authorgravatar for marc@tiehu.isMarc Tiehuis <marc@tiehu.is> 2017-10-27 03:00:23+13:00
committergravatar for andrew@ziglang.orgAndrew Kelley <andrew@ziglang.org> 2017-10-26 10:00:23-04:00
log66636381957f214f1acc22dcea01cb4cd1032649
treee74fdbb0df53d8c9d30fc04b3af4fc735732ecab
parentf4ca3482f1954eb0bdd0c378206d3a7b8e178f55

Improve invalid character error messages (#566)

See #544

2 files changed, 70 insertions(+), 9 deletions(-)

src/tokenizer.cpp+47-9
......@@ -416,6 +416,44 @@ static void handle_string_escape(Tokenize *t, uint8_t c) {
416416 }
417417}
418418
419static const char* get_escape_shorthand(uint8_t c) {
420 switch (c) {
421 case '\0':
422 return "\\0";
423 case '\a':
424 return "\\a";
425 case '\b':
426 return "\\b";
427 case '\t':
428 return "\\t";
429 case '\n':
430 return "\\n";
431 case '\v':
432 return "\\v";
433 case '\f':
434 return "\\f";
435 case '\r':
436 return "\\r";
437 default:
438 return nullptr;
439 }
440}
441
442static void invalid_char_error(Tokenize *t, uint8_t c) {
443 if (c == '\r') {
444 tokenize_error(t, "invalid carriage return, only '\\n' line endings are supported");
445 } else if (isprint(c)) {
446 tokenize_error(t, "invalid character: '%c'", c);
447 } else {
448 const char *sh = get_escape_shorthand(c);
449 if (sh) {
450 tokenize_error(t, "invalid character: '%s'", sh);
451 } else {
452 tokenize_error(t, "invalid character: '\\x%x'", c);
453 }
454 }
455}
456
419457void tokenize(Buf *buf, Tokenization *out) {
420458 Tokenize t = {0};
421459 t.out = out;
......@@ -580,7 +618,7 @@ void tokenize(Buf *buf, Tokenization *out) {
580618 t.state = TokenizeStateSawQuestionMark;
581619 break;
582620 default:
583 tokenize_error(&t, "invalid character: '%c'", c);
621 invalid_char_error(&t, c);
584622 }
585623 break;
586624 case TokenizeStateSawQuestionMark:
......@@ -890,7 +928,7 @@ void tokenize(Buf *buf, Tokenization *out) {
890928 t.state = TokenizeStateLineString;
891929 break;
892930 default:
893 tokenize_error(&t, "invalid character: '%c'", c);
931 invalid_char_error(&t, c);
894932 break;
895933 }
896934 break;
......@@ -919,7 +957,7 @@ void tokenize(Buf *buf, Tokenization *out) {
919957 break;
920958 case '\\':
921959 if (t.cur_tok->data.str_lit.is_c_str) {
922 tokenize_error(&t, "invalid character: '%c'", c);
960 invalid_char_error(&t, c);
923961 }
924962 t.state = TokenizeStateLineStringContinue;
925963 break;
......@@ -949,7 +987,7 @@ void tokenize(Buf *buf, Tokenization *out) {
949987 buf_append_char(&t.cur_tok->data.str_lit.str, '\n');
950988 break;
951989 default:
952 tokenize_error(&t, "invalid character: '%c'", c);
990 invalid_char_error(&t, c);
953991 break;
954992 }
955993 break;
......@@ -1073,7 +1111,7 @@ void tokenize(Buf *buf, Tokenization *out) {
10731111 handle_string_escape(&t, '\"');
10741112 break;
10751113 default:
1076 tokenize_error(&t, "invalid character: '%c'", c);
1114 invalid_char_error(&t, c);
10771115 }
10781116 break;
10791117 case TokenizeStateCharCode:
......@@ -1147,7 +1185,7 @@ void tokenize(Buf *buf, Tokenization *out) {
11471185 t.state = TokenizeStateStart;
11481186 break;
11491187 default:
1150 tokenize_error(&t, "invalid character: '%c'", c);
1188 invalid_char_error(&t, c);
11511189 }
11521190 break;
11531191 case TokenizeStateZero:
......@@ -1189,7 +1227,7 @@ void tokenize(Buf *buf, Tokenization *out) {
11891227 uint32_t digit_value = get_digit_value(c);
11901228 if (digit_value >= t.radix) {
11911229 if (is_symbol_char(c)) {
1192 tokenize_error(&t, "invalid character: '%c'", c);
1230 invalid_char_error(&t, c);
11931231 }
11941232 // not my char
11951233 t.pos -= 1;
......@@ -1233,7 +1271,7 @@ void tokenize(Buf *buf, Tokenization *out) {
12331271 uint32_t digit_value = get_digit_value(c);
12341272 if (digit_value >= t.radix) {
12351273 if (is_symbol_char(c)) {
1236 tokenize_error(&t, "invalid character: '%c'", c);
1274 invalid_char_error(&t, c);
12371275 }
12381276 // not my char
12391277 t.pos -= 1;
......@@ -1282,7 +1320,7 @@ void tokenize(Buf *buf, Tokenization *out) {
12821320 uint32_t digit_value = get_digit_value(c);
12831321 if (digit_value >= t.radix) {
12841322 if (is_symbol_char(c)) {
1285 tokenize_error(&t, "invalid character: '%c'", c);
1323 invalid_char_error(&t, c);
12861324 }
12871325 // not my char
12881326 t.pos -= 1;
test/compile_errors.zig+23
......@@ -2252,4 +2252,27 @@ pub fn addCases(cases: &tests.CompileErrorContext) {
22522252 \\}
22532253 ,
22542254 ".tmp_source.zig:9:13: error: type '&MyType' does not support field access");
2255
2256 cases.add("carriage return special case",
2257 "fn test() -> bool {\r\n" ++
2258 " true\r\n" ++
2259 "}\r\n"
2260 ,
2261 ".tmp_source.zig:1:20: error: invalid carriage return, only '\\n' line endings are supported");
2262
2263 cases.add("non-printable invalid character",
2264 "\xff\xfe" ++
2265 \\fn test() -> bool {\r
2266 \\ true\r
2267 \\}
2268 ,
2269 ".tmp_source.zig:1:1: error: invalid character: '\\xff'");
2270
2271 cases.add("non-printable invalid character with escape alternative",
2272 "fn test() -> bool {\n" ++
2273 "\ttrue\n" ++
2274 "}\n"
2275 ,
2276 ".tmp_source.zig:2:1: error: invalid character: '\\t'");
2277
22552278}