* Copyright 2013 CE Strake <strake888 at gmail.com>
*
* See http://pubs.opengroup.org/onlinepubs/9699919799/utilities/grep.html
+ *
+ * TODO: -ABC
-USE_GREP(NEWTOY(grep, "ZzEFHabhinorsvwclqe*f*m#x[!wx][!EFw]", TOYFLAG_BIN))
+USE_GREP(NEWTOY(grep, "C#B#A#ZzEFHabhinorsvwclqe*f*m#x[!wx][!EFw]", TOYFLAG_BIN))
USE_EGREP(OLDTOY(egrep, grep, TOYFLAG_BIN))
USE_FGREP(OLDTOY(fgrep, grep, TOYFLAG_BIN))
bool "grep"
default y
help
- usage: grep [-EFivwcloqsHbhn] [-m MAX] [-e REGEX]... [-f REGFILE] [FILE]...
+ usage: grep [-EFivwcloqsHbhn] [-A NUM] [-m MAX] [-e REGEX]... [-f REGFILE] [FILE]...
Show lines matching regular expressions. If no -e, first argument is
regular expression to match. With no files (or "-" filename) read stdin.
-f File containing regular expressions to match.
match type:
- -E extended regex syntax -F fixed (match literal string)
- -i case insensitive -m stop after this many lines matched
- -r recursive (on dir) -v invert match
- -w whole word (implies -E) -x whole line
- -z input NUL terminated
+ -A Show NUM lines after -B Show NUM lines before match
+ -C NUM lines context (A+B) -E extended regex syntax
+ -F fixed (literal match) -i case insensitive
+ -m match MAX many lines -r recursive (on dir)
+ -v invert match -w whole word (implies -E)
+ -x whole line -z input NUL terminated
display modes: (default: matched line)
-c count of matching lines -l show matching filenames
long m;
struct arg_list *f;
struct arg_list *e;
+ long a;
+ long b;
+ long c;
- struct arg_list *regex;
+ char indelim, outdelim;
)
+// Emit line with various potential prefixes and delimiter
+static void outline(char *line, char dash, char *name, long lcount, long bcount,
+ int trim)
+{
+ if (name && (toys.optflags&FLAG_H)) printf("%s%c", name, dash);
+ if (!line || (lcount && (toys.optflags&FLAG_n)))
+ printf("%ld%c", lcount, line ? dash : TT.outdelim);
+ if (bcount && (toys.optflags&FLAG_b)) printf("%ld%c", bcount-1, dash);
+ if (line) xprintf("%.*s%c", trim, line, TT.outdelim);
+}
+
+// Show matches in one file
static void do_grep(int fd, char *name)
{
+ struct double_list *dlb = 0;
FILE *file = fdopen(fd, "r");
- long offset = 0;
- int lcount = 0, mcount = 0, which = toys.optflags & FLAG_w ? 2 : 0;
- char indelim = '\n' * !(toys.optflags&FLAG_z),
- outdelim = '\n' * !(toys.optflags&FLAG_Z);
+ long lcount = 0, mcount = 0, offset = 0, after = 0, before = 0;
+ char *bars = 0;
if (!fd) name = "(standard input)";
if (!file) {
- perror_msg("%s", name);
+ perror_msg_raw(name);
return;
}
+ // Loop through lines of input
for (;;) {
char *line = 0, *start;
- regmatch_t matches[3];
+ regmatch_t matches;
size_t unused;
long len;
int mmatch = 0;
lcount++;
- if (0 > (len = getdelim(&line, &unused, indelim, file))) break;
- if (line[len-1] == indelim) line[len-1] = 0;
+ if (0 > (len = getdelim(&line, &unused, TT.indelim, file))) break;
+ if (line[len-1] == TT.indelim) line[len-1] = 0;
start = line;
- for (;;)
- {
+ // Loop through matches in this line
+ do {
int rc = 0, skip = 0;
+ // Handle non-regex matches
if (toys.optflags & FLAG_F) {
struct arg_list *seek, fseek;
char *s = 0;
fseek.arg = s = line;
break;
}
- if (toys.optflags & FLAG_i) {
- long ll = strlen(seek->arg);;
-
- // Alas, posix hasn't got strcasestr()
- for (s = line; *s; s++) if (!strncasecmp(s, seek->arg, ll)) break;
- if (!*s) s = 0;
- } else s = strstr(line, seek->arg);
+ if (toys.optflags & FLAG_i) s = strnstr(line, seek->arg);
+ else s = strstr(line, seek->arg);
if (s) break;
}
if (s) {
- matches[which].rm_so = (s-line);
- skip = matches[which].rm_eo = (s-line)+strlen(seek->arg);
+ matches.rm_so = (s-line);
+ skip = matches.rm_eo = (s-line)+strlen(seek->arg);
} else rc = 1;
} else {
- rc = regexec((regex_t *)toybuf, start, 3, matches,
+ rc = regexec((regex_t *)toybuf, start, 1, &matches,
start==line ? 0 : REG_NOTBOL);
- skip = matches[which].rm_eo;
+ skip = matches.rm_eo;
}
if (toys.optflags & FLAG_x)
- if (matches[which].rm_so || line[matches[which].rm_eo]) rc = 1;
+ if (matches.rm_so || line[matches.rm_eo]) rc = 1;
+
+ if (!rc && (toys.optflags & FLAG_w)) {
+ char c = 0;
+
+ if ((start+matches.rm_so)!=line) {
+ c = start[matches.rm_so-1];
+ if (!isalnum(c) && c != '_') c = 0;
+ }
+ if (!c) {
+ c = start[matches.rm_eo];
+ if (!isalnum(c) && c != '_') c = 0;
+ }
+ if (c) {
+ start += matches.rm_so+1;
+
+ continue;
+ }
+ }
if (toys.optflags & FLAG_v) {
if (toys.optflags & FLAG_o) {
- if (rc) skip = matches[which].rm_eo = strlen(start);
- else if (!matches[which].rm_so) {
+ if (rc) skip = matches.rm_eo = strlen(start);
+ else if (!matches.rm_so) {
start += skip;
continue;
- } else matches[which].rm_eo = matches[which].rm_so;
+ } else matches.rm_eo = matches.rm_so;
} else {
if (!rc) break;
- matches[which].rm_eo = strlen(start);
+ matches.rm_eo = strlen(start);
}
- matches[which].rm_so = 0;
+ matches.rm_so = 0;
} else if (rc) break;
+ // At least one line we didn't print since match while -ABC active
+ if (bars) {
+ xputs(bars);
+ bars = 0;
+ }
mmatch++;
toys.exitval = 0;
if (toys.optflags & FLAG_q) xexit();
if (toys.optflags & FLAG_l) {
- printf("%s%c", name, outdelim);
+ xprintf("%s%c", name, TT.outdelim);
free(line);
fclose(file);
return;
}
if (toys.optflags & FLAG_o)
- if (matches[which].rm_eo == matches[which].rm_so)
+ if (matches.rm_eo == matches.rm_so)
break;
if (!(toys.optflags & FLAG_c)) {
- if (toys.optflags & FLAG_H) printf("%s:", name);
- if (toys.optflags & FLAG_n) printf("%d:", lcount);
- if (toys.optflags & FLAG_b)
- printf("%ld:", offset + (start-line) +
- ((toys.optflags & FLAG_o) ? matches[which].rm_so : 0));
- if (!(toys.optflags & FLAG_o)) xprintf("%s%c", line, outdelim);
- else {
- xprintf("%.*s%c", matches[which].rm_eo - matches[which].rm_so,
- start + matches[which].rm_so, outdelim);
- }
+ long bcount = 1 + offset + (start-line) +
+ ((toys.optflags & FLAG_o) ? matches.rm_so : 0);
+
+ if (!(toys.optflags & FLAG_o)) {
+ while (dlb) {
+ struct double_list *dl = dlist_pop(&dlb);
+
+ outline(dl->data, '-', name, lcount-before, 0, -1);
+ free(dl->data);
+ free(dl);
+ before--;
+ }
+
+ outline(line, ':', name, lcount, bcount, -1);
+ if (TT.a) after = TT.a+1;
+ } else outline(start+matches.rm_so, ':', name, lcount, bcount,
+ matches.rm_eo-matches.rm_so);
}
start += skip;
- if (!(toys.optflags & FLAG_o) || !*start) break;
- }
+ if (!(toys.optflags & FLAG_o)) break;
+ } while (*start);
offset += len;
+ if (mmatch) mcount++;
+ else {
+ int discard = (after || TT.b);
+
+ if (after && --after) {
+ outline(line, '-', name, lcount, 0, -1);
+ discard = 0;
+ }
+ if (discard && TT.b) {
+ dlist_add(&dlb, line);
+ line = 0;
+ if (++before>TT.b) {
+ struct double_list *dl;
+
+ dl = dlist_pop(&dlb);
+ free(dl->data);
+ free(dl);
+ before--;
+ } else discard = 0;
+ }
+ // If we discarded a line while displaying context, show bars before next
+ // line (but don't show them now in case that was last match in file)
+ if (discard && mcount) bars = "--";
+ }
free(line);
- if (mmatch) mcount++;
if ((toys.optflags & FLAG_m) && mcount >= TT.m) break;
}
- if (toys.optflags & FLAG_c) {
- if (toys.optflags & FLAG_H) printf("%s:", name);
- xprintf("%d%c", mcount, outdelim);
- }
+ if (toys.optflags & FLAG_c) outline(0, ':', name, mcount, 0, -1);
// loopfiles will also close the fd, but this frees an (opaque) struct.
fclose(file);
if (TT.f) s = ss = xreadfile(al->arg, 0, 0);
else s = ss = al->arg;
+ // Split lines at \n, add individual lines to new list.
do {
ss = strchr(s, '\n');
if (ss) *(ss++) = 0;
list = new;
s = ss;
} while (ss && *s);
+
+ // Advance, when we run out of -f switch to -e.
al = al->next;
if (!al && TT.f) {
TT.f = 0;
TT.e = list;
if (!(toys.optflags & FLAG_F)) {
- int w = toys.optflags & FLAG_w;
char *regstr;
+ int i;
// Convert strings to one big regex
- if (w) len = 36;
for (al = TT.e; al; al = al->next)
len += strlen(al->arg)+1+!(toys.optflags & FLAG_E);
regstr = s = xmalloc(len);
- if (w) s = stpcpy(s, "(^|[^_[:alnum:]])(");
for (al = TT.e; al; al = al->next) {
s = stpcpy(s, al->arg);
if (!(toys.optflags & FLAG_E)) *(s++) = '\\';
*(s++) = '|';
}
*(s-=(1+!(toys.optflags & FLAG_E))) = 0;
- if (w) strcpy(s, ")($|[^_[:alnum:]])");
- w = regcomp((regex_t *)toybuf, regstr,
+ i = regcomp((regex_t *)toybuf, regstr,
((toys.optflags & FLAG_E) ? REG_EXTENDED : 0) |
((toys.optflags & FLAG_i) ? REG_ICASE : 0));
- if (w) {
- regerror(w, (regex_t *)toybuf, toybuf+sizeof(regex_t),
+ if (i) {
+ regerror(i, (regex_t *)toybuf, toybuf+sizeof(regex_t),
sizeof(toybuf)-sizeof(regex_t));
error_exit("bad REGEX: %s", toybuf);
}
{
char **ss = toys.optargs;
+ if (!TT.a) TT.a = TT.c;
+ if (!TT.b) TT.b = TT.c;
+
+ TT.indelim = '\n' * !(toys.optflags&FLAG_z);
+ TT.outdelim = '\n' * !(toys.optflags&FLAG_Z);
+
// Handle egrep and fgrep
- if (*toys.which->name == 'e' || (toys.optflags & FLAG_w))
- toys.optflags |= FLAG_E;
+ if (*toys.which->name == 'e') toys.optflags |= FLAG_E;
if (*toys.which->name == 'f') toys.optflags |= FLAG_F;
if (!TT.e && !TT.f) {
}
if (toys.optflags & FLAG_r) {
+ // Iterate through -r arguments. Use "." as default if none provided.
for (ss = *ss ? ss : (char *[]){".", 0}; *ss; ss++) {
if (!strcmp(*ss, "-")) do_grep(0, *ss);
else dirtree_read(*ss, do_grep_r);