From 73e3a644c1af8e28ca45c3b5169dee8acc06b179 Mon Sep 17 00:00:00 2001 From: Rob Landley Date: Sun, 1 Nov 2015 02:33:51 -0600 Subject: Change grep -w to checking matches after the fact rather than modifing regex. This lets '(x)\1' match, as reported by Isabella Parakiss. --- tests/grep.test | 11 ++++---- toys/posix/grep.c | 83 ++++++++++++++++++++++++++++++++++--------------------- 2 files changed, 58 insertions(+), 36 deletions(-) diff --git a/tests/grep.test b/tests/grep.test index 13b1af78..f3bc9a3b 100755 --- a/tests/grep.test +++ b/tests/grep.test @@ -42,8 +42,8 @@ testing "grep -o" "grep -o is input" "is\nis\nis\nis\n" \ testing "grep -s" "grep -hs hello asdf input 2>&1" "hello\n" "hello\n" "" testing "grep -v" "grep -v abc input" "1234123asdfas123123\n1ABa\n" \ "1234123asdfas123123\n1ABabc\nabc\n1ABa\nabcde" "" -testing "grep -w" "grep -w abc input" "abc\n" \ - "1234123asdfas123123\n1ABabc\nabc\n1ABa\nabcde" "" +testing "grep -w" "grep -w abc input" "abc\n123 abc\nabc 123\n123 abc 456\n" \ + "1234123asdfas123123\n1ABabc\nabc\n1ABa\nabcde\n123 abc\nabc 123\n123 abc 456\n" "" testing "grep -x" "grep -x abc input" "abc\n" \ "aabcc\nabc\n" "" @@ -84,15 +84,16 @@ testing "grep -r dir" "grep -r one sub | sort" "sub/one:one\nsub/two:one\n" \ "" "" rm -rf sub -# Not sure if -Fx '' should do? Posix is unclear, '' match every line but does -# it match every _byte_ in that line? -testing "grep -Fx ''" "grep -Fx '' input" "one one one\n" "one one one\n" "" +# -x exact match trumps -F's "empty string matches whole line" behavior +testing "grep -Fx ''" "grep -Fx '' input" "" "one one one\n" "" testing "grep -F ''" "grep -F '' input" "one one one\n" "one one one\n" "" testing "grep -F -e blah -e ''" "grep -F -e blah -e '' input" "one one one\n" \ "one one one\n" "" testing "grep -e blah -e ''" "grep -e blah -e '' input" "one one one\n" \ "one one one\n" "" testing "grep -w ''" "grep -w '' input" "" "one one one\n" "" +testing "grep -w '' 2" "grep -w '' input" "one two\n" "one two\n" "" +testing "grep -w \\1" "grep -wo '\\(x\\)\\1'" "xx\n" "" "xx" testing "grep -o ''" "grep -o '' input" "" "one one one\n" "" testing "grep backref" 'grep -e "a\(b\)" -e "b\(c\)\1"' "bcc\nab\n" \ "" "bcc\nbcb\nab\n" diff --git a/toys/posix/grep.c b/toys/posix/grep.c index ffc920c1..d44a92d6 100644 --- a/toys/posix/grep.c +++ b/toys/posix/grep.c @@ -3,6 +3,8 @@ * Copyright 2013 CE Strake * * 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_EGREP(OLDTOY(egrep, grep, TOYFLAG_BIN)) @@ -56,15 +58,14 @@ GLOBALS( long m; struct arg_list *f; struct arg_list *e; - - struct arg_list *regex; ) +// Show matches in one file static void do_grep(int fd, char *name) { FILE *file = fdopen(fd, "r"); long offset = 0; - int lcount = 0, mcount = 0, which = toys.optflags & FLAG_w ? 2 : 0; + int lcount = 0, mcount = 0; char indelim = '\n' * !(toys.optflags&FLAG_z), outdelim = '\n' * !(toys.optflags&FLAG_Z); @@ -75,9 +76,10 @@ static void do_grep(int fd, char *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; @@ -88,10 +90,11 @@ static void do_grep(int fd, char *name) 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; @@ -117,30 +120,48 @@ static void do_grep(int fd, char *name) } 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; mmatch++; @@ -153,7 +174,7 @@ static void do_grep(int fd, char *name) 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)) { @@ -161,17 +182,17 @@ static void do_grep(int fd, char *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)); + ((toys.optflags & FLAG_o) ? matches.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); + xprintf("%.*s%c", matches.rm_eo - matches.rm_so, + start + matches.rm_so, outdelim); } } start += skip; - if (!(toys.optflags & FLAG_o) || !*start) break; - } + if (!(toys.optflags & FLAG_o)) break; + } while (*start); offset += len; free(line); @@ -202,6 +223,7 @@ static void parse_regex(void) 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; @@ -211,6 +233,8 @@ static void parse_regex(void) 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; @@ -220,30 +244,27 @@ static void parse_regex(void) 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); } @@ -272,8 +293,7 @@ void grep_main(void) char **ss = toys.optargs; // 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) { @@ -294,6 +314,7 @@ void grep_main(void) } 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); -- cgit v1.2.3