comparison: src/json.c
src/json.c
- changeset 1156
- 96f16b5a0029
- parent 1152
- e4af44b488bc
- child 1158
- fa2811e9ab19
equal
deleted
inserted
replaced
| 1155:b77d56a27e9c | 1156:96f16b5a0029 |
|---|---|
| 351 } | 351 } |
| 352 | 352 |
| 353 return CX_JSON_INCOMPLETE_DATA; | 353 return CX_JSON_INCOMPLETE_DATA; |
| 354 } | 354 } |
| 355 | 355 |
| 356 // converts a unicode (up to U+FFFF) codepoint to utf8 | 356 // converts a Unicode codepoint to utf8 |
| 357 static int codepoint_to_utf8(uint32_t codepoint, char *output_buf) { | 357 static unsigned codepoint_to_utf8(uint32_t codepoint, char *output_buf) { |
| 358 if (codepoint <= 0x7F) { | 358 if (codepoint <= 0x7F) { |
| 359 *output_buf = (char)codepoint; | 359 *output_buf = (char)codepoint; |
| 360 return 1; | 360 return 1; |
| 361 } else if (codepoint <= 0x7FF) { | 361 } else if (codepoint <= 0x7FF) { |
| 362 output_buf[0] = (char)(0xC0 | ((codepoint >> 6) & 0x1F)); | 362 output_buf[0] = (char)(0xC0 | ((codepoint >> 6) & 0x1F)); |
| 373 output_buf[2] = (char)(0x80 | ((codepoint >> 6) & 0x3F)); | 373 output_buf[2] = (char)(0x80 | ((codepoint >> 6) & 0x3F)); |
| 374 output_buf[3] = (char)(0x80 | (codepoint & 0x3F)); | 374 output_buf[3] = (char)(0x80 | (codepoint & 0x3F)); |
| 375 return 4; | 375 return 4; |
| 376 } | 376 } |
| 377 | 377 |
| 378 return 0; | 378 return 0; // LCOV_EXCL_LINE |
| 379 } | 379 } |
| 380 | 380 |
| 381 // converts a utf16 surrogate pair to utf8 | 381 // converts a utf16 surrogate pair to utf8 |
| 382 static inline uint32_t utf16pair_to_codepoint(uint16_t c0, uint16_t c1) { | 382 static inline uint32_t utf16pair_to_codepoint(uint16_t c0, uint16_t c1) { |
| 383 return ((c0 - 0xD800) << 10) + (c1 - 0xDC00) + 0x10000; | 383 return ((c0 - 0xD800) << 10) + (c1 - 0xDC00) + 0x10000; |
| 396 char c = str.ptr[i]; | 396 char c = str.ptr[i]; |
| 397 if (u) { | 397 if (u) { |
| 398 u = false; | 398 u = false; |
| 399 if (c == 'n') { | 399 if (c == 'n') { |
| 400 c = '\n'; | 400 c = '\n'; |
| 401 } else if (c == '"') { | |
| 402 c = '"'; | |
| 401 } else if (c == 't') { | 403 } else if (c == 't') { |
| 402 c = '\t'; | 404 c = '\t'; |
| 403 } else if (c == 'r') { | 405 } else if (c == 'r') { |
| 404 c = '\r'; | 406 c = '\r'; |
| 405 } else if (c == '\\') { | 407 } else if (c == '\\') { |
| 409 } else if (c == 'f') { | 411 } else if (c == 'f') { |
| 410 c = '\f'; | 412 c = '\f'; |
| 411 } else if (c == 'b') { | 413 } else if (c == 'b') { |
| 412 c = '\b'; | 414 c = '\b'; |
| 413 } else if (c == 'u') { | 415 } else if (c == 'u') { |
| 414 if (i+4 < str.length - 1) { | 416 if (i + 4 < str.length - 1) { |
| 415 cxstring codepoint_str = { str.ptr + i + 1, 4}; | 417 cxstring ustr1 = { str.ptr + i + 1, 4}; |
| 416 uint32_t codepoint; | 418 uint16_t utf16a, utf16b; |
| 417 if (!cx_strtou32_lc(codepoint_str, &codepoint, 16, "")) { | 419 char utf8buf[4]; |
| 418 char utf8buf[4]; | 420 unsigned utf8len = 0; |
| 419 int utf8len = 0; | 421 if (!cx_strtou16_lc(ustr1, &utf16a, 16, "")) { |
| 420 if (codepoint >= 0xD800 && codepoint <= 0xDFFF) { | 422 uint32_t codepoint; |
| 423 if (utf16a >= 0xD800 && utf16a <= 0xDFFF) { | |
| 421 // character is encoded as a surrogate pair | 424 // character is encoded as a surrogate pair |
| 422 // get next 6 bytes | 425 // get next 6 bytes |
| 423 if (i + 10 < str.length - 1) { | 426 if (i + 10 < str.length - 1) { |
| 424 char *surrogate2 = str.ptr+i+5; | 427 if (*(str.ptr+i+5) == '\\' && *(str.ptr+i+6) == 'u') { |
| 425 if (surrogate2[0] == '\\' && surrogate2[1] == 'u') { | 428 cxstring ustr2 = { str.ptr+i+7, 4 }; |
| 426 cxstring c2_str = { surrogate2 + 2, 4 }; | 429 if (!cx_strtou16_lc(ustr2, &utf16b, 16, "") |
| 427 uint32_t c2; | 430 && utf16b >= 0xDC00 && utf16b <= 0xDFFF) { |
| 428 if (!cx_strtou32_lc(c2_str, &c2, 16, "")) { | 431 codepoint = utf16pair_to_codepoint(utf16a, utf16b); |
| 429 codepoint = utf16pair_to_codepoint((uint16_t)codepoint, c2); | |
| 430 utf8len = codepoint_to_utf8(codepoint, utf8buf); | 432 utf8len = codepoint_to_utf8(codepoint, utf8buf); |
| 431 i += 6; | 433 i += 10; |
| 432 } | 434 } |
| 433 } | 435 } |
| 434 } | 436 } |
| 435 } else { | 437 } else { |
| 436 // character is in the Basic Multilingual Plane | 438 // character is in the Basic Multilingual Plane |
| 437 // and encoded as a single utf16 char | 439 // and encoded as a single utf16 char |
| 440 codepoint = utf16a; | |
| 438 utf8len = codepoint_to_utf8(codepoint, utf8buf); | 441 utf8len = codepoint_to_utf8(codepoint, utf8buf); |
| 442 i += 4; | |
| 439 } | 443 } |
| 440 if(utf8len > 0) { | 444 } |
| 441 // add all bytes from utf8buf expect the last char | 445 if(utf8len > 0) { |
| 442 // to the result | 446 // add all bytes from utf8buf except the last char |
| 443 utf8len--; | 447 // to the result (last char will be added below) |
| 444 c = utf8buf[utf8len]; | 448 utf8len--; |
| 445 for(int x=0;x<utf8len;x++) { | 449 c = utf8buf[utf8len]; |
| 446 result.ptr[result.length++] = utf8buf[x]; | 450 for (unsigned x = 0; x < utf8len; x++) { |
| 447 } | 451 result.ptr[result.length++] = utf8buf[x]; |
| 448 } | 452 } |
| 449 i += 4; | 453 } else { |
| 454 // decoding failed, ignore the entire sequence | |
| 455 result.ptr[result.length++] = '\\'; | |
| 450 } | 456 } |
| 451 } | 457 } |
| 452 } | 458 } else { |
| 453 | 459 // TODO: discuss the behavior for unrecognized escape sequences |
| 454 // TODO: discuss the behavior for unrecognized escape sequences | 460 // most parsers throw an error here - we just ignore it |
| 455 // most parsers throw an error here | 461 result.ptr[result.length++] = '\\'; |
| 462 } | |
| 463 | |
| 456 result.ptr[result.length++] = c; | 464 result.ptr[result.length++] = c; |
| 457 } else { | 465 } else { |
| 458 if (c == '\\') { | 466 if (c == '\\') { |
| 459 u = true; | 467 u = true; |
| 460 } else { | 468 } else { |
