How to encode shapefiles from LATIN1 to UTF-8?

Question

Is there a way to encode shapefiles from LATIN1 to UTF-8?

score 29 · Answer 1 · edited Feb 11 '13 at 01:17

29

Assuming that you are on Linux and have access to a recent version of GDAL you can try the following (from this post) :

 export SHAPE_ENCODING="ISO-8859-1"
 ogr2ogr output.shp input -lco ENCODING=UTF-8

Note: LATIN1 should work too instead of ISO-8859-1.

In Windows, do NOT set the SHAPE_ENCODING, ogr2ogr does not recognize ISO-8859-1, nor LATIN1.

edited Feb 11 '13 at 01:17

flup

answered Oct 19 '11 at 12:36

unicoletti

Just to clarify, “recent” here seems to mean 1.9.x. I tried this with 1.8.1 and it did not work, tried again with 1.9.1 and it did. – Michal Migurski May 26 '12 at 04:59
Didn't work for me on Windows. – Pavel Radzivilovsky Nov 21 '12 at 21:16
Got it to work in Windows, have updated the answer. – flup Feb 10 '13 at 23:29
Do you know if I can look-up the list of valid values for SHAPE_ENCODING somewhere? – JJD Feb 28 '14 at 13:30
1

@JJD They are the same used by iconv: http://www.gnu.org/software/libiconv/ – unicoletti Mar 02 '14 at 15:46
Maybe on Windows one could use: SET SHAPE_ENCODING="ISO-8859-1" in the command-line before calling ogr2ogr? Just a hunch. – Rauni Feb 16 '16 at 14:49
@Rauni: I did, omitting the double quotes and it worked. Some time ago, but this still appears the only way to convince ogr2ogr that the input .dbf has LATIN1/WIN1252/ISO-8859-1 encoding. LATIN1 and WIN1252 did not work. Use ISO-8859-1 – Jan Jan 07 '23 at 14:18

score 17 · Accepted Answer · edited Oct 19 '11 at 20:23

17

You can do this with QGIS. First load your shapefiles then save as and coding select dropdown menu utf-8 and then save your file.

edited Oct 19 '11 at 20:23

whuber

answered Oct 19 '11 at 12:36

urcm

9

Word of caution: for this solution to work qgis must be started in a LATIN1/ISO-8859-1 environment when you open the input shapefile or it won't read it correctly. If you are working in an UTF8 enviroment the conversion might fail silently, corrupting the data, depending on the data in the shapefile. BTW, most OSes now default to UTF8. – unicoletti Oct 19 '11 at 15:48
1

Note that in some situations it may not work, I had this kind of issues: (http://osgeo-org.1560.n6.nabble.com/UTF-8-no-more-td4647456.html) – JaakL Jul 16 '12 at 05:48
I am trying to save single shape file from system to utf-8 but it is not working. http://gis.stackexchange.com/questions/219092/how-to-change-shapefile-data-source-encoding-directly – Muhammad Faizan Khan Nov 25 '16 at 10:09

score 5 · Answer 3 · edited Apr 13 '17 at 12:34

If my understanding of Shapefiles is correct, the part that is encoding-sensitive is the attribute database contained in the DBF file.

Here is an answer that helped me solve this issue using LibreOffice Calc/Base:

It has the additional benefit that you can open the DBF file using a distinct character set and thus find out what the current encoding actually is.

3 Answers3