Cannot import shapefile due to encoding problem

[miguelluaces]

A shapefile that could be imported without problems in shp2pgsql 2.1.7 cannot be imported in shp2pgsql 2.3.0. The execution fails with the following error:

Unable to convert data value to UTF-8 (iconv reports "Illegal byte sequence"). Current encoding is "WINDOWS-1252". Try "LATIN1" (Western European), or one of the values described at ​http://www.postgresql.org/docs/current/static/multibyte.html.

The commandline was:

d:\PostgreSQL\9.5\bin\shp2pgsql.exe -W UTF-8 establecimiento.shp

[miguelluaces]

Shapefile extracted from OpenStreetMap that worked without problems in shp2pgsql 2.1.7

[robe]

As the note says, try LATIN1.

This works fine for me with:

shp2pgsql.exe -W LATIN1 establecimiento.shp 

I can't see why UTF8 would ever work with this file in 2.1.7

[miguelluaces]

I should have written explicitly that the dataset is UTF-8, not LATIN1. If you open the dbf file in LibreOffice using Latin1 as encoding you can see that row number 3 is shown as

Muíño Daniel o Ferreiro

instead of :

Muíño Daniel o Ferreiro

What should I do now? Compile it myself and set a breakpoint to see which is the offending character? I would rather have the error messages improved to show me which is the offending row so that I can try to fix it.

And again, the same shapefile can be imported with shp2pgsql 2.1.7 with the same commandline without errors.