fix gbk encodings on windows (#17843)

This commit is contained in:
flywind
2021-04-26 20:48:25 +08:00
committed by GitHub
parent 68e522ecec
commit 01bca8cd4f
2 changed files with 26 additions and 0 deletions

View File

@@ -72,6 +72,7 @@ when defined(windows):
(875, "cp875"), # IBM EBCDIC Greek Modern
(932, "shift_jis"), # ANSI/OEM Japanese; Japanese (Shift-JIS)
(936, "gb2312"), # ANSI/OEM Simplified Chinese (PRC, Singapore); Chinese Simplified (GB2312)
(936, "gbk"), # Alias for GB2312 encoding
(949, "ks_c_5601-1987"), # ANSI/OEM Korean (Unified Hangul Code)
(950, "big5"), # ANSI/OEM Traditional Chinese (Taiwan; Hong Kong SAR, PRC); Chinese Traditional (Big5)
(1026, "IBM1026"), # IBM EBCDIC Turkish (Latin 5)

View File

@@ -0,0 +1,25 @@
import std/encodings
var fromGBK = open("utf-8", "gbk")
var toGBK = open("gbk", "utf-8")
var fromGB2312 = open("utf-8", "gb2312")
var toGB2312 = open("gb2312", "utf-8")
block:
let data = "\215\237\186\243\178\187\214\170\204\236\212\218\203\174\163\172\194\250\180\178\208\199\195\206\209\185\208\199\186\211"
doAssert fromGBK.convert(data) == "醉后不知天在水,满床星梦压星河"
block:
let data = "万两黄金容易得,知心一个也难求"
doAssert toGBK.convert(data) == "\205\242\193\189\187\198\189\240\200\221\210\215\181\195\163\172\214\170\208\196\210\187\184\246\210\178\196\209\199\243"
block:
let data = "\215\212\208\197\200\203\201\250\182\254\176\217\196\234\163\172\187\225\181\177\203\174\187\247\200\253\199\167\192\239"
doAssert fromGB2312.convert(data) == "自信人生二百年,会当水击三千里"
block:
let data = "谁怕?一蓑烟雨任平生"
doAssert toGB2312.convert(data) == "\203\173\197\194\163\191\210\187\203\242\209\204\211\234\200\206\198\189\201\250"