php实现的一个UTF8编码转Unicode的函数

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 {

12. iCode = strUtf8.charCodeAt(nOffset);
13. if ((iCode & 0×80) == 0) // 1 byte.
14. {
15. if ( nRemainingBytes break;

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

17. nOffset ++;
18. nRemainingBytes -= 1;
19. }
20. else if ((iCode & 0xE0) == 0xC0) // 2 bytes
21. {
22. iCode1 = strUtf8.charCodeAt(nOffset + 1);
23. if ( nRemainingBytes (iCode1 & 0xC0) != 0×80 ) // invalid pattern
24. {
25. break;
26. }

27. bstr += String.fromCharCode(((iCode & 0×3F) nOffset += 2;

28. nRemainingBytes -= 2;
29. }
30. else if ((iCode & 0xF0) == 0xE0) // 3 bytes
31. {
32. iCode1 = strUtf8.charCodeAt(nOffset + 1);
33. iCode2 = strUtf8.charCodeAt(nOffset + 2);
34. if ( nRemainingBytes (iCode1 & 0xC0) != 0×80 || // invalid pattern
35. (iCode2 & 0xC0) != 0×80 )
36. {
37. break;
38. }

39. bstr += String.fromCharCode(((iCode & 0×0F) ((iCode1 & 0×3F) (iCode2 & 0×3F));

40. nOffset += 3;
41. nRemainingBytes -= 3;
42. }
43. else // 4 or more bytes — unsupported
44. break;
45. }

46. if (nRemainingBytes != 0)

47. {
48. // bad UTF8 string.
49. return “”;
50. }

51. return bstr;

52. }
53. ?>

复制代码