Emit an error when repetition lower bound exceeds upper bound.

Historically this was allowed and re2c swapped the bounds. However, it most likely indicates an error in user code and there is only a single occurrence in the tests (and the test in an artificial one), so although the change is backwards incompatible there is low chance of breaking real-world code. This fixes second test case in the bug #394 "Stack overflow due to recursion in src/dfa/dead_rules.cc" (the actual fix is to limit DFA size but the test also has counted repetition with swapped bounds).
skvadrik · Jan 21, 2022 · 039c189 · 039c189
1 parent a3473fd
commit 039c189
Show file tree

Hide file tree

Showing 9 changed files with 2,453 additions and 2,519 deletions.
diff --git a/bootstrap/src/parse/lex.cc b/bootstrap/src/parse/lex.cc
diff --git a/bootstrap/src/parse/lex.h b/bootstrap/src/parse/lex.h
@@ -1,4 +1,4 @@
-/* Generated by re2c 2.2 on Fri Dec 24 21:36:58 2021 */
+/* Generated by re2c 2.2 on Fri Jan 21 22:49:06 2022 */
 
 #ifndef _RE2C_PARSE_LEX_
 #define _RE2C_PARSE_LEX_

diff --git a/src/debug/debug.h b/src/debug/debug.h
@@ -35,6 +35,7 @@ template<typename ctx_t>
 
 #include <assert.h>
 #include "src/util/c99_stdint.h"
+#include <stddef.h>
 #include <map>
 #include <set>
 #include <vector>

diff --git a/src/parse/ast.cc b/src/parse/ast.cc
@@ -1,5 +1,6 @@
 #include <limits>
 
+#include "src/debug/debug.h"
 #include "src/parse/ast.h"
 #include "src/util/free_list.h"
 
@@ -83,6 +84,7 @@ const AST *ast_cat(const AST *a1, const AST *a2)
 
 const AST *ast_iter(const AST *a, uint32_t n, uint32_t m)
 {
+    DASSERT(n <= m);
     AST *ast = new AST(a->loc, AST::ITER);
     ast->iter.ast = a;
     ast->iter.min = n;

diff --git a/src/parse/lex.re b/src/parse/lex.re
@@ -461,13 +461,15 @@ scan:
     }
 
     "{" [0-9]+ @p "," [0-9]+ "}" {
-        if (!s_to_u32_unsafe (tok + 1, p, yylval.bounds.min)) {
+        if (!s_to_u32_unsafe(tok + 1, p, yylval.bounds.min)) {
             msg.error(tok_loc(), "repetition lower bound overflow");
             exit(1);
-        }
-        if (!s_to_u32_unsafe (p + 1, cur - 1, yylval.bounds.max)) {
+        } else if (!s_to_u32_unsafe(p + 1, cur - 1, yylval.bounds.max)) {
             msg.error(tok_loc(), "repetition upper bound overflow");
             exit(1);
+        } else if (yylval.bounds.min > yylval.bounds.max) {
+            msg.error(tok_loc(), "repetition lower bound exceeds upper bound");
+            exit(1);
         }
         return TOKEN_CLOSESIZE;
     }

diff --git a/test/messages/repetition_swapped_bounds.c b/test/messages/repetition_swapped_bounds.c
@@ -0,0 +1 @@
+messages/repetition_swapped_bounds.re:5:7: error: repetition lower bound exceeds upper bound
diff --git a/test/messages/repetition_swapped_bounds.re b/test/messages/repetition_swapped_bounds.re
@@ -0,0 +1,6 @@
+// re2c $INPUT
+
+// error: lower bound exceeds upper bound
+/*!re2c
+    [a]{3,2} {}
+*/
diff --git a/test/repeater2.c b/test/repeater2.c
@@ -208,33 +208,29 @@
 #line 209 "repeater2.c"
 {
 	YYCTYPE yych;
-	if (YYLIMIT <= YYCURSOR) YYFILL(1);
+	if ((YYLIMIT - YYCURSOR) < 5) YYFILL(5);
 	yych = *YYCURSOR;
 	switch (yych) {
 		case 'a': goto yy31;
 		default: goto yy30;
 	}
 yy30:
 yy31:
-	++YYCURSOR;
-#line 39 "repeater2.re"
-	{}
-#line 223 "repeater2.c"
-}
-#line 40 "repeater2.re"
-
-
-
-#line 229 "repeater2.c"
-{
-	YYCTYPE yych;
-	if ((YYLIMIT - YYCURSOR) < 5) YYFILL(5);
-	yych = *YYCURSOR;
+	yych = *++YYCURSOR;
+	switch (yych) {
+		case 'a': goto yy32;
+		default: goto yy30;
+	}
+yy32:
+	yych = *++YYCURSOR;
 	switch (yych) {
 		case 'a': goto yy34;
 		default: goto yy33;
 	}
 yy33:
+#line 39 "repeater2.re"
+	{}
+#line 234 "repeater2.c"
 yy34:
 	yych = *++YYCURSOR;
 	switch (yych) {
@@ -244,118 +240,58 @@
 yy35:
 	yych = *++YYCURSOR;
 	switch (yych) {
-		case 'a': goto yy37;
-		default: goto yy36;
+		case 'a': goto yy36;
+		default: goto yy33;
 	}
 yy36:
-#line 43 "repeater2.re"
-	{}
-#line 254 "repeater2.c"
-yy37:
-	yych = *++YYCURSOR;
-	switch (yych) {
-		case 'a': goto yy38;
-		default: goto yy36;
-	}
-yy38:
-	yych = *++YYCURSOR;
-	switch (yych) {
-		case 'a': goto yy39;
-		default: goto yy36;
-	}
-yy39:
-	++YYCURSOR;
-	goto yy36;
-}
-#line 44 "repeater2.re"
-
-
-
-#line 275 "repeater2.c"
-{
-	YYCTYPE yych;
-	if ((YYLIMIT - YYCURSOR) < 5) YYFILL(5);
-	yych = *YYCURSOR;
-	switch (yych) {
-		case 'a': goto yy42;
-		default: goto yy41;
-	}
-yy41:
-yy42:
-	yych = *++YYCURSOR;
-	switch (yych) {
-		case 'a': goto yy43;
-		default: goto yy41;
-	}
-yy43:
-	yych = *++YYCURSOR;
-	switch (yych) {
-		case 'a': goto yy44;
-		default: goto yy41;
-	}
-yy44:
-	yych = *++YYCURSOR;
-	switch (yych) {
-		case 'a': goto yy45;
-		default: goto yy41;
-	}
-yy45:
-	yych = *++YYCURSOR;
-	switch (yych) {
-		case 'a': goto yy46;
-		default: goto yy41;
-	}
-yy46:
 	++YYCURSOR;
-#line 47 "repeater2.re"
-	{}
-#line 313 "repeater2.c"
+	goto yy33;
 }
-#line 48 "repeater2.re"
+#line 40 "repeater2.re"
 
 
 
-#line 319 "repeater2.c"
+#line 255 "repeater2.c"
 {
 	YYCTYPE yych;
 	if ((YYLIMIT - YYCURSOR) < 5) YYFILL(5);
 	yych = *YYCURSOR;
 	switch (yych) {
-		case 'a': goto yy49;
-		default: goto yy48;
+		case 'a': goto yy39;
+		default: goto yy38;
 	}
-yy48:
-yy49:
+yy38:
+yy39:
 	yych = *++YYCURSOR;
 	switch (yych) {
-		case 'a': goto yy50;
-		default: goto yy48;
+		case 'a': goto yy40;
+		default: goto yy38;
 	}
-yy50:
+yy40:
 	yych = *++YYCURSOR;
 	switch (yych) {
-		case 'a': goto yy51;
-		default: goto yy48;
+		case 'a': goto yy41;
+		default: goto yy38;
 	}
-yy51:
+yy41:
 	yych = *++YYCURSOR;
 	switch (yych) {
-		case 'a': goto yy52;
-		default: goto yy48;
+		case 'a': goto yy42;
+		default: goto yy38;
 	}
-yy52:
+yy42:
 	yych = *++YYCURSOR;
 	switch (yych) {
-		case 'a': goto yy53;
-		default: goto yy48;
+		case 'a': goto yy43;
+		default: goto yy38;
 	}
-yy53:
+yy43:
 	++YYCURSOR;
-#line 51 "repeater2.re"
+#line 43 "repeater2.re"
 	{}
-#line 357 "repeater2.c"
+#line 293 "repeater2.c"
 }
-#line 52 "repeater2.re"
+#line 44 "repeater2.re"
 
 repeater2.re:6:0: warning: control flow is undefined for strings that match '[\x0-\x60\x62-\xFF]', use default rule '*' [-Wundefined-control-flow]
 repeater2.re:10:0: warning: control flow is undefined for strings that match 
@@ -372,19 +308,11 @@ repeater2.re:22:0: warning: control flow is undefined for strings that match
 , use default rule '*' [-Wundefined-control-flow]
 repeater2.re:31:10: warning: rule matches empty string [-Wmatch-empty-string]
 repeater2.re:34:0: warning: control flow is undefined for strings that match '[\x0-\x60\x62-\xFF]', use default rule '*' [-Wundefined-control-flow]
-repeater2.re:38:0: warning: control flow is undefined for strings that match '[\x0-\x60\x62-\xFF]', use default rule '*' [-Wundefined-control-flow]
-repeater2.re:42:0: warning: control flow is undefined for strings that match 
-	'[\x0-\x60\x62-\xFF]'
-	'\x61 [\x0-\x60\x62-\xFF]'
-, use default rule '*' [-Wundefined-control-flow]
-repeater2.re:46:0: warning: control flow is undefined for strings that match 
+repeater2.re:38:0: warning: control flow is undefined for strings that match 
 	'[\x0-\x60\x62-\xFF]'
 	'\x61 [\x0-\x60\x62-\xFF]'
-	'\x61 \x61 [\x0-\x60\x62-\xFF]'
-	'\x61 \x61 \x61 [\x0-\x60\x62-\xFF]'
-	'\x61 \x61 \x61 \x61 [\x0-\x60\x62-\xFF]'
 , use default rule '*' [-Wundefined-control-flow]
-repeater2.re:50:0: warning: control flow is undefined for strings that match 
+repeater2.re:42:0: warning: control flow is undefined for strings that match 
 	'[\x0-\x60\x62-\xFF]'
 	'\x61 [\x0-\x60\x62-\xFF]'
 	'\x61 \x61 [\x0-\x60\x62-\xFF]'

diff --git a/test/repeater2.re b/test/repeater2.re
@@ -31,10 +31,6 @@
 	"a"{0,1} {}
 */
 
-/*!re2c
-	"a"{1,0} {}
-*/
-
 /*!re2c
 	"a"{1,1} {}
 */
@@ -43,10 +39,6 @@
 	"a"{2,5} {}
 */
 
-/*!re2c
-	"a"{5,2} {}
-*/
-
 /*!re2c
 	"a"{5,5} {}
 */