Updated '|' to mean word boundary and \N for nodent

compiler.c

@@ -162,6 +162,12 @@ vm_op_t *bpeg_simplepattern(file_t *f, const char *str)
         case '\\': {
             if (!*str || *str == '\n')
                 file_err(f, str, str, "There should be an escape sequence here after this backslash.");
+
+            if (matchchar(&str, 'N')) { // \N (nodent)
+                op->op = VM_NODENT;
+                break;
+            }
+
             op->len = 1;
             unsigned char e = unescapechar(str, &str);
             if (*str == '-') { // Escape range (e.g. \x00-\xFF)
@@ -397,7 +403,7 @@ vm_op_t *bpeg_simplepattern(file_t *f, const char *str)
             break;
         }
         // Special rules:
-        case '_': case '^': case '$': {
+        case '_': case '^': case '$': case '|': {
             if (matchchar(&str, c)) { // double __, ^^, $$
                 char tmp[3] = {c, c, '\0'};
                 op->args.s = strdup(tmp);
@@ -415,10 +421,6 @@ vm_op_t *bpeg_simplepattern(file_t *f, const char *str)
             op->op = VM_REF;
             break;
         }
-        case '|': {
-            op->op = VM_NODENT;
-            break;
-        }
         default: {
             // Reference
             if (isalpha(c)) {
@@ -488,6 +490,13 @@ vm_op_t *bpeg_stringpattern(file_t *f, const char *str)
             if (*str == '\\') {
                 if (!str[1] || str[1] == '\n')
                     file_err(f, str, str, "There should be an escape sequence or pattern here after this backslash.");
+                
+                if (matchchar(&str, 'N')) { // \N (nodent)
+                    interp = calloc(sizeof(vm_op_t), 1);
+                    interp->op = VM_NODENT;
+                    break;
+                }
+
                 const char *after_escape;
                 unsigned char e = unescapechar(&str[1], &after_escape);
                 if (e != str[1]) {

grammar.c

@@ -39,7 +39,7 @@ vm_op_t *load_grammar(grammar_t *g, file_t *f)
     while (*src) {
         const char *name = src;
         const char *name_end = after_name(name);
-        check(name_end > name, "Invalid name for definition");
+        check(name_end > name, "Invalid name for definition: %s", name);
         name = strndup(name, (size_t)(name_end-name));
         src = after_spaces(name_end);
         check(matchchar(&src, ':'), "Expected ':' in definition");

grammars/bpeg.bp

@@ -7,11 +7,12 @@ Def: @name=id _ `: __ (
     / @!={...>(`;/id_`:/$) => "Invalid definition: @0"})
 
 # This is used for command line arguments:
-String-pattern: ... % (`\ (escape-sequence / pat [`;]))
+String-pattern: ... % (Nodent / Escape / `\ pat [`;])
 
 pat: simple-pat !(__("!="/"==")) / suffixed-pat
-simple-pat: Upto-and / Dot / String / Char-range / Char / Escape-range / Escape / No
-    / Nodent / Repeat / Optional / After / Before / Capture / Replace / Ref / parens
+simple-pat: Upto-and / Dot / String / Char-range / Char / Nodent / Escape-range
+    / Escape / Repeat / Optional / After / Before / Capture / Replace
+    / Ref / parens
 
 suffixed-pat: (
       Eq-pat
@@ -39,7 +40,7 @@ escape-sequence: (
       / `x 2 (`0-9/`a-f/`A-F)
     )
 No: `! (_@pat / @!={=>"Expected a pattern after the exclamation mark"})
-Nodent: `|
+Nodent: `\ `N
 Upto-and: 2-3`. [_@first=simple-pat] [__`%__@second=simple-pat]
 Repeat: (
         @min='' `* @max=''
@@ -70,7 +71,7 @@ $:  !.
 ^^: !<$.
 ^:  !<.
 
-id: "^^" / "^" / "__" / "_" / "$$" / "$" / (`a-z/`A-Z) *(`a-z/`A-Z/`0-9/`-)
+id: "^^" / "^" / "__" / "_" / "$$" / "$" / "|" / (`a-z/`A-Z) *(`a-z/`A-Z/`0-9/`-)
 
 line-comment: `# .. $
 block-comment: "#("..")#" % block-comment

utils.c

@@ -44,6 +44,7 @@ const char *after_spaces(const char *str)
  */
 const char *after_name(const char *str)
 {
+    if (*str == '|') return &str[1];
     if (*str == '^' || *str == '_' || *str == '$') {
         return (str[1] == *str) ? &str[2] : &str[1];
     }