[QGIS-Developer] Loading many Excel-Files in QGIS with correct encoding

[Andrea Giudiceandrea]

Il 16/07/2021 11:51, Andreas Neumann ha scritto:
> Hi all and thanks for the discussion,
> 
> Even and I discovered that the .xls files are loading with correct 
> encoding in Linux but not on Windows which leads to the assumption that 
> freexl on Windows might not have been correctly compiled/linked against 
> iconv.

Hi Andreas,
usually on GNU/Linux distributions the "system" encoding is UTF-8, while 
on Windows is the "ANSI" encoding (e.g. Windows-1252 for Western Europe 
locales).

So I think that since XLS layers are imported in QGIS using the "system" 
encoding, and since FreeXL translates the strings values to UTF-8 and 
GDAL/OGR exposes such strings values as they are passed by FreeXL, then 
XLS tables are correctly displayed in QGIS on GNU/Linux systems, while 
are incorrectly displayed in QGIS on Windows systems.

You can test your xls files on Windows using ogrinfo in the OSGeo4W Shell:

ogrinfo path\to\file.xls sheet_name > path\to\file.txt

then you can check if file.txt will contain the string cell values UTF-8 
encoded.

Or you can use the ogr module from osgeo in the QGIS Python Console on 
Windows to open the XLS datasource and output the features fields 
(cells) values as strings: the cells values are correctly read by ogr 
(via FreeXL) from the XLS file and correctly displayed as UTF-8 strings.

AFAIK FreeXL depends by iconv libary and cannot be compiled and cannot 
work at all without it.

In fact, in my QGIS installations via OSGeo4W (old and new repository) 
both the FreeXL and iconv DLLs (FreeXL 1.0.2 / iconv 1.14 - FreeXL 1.0.6 
/ iconv 1.16) are present, with the former linked to the latter as 
dependency.

Regards.

Andrea