commit 1513c2b7662c1e7f475ce1c45b61f6a76d665c8b
Author: FRIGN <[email protected]>
Date: Sun Feb 8 21:24:22 2015 +0100
Refactor unexpand(1) code and manpage, adding tablist support
as already seen for expand(1), only twice as complicated.
diff --git a/README b/README
index 8400dad..daedbc3 100644
--- a/README
+++ b/README
@@ -73,7 +73,7 @@ The following tools are implemented ('*' == finished, '#' ==
UTF-8 support,
=* true yes none
=* tty yes none
=* uname yes none
-# unexpand yes none
+#* unexpand yes none
= uniq no -f, -s
= unlink yes none
= uudecode no -o
diff --git a/unexpand.1 b/unexpand.1
index 506d9ed..c5a44b9 100644
--- a/unexpand.1
+++ b/unexpand.1
@@ -1,30 +1,45 @@
-.Dd January 30, 2015
+.Dd February 8, 2015
.Dt UNEXPAND 1
.Os sbase
.Sh NAME
.Nm unexpand
-.Nd convert blanks to tabs
+.Nd unexpand spaces to tabs
.Sh SYNOPSIS
.Nm
.Op Fl a
-.Op Fl t Ar n
+.Op Fl t Ar tablist
.Op Ar file ...
.Sh DESCRIPTION
.Nm
-processes the named
-.Ar files
-or the standard input, writing the
-standard output with consecutive blanks (spaces and tabs) converted
-into tabs. Backspace characters are preserved into the output and
-decrement the column count for tab calculations.
+converts spaces to tabs in each
+.Ar file
+as specified in
+.Ar tablist .
+If no file is given,
+.Nm
+reads from stdin.
+.Pp
+Backspace characters are preserved and decrement the column count
+for tab calculations.
.Sh OPTIONS
.Bl -tag -width Ds
.It Fl a
-Convert blanks to tabs everywhere, not just at the start of lines.
-.It Fl t Ar n
-Set tab size to
-.Ar n
-spaces (default: 8).
+Convert spaces to tabs everywhere, not just at the start of lines.
+.It Fl t Ar tablist
+Specify tab size or tabstops.
+.Ar tablist
+is a list of one (in the former case) or multiple (in the latter case)
+strictly positive integers separated by ' ' or ','.
+.Pp
+The default
+.Ar tablist
+is "8".
.El
.Sh SEE ALSO
.Xr expand 1
+.Sh STANDARDS
+The
+.Nm
+utility is compliant with the
+.St -p1003.1-2008
+specification.
diff --git a/unexpand.c b/unexpand.c
index 839eac6..8f1fe07 100644
--- a/unexpand.c
+++ b/unexpand.c
@@ -1,5 +1,5 @@
/* See LICENSE file for copyright and license details. */
-#include <limits.h>
+#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <wchar.h>
@@ -7,114 +7,162 @@
#include "utf.h"
#include "util.h"
-static void unexpand(const char *, FILE *);
+static int aflag = 0;
+static size_t *tablist = NULL;
+static int tablistlen = 8;
-static int aflag = 0;
-static int tabsize = 8;
-
-static void
-usage(void)
+static size_t
+parselist(const char *s)
{
- eprintf("usage: %s [-a] [-t n] [file ...]\n", argv0);
+ size_t i;
+ char *p, *tmp;
+
+ tmp = estrdup(s);
+ for (i = 0; (p = strsep(&tmp, " ,")); i++) {
+ if (*p == '\0')
+ eprintf("empty field in tablist\n");
+ tablist = erealloc(tablist, (i + 1) * sizeof(*tablist));
+ tablist[i] = estrtonum(p, 1, MIN(LLONG_MAX, SIZE_MAX));
+ if (i > 0 && tablist[i - 1] >= tablist[i])
+ eprintf("tablist must be ascending\n");
+ }
+ tablist = erealloc(tablist, (i + 1) * sizeof(*tablist));
+ return i;
}
-int
-main(int argc, char *argv[])
+static void
+unexpandspan(size_t last, size_t col)
{
- FILE *fp;
- int ret = 0;
+ size_t off, i, j;
+ Rune r;
- ARGBEGIN {
- case 't':
- tabsize = estrtonum(EARGF(usage()), 0, INT_MAX);
- if (tabsize <= 0)
- eprintf("unexpand: invalid tabsize\n");
- /* Fallthrough: -t implies -a */
- case 'a':
- aflag = 1;
- break;
- default:
- usage();
- } ARGEND;
+ if (tablistlen == 1) {
+ i = 0;
+ off = last % tablist[i];
- if (argc == 0) {
- unexpand("<stdin>", stdin);
+ if ((col - last) + off >= tablist[i] && last < col)
+ last -= off;
+
+ r = '\t';
+ for (; last + tablist[i] <= col; last += tablist[i])
+ writerune("<stdout>", stdout, &r);
+ r = ' ';
+ for (; last < col; last++)
+ writerune("<stdout>", stdout, &r);
} else {
- for (; argc > 0; argc--, argv++) {
- if (!(fp = fopen(argv[0], "r"))) {
- weprintf("fopen %s:", argv[0]);
- ret = 1;
- continue;
- }
- unexpand(argv[0], fp);
- fclose(fp);
+ for (i = 0; i < tablistlen; i++)
+ if (col < tablist[i])
+ break;
+ for (j = 0; j < tablistlen; j++)
+ if (last < tablist[j])
+ break;
+ r = '\t';
+ for (; j < i; j++) {
+ writerune("<stdout>", stdout, &r);
+ last = tablist[j];
}
+ r = ' ';
+ for (; last < col; last++)
+ writerune("<stdout>", stdout, &r);
}
- return ret;
-}
-
-static void
-unexpandspan(unsigned int n, unsigned int col)
-{
- unsigned int off = (col-n) % tabsize;
- Rune r;
-
- if (n + off >= tabsize && n > 1)
- n += off;
- r = '\t';
- for (; n >= tabsize; n -= tabsize)
- writerune("<stdout>", stdout, &r);
- r = ' ';
- while (n--)
- writerune("<stdout>", stdout, &r);
}
static void
unexpand(const char *file, FILE *fp)
{
- unsigned int n = 0, col = 0;
Rune r;
+ size_t last = 0, col = 0, i;
int bol = 1;
- while (1) {
- if (!readrune(file, fp, &r))
- break;
-
+ while (readrune(file, fp, &r)) {
switch (r) {
case ' ':
- if (bol || aflag)
- n++;
+ if (!bol && !aflag)
+ last++;
col++;
break;
case '\t':
- if (bol || aflag)
- n += tabsize - col % tabsize;
- col += tabsize - col % tabsize;
+ if (tablistlen == 1) {
+ if (!bol && !aflag)
+ last += tablist[0] - col % tablist[0];
+ col += tablist[0] - col % tablist[0];
+ } else {
+ for (i = 0; i < tablistlen; i++)
+ if (col < tablist[i])
+ break;
+ if (!bol && !aflag)
+ last = tablist[i];
+ col = tablist[i];
+ }
break;
case '\b':
if (bol || aflag)
- unexpandspan(n, col);
+ unexpandspan(last, col);
col -= (col > 0);
- n = 0;
+ last = col;
bol = 0;
break;
case '\n':
if (bol || aflag)
- unexpandspan(n, col);
- n = col = 0;
+ unexpandspan(last, col);
+ last = col = 0;
bol = 1;
break;
default:
if (bol || aflag)
- unexpandspan(n, col);
- n = 0;
- col++;
+ unexpandspan(last, col);
+ last = ++col;
bol = 0;
+ break;
}
if ((r != ' ' && r != '\t') || (!aflag && !bol))
writerune("<stdout>", stdout, &r);
}
- if (n > 0 && (bol || aflag))
- unexpandspan(n, col);
+ if (last < col && (bol || aflag))
+ unexpandspan(last, col);
+}
+
+static void
+usage(void)
+{
+ eprintf("usage: %s [-a] [-t tablist] [file ...]\n", argv0);
+}
+
+int
+main(int argc, char *argv[])
+{
+ FILE *fp;
+ int ret = 0;
+ char *tl = "8";
+
+ ARGBEGIN {
+ case 't':
+ tl = EARGF(usage());
+ if (!*tl)
+ eprintf("tablist cannot be empty\n");
+ /* Fallthrough: -t implies -a */
+ case 'a':
+ aflag = 1;
+ break;
+ default:
+ usage();
+ } ARGEND;
+
+ tablistlen = parselist(tl);
+
+ if (argc == 0)
+ unexpand("<stdin>", stdin);
+ else {
+ for (; argc > 0; argc--, argv++) {
+ if (!(fp = fopen(argv[0], "r"))) {
+ weprintf("fopen %s:", argv[0]);
+ ret = 1;
+ continue;
+ }
+ unexpand(argv[0], fp);
+ fclose(fp);
+ }
+ }
+ return ret;
}
