Bug report #2496
Error trying to make distance matrix for "ID fields" with non ascii characters
| Status: | Closed | ||
|---|---|---|---|
| Priority: | Low | ||
| Assignee: |  cfarmer - | ||
| Category: | Python plugins | ||
| Affected QGIS version: | Regression?: | No | |
| Operating System: | All | Easy fix?: | No |
| Pull Request or Patch supplied: | Resolution: | fixed | |
| Crashes QGIS or corrupts data: | Copied to github as #: | 12556 |
Description
1. Make point layer from admin.shp with Mean Coordinates tool
2. Using created point layer and poi-osm, try to use Distance Matrix tool using CATEGORY as a target field.
I'm getting this error, probably because CATEGORY contains values in cyrrilics.
See shape-files attached.
An error has occured while executing Python code:
Traceback (most recent call last):
File "c:\\OSGeo4w\\apps\\qgis-dev\\python\\plugins\\fTools\\tools\\doPointDistance.py", line 105, in accept
self.compute(point1, point2, field1, field2, outPath, matType, nearest, self.progressBar)
File "c:\\OSGeo4w\\apps\\qgis-dev\\python\\plugins\\fTools\\tools\\doPointDistance.py", line 150, in compute
self.linearMatrix(writer, provider1, provider2, index1, index2, nearest, distArea, matType, sindex, progressBar)
File "c:\\OSGeo4w\\apps\\qgis-dev\\python\\plugins\\fTools\\tools\\doPointDistance.py", line 204, in linearMatrix
if matType == "Linear": writer.writerow([unicode(inID), unicode(outID), float(dist)])
[[UnicodeEncodeError]]: 'ascii' codec can't encode characters in position 0-3: ordinal not in range(128)
shape.zip (74.1 KB)
Screenshot.png (52 KB)
Screenshot-1.png (72.8 KB)
fix_2496.diff 
- Proposed patch
(1.24 KB)
Associated revisions
Updates distance matrix tool to allow non ascii characters in id fields. This is based on a workaround (from http://docs.python.org/library/csv.html), as the default Python csv function(s) do not support nonascii characters. Fixes #2496
git-svn-id: http://svn.osgeo.org/qgis/trunk/qgis@13019 c8812cc2-4d05-0410-92ff-de0c093fc19c
Updates distance matrix tool to allow non ascii characters in id fields. This is based on a workaround (from http://docs.python.org/library/csv.html), as the default Python csv function(s) do not support nonascii characters. Fixes #2496
git-svn-id: http://svn.osgeo.org/qgis/trunk@13019 c8812cc2-4d05-0410-92ff-de0c093fc19c
now supports fields containing non-ascii chars. Fixes #2496. Patch from alexbruy
git-svn-id: http://svn.osgeo.org/qgis/trunk/qgis@13478 c8812cc2-4d05-0410-92ff-de0c093fc19c
now supports fields containing non-ascii chars. Fixes #2496. Patch from alexbruy
git-svn-id: http://svn.osgeo.org/qgis/trunk@13478 c8812cc2-4d05-0410-92ff-de0c093fc19c
History
#1
Updated by Giovanni Manghi over 14 years ago
Hi,
I'm trying to replicate here under linux but I don't see any Cyrillic char in the category column of the poi-osm layer you posted.
So... with the attached data making the matrix using just poi-osm works fine, and also making the matrix between poi-osm and the "mean points".
The real problem seems that the distance matrix GUI shows a wrong list of fields for the "input unique ID field" and/or for the "target unique ID field" when the input/target point layer is the "mean" one:
in fact it shows his "UID" field (the "real" type fields are not considered by this tool) but also the fields of the poi-osm are available. See screenshot.
If you choose the "wrong" unique ID field for the "mean" layer (belonging to the other layer) then a python error is thrown (see 2nd screenshot).
#2
Updated by Maxim Dubinin over 14 years ago
I'm trying to replicate here under linux but I don't see any Cyrillic char in the category column of the poi-osm layer you posted.
Hmm, please check this screenshot, it does have cyrrilics
http://gis-lab.info/images/screenshots/20100306-bu0-101kb.jpg
Please note my setup for this tool also:
http://gis-lab.info/images/screenshots/20100306-3mo-37kb.jpg
I guess there are multiple issues here.
#3
Updated by Giovanni Manghi over 14 years ago
Hmm, please check this screenshot, it does have cyrrilics
http://gis-lab.info/images/screenshots/20100306-bu0-101kb.jpg
Hmmm... then probably depends about the encondig I used to open the vector.
Please note my setup for this tool also:
http://gis-lab.info/images/screenshots/20100306-3mo-37kb.jpg
Ok, so the issues I described are confirmed. We can open a new ticket or change the title of this one.
#4
Updated by Giovanni Manghi over 14 years ago
The same happens with non ascii chars, like "ç" or "õ". The error is the same
An error has occured while executing Python code:
Traceback (most recent call last):
File "/usr/share/qgis/python/plugins/fTools/tools/doPointDistance.py", line 105, in accept
self.compute(point1, point2, field1, field2, outPath, matType, nearest, self.progressBar)
File "/usr/share/qgis/python/plugins/fTools/tools/doPointDistance.py", line 152, in compute
self.regularMatrix(writer, provider1, provider2, index1, index2, nearest, distArea, sindex, progressBar)
File "/usr/share/qgis/python/plugins/fTools/tools/doPointDistance.py", line 173, in regularMatrix
writer.writerow(data)
[[UnicodeEncodeError]]: 'ascii' codec can't encode characters in position 3-4: ordinal not in range(128)
#5
Updated by cfarmer - over 14 years ago
- Resolution set to fixed
- Status changed from Open to Closed
A fix has been made in 13019. This fix is based on a workaround (from http://docs.python.org/library/csv.html), as the default Python csv function(s) do not support non-ascii characters. As such, please test and confirm that the revision has indeed fixed the problem.
-Carson
#6
Updated by Giovanni Manghi over 14 years ago
Replying to [comment:5 cfarmer]:
A fix has been made in 13019. This fix is based on a workaround (from http://docs.python.org/library/csv.html), as the default Python csv function(s) do not support non-ascii characters. As such, please test and confirm that the revision has indeed fixed the problem.
-Carson
the problem seem solved, thanks!
The problem regarding the wrong columns being shown is still there, but I believe that a different ticket has been filed meanwhile.
#7
Updated by Alexander Bruy over 14 years ago
- Status changed from Closed to Feedback
- Resolution deleted (
fixed)
Problem with non ASCII characters in fields values still exists. Tested under Linux with UTF-8 locale and QGIS 13404. Here is error log
Traceback (most recent call last):
File "/usr/local/qgis-svn/share/qgis/python/plugins/fTools/tools/doPointDistance.py", line 143, in accept
self.compute(point1, point2, field1, field2, outPath, matType, nearest, self.progressBar)
File "/usr/local/qgis-svn/share/qgis/python/plugins/fTools/tools/doPointDistance.py", line 188, in compute
self.linearMatrix(writer, provider1, provider2, index1, index2, nearest, distArea, matType, sindex, progressBar)
File "/usr/local/qgis-svn/share/qgis/python/plugins/fTools/tools/doPointDistance.py", line 242, in linearMatrix
if matType == "Linear": writer.writerow([unicode(inID), unicode(outID), float(dist)])
File "/usr/local/qgis-svn/share/qgis/python/plugins/fTools/tools/doPointDistance.py", line 64, in writerow
self.writer.writerow(row)
[[UnicodeEncodeError]]: 'ascii' codec can't encode characters in position 0-10: ordinal not in range(128)
Patch added
#8
Updated by cfarmer - over 14 years ago
- Status changed from Feedback to Closed
- Resolution set to fixed
Patch applied in b2dc3549 (SVN r13479).
Carson