fixed ipv4 for general dot-decimal notation

README.md

@@ -129,7 +129,7 @@ All `LOG_EVENTS` in the code that are lower on this list will be ommitted (in co
 I have included a couple toy parsers generated with <b>peggy</b> in `examples/`. These are not intended to be stand alone project as they are not in the best of condition and of limited practical use--the possible exception being the C parser. They are meant to show different ways to use <b>peggy</b>. If you have any interesting use cases or ideas, let me know.
 
 ### ipv4
-This validates a string as ipv4. This is only included because it is the simplest possible grammar I could think of to show how to build a parser with <b>peggy</b>. It is of no other use.
+This validates a string in dotted decimal notation as ipv4. This is only included because it is the simplest possible grammar I could think of to show how to build a parser with <b>peggy</b>. It is of no other use.
 
 ### csv 
 A simple csv parser. Generates a `CSVData` struct from an input file. Data is not converted to any data types but kept as strings. This example is more to demonstrate the linearity of the parser. The build command `make test_all` will parse the files in `examples/csv/sample_files` and time the results.
@@ -176,7 +176,7 @@ int main(int narg, char ** args) {
         initialize the parser for the given grammar. This is a generic parser that will
         simply provide a token list an AST
         */
-        Parser_init(&parser, (Rule *)&token, (Rule *)&root, MYDSL_NRULES);
+        Parser_init(&parser, (Rule *)&mydsl_token, (Rule *)&mydsl_root, MYDSL_NRULES, 0);
         // optional for logging. default is "stdout" with LOG_LEVEL_ERROR
         Parser_set_log_file(&parser, "stdout", LOG_LEVEL_WARN);  
         // LOG_LEVEL_* above are the same that can be specified at build time (see above)

examples/ipv4/ipv4.grmr

@@ -1,9 +1,9 @@
 // ipv4.grmr
-export = ipv4                // redundant but allows change of file name
-import = ipv4parser          // check_ipv4 is declared in ipv4parser.h
+export = ipv4               // redundant but allows change of file name
+import = ipv4parser         // check_ipv4 is declared in ipv4parser.h
 
 punctuator: '.'             // needed to be able to use '.' in ipv4
 digit: "[0-9]"              
-octet: digit{1,3}           // 0 digits and more than 3 are definitely errors
+octet: digit+               // 0 digits and more than 3 are definitely errors
 token: punctuator | digit
 ipv4(check_ipv4): '.'.octet // check_ipv4 as a build transform function triggers AST traversal

examples/ipv4/ipv4parser.c

@@ -1,6 +1,8 @@
 // ipv4parser.c
 #include <stdio.h>
+#include <stdlib.h>
 #include <string.h>
+#include <limits.h>
 
 #include <peggy/astnode.h>
 #include <peggy/rule.h>
@@ -12,36 +14,54 @@
 
 ASTNode ASTNode_syntax_error = {0}; // used to signal a failure that has already been addressed
 
-int check_octet(Parser * parser, ASTNode * node, unsigned char loc) {
+size_t max_ull_len = 0;
+
+int check_octet(Parser * parser, ASTNode * node, unsigned char loc, size_t max) {
     // only success of the subrule will make it here. Do not have to check node for failure, node == NULL is a more serious failure so assume it is non-null
-    Token * toks = node->token_start;
-    char const * const str = toks->string;
+    char const * const str = node->token_start->string;
     size_t len = ASTNode_string_length(node);
-    // if the octet is < 3 decimals or numerically <= 255, it is valid
-    //printf("octet (%zu)(%zu)(%zu): %.*s\n", len, node->str_length, node->nchildren, (int)len, str);
-    if (len && (len < 3 || str[0] < '2' || (str[0] == '2' && (str[1] < '5' || (str[1] == '5' && str[2] <= '5'))))) {
-        return 0;
+    if (len > max_ull_len) { // if length of string is longer than an unsigned long long can possibly be, fail
+        printf("ipv4 failed. invalid octet at %hu (%.*s). too many characters for valid number\n", loc, (int)len, str);
+        return 1;
+    }
+    unsigned long long val = strtoull(str, NULL, 10);
+    if (val >= max) {
+        printf("ipv4 failed. invalid octet at %hu (%.*s). must be in range [0-%zu)\n", loc, (int)len, str, max);
+        return 1;
     }
-    printf("ipv4 failed. invalid octet at %hu (%.*s). must be in range 0-255\n", loc, (int)len, str);
-    return 1;
+    return 0;
 }
 ASTNode * check_ipv4(Production * octet, Parser * parser, ASTNode * node) {
     // if the parser has already failed or resulted in something other than than 3 '.' + 4 octets = 7, fail
-    if (node->nchildren != 7 || Parser_is_fail_node(parser, node)) { 
-        printf("%s is NOT valid ipv4. %zu octets not found, need 4.\n", parser->token_head->next->string, (node->nchildren + 1) / 2);
+    if (node->nchildren < 3 || node->nchildren > 7 || Parser_is_fail_node(parser, node)) { 
+        printf("%s is NOT valid ipv4. %zu octets not found, need 2-4.\n", parser->token_head->next->string, (node->nchildren + 1) / 2);
         return &ASTNode_syntax_error;
     }
 
-    for (size_t i = 0; i < node->nchildren; i++) {
-        if (check_octet(parser, node->children[i], i + 1)) {
+    unsigned char noctets = (node->nchildren >> 1) + 1;
+    // maximum value for the 
+    size_t max = 1 << (8 * (5 - noctets));
+    unsigned char i = 0;
+    while (i < noctets - 1) {
+        if (check_octet(parser, node->children[2 * i], i, 256)) {
+            return &ASTNode_syntax_error;
+        }
+        i++;
+    }
+    while (i < noctets) {
+        if (check_octet(parser, node->children[2 * i], i, max)) {
             return &ASTNode_syntax_error;
         }
+        max >>= 8;
+        i++;
     }
     printf("%s is valid ipv4\n", parser->token_head->next->string);
     return node;
 }
 
 int main(int narg, char ** args) {
+    char buffer[256];
+    max_ull_len = snprintf(buffer, 256, "%llu", ULLONG_MAX);
     if (narg > 1) {
         Parser parser = {._class = &Parser_class};
         Parser_init(&parser, (Rule *)&ipv4_token, (Rule *)&ipv4_ipv4, IPV4_NRULES, 0);