comparison: src/json.c
src/json.c
- changeset 1152
- e4af44b488bc
- parent 1151
- 60113356a7de
- child 1156
- 96f16b5a0029
equal
deleted
inserted
replaced
| 1151:60113356a7de | 1152:e4af44b488bc |
|---|---|
| 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 static int codepoint_to_utf8(uint32_t codepoint, char *output_buf) { | 357 static int codepoint_to_utf8(uint32_t codepoint, char *output_buf) { |
| 357 if (codepoint <= 0x7F) { | 358 if (codepoint <= 0x7F) { |
| 358 *output_buf = (char)codepoint; | 359 *output_buf = (char)codepoint; |
| 359 return 1; | 360 return 1; |
| 360 } else if (codepoint <= 0x7FF) { | 361 } else if (codepoint <= 0x7FF) { |
| 364 } else if (codepoint <= 0xFFFF) { | 365 } else if (codepoint <= 0xFFFF) { |
| 365 output_buf[0] = (char)(0xE0 | ((codepoint >> 12) & 0x0F)); | 366 output_buf[0] = (char)(0xE0 | ((codepoint >> 12) & 0x0F)); |
| 366 output_buf[1] = (char)(0x80 | ((codepoint >> 6) & 0x3F)); | 367 output_buf[1] = (char)(0x80 | ((codepoint >> 6) & 0x3F)); |
| 367 output_buf[2] = (char)(0x80 | (codepoint & 0x3F)); | 368 output_buf[2] = (char)(0x80 | (codepoint & 0x3F)); |
| 368 return 3; | 369 return 3; |
| 370 } else if (codepoint <= 0x10FFFF) { | |
| 371 output_buf[0] = (char)(0xF0 | ((codepoint >> 18) & 0x07)); | |
| 372 output_buf[1] = (char)(0x80 | ((codepoint >> 12) & 0x3F)); | |
| 373 output_buf[2] = (char)(0x80 | ((codepoint >> 6) & 0x3F)); | |
| 374 output_buf[3] = (char)(0x80 | (codepoint & 0x3F)); | |
| 375 return 4; | |
| 369 } | 376 } |
| 370 | 377 |
| 371 return 0; | 378 return 0; |
| 379 } | |
| 380 | |
| 381 // converts a utf16 surrogate pair to utf8 | |
| 382 static inline uint32_t utf16pair_to_codepoint(uint16_t c0, uint16_t c1) { | |
| 383 return ((c0 - 0xD800) << 10) + (c1 - 0xDC00) + 0x10000; | |
| 372 } | 384 } |
| 373 | 385 |
| 374 static cxmutstr unescape_string(const CxAllocator *a, cxmutstr str) { | 386 static cxmutstr unescape_string(const CxAllocator *a, cxmutstr str) { |
| 375 // note: this function expects that str contains the enclosing quotes! | 387 // note: this function expects that str contains the enclosing quotes! |
| 376 | 388 |
| 400 c = '\b'; | 412 c = '\b'; |
| 401 } else if (c == 'u') { | 413 } else if (c == 'u') { |
| 402 if (i+4 < str.length - 1) { | 414 if (i+4 < str.length - 1) { |
| 403 cxstring codepoint_str = { str.ptr + i + 1, 4}; | 415 cxstring codepoint_str = { str.ptr + i + 1, 4}; |
| 404 uint32_t codepoint; | 416 uint32_t codepoint; |
| 405 if(!cx_strtou32_lc(codepoint_str, &codepoint, 16, "")) { | 417 if (!cx_strtou32_lc(codepoint_str, &codepoint, 16, "")) { |
| 406 char utf8buf[4]; | 418 char utf8buf[4]; |
| 407 int utf8len = codepoint_to_utf8(codepoint, utf8buf); | 419 int utf8len = 0; |
| 420 if (codepoint >= 0xD800 && codepoint <= 0xDFFF) { | |
| 421 // character is encoded as a surrogate pair | |
| 422 // get next 6 bytes | |
| 423 if (i + 10 < str.length - 1) { | |
| 424 char *surrogate2 = str.ptr+i+5; | |
| 425 if (surrogate2[0] == '\\' && surrogate2[1] == 'u') { | |
| 426 cxstring c2_str = { surrogate2 + 2, 4 }; | |
| 427 uint32_t c2; | |
| 428 if (!cx_strtou32_lc(c2_str, &c2, 16, "")) { | |
| 429 codepoint = utf16pair_to_codepoint((uint16_t)codepoint, c2); | |
| 430 utf8len = codepoint_to_utf8(codepoint, utf8buf); | |
| 431 i += 6; | |
| 432 } | |
| 433 } | |
| 434 } | |
| 435 } else { | |
| 436 // character is in the Basic Multilingual Plane | |
| 437 // and encoded as a single utf16 char | |
| 438 utf8len = codepoint_to_utf8(codepoint, utf8buf); | |
| 439 } | |
| 408 if(utf8len > 0) { | 440 if(utf8len > 0) { |
| 409 // add all bytes from utf8buf expect the last char | 441 // add all bytes from utf8buf expect the last char |
| 410 // to the result | 442 // to the result |
| 411 utf8len--; | 443 utf8len--; |
| 412 c = utf8buf[utf8len]; | 444 c = utf8buf[utf8len]; |
| 413 for(int i=0;i<utf8len;i++) { | 445 for(int x=0;x<utf8len;x++) { |
| 414 result.ptr[result.length++] = utf8buf[i]; | 446 result.ptr[result.length++] = utf8buf[x]; |
| 415 } | 447 } |
| 416 } | 448 } |
| 417 i += 4; | 449 i += 4; |
| 418 } | 450 } |
| 419 } | 451 } |
