Libreoffice convert writer to text Windows , Linux

English
#1

When I run the following command on Windows
soffice.exe --headless --convert-to txt:Text --outdir C:\C#\odt2txt “R:\Greek\Greek Lemon Roast Chicken.odt”

The text utf-8 characters are corrupted.

When I run the following command on Linux
libreoffice --headless --convert-to txt:Text --outdir C:\C#\odt2txt “R:\Greek\Greek Lemon Roast Chicken.odt”

The utf-8 characters are correct. The Region on Windows was changed to UTF-8, this made no difference. When I run libreoffice on linux under python, the output text files are correct. When I run libreoffice on linux under dotnet c# the text output is corrupted.

Why is the text corrupted on Windows and c#. I much prefer programming with c#, but will use python if I have to.

Thank you.

#4

I was using notepad++, for some reason does not work with utf8. After changing the Windows region to utf8, the characters display normally in Notepad.

#2

In which way? What is the original text, and what exactly (byte values) is saved? May it be that it saves in UTF-16 instead of UTF-8, because that is default text encoding for Unicode on Windows? (not using Windows myself so I can’t really tell).

#3

utf8 is a compatable subset of utf16, should make no difference. unicode characters corrupted are

00E9, 2309, 2303.

I was using notepad++, for some reason does not work with utf8. After changing the Windows region to utf8, the characters display normally in Notepad.

#5

No, it is absolutely not. UTF-8 and UTF-16 are different encodings of Unicode characters, specifically UTF-16 uses two-byte values, 16-bit, hence UTF-16, and UTF-8 uses one-byte values, 8-bit, hence UTF-8. Both encodings are capable to encode the full Unicode character range.

1 Like
#6

Notepad++ is working with UTF8 and UTF16 maybe the auto-detect was misleading. BOM set?

Impressum (Legal Info) | Datenschutzerklärung (Privacy Policy)   Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.