Lazarus
Announcements => Third party => Topic started by: AlexTP on September 16, 2019, 12:39:22 pm
-
EncConv is simplified code of LConvEncoding from Lazarus.
it uses the same codepage tables.
changes:
- type of encoding ID is now enum
- removed using of FPC string convertion API, removed using of Unix iconv;
so unit must work the same way on all platforms and FPC versions
- removed optional changing of CodePage for resulting string
- renamed most of functions, except simple functions like UTF8ToNNN, NNNToUTF8
- added EncConvErrorMode global var, which has 4 possible values
(LConvEncoding global var had 2 values):
raise exception, skip chars, replace chars with '?', return empty string
- removed parameter Encoded of functions
author: Alexey Torgashin
https://github.com/Alexey-T/EncConv
-
Looks like Pre FPC 3.0 code. It stuffs encodings into bytes without setting the encoding.
These are not conventions that are used anymore.
-
The code even fails in trunk.
Alex what version are you using?
-
@marcov
Code was needed for ATSynEdit, which holds list of string (TFPGList<StringItem> which has string). so FPC 3 codepages were not needed for me.
@Thaddy
I tested it on Linux in FPC 3.2 and 3.3 2019/04/26.
-
Updated FPC 3.3 from svn today. Again all works.
Demo must show "ok" for tests 1,2 and show "failed" for test3 (it must fail by design).
-
Updated the component:
+ add: support encoding iso-8859-7
+ add: support encoding iso-8859-10
+ add: support encoding iso-8859-13
+ add: support encoding iso-8859-5
+ add: support encoding iso-8859-14
+ add: support encoding iso-8859-16
I also send the patches of these encodings to LConvEncoding, but for those ppl who don't use Lazarus trunk, it's not possible to use these last changes.
Last 3 encodings ( -10, -7 and -13 ) are not merged to Laz.
So use EncConv.
-
+ add: support encoding iso-8859-3.
Anybody needs more ISO-8859 encodings?
Any other encodings (1-byte codepages)?