		continue;
	    default:
		continue;
	}
	break;
    }
    /* Strip trailing whitespace. */
    while (to > str && (to[-1] == ' ' || to[-1] == '\t'))
	--to;
    *to = 0;
    return str;
}

/* Copy a hex string to the output as a binary token. */
private void
hex_string_to_binary(FILE *out, FILE *in, bool to_c)
{
#define MAX_STR 0xffff	/* longest possible PostScript string token */
    byte *strbuf = (byte *)malloc(MAX_STR);
    byte *q = strbuf;
    int c, digit;
    bool which = false;
    int len;
    byte prefix[3];

    if (strbuf == 0) {
	fprintf(stderr, "Unable to allocate string token buffer.\n");
	exit(1);
    }
    while ((c = getc(in)) >= 0) {
	if (isxdigit(c)) {
	    c -= (isdigit(c) ? '0' : islower(c) ? 'a' : 'A');
	    if ((which = !which)) {
		if (q == strbuf + MAX_STR) {
		    fprintf(stderr, "Can't handle string token > %d bytes.\n",
			    MAX_STR);
		    exit(1);
		}
		*q++ = c << 4;
	    } else
		q[-1] += c;
	} else if (isspace(c))
	    continue;
	else if (c == '>')
	    break;
	else
	    fprintf(stderr, "Unknown character in ASCIIHex string: %c\n", c);
    }
    len = q - strbuf;
    if (len <= 255) {
	prefix[0] = 142;
	prefix[1] = (byte)len;
	ws(out, prefix, 2, to_c);
    } else {
	prefix[0] = 143;
	prefix[1] = (byte)(len >> 8);
	prefix[2] = (byte)len;
	ws(out, prefix, 3, to_c);
    }
    ws(out, strbuf, len, to_c);
    free((char *)strbuf);
#undef MAX_STR
}

/* Merge a file from input to output. */
private void
flush_buf(FILE * out, char *buf, bool to_c)
{
    if (buf[0] != 0) {
	wl(out, buf, to_c);
	buf[0] = 0;
    }
}
private void
mergefile(const char *prefix, const char *inname, FILE * in, FILE * config,
	  FILE * out, bool to_c)
{
    char line[LINE_SIZE + 1];
    char buf[LINE_SIZE + 1];
    char *str;
    int level = 1;
    bool first = true;

    buf[0] = 0;
    while (rl(in, line, LINE_SIZE)) {
	char psname[LINE_SIZE + 1];
	int nlines;

	if (!strncmp(line, "%% Replace ", 11) &&
	    sscanf(line + 11, "%d %s", &nlines, psname) == 2
	    ) {
	    flush_buf(out, buf, to_c);
	    while (nlines-- > 0)
		rl(in, line, LINE_SIZE);
	    if (psname[0] == '(') {
		FILE *ps;

		psname[strlen(psname) - 1] = 0;
		ps = prefix_open(prefix, psname + 1);
		if (ps == 0) {
		    fprintf(stderr, "Cannot open %s for reading.\n", psname + 1);
		    exit(1);
		}
		mergefile(prefix, psname + 1, ps, config, out, to_c);
	    } else if (!strcmp(psname, "INITFILES")) {
		/*
		 * We don't want to bind config.h into geninit, so
		 * we parse it ourselves at execution time instead.
		 */
		rewind(config);
		while (rl(config, psname, LINE_SIZE))
		    if (!strncmp(psname, "psfile_(\"", 9)) {
			FILE *ps;
			char *quote = strchr(psname + 9, '"');

			*quote = 0;
			ps = prefix_open(prefix, psname + 9);
			if (ps == 0) {
			    fprintf(stderr, "Cannot open %s for reading.\n", psname + 9);
			    exit(1);
			}
			mergefile(prefix, psname + 9, ps, config, out, to_c);
		    }
	    } else {
		fprintf(stderr, "Unknown %%%% Replace %d %s\n",
			nlines, psname);
		exit(1);
	    }
	} else if (!strcmp(line, "currentfile closefile")) {
	    /* The rest of the file is debugging code, stop here. */
	    break;
	} else {
	    int len;

	    str = doit(line);
	    if (str == 0)
		continue;
	    len = strlen(str);
	    if (first && len >= 2 && str[len - 2] == '<' &&
		(str[len - 1] == '<' || str[len - 1] == '~')
		) {
		wl(out, "currentobjectformat 1 setobjectformat\n", to_c);
		level = 2;
	    }
	    if (level > 1 && len > 0 && str[len - 1] == '<' &&
		(len < 2 || str[len - 2] != '<')
		) {
		/* Convert a hex string to a binary token. */
		flush_buf(out, buf, to_c);
		str[len - 1] = 0;
		wl(out, str, to_c);
		hex_string_to_binary(out, in, to_c);
		continue;
	    }
	    if (buf[0] != '%' &&	/* i.e. not special retained comment */
		strlen(buf) + len < LINE_SIZE &&
		(strchr("(/[]{}", str[0]) ||
		 (buf[0] != 0 && strchr(")[]{}", buf[strlen(buf) - 1])))
		)
		strcat(buf, str);
	    else {
		flush_buf(out, buf, to_c);
		strcpy(buf, str);
	    }
	    first = false;
	}
    }
    flush_buf(out, buf, to_c);
    if (level > 1)
	wl(out, "setobjectformat", to_c);
    fprintf(stderr, "%s: %ld bytes, output pos = %ld\n",
	    inname, ftell(in), ftell(out));
    fclose(in);
}

/* Merge and produce a C file. */
private void
merge_to_c(const char *prefix, const char *inname, FILE * in, FILE * config,
	   FILE * out)
{
    char line[LINE_SIZE + 1];

    fputs("/*\n", out);
    while ((rl(in, line, LINE_SIZE), line[0]))
	fprintf(out, "%s\n", line);
    fputs("*/\n", out);
    fputs("\n", out);
    fputs("/* Pre-compiled interpreter initialization string. */\n", out);
    fputs("\n", out);
    fputs("const unsigned char gs_init_string[] = {\n", out);
    mergefile(prefix, inname, in, config, out, true);
    fputs("10};\n", out);
    fputs("const unsigned int gs_init_string_sizeof = sizeof(gs_init_string);\n", out);
}

/* Merge and produce a PostScript file. */
private void
merge_to_ps(const char *prefix, const char *inname, FILE * in, FILE * config,
	    FILE * out)
{
    char line[LINE_SIZE + 1];

    while ((rl(in, line, LINE_SIZE), line[0]))
	fprintf(out, "%s\n", line);
    mergefile(prefix, inname, in, config, out, false);
}