GB18030Encoder.js 3.4 KB

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465666768697071727374757677
  1. import { encoderError } from "../../encoding/encodings";
  2. import { finished } from "../../encoding/finished";
  3. import { index, indexGB18030RangesPointerFor, indexPointerFor } from "../../encoding/indexes";
  4. import { end_of_stream, isASCIICodePoint } from "../../encoding/terminology";
  5. /**
  6. * @constructor
  7. * @implements {Encoder}
  8. * @param {{fatal: boolean}} options
  9. * @param {boolean=} gbk_flag
  10. */
  11. export class GB18030Encoder {
  12. constructor(options, gbk_flag = undefined) {
  13. this.gbk_flag = gbk_flag;
  14. this.fatal = options.fatal;
  15. // gb18030's decoder has an associated gbk flag (initially unset).
  16. }
  17. /**
  18. * @param {Stream} stream Input stream.
  19. * @param {number} code_point Next code point read from the stream.
  20. * @return {(number|!Array.<number>)} Byte(s) to emit.
  21. */
  22. handler(stream, code_point) {
  23. // 1. If code point is end-of-stream, return finished.
  24. if (code_point === end_of_stream)
  25. return finished;
  26. // 2. If code point is an ASCII code point, return a byte whose
  27. // value is code point.
  28. if (isASCIICodePoint(code_point))
  29. return code_point;
  30. // 3. If code point is U+E5E5, return error with code point.
  31. if (code_point === 0xE5E5)
  32. return encoderError(code_point);
  33. // 4. If the gbk flag is set and code point is U+20AC, return
  34. // byte 0x80.
  35. if (this.gbk_flag && code_point === 0x20AC)
  36. return 0x80;
  37. // 5. Let pointer be the index pointer for code point in index
  38. // gb18030.
  39. let pointer = indexPointerFor(code_point, index('gb18030'));
  40. // 6. If pointer is not null, run these substeps:
  41. if (pointer !== null) {
  42. // 1. Let lead be Math.floor(pointer / 190) + 0x81.
  43. const lead = Math.floor(pointer / 190) + 0x81;
  44. // 2. Let trail be pointer % 190.
  45. const trail = pointer % 190;
  46. // 3. Let offset be 0x40 if trail is less than 0x3F and 0x41 otherwise.
  47. const offset = trail < 0x3F ? 0x40 : 0x41;
  48. // 4. Return two bytes whose values are lead and trail + offset.
  49. return [lead, trail + offset];
  50. }
  51. // 7. If gbk flag is set, return error with code point.
  52. if (this.gbk_flag)
  53. return encoderError(code_point);
  54. // 8. Set pointer to the index gb18030 ranges pointer for code
  55. // point.
  56. pointer = indexGB18030RangesPointerFor(code_point);
  57. // 9. Let byte1 be Math.floor(pointer / 10 / 126 / 10).
  58. const byte1 = Math.floor(pointer / 10 / 126 / 10);
  59. // 10. Set pointer to pointer − byte1 × 10 × 126 × 10.
  60. pointer = pointer - byte1 * 10 * 126 * 10;
  61. // 11. Let byte2 be Math.floor(pointer / 10 / 126).
  62. const byte2 = Math.floor(pointer / 10 / 126);
  63. // 12. Set pointer to pointer − byte2 × 10 × 126.
  64. pointer = pointer - byte2 * 10 * 126;
  65. // 13. Let byte3 be Math.floor(pointer / 10).
  66. const byte3 = Math.floor(pointer / 10);
  67. // 14. Let byte4 be pointer − byte3 × 10.
  68. const byte4 = pointer - byte3 * 10;
  69. // 15. Return four bytes whose values are byte1 + 0x81, byte2 +
  70. // 0x30, byte3 + 0x81, byte4 + 0x30.
  71. return [byte1 + 0x81,
  72. byte2 + 0x30,
  73. byte3 + 0x81,
  74. byte4 + 0x30];
  75. }
  76. }
  77. //# sourceMappingURL=GB18030Encoder.js.map