Re: [JPP-Devel] csv plugin merging?
Michael, do you keep in mind to update the readme.txt accordingly if you modify PLUS? thanks.. ede On 11.08.2012 22:18, Michaël Michaud wrote: > Hi Stefan, > > I made small changes to the csv driver and uploaded the sources in the > plugin repository > > - now auto mode import empty GeometryCollection if no geometry column is > detected > - options mode import empty GeometryCollection if x column or y column > are set to '-' > (dash character, after the sixteenth column) > > - I did not do anything for data type detection. Feel free to patch. > > - the jar file is in the dist directory (csv-driver-0.6.5.jar). > I'll put it in the ext directory for OJ PLUS version after you have > given it a try > > Michaël > >> Hi Michael, >> >>> I suggest to add the feature you ask in the automatic mode : >>> no geometry found -> create empty GeometryCollection >> ok, sounds good, but that should be documented somewhere. >> However, I used the manual settings option, just to get sure. So having >> it there maybe helpful too. >> >>> What do you mean ? I can save a layer with no feature in JML and I can >>> save a layer >>> with a single GEOMETRYCOLLECTION EMPTY. >> I created null geometry points for the csv with no geometry info using: >> >> Coordinate cord = null; >> Geometry defaultGeometry = new GeometryFactory().createPoint(cord); >> >> maybe that was the mistake. I will change it to empty >> GeometryCollection. Thanks for the hint! >> >>> Did you try to save a feature with a null geometry ? How did you create it ? >>> I don't think we should authorize it. >> mhm.. not sure? >> May point is that I think it would be a good idea to allow handling of >> non geometry datasets (= tables) and saving them to JML? but the empty >> GeomentryCollection would solve that (more or less). >> just a thought >> >>> Yes, it is a possible improvement. There are some reasons why it is not >>> already done : >> yes.. I see the reasons too. >> Actually, I am not sure if the type casting is a good way too... need to >> check what happens in the case that a column has no entries in some >> places but "0" can be there too. I.e. avoid the conversion "null" => 0. >> >> -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ ___ Jump-pilot-devel mailing list Jump-pilot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel
Re: [JPP-Devel] csv plugin merging?
Hi Ede, > Michael, > > do you keep in mind to update the readme.txt accordingly if you modify PLUS? Thanks for the reminder, I'll add it. Michaël > > thanks.. ede > > > On 11.08.2012 22:18, Michaël Michaud wrote: >> Hi Stefan, >> >> I made small changes to the csv driver and uploaded the sources in the >> plugin repository >> >> - now auto mode import empty GeometryCollection if no geometry column is >> detected >> - options mode import empty GeometryCollection if x column or y column >> are set to '-' >> (dash character, after the sixteenth column) >> >> - I did not do anything for data type detection. Feel free to patch. >> >> - the jar file is in the dist directory (csv-driver-0.6.5.jar). >> I'll put it in the ext directory for OJ PLUS version after you have >> given it a try >> >> Michaël >> >>> Hi Michael, >>> I suggest to add the feature you ask in the automatic mode : no geometry found -> create empty GeometryCollection >>> ok, sounds good, but that should be documented somewhere. >>> However, I used the manual settings option, just to get sure. So having >>> it there maybe helpful too. >>> What do you mean ? I can save a layer with no feature in JML and I can save a layer with a single GEOMETRYCOLLECTION EMPTY. >>> I created null geometry points for the csv with no geometry info using: >>> >>> Coordinate cord = null; >>> Geometry defaultGeometry = new GeometryFactory().createPoint(cord); >>> >>> maybe that was the mistake. I will change it to empty >>> GeometryCollection. Thanks for the hint! >>> Did you try to save a feature with a null geometry ? How did you create it ? I don't think we should authorize it. >>> mhm.. not sure? >>> May point is that I think it would be a good idea to allow handling of >>> non geometry datasets (= tables) and saving them to JML? but the empty >>> GeomentryCollection would solve that (more or less). >>> just a thought >>> Yes, it is a possible improvement. There are some reasons why it is not already done : >>> yes.. I see the reasons too. >>> Actually, I am not sure if the type casting is a good way too... need to >>> check what happens in the case that a column has no entries in some >>> places but "0" can be there too. I.e. avoid the conversion "null" => 0. >>> >>> > -- > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > ___ > Jump-pilot-devel mailing list > Jump-pilot-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel > > -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ ___ Jump-pilot-devel mailing list Jump-pilot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel
[JPP-Devel] new OffsetCurve plugin for OpenJUMP
Hi all, Just added this plugin (should be available in next NB). Firstly, I thought it was available in JTS, but the JTS OffsetCurveBuilder produce a rough offset curve often containing self-intersection and needs to be post-processed to get something more usable. Here is the process : - compute rough offset curve (distance=d) - node the result (with UnaryUnionOp) - remove components located at a distance < d from the source (this is the tricky part as the computed offset curve itself may be slightly inside the distance - I estimated the error and took 1 more percent to be sur no valid component are discarded, but it may still fails in some situations like using join mitre with a small limit or using a fixed precision model) - merge the remaining linestrings Here are some results -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/___ Jump-pilot-devel mailing list Jump-pilot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel
Re: [JPP-Devel] new OffsetCurve plugin for OpenJUMP
Hi, What about non simple lines and multilinestrings ? I just added a patch to return the rough offset curve in case of non simple input line. For multilinestrings, offset are computed individually for each component, then combined. Not sure if this is the best semantic for these corner cases, any suggestion is welcome. Michaël In the upper right corner, I added a non simple line, and in the lower right corner, the same geometry after noding (producing a multi linestring) Le 13/08/2012 00:01, Michaël Michaud a écrit : Hi all, Just added this plugin (should be available in next NB). Firstly, I thought it was available in JTS, but the JTS OffsetCurveBuilder produce a rough offset curve often containing self-intersection and needs to be post-processed to get something more usable. Here is the process : - compute rough offset curve (distance=d) - node the result (with UnaryUnionOp) - remove components located at a distance < d from the source (this is the tricky part as the computed offset curve itself may be slightly inside the distance - I estimated the error and took 1 more percent to be sur no valid component are discarded, but it may still fails in some situations like using join mitre with a small limit or using a fixed precision model) - merge the remaining linestrings -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ ___ Jump-pilot-devel mailing list Jump-pilot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel -- Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/___ Jump-pilot-devel mailing list Jump-pilot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel
