utils.c

/* $Id$ */
/**************************************************************************
 *   utils.c                                                              *
 *                                                                        *
 *   Copyright (C) 1999-2004 Chris Allegretta                             *
 *   Copyright (C) 2005 David Lawrence Ramsey                             *
 *   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, 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., 51 Franklin St, Fifth Floor, Boston, MA            *
 *   02110-1301, USA.                                                     *
 *                                                                        *
 **************************************************************************/

#include "proto.h"

#include <string.h>
#include <stdio.h>
#include <unistd.h>
#include <pwd.h>
#include <ctype.h>
#include <errno.h>

/* Return the number of decimal digits in n. */
int digits(size_t n)
{
    int i = 1;

    while (n > 10) {
	n /= 10;
	i++;
    }

    return i;
}

/* Return the user's home directory.  We use $HOME, and if that fails,
 * we fall back on getpwuid(). */
void get_homedir(void)
{
    if (homedir == NULL) {
	const char *homenv = getenv("HOME");

	if (homenv == NULL) {
	    const struct passwd *userage = getpwuid(geteuid());

	    if (userage != NULL)
		homenv = userage->pw_dir;
	}
	homedir = mallocstrcpy(NULL, homenv);
    }
}

/* Read a ssize_t from str, and store it in *val (if val is not NULL).
 * On error, we return FALSE and don't change *val.  Otherwise, we
 * return TRUE. */
bool parse_num(const char *str, ssize_t *val)
{
    char *first_error;
    ssize_t j;

    assert(str != NULL);

    j = (ssize_t)strtol(str, &first_error, 10);

    if (errno == ERANGE || *str == '\0' || *first_error != '\0')
	return FALSE;

    if (val != NULL)
	*val = j;

    return TRUE;
}

/* Read two ssize_t's, separated by a comma, from str, and store them in
 * *line and *column (if they're not both NULL).  Return FALSE on error,
 * or TRUE otherwise. */
bool parse_line_column(const char *str, ssize_t *line, ssize_t *column)
{
    bool retval = TRUE;
    const char *comma;

    assert(str != NULL);

    comma = strchr(str, ',');

    if (comma != NULL && column != NULL) {
	if (!parse_num(comma + 1, column))
	    retval = FALSE;
    }

    if (line != NULL) {
	if (comma != NULL) {
	    char *str_line = mallocstrncpy(NULL, str, comma - str + 1);
	    str_line[comma - str] = '\0';

	    if (str_line[0] != '\0' && !parse_num(str_line, line))
		retval = FALSE;

	    free(str_line);
	} else if (!parse_num(str, line))
	    retval = FALSE;
    }

    return retval;
}

/* Fix the memory allocation for a string. */
void align(char **str)
{
    assert(str != NULL);

    if (*str != NULL)
	*str = charealloc(*str, strlen(*str) + 1);
}

/* Null a string at a certain index and align it. */
void null_at(char **data, size_t index)
{
    assert(data != NULL);

    *data = charealloc(*data, index + 1);
    (*data)[index] = '\0';
}

/* For non-null-terminated lines.  A line, by definition, shouldn't
 * normally have newlines in it, so encode its nulls as newlines. */
void unsunder(char *str, size_t true_len)
{
    assert(str != NULL);

    for (; true_len > 0; true_len--, str++) {
	if (*str == '\0')
	    *str = '\n';
    }
}

/* For non-null-terminated lines.  A line, by definition, shouldn't
 * normally have newlines in it, so decode its newlines into nulls. */
void sunder(char *str)
{
    assert(str != NULL);

    for (; *str != '\0'; str++) {
	if (*str == '\n')
	    *str = '\0';
    }
}

/* These functions, ngetline() (originally getline()) and ngetdelim()
 * (originally getdelim()), were adapted from GNU mailutils 0.5
 * (mailbox/getline.c).  Here is the notice from that file, after
 * converting to the GPL via LGPL clause 3, and with the Free Software
 * Foundation's address updated:
 *
 * GNU Mailutils -- a suite of utilities for electronic mail
 * Copyright (C) 1999, 2000, 2001, 2002 Free Software Foundation, Inc.
 *
 * This library 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 library 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 library; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
 * 02110-1301, USA. */

#if !defined(NANO_TINY) && defined(ENABLE_NANORC)
#ifndef HAVE_GETLINE
/* This function is equivalent to getline(). */
ssize_t ngetline(char **lineptr, size_t *n, FILE *stream)
{
    return getdelim(lineptr, n, '\n', stream);
}
#endif

#ifndef HAVE_GETDELIM
/* This function is equivalent to getdelim(). */
ssize_t ngetdelim(char **lineptr, size_t *n, int delim, FILE *stream)
{
    size_t indx = 0;
    int c;

    /* Sanity checks. */
    assert(lineptr != NULL && n != NULL && stream != NULL);

    /* Allocate the line the first time. */
    if (*lineptr == NULL) {
	*lineptr = charalloc(MAX_BUF_SIZE);
	*n = MAX_BUF_SIZE;
    }

    while ((c = getc(stream)) != EOF) {
	/* Check if more memory is needed. */
	if (indx >= *n) {
	    *lineptr = charealloc(*lineptr, *n + MAX_BUF_SIZE);
	    *n += MAX_BUF_SIZE;
	}

	/* Put the result in the line. */
	(*lineptr)[indx++] = (char)c;

	/* Bail out. */
	if (c == delim)
	    break;
    }

    /* Make room for the null character. */
    if (indx >= *n) {
	*lineptr = charealloc(*lineptr, *n + MAX_BUF_SIZE);
	*n += MAX_BUF_SIZE;
    }

    /* Null terminate the buffer. */
    null_at(lineptr, indx++);
    *n = indx;

    /* The last line may not have the delimiter.  We have to return what
     * we got, and the error will be seen on the next iteration. */
    return (c == EOF && (indx - 1) == 0) ? -1 : indx - 1;
}
#endif
#endif /* !NANO_TINY && ENABLE_NANORC */

#ifdef HAVE_REGEX_H
#ifdef BROKEN_REGEXEC
/* Work around a potential segfault in glibc 2.2.3's regexec(). */
int safe_regexec(const regex_t *preg, const char *string, size_t nmatch,
	regmatch_t pmatch[], int eflags)
{
    if (string != NULL && *string != '\0')
	return regexec(preg, string, nmatch, pmatch, eflags);

    return REG_NOMATCH;
}
#endif

/* Do the compiled regex in preg and the regex in string match the
 * beginning or end of a line? */
bool regexp_bol_or_eol(const regex_t *preg, const char *string)
{
    return (regexec(preg, string, 0, NULL, 0) == 0 &&
	regexec(preg, string, 0, NULL, REG_NOTBOL | REG_NOTEOL) ==
	REG_NOMATCH);
}
#endif /* HAVE_REGEX_H */

#ifndef DISABLE_SPELLER
/* Is the word starting at position pos in buf a whole word? */
bool is_whole_word(size_t pos, const char *buf, const char *word)
{
    char *p = charalloc(mb_cur_max()), *r = charalloc(mb_cur_max());
    size_t word_end = pos + strlen(word);
    bool retval;

    assert(buf != NULL && pos <= strlen(buf) && word != NULL);

    parse_mbchar(buf + move_mbleft(buf, pos), p, NULL);
    parse_mbchar(buf + word_end, r, NULL);

    /* If we're at the beginning of the line or the character before the
     * word isn't a non-punctuation "word" character, and if we're at
     * the end of the line or the character after the word isn't a
     * non-punctuation "word" character, we have a whole word. */
    retval = (pos == 0 || !is_word_mbchar(p, FALSE)) &&
	(word_end == strlen(buf) || !is_word_mbchar(r, FALSE));

    free(p);
    free(r);

    return retval;
}
#endif /* !DISABLE_SPELLER */

/* If we are searching backwards, we will find the last match that
 * starts no later than start.  Otherwise we find the first match
 * starting no earlier than start.  If we are doing a regexp search, we
 * fill in the global variable regmatches with at most 9 subexpression
 * matches.  Also, all .rm_so elements are relative to the start of the
 * whole match, so regmatches[0].rm_so == 0. */
const char *strstrwrapper(const char *haystack, const char *needle,
	const char *start)
{
    /* start can be 1 character before the start or after the end of the
     * line.  In either case, we just say no match was found. */
    if ((start > haystack && *(start - 1) == '\0') || start < haystack)
	return NULL;

    assert(haystack != NULL && needle != NULL && start != NULL);

#ifdef HAVE_REGEX_H
    if (ISSET(USE_REGEXP)) {
#ifndef NANO_TINY
	if (ISSET(BACKWARDS_SEARCH)) {
	    if (regexec(&search_regexp, haystack, 1, regmatches,
		0) == 0 && haystack + regmatches[0].rm_so <= start) {
		const char *retval = haystack + regmatches[0].rm_so;

		/* Search forward until there are no more matches. */
		while (regexec(&search_regexp, retval + 1, 1,
			regmatches, REG_NOTBOL) == 0 &&
			retval + regmatches[0].rm_so + 1 <= start)
		    retval += regmatches[0].rm_so + 1;
		/* Finally, put the subexpression matches in global
		 * variable regmatches.  The REG_NOTBOL flag doesn't
		 * matter now. */
		regexec(&search_regexp, retval, 10, regmatches, 0);
		return retval;
	    }
	} else
#endif /* !NANO_TINY */
	if (regexec(&search_regexp, start, 10, regmatches,
		(start > haystack) ? REG_NOTBOL : 0) == 0) {
	    const char *retval = start + regmatches[0].rm_so;

	    regexec(&search_regexp, retval, 10, regmatches, 0);
	    return retval;
	}
	return NULL;
    }
#endif /* HAVE_REGEX_H */
#if !defined(NANO_TINY) || !defined(DISABLE_SPELLER)
    if (ISSET(CASE_SENSITIVE)) {
#ifndef NANO_TINY
	if (ISSET(BACKWARDS_SEARCH))
	    return revstrstr(haystack, needle, start);
	else
#endif
	    return strstr(start, needle);
    }
#endif /* !DISABLE_SPELLER || !NANO_TINY */
#ifndef NANO_TINY
    else if (ISSET(BACKWARDS_SEARCH))
	return mbrevstrcasestr(haystack, needle, start);
#endif
    return mbstrcasestr(start, needle);
}

/* This is a wrapper for the perror() function.  The wrapper takes care
 * of curses, calls perror() (which writes to stderr), and then
 * refreshes the screen.  Note that nperror() causes the window to
 * flicker once. */
void nperror(const char *s)
{
    /* Leave curses mode and go to the terminal. */
    if (endwin() != ERR) {
	perror(s);		/* Print the error. */
	total_refresh();	/* Return to curses and refresh. */
    }
}

/* This is a wrapper for the malloc() function that properly handles
 * things when we run out of memory.  Thanks, BG, many people have been
 * asking for this... */
void *nmalloc(size_t howmuch)
{
    void *r = malloc(howmuch);

    if (r == NULL && howmuch != 0)
	die(_("nano is out of memory!"));

    return r;
}

/* This is a wrapper for the realloc() function that properly handles
 * things when we run out of memory. */
void *nrealloc(void *ptr, size_t howmuch)
{
    void *r = realloc(ptr, howmuch);

    if (r == NULL && howmuch != 0)
	die(_("nano is out of memory!"));

    return r;
}

/* Copy the first n characters of one malloc()ed string to another
 * pointer.  Should be used as: "dest = mallocstrncpy(dest, src,
 * n);". */
char *mallocstrncpy(char *dest, const char *src, size_t n)
{
    if (src == NULL)
	src = "";

    if (src != dest)
	free(dest);

    dest = charalloc(n);
    strncpy(dest, src, n);

    return dest;
}

/* Copy one malloc()ed string to another pointer.  Should be used as:
 * "dest = mallocstrcpy(dest, src);". */
char *mallocstrcpy(char *dest, const char *src)
{
    return mallocstrncpy(dest, src, (src == NULL) ? 1 :
	strlen(src) + 1);
}

/* Free the malloc()ed string at dest and return the malloc()ed string
 * at src.  Should be used as: "answer = mallocstrassn(answer,
 * real_dir_from_tilde(answer));". */
char *mallocstrassn(char *dest, char *src)
{
    free(dest);
    return src;
}

/* nano scrolls horizontally within a line in chunks.  Return the column
 * number of the first character displayed in the edit window when the
 * cursor is at the given column.  Note that (0 <= column -
 * get_page_start(column) < COLS). */
size_t get_page_start(size_t column)
{
    if (column == 0 || column < COLS - 1)
	return 0;
    else if (COLS > 8)
	return column - 7 - (column - 7) % (COLS - 8);
    else
	return column - (COLS - 2);
}

/* Return the placewewant associated with current_x, i.e, the zero-based
 * column position of the cursor.  The value will be no smaller than
 * current_x. */
size_t xplustabs(void)
{
    return strnlenpt(openfile->current->data, openfile->current_x);
}

/* Return the index in s of the character displayed at the given column,
 * i.e, the largest value such that strnlenpt(s, actual_x(s, column)) <=
 * column. */
size_t actual_x(const char *s, size_t column)
{
    size_t i = 0;
	/* The position in s, returned. */
    size_t len = 0;
	/* The screen display width to s[i]. */

    assert(s != NULL);

    while (*s != '\0') {
	int s_len = parse_mbchar(s, NULL, &len);

	if (len > column)
	    break;

	i += s_len;
	s += s_len;
    }

    return i;
}

/* A strnlen() with tabs and multicolumn characters factored in, similar
 * to xplustabs().  How many columns wide are the first maxlen characters
 * of s? */
size_t strnlenpt(const char *s, size_t maxlen)
{
    size_t len = 0;
	/* The screen display width to s[i]. */

    if (maxlen == 0)
	return 0;

    assert(s != NULL);

    while (*s != '\0') {
	int s_len = parse_mbchar(s, NULL, &len);

	s += s_len;

	if (maxlen <= s_len)
	    break;

	maxlen -= s_len;
    }

    return len;
}

/* A strlen() with tabs and multicolumn characters factored in, similar
 * to xplustabs().  How many columns wide is s? */
size_t strlenpt(const char *s)
{
    return strnlenpt(s, (size_t)-1);
}

/* Append a new magicline to filebot. */
void new_magicline(void)
{
    openfile->filebot->next = (filestruct *)nmalloc(sizeof(filestruct));
    openfile->filebot->next->data = mallocstrcpy(NULL, "");
    openfile->filebot->next->prev = openfile->filebot;
    openfile->filebot->next->next = NULL;
    openfile->filebot->next->lineno = openfile->filebot->lineno + 1;
    openfile->filebot = openfile->filebot->next;
    openfile->totsize++;
}

#ifndef NANO_TINY
/* Remove the magicline from filebot, if there is one and it isn't the
 * only line in the file.  Assume that edittop and current are not at
 * filebot. */
void remove_magicline(void)
{
    if (openfile->filebot->data[0] == '\0' &&
	openfile->filebot != openfile->fileage) {
	assert(openfile->filebot != openfile->edittop && openfile->filebot != openfile->current);

	openfile->filebot = openfile->filebot->prev;
	free_filestruct(openfile->filebot->next);
	openfile->filebot->next = NULL;
	openfile->totsize--;
    }
}

/* Set top_x and bot_x to the top and bottom x-coordinates of the mark,
 * respectively, based on the locations of top and bot.  If
 * right_side_up isn't NULL, set it to TRUE If the mark begins with
 * (mark_begin, mark_begin_x) and ends with (current, current_x), or
 * FALSE otherwise. */
void mark_order(const filestruct **top, size_t *top_x, const filestruct
	**bot, size_t *bot_x, bool *right_side_up)
{
    assert(top != NULL && top_x != NULL && bot != NULL && bot_x != NULL);

    if ((openfile->current->lineno == openfile->mark_begin->lineno &&
	openfile->current_x > openfile->mark_begin_x) ||
	openfile->current->lineno > openfile->mark_begin->lineno) {
	*top = openfile->mark_begin;
	*top_x = openfile->mark_begin_x;
	*bot = openfile->current;
	*bot_x = openfile->current_x;
	if (right_side_up != NULL)
	    *right_side_up = TRUE;
    } else {
	*bot = openfile->mark_begin;
	*bot_x = openfile->mark_begin_x;
	*top = openfile->current;
	*top_x = openfile->current_x;
	if (right_side_up != NULL)
	    *right_side_up = FALSE;
    }
}
#endif

/* Calculate the number of characters between begin and end, and return
 * it. */
size_t get_totsize(const filestruct *begin, const filestruct *end)
{
    size_t totsize = 0;
    const filestruct *f;

    /* Go through the lines from begin to end->prev, if we can. */
    for (f = begin; f != end && f != NULL; f = f->next) {
	/* Count the number of characters on this line. */
	totsize += mbstrlen(f->data);

	/* Count the newline if we have one. */
	if (f->next != NULL)
	    totsize++;
    }

    /* Go through the line at end, if we can. */
    if (f != NULL) {
	/* Count the number of characters on this line. */
	totsize += mbstrlen(f->data);

	/* Count the newline if we have one. */
	if (f->next != NULL)
	    totsize++;
    }

    return totsize;
}

#ifndef NDEBUG
/* Return what the current line number should be, starting at edittop
 * and ending at fileptr. */
int check_linenumbers(const filestruct *fileptr)
{
    int check_line = 0;
    const filestruct *filetmp;

    for (filetmp = openfile->edittop; filetmp != fileptr;
	filetmp = filetmp->next)
	check_line++;

    return check_line;
}
#endif /* !NDEBUG */

#ifdef DEBUG
/* Dump the filestruct inptr to stderr. */
void dump_filestruct(const filestruct *inptr)
{
    if (inptr == openfile->fileage)
	fprintf(stderr, "Dumping file buffer to stderr...\n");
    else if (inptr == cutbuffer)
	fprintf(stderr, "Dumping cutbuffer to stderr...\n");
    else
	fprintf(stderr, "Dumping a buffer to stderr...\n");

    while (inptr != NULL) {
	fprintf(stderr, "(%ld) %s\n", (long)inptr->lineno, inptr->data);
	inptr = inptr->next;
    }
}

/* Dump the current buffer's filestruct to stderr in reverse. */
void dump_filestruct_reverse(void)
{
    const filestruct *fileptr = openfile->filebot;

    while (fileptr != NULL) {
	fprintf(stderr, "(%ld) %s\n", (long)fileptr->lineno,
		fileptr->data);
	fileptr = fileptr->prev;
    }
}
#endif /* DEBUG */