pike.c: handle 0 case in repetition

README

@@ -32,7 +32,6 @@ so that the user does not need to waste time taking strlen()
 * Support for quoted chars in regex. Escapes in brackets.
 * Support for ^, $ assertions in regex.
 * Support for repetition operator {n} and {n,m} and {n,}.
-- Note: cases with 0 are not handled, avoid them, they can easily be replaced.
 * Support for Unicode (UTF-8).
 * Unlike other engines, the output is byte level offset. (Which is more useful)
 * Support for non capture group ?:

pike.c

@@ -268,7 +268,7 @@ static int compilecode(const char *re_loc, rcode *prog, int sizecode)
 			}
 			break;
 		case '{':;
-			int maxcnt = 0, mincnt = 0, i = 0, size = PC - term;
+			int i, maxcnt = 0, mincnt = 0, size = PC - term, nojmp = 0;
 			re++;
 			while (isdigit((unsigned char) *re))
 				mincnt = mincnt * 10 + *re++ - '0';
@@ -279,12 +279,21 @@ static int compilecode(const char *re_loc, rcode *prog, int sizecode)
 					EMIT(PC+1, REL(PC, PC - size));
 					PC += 2;
 					maxcnt = mincnt;
+					nojmp = 1;
 				}
 				while (isdigit((unsigned char) *re))
 					maxcnt = maxcnt * 10 + *re++ - '0';
 			} else
 				maxcnt = mincnt;
-			for (; i < mincnt-1; i++) {
+			if (!mincnt && !maxcnt) {
+				zcase:
+			        INSERT_CODE(term, 2, PC);
+			        EMIT(term, nojmp ? SPLIT : JMP);
+			        EMIT(term + 1, REL(term, PC));
+			        term = PC;
+			        break;
+			}
+			for (i = 0; i < mincnt-1; i++) {
 				if (code)
 					memcpy(&code[PC], &code[term], size*sizeof(int));
 				PC += size;
@@ -296,6 +305,10 @@ static int compilecode(const char *re_loc, rcode *prog, int sizecode)
 					memcpy(&code[PC], &code[term], size*sizeof(int));
 				PC += size;
 			}
+			if (!mincnt && maxcnt) {
+				nojmp = 1;
+				goto zcase;
+			}
 			break;
 		case '?':
 			if (PC == term)

test.sh

@@ -1,6 +1,20 @@
 #!/bin/sh
 
 regex="\
+a{0}
+(aaaa){0,}
+(aaaa){0,0}
+(aaa+a){0,}
+(aaaa){0,0}|a
+(aaaa){0,0}|abc
+(aaaa){0,}|bc
+(aaaa){0,2}|bc
+(aaaa){0,2}|bc
+(aaaa){0,2}|bc
+bc{0}|(aaaa){0,2}
+(bc{0,})+|(aaaa){0,2}
+(bc{0,}){3,5}|(aaaa){0,2}
+(bc{0,}){3,5}|(aaaa){0,2}
 abc
 cde
 abc*
@@ -195,6 +209,20 @@ aaaaa(aa)aa(aa(a)a)?aa
 (((?:(?:(?:ffffff(a)?ffff)+)+?)*)*)+
 "
 input="\
+aaaaaaaaaa
+aaaaaaaaaa
+aaaaaaaaaa
+aaaaaaaaaa
+aaaaaaaaaa
+aaaaaaaaaa
+aaaaaaaaaa
+aaaaaa
+aaaa
+aaa
+ccc
+bbbbccc
+bbbbbccccc
+bbbbbbccccc
 abcdef
 abcdef
 abdef
@@ -389,6 +417,20 @@ fffffaffffffffff
 fffffffffffffffffffffffff
 "
 expect="\
+(0,0)
+(0,8)(4,8)
+(0,0)(?,?)
+(0,10)(0,10)
+(0,0)(?,?)
+(0,0)(?,?)
+(0,8)(4,8)
+(0,4)(0,4)
+(0,4)(0,4)
+(0,0)(?,?)
+(0,0)(?,?)
+(0,7)(3,7)(?,?)
+(0,10)(4,10)(?,?)
+(0,5)(4,5)(?,?)
 (0,3)
 (2,5)
 (0,2)