hare

commit b7f1d4226cb5e5ba3e2852294e8ec2a52d21129a
parent 475e50fe3d81bb50983e6b122e7a3629967b6cd2
Author: Sebastian <sebastian@sebsite.pw>
Date:   Wed, 13 Apr 2022 20:33:48 -0400

ascii: abort strcasecmp when non-ASCII encountered

Signed-off-by: Sebastian <sebastian@sebsite.pw>

Diffstat:
M
ascii/strcmp.ha
 | 
28
+++++++---------------------
M
format/xml/parser.ha
 | 
20
++++++++------------
M
net/dns/decode.ha
 | 
10
++++------

3 files changed, 19 insertions(+), 39 deletions(-)
diff --git a/ascii/strcmp.ha b/ascii/strcmp.ha
@@ -5,25 +5,18 @@ use strings;
 
 // Compares two strings by their ASCII sort order, treating all capital letters
 // as their lowercase counterpart (i.e. a case-insensitive comparison is
-// performed). If either string is not entirely composed of ASCII characters,
-// void is returned. Otherwise, zero is returned if the strings are equal, a
-// negative value if a is less than b, or a positive value if a is greater than
-// b.
-export fn strcasecmp(a: str, b: str) (int | void) = {
+// performed). Zero is returned if the strings are equal, a negative value if a
+// is less than b, or a positive value if a is greater than b. Aborts if a
+// non-ASCII byte is encountered.
+export fn strcasecmp(a: str, b: str) int = {
 	let a = strings::iter(a), b = strings::iter(b);
 	for (true) {
 		let ra = match (strings::next(&a)) {
 		case void =>
 			match (strings::next(&b)) {
 			case void =>
-				return 0;
+				break;
 			case rune =>
-				for (true) match (strings::next(&b)) {
-				case void =>
-					break;
-				case let r: rune =>
-					if (!valid(r)) return;
-				};
 				return -1;
 			};
 		case let r: rune =>
@@ -31,24 +24,17 @@ export fn strcasecmp(a: str, b: str) (int | void) = {
 		};
 		let rb = match (strings::next(&b)) {
 		case void =>
-			for (true) match (strings::next(&a)) {
-			case void =>
-				break;
-			case let r: rune =>
-				if (!valid(r)) return;
-			};
 			return 1;
 		case let r: rune =>
 			yield r;
 		};
-		if (!valid(ra) || !valid(rb)) {
-			return;
-		};
+		assert(valid(ra) && valid(rb));
 		let ra = tolower(ra), rb = tolower(rb);
 		if (ra != rb) {
 			return ra: u32: int - rb: u32: int;
 		};
 	};
+	return 0;
 };
 
 @test fn strcmp() void = {
diff --git a/format/xml/parser.ha b/format/xml/parser.ha
@@ -467,13 +467,11 @@ fn prolog(par: *parser) (void | error) = {
 			return par.line: syntaxerr;
 		};
 		// XXX: Deliberate omission: all values other than utf-8
-		match (ascii::strcasecmp(attr.1, "utf-8")) {
-		case void =>
+		if (!ascii::validstr(attr.1)) {
+			return utf8::invalid;
+		};
+		if (ascii::strcasecmp(attr.1, "utf-8") != 0) {
 			return utf8::invalid;
-		case let n: int =>
-			if (n != 0) {
-				return utf8::invalid;
-			};
 		};
 	};
 
@@ -491,13 +489,11 @@ fn prolog(par: *parser) (void | error) = {
 			return par.line: syntaxerr;
 		};
 		// XXX: Deliberate omission: non-standalone documents
-		match (ascii::strcasecmp(attr.1, "yes")) {
-		case void =>
+		if (!ascii::validstr(attr.1)) {
+			return par.line: syntaxerr;
+		};
+		if (ascii::strcasecmp(attr.1, "yes") != 0) {
 			return par.line: syntaxerr;
-		case let n: int =>
-			if (n != 0) {
-				return par.line: syntaxerr;
-			};
 		};
 	};
 
diff --git a/net/dns/decode.ha b/net/dns/decode.ha
@@ -109,15 +109,13 @@ fn decode_name(dec: *decoder) ([]str | format) = {
 			break;
 		};
 
-		const name = dec.cur[..z];
+		const name = strings::fromutf8(dec.cur[..z]);
 		dec.cur = dec.cur[z..];
-		for (let i = 0z; i < len(name); i += 1) {
-			if (!ascii::valid(name[i]: u32: rune)) {
-				return format;
-			};
+		if (!ascii::validstr(name)) {
+			return format;
 		};
 
-		append(names, strings::dup(strings::fromutf8(name)));
+		append(names, strings::dup(name));
 	};
 	return names;
 };