Thu, 22 Jan 2015 13:21:57 +0100
Incorporate requested changes from Mozilla in review:
https://bugzilla.mozilla.org/show_bug.cgi?id=1123480#c6
michael@0 | 1 | /* This Source Code Form is subject to the terms of the Mozilla Public |
michael@0 | 2 | * License, v. 2.0. If a copy of the MPL was not distributed with this |
michael@0 | 3 | * file, You can obtain one at http://mozilla.org/MPL/2.0/. */ |
michael@0 | 4 | // ======================================================================= |
michael@0 | 5 | // Original Author: Yueheng Xu |
michael@0 | 6 | // email: yueheng.xu@intel.com |
michael@0 | 7 | // phone: (503)264-2248 |
michael@0 | 8 | // Intel Corporation, Oregon, USA |
michael@0 | 9 | // Last Update: September 7, 1999 |
michael@0 | 10 | // Revision History: |
michael@0 | 11 | // 09/07/1999 - initial version. |
michael@0 | 12 | // 09/28/1999 - changed leftbyte and rightbyte from char to unsigned char |
michael@0 | 13 | // in struct DByte |
michael@0 | 14 | // 04/10/1999 - changed leftbyte. rightbyte to uint8_t in struct DByte; |
michael@0 | 15 | // added table UnicodeToGBKTable[0x5200] |
michael@0 | 16 | // |
michael@0 | 17 | // 05/16/2000 - added gUnicodeToGBKTableInitialized flag for optimization |
michael@0 | 18 | // ====================================================================================== |
michael@0 | 19 | // Table GBKToUnicode[] maps the GBK code to its unicode. |
michael@0 | 20 | // The mapping data of this GBK table is obtained from |
michael@0 | 21 | // ftp://ftp.unicode.org/Public/MAPPINGS/VENDORS/MICSFT/WINDOWS/CP936.TXT |
michael@0 | 22 | // Frank Tang of Netscape wrote the original perl tool to re-align the |
michael@0 | 23 | // mapping data into an 8-item per line format ( i.e. file cp936map.txt ). |
michael@0 | 24 | // |
michael@0 | 25 | // The valid GBK charset range: left byte is [0x81, 0xfe], right byte are |
michael@0 | 26 | // [0x40, 0x7e] and [0x80, 0xfe]. But for the convenience of index |
michael@0 | 27 | // calculation, the table here has a single consecutive range of |
michael@0 | 28 | // [0x40, 0xfe] for the right byte. Those invalid chars whose right byte |
michael@0 | 29 | // is 0x7f will be mapped to undefined unicode 0xFFFF. |
michael@0 | 30 | // |
michael@0 | 31 | // |
michael@0 | 32 | // Table UnicodeToGBK[] maps the unicode to GBK code. To reduce memory usage, we |
michael@0 | 33 | // only do Unicode to GBK table mapping for unicode between 0x4E00 and 0xA000; |
michael@0 | 34 | // Others let converter to do search from table GBKToUnicode[]. If we want further |
michael@0 | 35 | // trade memory for performance, we can let more unicode to do table mapping to get |
michael@0 | 36 | // its GBK instead of searching table GBKToUnicode[]. |
michael@0 | 37 | #ifndef _GBKU_H__ |
michael@0 | 38 | #define _GBKU_H__ |
michael@0 | 39 | |
michael@0 | 40 | |
michael@0 | 41 | #define UCS2_NO_MAPPING ((char16_t) 0xfffd) |
michael@0 | 42 | #define UINT8_IN_RANGE(a, b, c) \ |
michael@0 | 43 | (((uint8_t)(a) <= (uint8_t)(b))&&((uint8_t)(b) <= (uint8_t)(c))) |
michael@0 | 44 | #define UNICHAR_IN_RANGE(a, b, c) \ |
michael@0 | 45 | (((char16_t)(a) <= (char16_t)(b))&&((char16_t)(b) <= (char16_t)(c))) |
michael@0 | 46 | #define CAST_CHAR_TO_UNICHAR(a) ((char16_t)((unsigned char)(a))) |
michael@0 | 47 | #define CAST_UNICHAR_TO_CHAR(a) ((char)a) |
michael@0 | 48 | |
michael@0 | 49 | #define IS_ASCII(a) (0==(0xff80 & (a))) |
michael@0 | 50 | #define IS_GBK_EURO(c) ((char)0x80 == (c)) |
michael@0 | 51 | #define UCS2_EURO ((char16_t) 0x20ac) |
michael@0 | 52 | |
michael@0 | 53 | #include "nsGBKConvUtil.h" |
michael@0 | 54 | |
michael@0 | 55 | #endif /* _GBKU_H__ */ |