/*
 * Copyright (c) 1983, 1993
 *	The Regents of the University of California.  All rights reserved.
 *
 * This code is derived from software contributed to Berkeley by
 * Asa Romberger and Jerry Berkman.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. Neither the name of the University nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 */

#include <sys/cdefs.h>
#ifndef lint
__COPYRIGHT("@(#) Copyright (c) 1983, 1993\
 The Regents of the University of California.  All rights reserved.");
#endif /* not lint */

#ifndef lint
#if 0
static char sccsid[] = "from: @(#)fsplit.c	8.1 (Berkeley) 6/6/93";
#else
__RCSID("$NetBSD: fsplit.c,v 1.26 2008/11/23 09:13:20 dholland Exp $");
#endif
#endif /* not lint */

#include <sys/types.h>
#include <sys/stat.h>

#include <assert.h>
#include <ctype.h>
#include <err.h>
#include <stdbool.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>

/*
 *	usage:		fsplit [-e efile] ... [file]
 *
 *	split single file containing source for several fortran programs
 *		and/or subprograms into files each containing one
 *		subprogram unit.
 *	each separate file will be named using the corresponding subroutine,
 *		function, block data or program name if one is found; otherwise
 *		the name will be of the form mainNNN.f or blkdtaNNN.f .
 *		If a file of that name exists, it is saved in a name of the
 *		form zzz000.f .
 *	If -e option is used, then only those subprograms named in the -e
 *		option are split off; e.g.:
 *			fsplit -esub1 -e sub2 prog.f
 *		isolates sub1 and sub2 in sub1.f and sub2.f.  The space
 *		after -e is optional.
 *
 *	Modified Feb., 1983 by Jerry Berkman, Computing Services, U.C. Berkeley.
 *		- added comments
 *		- more function types: double complex, character*(*), etc.
 *		- fixed minor bugs
 *		- instead of all unnamed going into zNNN.f, put mains in
 *		  mainNNN.f, block datas in blkdtaNNN.f, dups in zzzNNN.f .
 */

#define BSZ 512
static char buf[BSZ];
static FILE *ifp;

static char x[] = "zzz000.f";
static char mainp[] = "main000.f";
static char blkp[] = "blkdta000.f";

static void badparms(void);
static const char *functs(const char *);
static int getline(void);
static void get_name(char *, int);
static int lend(void);
static int lname(char *, size_t);
static const char *look(const char *, const char *);
static int saveit(const char *);
static int scan_name(char *, size_t, const char *);
static const char *skiplab(const char *);
static const char *skipws(const char *);

struct extract {
	bool found;
	char *name;
};

#define MAXEXTONLY 100
static struct extract extonly[MAXEXTONLY];
static int numextonly = 0;

int
main(int argc, char **argv)
{
	FILE *ofp;	/* output file */
	int rv;		/* 1 if got card in output file, 0 otherwise */
	int nflag;	/* 1 if got name of subprog., 0 otherwise */
	int retval, i, ch;
	char name[80];

	while ((ch = getopt(argc, argv, "e:")) != -1) {
		switch (ch) {
		    case 'e':
			if (numextonly >= MAXEXTONLY) {
				errx(1, "Too many -e options");
			}
			extonly[numextonly].name = optarg;
			extonly[numextonly].found = false;
			numextonly++;
			break;
		    default:
			badparms();
			break;
		}
	}

	if (argc > 2) {
		badparms();
	} else if (argc == 2) {
		if ((ifp = fopen(argv[1], "r")) == NULL) {
			err(1, "%s", argv[1]);
		}
	} else {
		ifp = stdin;
	}

	for (;;) {
		/*
		 * Look for a temp file that doesn't correspond to an
		 * existing file.
		 */

		get_name(x, 3);
		ofp = fopen(x, "w");
		if (ofp == NULL) {
			err(1, "%s", x);
		}
		nflag = 0;
		rv = 0;
		while (getline() > 0) {
			rv = 1;
			fprintf(ofp, "%s", buf);
			/* look for an 'end' statement */
			if (lend()) {
				break;
			}
			/* if no name yet, try and find one */
			if (nflag == 0) {
				nflag = lname(name, sizeof(name));
			}
		}
		fclose(ofp);
		if (rv == 0) {
			/* no lines in file, forget the file */
			unlink(x);
			retval = 0;
			for (i = 0; i < numextonly; i++) {
				if (!extonly[i].found) {
					retval = 1;
					warnx("%s not found", extonly[i].name);
				}
			}
			exit(retval);
		}
		if (nflag) {
			/* rename the file */
			if (saveit(name)) {
				struct stat sbuf;

				if (stat(name, &sbuf) < 0) {
					if (rename(x, name) < 0) {
						warn("%s: rename", x);
						printf("%s left in %s\n",
						    name, x);
					} else {
						printf("%s\n", name);
					}
					continue;
				} else if (strcmp(name, x) == 0) {
					printf("%s\n", x);
					continue;
				}
				printf("%s already exists, put in %s\n",
				    name, x);
				continue;
			} else {
				unlink(x);
				continue;
			}
		}
		if (numextonly == 0) {
			printf("%s\n", x);
		} else {
			unlink(x);
		}
	}
}

static void
badparms(void)
{
	err(1, "Usage: fsplit [-e efile] ... [file]");
}

static int
saveit(const char *name)
{
	int i;
	char fname[50];
	size_t fnamelen;

	if (numextonly == 0) {
		return 1;
	}
	strlcpy(fname, name, sizeof(fname));
	fnamelen = strlen(fname);
	assert(fnamelen > 2);
	assert(fname[fnamelen-2] = '.');
	assert(fname[fnamelen-1] = 'f');
	fname[fnamelen-2] = '\0';

	for (i = 0; i < numextonly; i++) {
		if (strcmp(fname, extonly[i].name) == 0) {
			extonly[i].found = true;
			return 1;
		}
	}
	return 0;
}

static void
get_name(char *name, int letters)
{
	struct stat sbuf;
	char *ptr;

	while (stat(name, &sbuf) >= 0) {
		for (ptr = name + letters + 2; ptr >= name + letters; ptr--) {
			(*ptr)++;
			if (*ptr <= '9')
				break;
			*ptr = '0';
		}
		if (ptr < name + letters) {
			errx(1, "Ran out of file names.\n");
		}
	}
}

static int
getline(void)
{
	char *ptr;

	for (ptr = buf; ptr < &buf[BSZ]; ) {
		*ptr = getc(ifp);
		if (feof(ifp))
			return -1;
		if (*ptr++ == '\n') {
			*ptr = '\0';
			return 1;
		}
	}
	while (getc(ifp) != '\n' && feof(ifp) == 0) {
		/* nothing */
	}
	warnx("Line truncated to %d characters.", BSZ);
	return 1;
}

/*
 * Return 1 for 'end' alone on card (up to col. 72), 0 otherwise.
 */
static int
lend(void)
{
	const char *p;

	if ((p = skiplab(buf)) == 0) {
		return 0;
	}
	p = skipws(p);
	if (*p != 'e' && *p != 'E') {
		return 0;
	}
	p++;
	p = skipws(p);
	if (*p != 'n' && *p != 'N') {
		return 0;
	}
	p++;
	p = skipws(p);
	if (*p != 'd' && *p != 'D') {
		return 0;
	}
	p++;
	p = skipws(p);
	if (p - buf >= 72 || *p == '\n') {
		return 1;
	}
	return 0;
}

/*
 * check for keywords for subprograms
 * return 0 if comment card, 1 if found
 * name and put in arg string. invent name for unnamed
 * block datas and main programs.
 */
static int
lname(char *s, size_t l)
{
#define LINESIZE 80
	const char *ptr, *p;
	char line[LINESIZE], *iptr = line;

	/* first check for comment cards */
	if (buf[0] == 'c' || buf[0] == 'C' || buf[0] == '*') {
		return 0;
	}
	ptr = skipws(buf);
	if (*ptr == '\n') {
		return 0;
	}

	ptr = skiplab(buf);
	if (ptr == NULL) {
		return 0;
	}

	/*  copy to buffer and converting to lower case */
	p = ptr;
	while (*p && p <= &buf[71] ) {
	   *iptr = tolower((unsigned char)*p);
	   iptr++;
	   p++;
	}
	*iptr = '\n';

	if ((ptr = look(line, "subroutine")) != NULL ||
	    (ptr = look(line, "function")) != NULL ||
	    (ptr = functs(line)) != NULL) {
		if (scan_name(s, l, ptr)) {
			return 1;
		}
		strlcpy(s, x, l);
	} else if ((ptr = look(line, "program")) != NULL) {
		if (scan_name(s, l, ptr)) {
			return 1;
		}
		get_name(mainp, 4);
		strlcpy(s, mainp, l);
	} else if ((ptr = look(line, "blockdata")) != NULL) {
		if (scan_name(s, l, ptr)) {
			return 1;
		}
		get_name(blkp, 6);
		strlcpy(s, blkp, l);
	} else if ((ptr = functs(line)) != NULL) {
		if (scan_name(s, l, ptr)) {
			return 1;
		}
		strlcpy(s, x, l);
	} else {
		get_name(mainp, 4);
		strlcpy(s, mainp, l);
	}
	return 1;
}

static int
scan_name(char *s, size_t smax, const char *ptr)
{
	char *sptr;
	size_t sptrmax;

	/* scan off the name */
	ptr = skipws(ptr);
	sptr = s;
	sptrmax = smax - 3;
	while (*ptr != '(' && *ptr != '\n') {
		if (*ptr != ' ' && *ptr != '\t' && *ptr != '/') {
			if (sptrmax == 0) {
				/* Not sure this is the right thing, so warn */
				warnx("Output name too long; truncated");
				break;
			}
			*sptr++ = *ptr;
			sptrmax--;
		}
		ptr++;
	}

	if (sptr == s) {
		return 0;
	}

	*sptr++ = '.';
	*sptr++ = 'f';
	*sptr++ = '\0';
	return 1;
}

/*
 * look for typed functions such as: real*8 function,
 * character*16 function, character*(*) function
 */
static const char *
functs(const char *p)
{
        const char *ptr;

        if ((ptr = look(p, "character")) != NULL ||
	    (ptr = look(p, "logical")) != NULL ||
	    (ptr = look(p, "real")) != NULL ||
	    (ptr = look(p, "integer")) != NULL ||
	    (ptr = look(p, "doubleprecision")) != NULL ||
	    (ptr = look(p, "complex")) != NULL ||
	    (ptr = look(p, "doublecomplex")) != NULL) {
                while (*ptr == ' ' || *ptr == '\t' || *ptr == '*'
		    || (*ptr >= '0' && *ptr <= '9')
		    || *ptr == '(' || *ptr == ')') {
			ptr++;
		}
		ptr = look(ptr, "function");
		return ptr;
	}
        else {
                return NULL;
	}
}

/*
 * if first 6 col. blank, return ptr to col. 7,
 * if blanks and then tab, return ptr after tab,
 * else return NULL (labelled statement, comment or continuation)
 */
static const char *
skiplab(const char *p)
{
	const char *ptr;

	for (ptr = p; ptr < &p[6]; ptr++) {
		if (*ptr == ' ')
			continue;
		if (*ptr == '\t') {
			ptr++;
			break;
		}
		return NULL;
	}
	return ptr;
}

/*
 * return NULL if m doesn't match initial part of s;
 * otherwise return ptr to next char after m in s
 */
static const char *
look(const char *s, const char *m)
{
	const char *sp, *mp;

	sp = s; mp = m;
	while (*mp) {
		sp = skipws(sp);
		if (*sp++ != *mp++)
			return NULL;
	}
	return sp;
}

static const char *
skipws(const char *p)
{
	while (*p == ' ' || *p == '\t') {
		p++;
	}
	return p;
}