当前位置：Gxlcms > PHP教程 > php实现的一个UTF8编码转Unicode的函数

php实现的一个UTF8编码转Unicode的函数

时间：2021-07-01 10:21:17 帮助过：22人阅读

  3. function Utf8ToUnicode(strUtf8)

  4. {
  5. var bstr = “”;
  6. var nTotalChars = strUtf8.length; // total chars to be processed.
  7. var nOffset = 0; // processing point on strUtf8
  8. var nRemainingBytes = nTotalChars; // how many bytes left to be converted
  9. var nOutputPosition = 0;
  10. var iCode, iCode1, iCode2; // the value of the unicode.

  11. while (nOffset < nTotalChars)

  12. {
  13. iCode = strUtf8.charCodeAt(nOffset);
  14. if ((iCode & 0×80) == 0) // 1 byte.
  15. {
  16. if ( nRemainingBytes < 1 ) // not enough data
  17. break;

  18. bstr += String.fromCharCode(iCode & 0×7F);

  19. nOffset ++;
  20. nRemainingBytes -= 1;
  21. }
  22. else if ((iCode & 0xE0) == 0xC0) // 2 bytes
  23. {
  24. iCode1 = strUtf8.charCodeAt(nOffset + 1);
  25. if ( nRemainingBytes < 2 || // not enough data
  26. (iCode1 & 0xC0) != 0×80 ) // invalid pattern
  27. {
  28. break;
  29. }

  30. bstr += String.fromCharCode(((iCode & 0×3F) << 6) | ( iCode1 & 0×3F));

  31. nOffset += 2;
  32. nRemainingBytes -= 2;
  33. }
  34. else if ((iCode & 0xF0) == 0xE0) // 3 bytes
  35. {
  36. iCode1 = strUtf8.charCodeAt(nOffset + 1);
  37. iCode2 = strUtf8.charCodeAt(nOffset + 2);
  38. if ( nRemainingBytes < 3 || // not enough data
  39. (iCode1 & 0xC0) != 0×80 || // invalid pattern
  40. (iCode2 & 0xC0) != 0×80 )
  41. {
  42. break;
  43. }

  44. bstr += String.fromCharCode(((iCode & 0×0F) << 12) |

  45. ((iCode1 & 0×3F) << 6) |
  46. (iCode2 & 0×3F));
  47. nOffset += 3;
  48. nRemainingBytes -= 3;
  49. }
  50. else // 4 or more bytes — unsupported
  51. break;
  52. }

  53. if (nRemainingBytes != 0)

  54. {
  55. // bad UTF8 string.
  56. return “”;
  57. }

  58. return bstr;

  59. }
  60. ?>

人气教程排行

本站所有资源全部来源于网络，若本站发布的内容侵害到您的隐私或者利益，请联系我们删除！