Add missing (GBK, for example) character sets to our ICU data file.

extern/icu/Readme.txt

@@ -2,13 +2,23 @@ icu.exe is a self-extract pre-built (by us) IBM ICU 52.1 library.
 
 The sources was downloaded from http://site.icu-project.org/download.
 
-Two customized data files was created using the tool at http://apps.icu-project.org/datacustom.
+Two customized data files was created using the tool at http://apps.icu-project.org/datacustom/ICUData52.html.
 
 One was created empty and saved to icudt52l_empty.dat. This file is used when building ICU, to
 make icudt a little stub. The file was saved in ICU sources directory at <icu>\source\data\in
 with name icudt52l.dat, before start ICU building process.
 
 The other is the data file we use in runtime. It was created selecting the options:
-- Charset Mapping Tables: ibm-943_P15A-2003
-- Collators
-- Base data
+- Charset Mapping Tables:
+    gb18030
+    ibm-1386_P100-2001
+    ibm-942_P12A-1999
+    ibm-943_P130-1999
+    ibm-943_P15A-2003
+    icu-internal-25546
+    windows-874-2000
+    windows-936-2000
+
+- Collators: all
+
+- Base data: all