Implement an --invert/-v for string match, like grep -v.

Only lines that do not match the pattern are shown.

doc_src/string.txt

@@ -12,7 +12,7 @@ string trim [(-l | --left)] [(-r | --right)] [(-c | --chars CHARS)]
             [(-q | --quiet)] [STRING...]
 string escape [(-n | --no-quoted)] [STRING...]
 string match [(-a | --all)] [(-i | --ignore-case)] [(-r | --regex)]
-             [(-n | --index)] [(-q | --quiet)] PATTERN [STRING...]
+             [(-n | --index)] [(-q | --quiet)] [(-v | --invert)] PATTERN [STRING...]
 string replace [(-a | --all)] [(-i | --ignore-case)] [(-r | --regex)]
                [(-q | --quiet)] PATTERN REPLACEMENT [STRING...]
 \endfish
@@ -44,7 +44,7 @@ The following subcommands are available:
 
 - `escape` escapes each STRING such that it can be passed back to `eval` to produce the original argument again. By default, all special characters are escaped, and quotes are used to simplify the output when possible. If `-n` or `--no-quote` is given, the simplifying quoted format is not used. Exit status: 0 if at least one string was escaped, or 1 otherwise.
 
-- `match` tests each STRING against PATTERN and prints matching substrings. Only the first match for each STRING is reported unless `-a` or `--all` is given, in which case all matches are reported. Matching can be made case-insensitive with `-i` or `--ignore-case`. If `-n` or `--index` is given, each match is reported as a 1-based start position and a length. By default, PATTERN is interpreted as a glob pattern matched against each entire STRING argument. If `-r` or `--regex` is given, PATTERN is interpreted as a Perl-compatible regular expression. For a regular expression containing capturing groups, multiple items will be reported for each match, one for the entire match and one for each capturing group. Exit status: 0 if at least one match was found, or 1 otherwise.
+- `match` tests each STRING against PATTERN and prints matching substrings. Only the first match for each STRING is reported unless `-a` or `--all` is given, in which case all matches are reported. Matching can be made case-insensitive with `-i` or `--ignore-case`. If `-n` or `--index` is given, each match is reported as a 1-based start position and a length. By default, PATTERN is interpreted as a glob pattern matched against each entire STRING argument. If `-r` or `--regex` is given, PATTERN is interpreted as a Perl-compatible regular expression. For a regular expression containing capturing groups, multiple items will be reported for each match, one for the entire match and one for each capturing group. If --invert or -v is used the selected lines will be only those which do not match the given glob pattern or regular expression.  Exit status: 0 if at least one match was found, or 1 otherwise.
 
 - `replace` is similar to `match` but replaces non-overlapping matching substrings with a replacement string and prints the result. By default, PATTERN is treated as a literal substring to be matched. If `-r` or `--regex` is given, PATTERN is interpreted as a Perl-compatible regular expression, and REPLACEMENT can contain C-style escape sequences like `\t` as well as references to capturing groups by number or name as `$n` or `${n}`. Exit status: 0 if at least one replacement was performed, or 1 otherwise.
 
@@ -120,6 +120,14 @@ The following subcommands are available:
 
 >_ echo 'ok?' | string match '*\\?'
 >_ <outp>ok?</outp>
+
+>_ string match -r -v "c.*[12]" {cat,dog}(seq 1 4)
+<outp>dog1</outp>
+<outp>dog2</outp>
+<outp>cat3</outp>
+<outp>dog3</outp>
+<outp>cat4</outp>
+<outp>dog4</outp>
 \endfish
 
 \subsection string-example-match-regex Match Regex Examples

src/builtin_string.cpp

@@ -304,9 +304,10 @@ struct match_options_t
     bool all;
     bool ignore_case;
     bool index;
+    bool invert_match;
     bool quiet;
 
-    match_options_t(): all(false), ignore_case(false), index(false), quiet(false) { }
+    match_options_t(): all(false), ignore_case(false), index(false), invert_match(false), quiet(false) { }
 };
 
 class string_matcher_t
@@ -328,17 +329,15 @@ public:
 
 class wildcard_matcher_t: public string_matcher_t
 {
+private:
     wcstring wcpattern;
-
 public:
     wildcard_matcher_t(const wchar_t * /*argv0*/, const wchar_t *pattern, const match_options_t &opts, io_streams_t &streams)
-        : string_matcher_t(opts, streams)
+    : string_matcher_t(opts, streams), wcpattern(parse_util_unescape_wildcards(pattern))
     {
-        wcpattern = parse_util_unescape_wildcards(pattern);
-
         if (opts.ignore_case)
         {
-            for (int i = 0; i < wcpattern.length(); i++)
+            for (size_t i = 0; i < wcpattern.length(); i++)
             {
                 wcpattern[i] = towlower(wcpattern[i]);
             }
@@ -352,10 +351,11 @@ public:
         // Note: --all is a no-op for glob matching since the pattern is always
         // matched against the entire argument
         bool match;
+
         if (opts.ignore_case)
         {
             wcstring s = arg;
-            for (int i = 0; i < s.length(); i++)
+            for (size_t i = 0; i < s.length(); i++)
             {
                 s[i] = towlower(s[i]);
             }
@@ -365,13 +365,11 @@ public:
         {
             match = wildcard_match(arg, wcpattern, false);
         }
-        if (match)
+        if (match ^ opts.invert_match)
         {
             total_matched++;
-        }
+
             if (!opts.quiet)
-        {
-            if (match)
             {
                 if (opts.index)
                 {
@@ -458,28 +456,38 @@ class pcre2_matcher_t: public string_matcher_t
         // Return values: -1 = error, 0 = no match, 1 = match
         if (pcre2_rc == PCRE2_ERROR_NOMATCH)
         {
-            return 0;
+            if (opts.invert_match && !opts.quiet)
+            {
+                streams.out.append(arg);
+                streams.out.push_back(L'\n');
             }
-        if (pcre2_rc < 0)
+
+            return opts.invert_match ? 1 : 0;
+        }
+        else if (pcre2_rc < 0)
         {
             string_error(streams, _(L"%ls: Regular expression match error: %ls\n"),
                     argv0, pcre2_strerror(pcre2_rc).c_str());
             return -1;
         }
-        if (pcre2_rc == 0)
+        else if (pcre2_rc == 0)
         {
             // The output vector wasn't big enough. Should not happen.
             string_error(streams, _(L"%ls: Regular expression internal error\n"), argv0);
             return -1;
         }
+
+        else if (opts.invert_match)
+            return 0;
+
         PCRE2_SIZE *ovector = pcre2_get_ovector_pointer(regex.match);
+
         for (int j = 0; j < pcre2_rc; j++)
         {
             PCRE2_SIZE begin = ovector[2*j];
             PCRE2_SIZE end = ovector[2*j + 1];
-            if (!opts.quiet)
+
-            {
+            if (begin != PCRE2_UNSET && end != PCRE2_UNSET && !opts.quiet)
-                if (begin != PCRE2_UNSET && end != PCRE2_UNSET)
             {
                 if (opts.index)
                 {
@@ -489,11 +497,12 @@ class pcre2_matcher_t: public string_matcher_t
                 {
                     streams.out.append(wcstring(&arg[begin], end - begin));
                 }
-                    streams.out.append(L'\n');
+                streams.out.push_back(L'\n');
             }
         }
-        }
+
-        return 1;
+        return opts.invert_match ? 0 : 1;
+
     }
 
 public:
@@ -525,7 +534,7 @@ public:
             // pcre2 match error
             return false;
         }
-        if (rc == 0)
+        else if (rc == 0)
         {
             // no match
             return true;
@@ -533,6 +542,11 @@ public:
         matched++;
         total_matched++;
 
+        if (opts.invert_match)
+        {
+            return true;
+        }
+
         // Report any additional matches
         PCRE2_SIZE *ovector = pcre2_get_ovector_pointer(regex.match);
         while (opts.all || matched == 0)
@@ -573,12 +587,13 @@ public:
 
 static int string_match(parser_t &parser, io_streams_t &streams, int argc, wchar_t **argv)
 {
-    const wchar_t *short_options = L"ainqr";
+    const wchar_t *short_options = L"ainvqr";
     const struct woption long_options[] =
     {
         { L"all", no_argument, 0, 'a'},
         { L"ignore-case", no_argument, 0, 'i'},
         { L"index", no_argument, 0, 'n'},
+        { L"invert", no_argument, 0, 'v'},
         { L"quiet", no_argument, 0, 'q'},
         { L"regex", no_argument, 0, 'r'},
         { 0, 0, 0, 0 }
@@ -612,6 +627,10 @@ static int string_match(parser_t &parser, io_streams_t &streams, int argc, wchar
                 opts.index = true;
                 break;
 
+            case 'v':
+                opts.invert_match = true;
+                break;
+
             case 'q':
                 opts.quiet = true;
                 break;
@@ -750,7 +769,7 @@ class regex_replacer_t: public string_replacer_t
     compiled_regex_t regex;
     wcstring replacement;
 
-    wcstring interpret_escapes(const wchar_t *orig)
+    static wcstring interpret_escapes(const wchar_t *orig)
     {
         wcstring result;
 
@@ -782,6 +801,7 @@ public:
 
     bool replace_matches(const wchar_t *arg)
     {
+
         // A return value of true means all is well (even if no replacements
         // were performed), false indicates an unrecoverable error.
         if (regex.code == 0)
@@ -1091,9 +1111,9 @@ static int string_split(parser_t &parser, io_streams_t &streams, int argc, wchar
     // If we are from the right, split_about gave us reversed strings, in reversed order!
     if (right)
     {
-        for (size_t i=0; i < splits.size(); i++)
+        for (size_t j = 0; j < splits.size(); j++)
         {
-            std::reverse(splits[i].begin(), splits[i].end());
+            std::reverse(splits[j].begin(), splits[j].end());
         }
         std::reverse(splits.begin(), splits.end());
     }
@@ -1339,6 +1359,7 @@ static const struct string_subcommand
     const wchar_t *name;
     int (*handler)(parser_t &, io_streams_t &, int argc, wchar_t **argv);
 }
+
 string_subcommands[] =
 {
     { L"escape", &string_escape },

tests/string.in

@@ -1,5 +1,26 @@
 # tests for string builtin
 # mostly taken from examples
+
+string match -r -v "c.*" dog can cat diz; and echo "exit 0"
+
+string match -q -r -v "c.*" dog can cat diz; and echo "exit 0"
+
+string match -v "c*" dog can cat diz; and echo "exit 0"
+
+string match -q -v "c*" dog can cat diz; and echo "exit 0"
+
+string match -v "d*" dog dan dat diz; or echo "exit 1"
+
+string match -q -v "d*" dog dan dat diz; or echo "exit 1"
+
+string match -r -v x y; and echo "exit 0"
+
+string match -r -v x x; or echo "exit 1"
+
+string match -q -r -v x y; and echo "exit 0"
+
+string match -q -r -v x x; or echo "exit 1"
+
 string length 'hello, world'
 
 string length -q ""; and echo not zero length
@@ -63,3 +84,7 @@ string match -r '[' 'a[sd' 2>/dev/null; or echo "invalid expression error"
 string invalidarg 2>/dev/null; or echo "invalid argument error"
 
 string length 2>/dev/null; or echo "missing argument returns 0"
+
+string match -r -v "[dcantg].*" dog can cat diz; or echo "no regexp invert match"
+
+string match -v "???" dog can cat diz; or echo "no glob invert match"

tests/string.out

@@ -1,3 +1,18 @@
+dog
+diz
+exit 0
+exit 0
+dog
+diz
+exit 0
+exit 0
+exit 1
+exit 1
+y
+exit 0
+exit 1
+exit 0
+exit 1
 12
 ab
 bc
@@ -45,3 +60,5 @@ aabb
 invalid expression error
 invalid argument error
 missing argument returns 0
+no regexp invert match
+no glob invert match