Add rules for nth and only-tags. Address #1503

These rules allow selecting from a number of sub-captures
while dropping the rest. `nth` is more succinct in many cases, but `only-tags` is
more general and corresponds to an internal mechanism already present.

Author: bakpakin

src/core/filewatch.c

@@ -588,6 +588,7 @@ JANET_CORE_FN(cfun_filewatch_make,
         "* `:wd-path` -- the string path for watched directory of file. For files, will be the same as `:file-name`, and for directories, will be the same as `:dir-name`.\n\n"
         "* `:cookie` -- a randomized integer used to associate related events, such as :moved-from and :moved-to events.\n\n"
         "") {
+    janet_sandbox_assert(JANET_SANDBOX_FS_READ);
     janet_arity(argc, 1, -1);
     JanetChannel *channel = janet_getchannel(argv, 0);
     JanetWatcher *watcher = janet_abstract(&janet_filewatch_at, sizeof(JanetWatcher));

src/core/peg.c

@@ -465,6 +465,16 @@ static const uint8_t *peg_rule(
             return result;
         }
 
+        case RULE_ONLY_TAGS: {
+            CapState cs = cap_save(s);
+            down1(s);
+            const uint8_t *result = peg_rule(s, s->bytecode + rule[1], text);
+            up1(s);
+            if (!result) return NULL;
+            cap_load_keept(s, cs);
+            return result;
+        }
+
         case RULE_GROUP: {
             uint32_t tag = rule[2];
             int oldmode = s->mode;
@@ -486,6 +496,29 @@ static const uint8_t *peg_rule(
             return result;
         }
 
+        case RULE_NTH: {
+            uint32_t nth = rule[1];
+            uint32_t tag = rule[3];
+            int oldmode = s->mode;
+            CapState cs = cap_save(s);
+            s->mode = PEG_MODE_NORMAL;
+            down1(s);
+            const uint8_t *result = peg_rule(s, s->bytecode + rule[2], text);
+            up1(s);
+            s->mode = oldmode;
+            if (!result) return NULL;
+            int32_t num_sub_captures = s->captures->count - cs.cap;
+            Janet cap;
+            if (num_sub_captures > (int32_t) nth) {
+                cap = s->captures->data[cs.cap + nth];
+            } else {
+                return NULL;
+            }
+            cap_load_keept(s, cs);
+            pushcap(s, cap, tag);
+            return result;
+        }
+
         case RULE_SUB: {
             const uint8_t *text_start = text;
             const uint32_t *rule_window = s->bytecode + rule[1];
@@ -1061,6 +1094,9 @@ static void spec_thru(Builder *b, int32_t argc, const Janet *argv) {
 static void spec_drop(Builder *b, int32_t argc, const Janet *argv) {
     spec_onerule(b, argc, argv, RULE_DROP);
 }
+static void spec_only_tags(Builder *b, int32_t argc, const Janet *argv) {
+    spec_onerule(b, argc, argv, RULE_ONLY_TAGS);
+}
 
 /* Rule of the form [rule, tag] */
 static void spec_cap1(Builder *b, int32_t argc, const Janet *argv, uint32_t op) {
@@ -1084,6 +1120,15 @@ static void spec_unref(Builder *b, int32_t argc, const Janet *argv) {
     spec_cap1(b, argc, argv, RULE_UNREF);
 }
 
+static void spec_nth(Builder *b, int32_t argc, const Janet *argv) {
+    peg_arity(b, argc, 2, 3);
+    Reserve r = reserve(b, 4);
+    uint32_t nth = peg_getnat(b, argv[0]);
+    uint32_t rule = peg_compile1(b, argv[1]);
+    uint32_t tag = (argc == 3) ? emit_tag(b, argv[2]) : 0;
+    emit_3(r, RULE_NTH, nth, rule, tag);
+}
+
 static void spec_capture_number(Builder *b, int32_t argc, const Janet *argv) {
     peg_arity(b, argc, 1, 3);
     Reserve r = reserve(b, 4);
@@ -1262,7 +1307,9 @@ static const SpecialPair peg_specials[] = {
     {"line", spec_line},
     {"look", spec_look},
     {"not", spec_not},
+    {"nth", spec_nth},
     {"number", spec_capture_number},
+    {"only-tags", spec_only_tags},
     {"opt", spec_opt},
     {"position", spec_position},
     {"quote", spec_capture},
@@ -1619,6 +1666,7 @@ static void *peg_unmarshal(JanetMarshalContext *ctx) {
                 break;
             case RULE_ERROR:
             case RULE_DROP:
+            case RULE_ONLY_TAGS:
             case RULE_NOT:
             case RULE_TO:
             case RULE_THRU:
@@ -1632,6 +1680,12 @@ static void *peg_unmarshal(JanetMarshalContext *ctx) {
                 if (rule[1] > JANET_MAX_READINT_WIDTH) goto bad;
                 i += 3;
                 break;
+            case RULE_NTH:
+                /* [nth, rule, tag] */
+                if (rule[2] >= blen) goto bad;
+                op_flags[rule[2]] |= 0x01;
+                i += 4;
+                break;
             default:
                 goto bad;
         }

src/include/janet.h

@@ -2180,7 +2180,9 @@ typedef enum {
     RULE_UNREF,        /* [rule, tag] */
     RULE_CAPTURE_NUM,  /* [rule, tag] */
     RULE_SUB,          /* [rule, rule] */
-    RULE_SPLIT         /* [rule, rule] */
+    RULE_SPLIT,        /* [rule, rule] */
+    RULE_NTH,          /* [nth, rule, tag] */
+    RULE_ONLY_TAGS,    /* [rule] */
 } JanetPegOpcod;
 
 typedef struct {

test/suite-peg.janet

@@ -664,6 +664,8 @@
   @[]) "peg if not")
 
 (defn test [name peg input expected]
+  (assert-no-error "compile peg" (peg/compile peg))
+  (assert-no-error "marshal/unmarshal peg" (-> peg marshal unmarshal))
   (assert (deep= (peg/match peg input) expected) name))
 
 (test "sub: matches the same input twice"
@@ -756,5 +758,19 @@
   "a,b,c"
   @["a" "b" "c"])
 
+(test "nth 1"
+  ~{:prefix (number :d+ nil :n)
+    :word '(lenprefix (-> :n) :w)
+    :main (some (nth 1 (* :prefix ":" :word)))}
+  "5:apple6:banana6:cherry"
+  @["apple" "banana" "cherry"])
+
+(test "only-tags 1"
+  ~{:prefix (number :d+ nil :n)
+    :word (capture (lenprefix (-> :n) :w) :W)
+    :main (some (* (only-tags (* :prefix ":" :word)) (-> :W)))}
+  "5:apple6:banana6:cherry"
+  @["apple" "banana" "cherry"])
+
 (end-suite)