Hi Ticker, okay, let's see where this goes... Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Dienstag, 11. Februar 2020 16:36 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Gerd To take advantage of efficiency gains based on knowing what is being asked for, ie: - do the polygons need to be merged or can we do one-by-one. - can we answer correctly even of !W.isComplete(). - can we stop early, eg ANY as soon as part is IN, ALL as soon as part is OUT, etc the top layers of code need to be with the method and its associated knowledge. I don't see any point in simply moving this into IsInUtil. I was going to take calcInsideness next and divide it into some library bits remaining in IsInUtil, and logic equivalent to the rest in IsInFunction. If you consider this is not the way to proceed, then I'd still like patch applied anyway, firstly because it contains other changes unrelated to this, secondly so that the code exists in SVN. I'll then immediately do another patch that removes the canStop logic etc and the POINT code that migrated here and restore it to just testing the composite flags. Ticker On Tue, 2020-02-11 at 14:11 +0000, Gerd Petermann wrote:

Hi all,

sorry, something is broken in the mail system. The latest posts don't appear on nabble (1) and my answers to existing threads are rejected.

@Ticker: I'd prefer to do the merge and I think it should be done in IsInUtil. Your patch is_in-function_v7.patch goes in the opposite direction. Not sure what you plan to do now?

Gerd

(1) http://gis.19327.n8.nabble.com/Mkgmap-Development-f5324443.html _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Ticker, whatever you plan to do. I moved the code to the lib because it is easier to write a unit test. I would not invest much time to avoid a few tests which only happen in very rare cases. Makes testing more complicated and code less readable. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Dienstag, 11. Februar 2020 16:43 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Gerd I assume you mean towards moving some logic from IsInUtil to IsInFunction. Ticker On Tue, 2020-02-11 at 15:41 +0000, Gerd Petermann wrote:

Hi Ticker,

okay, let's see where this goes...

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Dienstag, 11. Februar 2020 16:36 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

To take advantage of efficiency gains based on knowing what is being asked for, ie:

- do the polygons need to be merged or can we do one-by-one. - can we answer correctly even of !W.isComplete(). - can we stop early, eg ANY as soon as part is IN, ALL as soon as part is OUT, etc

the top layers of code need to be with the method and its associated knowledge. I don't see any point in simply moving this into IsInUtil.

I was going to take calcInsideness next and divide it into some library bits remaining in IsInUtil, and logic equivalent to the rest in IsInFunction.

If you consider this is not the way to proceed, then I'd still like patch applied anyway, firstly because it contains other changes unrelated to this, secondly so that the code exists in SVN. I'll then immediately do another patch that removes the canStop logic etc and the POINT code that migrated here and restore it to just testing the composite flags.

Ticker

On Tue, 2020-02-11 at 14:11 +0000, Gerd Petermann wrote:

...

Hi all,

sorry, something is broken in the mail system. The latest posts don't appear on nabble (1) and my answers to existing threads are rejected.

@Ticker: I'd prefer to do the merge and I think it should be done in IsInUtil. Your patch is_in-function_v7.patch goes in the opposite direction. Not sure what you plan to do now?

Gerd

(1) http://gis.19327.n8.nabble.com/Mkgmap-Development-f5324443.html _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

---

mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Ticker, did you run the unit tests? This should download a newer version of the samples. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Mittwoch, 12. Februar 2020 16:23 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Gerd Here it is - changes are: - Some restructuring with early stopping where possible. - Merging polygons for POINT IN/ON test so a point on shared boundary becomes IN rather than ON. - Not merging polygons when no need. - Make the function cacheable, so that there is negligible cost to the second call: highway=path & is_in(landuse, residential, all)=true [0xAA] highway=path & is_in(landuse, residential, all)=false [0xBB] - Improved the layout of documentation in the Style Manual. - Fixed quite a few problems. I've left quite a lot of debug in for the moment, I think there will still be work to do. It gives correct answers to 'point-on.osm'. I haven't worked through is -in-hook-sample-v3.osm yet because I wanted to get this revision out to replace faults in the previous versions. Ticker On Tue, 2020-02-11 at 15:49 +0000, Gerd Petermann wrote:

Hi Ticker,

whatever you plan to do. I moved the code to the lib because it is easier to write a unit test. I would not invest much time to avoid a few tests which only happen in very rare cases. Makes testing more complicated and code less readable.

Gerd

Hi Ticker, I found it difficult to test the real style function, that's why I moved all the logic out of it. I am looking forward to your solution. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Mittwoch, 12. Februar 2020 16:47 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Gerd The re-structuring makes this difficult. I propose a function in IsInUtilTest with the same interface as calcInsideness from IsInUtil that somehow drives the real function IsIn Function to collect and build the IN/ON/OUT intflag. Ticker On Wed, 2020-02-12 at 15:28 +0000, Gerd Petermann wrote:

Hi Ticker,

did you run the unit tests? This should download a newer version of the samples.

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Mittwoch, 12. Februar 2020 16:23 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

Here it is - changes are:

- Some restructuring with early stopping where possible.

- Merging polygons for POINT IN/ON test so a point on shared boundary becomes IN rather than ON.

- Not merging polygons when no need.

- Make the function cacheable, so that there is negligible cost to the second call: highway=path & is_in(landuse, residential, all)=true [0xAA] highway=path & is_in(landuse, residential, all)=false [0xBB]

- Improved the layout of documentation in the Style Manual.

- Fixed quite a few problems.

I've left quite a lot of debug in for the moment, I think there will still be work to do.

It gives correct answers to 'point-on.osm'. I haven't worked through is -in-hook-sample-v3.osm yet because I wanted to get this revision out to replace faults in the previous versions.

Ticker

On Tue, 2020-02-11 at 15:49 +0000, Gerd Petermann wrote:

...

Hi Ticker,

whatever you plan to do. I moved the code to the lib because it is easier to write a unit test. I would not invest much time to avoid a few tests which only happen in very rare cases. Makes testing more complicated and code less readable.

Gerd

mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Gerd Here is patch with: - IsInUtilTest driver using the is_in function with methods. - a slight re-formulation to the 'any_in_or_in' method. - updates to style manual for above. - some fixes and improvements relating to holes, ON etc. There is one more bit of logic that is needed for polygons: after the polygon has been found to be fully within a shape, and it isn't in a hole, check that a hole isn't within it. I'll so this in the next day or so. I'm having trouble with b14 (building is inner to a multi-polygon): Without merging the cut outer, isLineInShape for the b14 polygon returns IN/ON/OUT for the first component - I was expecting ON/OUT. With merging, it returns IN for the merged shape (good), but IN/ON for the hole - I was expecting ON. Ticker On Wed, 2020-02-12 at 15:55 +0000, Gerd Petermann wrote:

Hi Ticker,

I found it difficult to test the real style function, that's why I moved all the logic out of it. I am looking forward to your solution.

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Mittwoch, 12. Februar 2020 16:47 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

The re-structuring makes this difficult.

I propose a function in IsInUtilTest with the same interface as calcInsideness from IsInUtil that somehow drives the real function IsIn Function to collect and build the IN/ON/OUT intflag.

Ticker

On Wed, 2020-02-12 at 15:28 +0000, Gerd Petermann wrote:

...

Hi Ticker,

did you run the unit tests? This should download a newer version of the samples.

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Mittwoch, 12. Februar 2020 16:23 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

Here it is - changes are:

- Some restructuring with early stopping where possible.

- Merging polygons for POINT IN/ON test so a point on shared boundary becomes IN rather than ON.

- Not merging polygons when no need.

- Make the function cacheable, so that there is negligible cost to the second call: highway=path & is_in(landuse, residential, all)=true [0xAA] highway=path & is_in(landuse, residential, all)=false [0xBB]

- Improved the layout of documentation in the Style Manual.

- Fixed quite a few problems.

I've left quite a lot of debug in for the moment, I think there will still be work to do.

It gives correct answers to 'point-on.osm'. I haven't worked through is -in-hook-sample-v3.osm yet because I wanted to get this revision out to replace faults in the previous versions.

Ticker

On Tue, 2020-02-11 at 15:49 +0000, Gerd Petermann wrote:

...

Hi Ticker,

whatever you plan to do. I moved the code to the lib because it is easier to write a unit test. I would not invest much time to avoid a few tests which only happen in very rare cases. Makes testing more complicated and code less readable.

Gerd

---

mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Gerd The new meaning of any_in_or_on is for handing a line where some is OUT, none is IN and touching the edge is irrelevant and is tested as ()=false, with the +ve logic being "some-IN or all-ON". Can't stop on ON because need to have IN component. This was the the meaning I had originally intended, but at some point forgot. It is the reciprocal of the method for a line where some is IN and touching the edge makes no difference. Various timing tests on a 2 tile area with max-jobs=1, my normal options and a function invocation in lines. These are with my latest code: highway=* & length()<200 {add maxspeed="20 mph"} Total time taken: 1 minute 12 seconds highway=* & is_in(landuse,residential,any)=true {add maxspeed="20 mph"} Total time taken: 1 minute 14 seconds ..., all)... Total time taken: 1 minute 23 seconds ..., on)... Total time taken: 1 minute 21 seconds Taking out the automatic "stop early" code, timings are: highway=* & is_in(landuse,residential,any)=true {add maxspeed="20 mph"} Total time taken: 1 minute 17 seconds ..., all)... Total time taken: 1 minute 24 seconds ..., on)... Total time taken: 1 minute 23 seconds ie the "stop-early" gives a slight improvement for "all"/"on" and a bigger improvement for "any" I feel the "stop-early" mechanism is well hidden from the much more complex coding of isLineInShape() and how the results of this should be interpreted with multiple shapes and holes, etc, and so it is worth-while. I'm getting strange results for b14 just from the 2 parts of the outer after MP cutting, without the Java2D logic in mergePolygons being invoked. I'll investigate. Ticker On Mon, 2020-02-17 at 11:33 +0000, Gerd Petermann wrote:

Hi Ticker,

I don't understand the new meaning of any_in_or_on. Why can't you stop early when ON was found?

I still think that the "stop early" code is just adding complexity. Do you have a sample that shows how this improves performance?

Reg. unexpected results when joining shapes: Welcome to the club! Some not so obvious things: - The java area code sometimes removes points which are on straight lines. - a node calculated with e.g. makeBetweenPoint is only very close to the place where it should be

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Montag, 17. Februar 2020 11:31 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

Here is patch with: - IsInUtilTest driver using the is_in function with methods. - a slight re-formulation to the 'any_in_or_in' method. - updates to style manual for above. - some fixes and improvements relating to holes, ON etc.

There is one more bit of logic that is needed for polygons: after the polygon has been found to be fully within a shape, and it isn't in a hole, check that a hole isn't within it. I'll so this in the next day or so.

I'm having trouble with b14 (building is inner to a multi-polygon): Without merging the cut outer, isLineInShape for the b14 polygon returns IN/ON/OUT for the first component - I was expecting ON/OUT. With merging, it returns IN for the merged shape (good), but IN/ON for the hole - I was expecting ON.

Ticker

On Wed, 2020-02-12 at 15:55 +0000, Gerd Petermann wrote:

...

Hi Ticker,

I found it difficult to test the real style function, that's why I moved all the logic out of it. I am looking forward to your solution.

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Mittwoch, 12. Februar 2020 16:47 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

The re-structuring makes this difficult.

I propose a function in IsInUtilTest with the same interface as calcInsideness from IsInUtil that somehow drives the real function IsIn Function to collect and build the IN/ON/OUT intflag.

Ticker

On Wed, 2020-02-12 at 15:28 +0000, Gerd Petermann wrote:

...

Hi Ticker,

did you run the unit tests? This should download a newer version of the samples.

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Mittwoch, 12. Februar 2020 16:23 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

Here it is - changes are:

- Some restructuring with early stopping where possible.

- Merging polygons for POINT IN/ON test so a point on shared boundary becomes IN rather than ON.

- Not merging polygons when no need.

- Make the function cacheable, so that there is negligible cost to the second call: highway=path & is_in(landuse, residential, all)=true [0xAA] highway=path & is_in(landuse, residential, all)=false [0xBB]

- Improved the layout of documentation in the Style Manual.

- Fixed quite a few problems.

I've left quite a lot of debug in for the moment, I think there will still be work to do.

It gives correct answers to 'point-on.osm'. I haven't worked through is -in-hook-sample-v3.osm yet because I wanted to get this revision out to replace faults in the previous versions.

Ticker

On Tue, 2020-02-11 at 15:49 +0000, Gerd Petermann wrote:

...

Hi Ticker,

whatever you plan to do. I moved the code to the lib because it is easier to write a unit test. I would not invest much time to avoid a few tests which only happen in very rare cases. Makes testing more complicated and code less readable.

Gerd

---

mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Gerd A simpler way of expressing it in the Style Manual would be: +any_in_or_on+ - if is_in(...,any_in_or_on)=false, part is outside, none is inside. What do you think? An alternative would be to re-phrase the keyword as something like 'some-out-none-in', testing for =true instead. But, with the function being called 'is_in', I wanted to have the methods expressing IN'ness as far as possible. Rec. early-stop. It has ~4% improvement for 'any' in my scenario, ~2.5% for 'on' and ~1% for all. It just seems easy and sensible to stop when, if asking for 'any', a part is found IN, or, if asking for 'all', a part is found OUT, etc Ticker On Mon, 2020-02-17 at 17:13 +0000, Gerd Petermann wrote:

Hi Ticker,

reg. result: Please find a better explanation for this: "This is useful for the negative - is_in(...,any_in_or_on)=false - for processing a line that is outside the polgon(s) but might touch an edge." For me this sounds plain wrong. It would be okay without "but might touch an edge".

reg. performance: I did not mean to remove early stop completely, I just don't believe that it is worth to distinguish the different methods. My original code also stops early when result is 7 (IN+ON+OUT).

reg. special case b14: one reason why I don't like my code. It's frustrating. OTOH nobody seems to care about the results for polygons.

Gerd

Hi Gerd The two most common cases that I imagine users will want to match are lines inside the polygon(s) and lines outside. Inside lines will frequently meet the edge, and I don't think these should fail the inside test. Outside lines will, less frequently, meet the edge, and I don't think these should fail the outside test. Often there will be 2 lines, joining at edge of the polygon, that make a continuous highway that is the entrance to the polygon. Both of these will have an ON component. So, the flag test for inside lines should be "hasIn && !hasOut" and for outside lines should be "hasOut && !hasIn". Expressing the outside test as the negative, in keeping with the is_in() methods expressing in-ness, it becomes "hasIn || !hasOut". I had expressed this as "hasIn || !(hasIn || hasOut)", because "!(hasIn || hasOut)" is all_ON, and this matches the method being called, ambiguously, ANY_IN_OR_ON (ON_OR_ANY_IN would have been better) but they are logically equivalent. This expressing as in-ness seems to be the cause of these problems we are discussing and maybe now is the time to abandon it. I suggest replacing ANY_IN_OR_ON with SOME_OUT_NONE_IN, giving it the method string "none", like SOME_IN_NONE_OUT is referenced as "all". Ticker On Mon, 2020-02-17 at 19:16 +0000, Gerd Petermann wrote:

Hi Ticker,

I would except +any_in_or_on+ - if is_in(...,any_in_or_on)=false, all is outside, none is in or on the edge.

Or name it +all_out+ - if is_in(...,all_out)=true, all is outside, none is in or on the edge.

A method that returns true when all is either out or on the edge could be called +out-or-on+

Gerd

Gerd Petermann wrote

Ticker Berkin wrote

...

I suggest replacing ANY_IN_OR_ON with SOME_OUT_NONE_IN, giving it the method string "none", like SOME_IN_NONE_OUT is referenced as "all".

Yes, much better.

Thinking again about it. What would be the difference between is_in(..all)=true and is_in(..none)=false? My understanding is that we want to distinguish 5 or 6 cases. Three methods returning true or false should be enough for that. Gerd -- Sent from: http://gis.19327.n8.nabble.com/Mkgmap-Development-f5324443.html

Hi Gerd Here is a replacement for is_in-function_v10.patch with the additions: - method ..,none)=true to replace ..,any_in_or_on)=false. - revised documentation. - detection of holes within the rule.polygon. Will set OUT, ie stop 'all' being true, and, if exactly ON, not set IN, ie stop 'any' being true. - removal the kind parameters and related logic from IsInUtil; the IsInFunction driver has better understanding of how POLYGON processing should differ from POLYLINE. - restructure the MethodArg enum to include all the method dependent logic. Test case b14 (rule.polygon being the hole in the multiPolygon) should be OUT, or maybe OUT|ON but the test driver only understands ALL or ANY for polygons. It still fails because isLineInShape returns IN|ON|OUT for the rule.polygon against one of the cut simple polygons, where it should be ON|OUT Ticker On Tue, 2020-02-18 at 00:23 -0700, Gerd Petermann wrote:

Ticker Berkin wrote

...

I suggest replacing ANY_IN_OR_ON with SOME_OUT_NONE_IN, giving it the method string "none", like SOME_IN_NONE_OUT is referenced as "all".

Yes, much better.

Gerd

-- Sent from: http://gis.19327.n8.nabble.com/Mkgmap-Development-f5324443.html _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Ticker, sorry, presssed send too early. Reg. documentation: I think it is logically correct now :) I think the sentence "The method +all_in_or_on+ additionally matches lines are only on the edge of the polygon." should be changed to "Compared to +all+ the method +all_in_or_on+ additionally matches lines which are only on the edge of the polygon." Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Gerd Petermann <gpetermann_muenchen@hotmail.com> Gesendet: Donnerstag, 20. Februar 2020 16:41 An: Ticker Berkin; Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Ticker, I see that you overwrite the expected value stored in the test data in the unit test. Please don't do this. If you think that the expected value in is-in-samples.osm is wrong we should discuss the test data. In my eyes b14 clearly has points on the edge (as it is part of the edge) and is out. If you think the expected results are correct but your new code doesn't allow to test them because of the early stop code please add a new tag to each object or maybe create a new file. The unit test file is meant to document what the code does. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 20. Februar 2020 14:11 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Gerd Here is a replacement for is_in-function_v10.patch with the additions: - method ..,none)=true to replace ..,any_in_or_on)=false. - revised documentation. - detection of holes within the rule.polygon. Will set OUT, ie stop 'all' being true, and, if exactly ON, not set IN, ie stop 'any' being true. - removal the kind parameters and related logic from IsInUtil; the IsInFunction driver has better understanding of how POLYGON processing should differ from POLYLINE. - restructure the MethodArg enum to include all the method dependent logic. Test case b14 (rule.polygon being the hole in the multiPolygon) should be OUT, or maybe OUT|ON but the test driver only understands ALL or ANY for polygons. It still fails because isLineInShape returns IN|ON|OUT for the rule.polygon against one of the cut simple polygons, where it should be ON|OUT Ticker On Tue, 2020-02-18 at 00:23 -0700, Gerd Petermann wrote:

Ticker Berkin wrote

...

I suggest replacing ANY_IN_OR_ON with SOME_OUT_NONE_IN, giving it the method string "none", like SOME_IN_NONE_OUT is referenced as "all".

Yes, much better.

Gerd

-- Sent from: http://gis.19327.n8.nabble.com/Mkgmap-Development-f5324443.html _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Ticker, patch is commited. It is a bit difficult for me because you change a lot of things and the unit test fails. I just want to make sure that we have something testable in the end. It is already difficult enough to understand what is documented. I think the tests are not always working because the result of Coord.makeBetweenPoint is rounded. That means a point calculated with it is typically not ON the line between the given points. A possible solution would be to change all the code in IsInUtil to use double values and rewrite e.g. makeBetweenPoint and the other used methods to keep the double precision. Still, it might fail when java area code comes in because that always uses a flat earth calculation. When I understood that I felt indeed a bit relunctant. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 20. Februar 2020 17:45 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Gerd I don't think the test data 'expected' values are wrong, it is just that they are more specific than the 'method' mechanism allows to be differentiated; eg a polygon can only be tested for ALL in or ANY in. At the moment I feel you have a reluctance about the whole concept of the methods. Once the principle is accepted, I'll go through the test data and add, as another tag, the list of methods that should match the element, then change the test driver to check that these match and the other applicable methods don't. Reg. b14: It isn't the stop-early code that causes the problems, isLineInShape is not giving the correct answer for a simple polygon produced by the MP cutter. It would be quite easy to introduce some POLYGON 'on' methods, that match the outer, inner or either edge of a polygon, but maybe this could wait until there is a call for it. Next mail: I'll change the sentence as you suggest. Please can you commit the patch as it stands; it has a lot of good stuff in it. Then I can do the IsInUtilTest and test data changes as the next stage. It's also handy to see how the Style Manual looks after each build into the download area, because I don't know how to generate it and am just guessing at the formatting. Thank you Ticker On Thu, 2020-02-20 at 15:41 +0000, Gerd Petermann wrote:

Hi Ticker,

I see that you overwrite the expected value stored in the test data in the unit test. Please don't do this. If you think that the expected value in is-in-samples.osm is wrong we should discuss the test data. In my eyes b14 clearly has points on the edge (as it is part of the edge) and is out.

If you think the expected results are correct but your new code doesn't allow to test them because of the early stop code please add a new tag to each object or maybe create a new file. The unit test file is meant to document what the code does.

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 20. Februar 2020 14:11 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

Here is a replacement for is_in-function_v10.patch with the additions:

- method ..,none)=true to replace ..,any_in_or_on)=false.

- revised documentation.

- detection of holes within the rule.polygon. Will set OUT, ie stop 'all' being true, and, if exactly ON, not set IN, ie stop 'any' being true.

- removal the kind parameters and related logic from IsInUtil; the IsInFunction driver has better understanding of how POLYGON processing should differ from POLYLINE.

- restructure the MethodArg enum to include all the method dependent logic.

Test case b14 (rule.polygon being the hole in the multiPolygon) should be OUT, or maybe OUT|ON but the test driver only understands ALL or ANY for polygons. It still fails because isLineInShape returns IN|ON|OUT for the rule.polygon against one of the cut simple polygons, where it should be ON|OUT

Ticker

On Tue, 2020-02-18 at 00:23 -0700, Gerd Petermann wrote:

...

Ticker Berkin wrote

...

I suggest replacing ANY_IN_OR_ON with SOME_OUT_NONE_IN, giving it the method string "none", like SOME_IN_NONE_OUT is referenced as "all".

Yes, much better.

Gerd

-- Sent from: http://gis.19327.n8.nabble.com/Mkgmap-Development-f5324443.html _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Ticker, I use GpxCreator class for debugging. It produces output pairs of files with garmin 24 bit precision as well as the highprec 30 bits. Helps a lot to understand how OSM values are rounded and were calculated points are placed when you load them in JOSM. Gerd ________________________________________ Von: Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 20. Februar 2020 18:37 An: Gerd Petermann; Development list for mkgmap Betreff: Re: AW: [mkgmap-dev] Work on is_in branch Hi Gerd Attached is a quick patch that cause b14 to give the correct answer for the 'any' method and hence pass the test; merge the 2 polygons and then it processes 1 outer, 1 hole with the expected results When my mind is up to it, I'll try and work out what happens during the isLineInShape processing. The hole after merging appears to retain the same upper and lower mid-points from the cutting and matches the line, with the values as output like (not sure what the precision is here): line [2492250/449714, 2492167/450038, 2492073/449970, 2492155/449646, 2492250/449714] ie the inner b14 polygon [2491978/449872, 2492086/449391, 2492384/449581, 2492319/449872, 2492209/449872, 2492250/449714, 2492155/449646, 2492097/449872, 2491978/449872] ie 1/2 of the MP from cutting. this gives IN|ON|OUT, should be ON|OUT hole [2492073/449970, 2492167/450038, 2492209/449872, 2492250/449714, 2492155/449646, 2492097/449872, 2492073/449970] after the java2d merging - this gives ON Ticker On Thu, 2020-02-20 at 17:04 +0000, Gerd Petermann wrote:

Hi Ticker,

patch is commited. It is a bit difficult for me because you change a lot of things and the unit test fails. I just want to make sure that we have something testable in the end. It is already difficult enough to understand what is documented.

I think the tests are not always working because the result of Coord.makeBetweenPoint is rounded. That means a point calculated with it is typically not ON the line between the given points. A possible solution would be to change all the code in IsInUtil to use double values and rewrite e.g. makeBetweenPoint and the other used methods to keep the double precision. Still, it might fail when java area code comes in because that always uses a flat earth calculation. When I understood that I felt indeed a bit relunctant.

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 20. Februar 2020 17:45 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

I don't think the test data 'expected' values are wrong, it is just that they are more specific than the 'method' mechanism allows to be differentiated; eg a polygon can only be tested for ALL in or ANY in.

At the moment I feel you have a reluctance about the whole concept of the methods. Once the principle is accepted, I'll go through the test data and add, as another tag, the list of methods that should match the element, then change the test driver to check that these match and the other applicable methods don't.

Reg. b14: It isn't the stop-early code that causes the problems, isLineInShape is not giving the correct answer for a simple polygon produced by the MP cutter.

It would be quite easy to introduce some POLYGON 'on' methods, that match the outer, inner or either edge of a polygon, but maybe this could wait until there is a call for it.

Next mail: I'll change the sentence as you suggest.

Please can you commit the patch as it stands; it has a lot of good stuff in it. Then I can do the IsInUtilTest and test data changes as the next stage. It's also handy to see how the Style Manual looks after each build into the download area, because I don't know how to generate it and am just guessing at the formatting.

Thank you Ticker

Hi Ticker, see r4457. I see one potential problem with your test driver: It does not always find a problem when the expected value in the test data is wrong. E.g. when I change the expected value for w2 from 1 to the wrong value 3 the test doesn't fail. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Montag, 24. Februar 2020 13:50 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Gerd Patch attached that: - rewords the sentence is the Style Manual and changes the highlighting; I need to check the next build/download to see if this is clearer. - fixes polygon 'any' method to also return true if exactly ON. - merge polygons for 'any' so that line on shared boundary is "in" rather than "on". - change the test driver to try all methods relevant to the element, checking they return true/false as appropriate. I decided that, rather than introducing a new tag saying which methods should match, it was clearer to use the 'expected' tag value as a description of how the element interacted with the polygons and generate the methods that should match from this and the non-matching from a list if all methods. Ticker On Thu, 2020-02-20 at 16:45 +0000, Ticker Berkin wrote: Hi Gerd I don't think the test data 'expected' values are wrong, it is just that they are more specific than the 'method' mechanism allows to be differentiated; eg a polygon can only be tested for ALL in or ANY in. At the moment I feel you have a reluctance about the whole concept of the methods. Once the principle is accepted, I'll go through the test data and add, as another tag, the list of methods that should match the element, then change the test driver to check that these match and the other applicable methods don't. Reg. b14: It isn't the stop-early code that causes the problems, isLineInShape is not giving the correct answer for a simple polygon produced by the MP cutter. It would be quite easy to introduce some POLYGON 'on' methods, that match the outer, inner or either edge of a polygon, but maybe this could wait until there is a call for it. Next mail: I'll change the sentence as you suggest. Please can you commit the patch as it stands; it has a lot of good stuff in it. Then I can do the IsInUtilTest and test data changes as the next stage. It's also handy to see how the Style Manual looks after each build into the download area, because I don't know how to generate it and am just guessing at the formatting. Thank you Ticker On Thu, 2020-02-20 at 15:41 +0000, Gerd Petermann wrote: Hi Ticker, I see that you overwrite the expected value stored in the test data in the unit test. Please don't do this. If you think that the expected value in is-in-samples.osm is wrong we should discuss the test data. In my eyes b14 clearly has points on the edge (as it is part of the edge) and is out. If you think the expected results are correct but your new code doesn't allow to test them because of the early stop code please add a new tag to each object or maybe create a new file. The unit test file is meant to document what the code does. Gerd

Hi Ticker, I think what I tried to point out is that we have lost some precision reg. the unit tests. On the other hand you are probably right that it doesn't matter for the mkgmap user. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Montag, 24. Februar 2020 14:54 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Gerd It isn't really a problem of the test driver; it is a consequence of not having a method that distinguishes between an inner line that touches the edge and one that doesn't, ditto for outer lines. If there is a strong case for methods that do distinguish these cases and they get implemented, then the new methods just needed to be added to the allLineMethods list and expected #1 and #4 maps. Then the tests will fail if you swap 1 & 3 or 4 & 6. Ticker On Mon, 2020-02-24 at 13:11 +0000, Gerd Petermann wrote:

Hi Ticker,

see r4457.

I see one potential problem with your test driver: It does not always find a problem when the expected value in the test data is wrong. E.g. when I change the expected value for w2 from 1 to the wrong value 3 the test doesn't fail.

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Montag, 24. Februar 2020 13:50 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

Patch attached that:

- rewords the sentence is the Style Manual and changes the highlighting; I need to check the next build/download to see if this is clearer.

- fixes polygon 'any' method to also return true if exactly ON.

- merge polygons for 'any' so that line on shared boundary is "in" rather than "on".

- change the test driver to try all methods relevant to the element, checking they return true/false as appropriate. I decided that, rather than introducing a new tag saying which methods should match, it was clearer to use the 'expected' tag value as a description of how the element interacted with the polygons and generate the methods that should match from this and the non-matching from a list if all methods.

Ticker

On Thu, 2020-02-20 at 16:45 +0000, Ticker Berkin wrote: Hi Gerd

I don't think the test data 'expected' values are wrong, it is just that they are more specific than the 'method' mechanism allows to be differentiated; eg a polygon can only be tested for ALL in or ANY in.

At the moment I feel you have a reluctance about the whole concept of the methods. Once the principle is accepted, I'll go through the test data and add, as another tag, the list of methods that should match the element, then change the test driver to check that these match and the other applicable methods don't.

Reg. b14: It isn't the stop-early code that causes the problems, isLineInShape is not giving the correct answer for a simple polygon produced by the MP cutter.

It would be quite easy to introduce some POLYGON 'on' methods, that match the outer, inner or either edge of a polygon, but maybe this could wait until there is a call for it.

Next mail: I'll change the sentence as you suggest.

Please can you commit the patch as it stands; it has a lot of good stuff in it. Then I can do the IsInUtilTest and test data changes as the next stage. It's also handy to see how the Style Manual looks after each build into the download area, because I don't know how to generate it and am just guessing at the formatting.

Thank you Ticker

On Thu, 2020-02-20 at 15:41 +0000, Gerd Petermann wrote: Hi Ticker,

I see that you overwrite the expected value stored in the test data in the unit test. Please don't do this. If you think that the expected value in is-in-samples.osm is wrong we should discuss the test data. In my eyes b14 clearly has points on the edge (as it is part of the edge) and is out.

If you think the expected results are correct but your new code doesn't allow to test them because of the early stop code please add a new tag to each object or maybe create a new file. The unit test file is meant to document what the code does.

Gerd

_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Ticker, no idea how the tool chain for the pdf works. There are some commented code blocks and Eclipse and SonarLint warn about several missing default statements, e.g. Complete cases by adding the missing enum constants or add a default case to this switch. IsInFunction.java mkgmap/src/uk/me/parabola/mkgmap/osmstyle/function line 164 SonarLint On-The-Fly Issue Reg. the TODO comment // problem with test b14 on the cut polygons and isLineInShape that goes away when merged. TODO: investigate sometime Maybe the reason is that we create a Coord instance at the place where the polygon is split. Due to the rounding errors this point can be a 1-2 mm inside or outside the original inner polygon. Merging should not change the result unless the added point is removed by the merge. In that case I would assume that there were no rounding errors. Some log statements might be removed or changed to debug level? log.info("done", System.identityHashCode(this), hasIn, hasOn, hasOut); Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Dienstag, 25. Februar 2020 10:27 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Gerd I think this is about ready for release. There is a slight problem with the layout in the Style Manual in that the width of "is_in(tag,value,method)" causes it to run into the Node/Way/Relation flags. If there was a way to put the function args down the first column it would fix it, but I have no idea of the rules of the formatting language. What are the tools used to transform doc/styles/*.txt to the style-manual.pdf? I'm not going to be able to do any complex line->polygon in/on/out debugging in the next few days and it seems to work well for most cases. Ticker On Mon, 2020-02-24 at 12:50 +0000, Ticker Berkin wrote:

Hi Gerd

Patch attached that:

- rewords the sentence is the Style Manual and changes the highlighting; I need to check the next build/download to see if this is clearer.

- fixes polygon 'any' method to also return true if exactly ON.

- merge polygons for 'any' so that line on shared boundary is "in" rather than "on".

- change the test driver to try all methods relevant to the element, checking they return true/false as appropriate. I decided that, rather than introducing a new tag saying which methods should match, it was clearer to use the 'expected' tag value as a description of how the element interacted with the polygons and generate the methods that should match from this and the non-matching from a list if all methods.

Ticker

_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Ticker, I've looked at b14 again. The mp-cutting produces a point which is a small bit inside the real hole (~1.6 mm). I've attached the gpx files. e_hp is the tested segmet of the element b14, s_hp is the tested segment created by cutting, shape_hp is the complete shape produced by the mp-cutter. The southern end crosses e_hp. This small overlap is ignored when the merged shape is analyzed. Responsible for this is the method isOnOrCloseToEdgeOfShape(), I think I wrote it exactly for this problem but it is just a work-around for the problems produced by the polygon cutting. Attached is a patch which might solve the problem, didn't test it much. It increases the allowed distance for "ON" from 1 to 2 mm and handles this special case. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 27. Februar 2020 14:01 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Gerd Changes: - Added comment in style-manual.txt about tools needed and how to generate the pdf. These are asciidoc, fop, python-pygments and mkgmap -pygments. - Adjusted the layout of "Table 4.2. Style functions" so that is_in(...) doesn't run into the flags and and slightly changed the highlighting again. - Changed log.info to log.debug. - Added default: throw... to a couple of switch statements. I disagree with SonarLint that this is always good practice; here it is just pointless clutter. I'm not sure what the problem is with commented code blocks. I left @override value() {...} in as commented because it clarifies what needs to change between a CachedFunction and a StyleFunction. doc/styles/main.txt should be deleted from svn; it has been superseded by style-manual.txt. Reg. b14. It might be slight rounding errors, but the hole generated from the merge of the 2 halves of the outer retains the 2 cut-points and this does match the inner polygon, whereas, for the inner polygon, isLineInShape gives IN/ON/OUT against one of the halves of the outer. I'd have expected the problems to be the other way around. Ticker On Wed, 2020-02-26 at 08:46 +0000, Gerd Petermann wrote:

Hi Ticker,

no idea how the tool chain for the pdf works.

There are some commented code blocks and Eclipse and SonarLint warn about several missing default statements, e.g. Complete cases by adding the missing enum constants or add a default case to this switch. IsInFunction.java mkgmap/src/uk/me/parabola/mkgmap/osmstyle/function line 164 SonarLint On-The-Fly Issue

Reg. the TODO comment // problem with test b14 on the cut polygons and isLineInShape that goes away when merged. TODO: investigate sometime Maybe the reason is that we create a Coord instance at the place where the polygon is split. Due to the rounding errors this point can be a 1-2 mm inside or outside the original inner polygon. Merging should not change the result unless the added point is removed by the merge. In that case I would assume that there were no rounding errors.

Some log statements might be removed or changed to debug level? log.info("done", System.identityHashCode(this), hasIn, hasOn, hasOut);

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Dienstag, 25. Februar 2020 10:27 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

I think this is about ready for release.

There is a slight problem with the layout in the Style Manual in that the width of "is_in(tag,value,method)" causes it to run into the Node/Way/Relation flags. If there was a way to put the function args down the first column it would fix it, but I have no idea of the rules of the formatting language. What are the tools used to transform doc/styles/*.txt to the style-manual.pdf?

I'm not going to be able to do any complex line->polygon in/on/out debugging in the next few days and it seems to work well for most cases.

Ticker

On Mon, 2020-02-24 at 12:50 +0000, Ticker Berkin wrote:

...

Hi Gerd

Patch attached that:

- rewords the sentence is the Style Manual and changes the highlighting; I need to check the next build/download to see if this is clearer.

- fixes polygon 'any' method to also return true if exactly ON.

- merge polygons for 'any' so that line on shared boundary is "in" rather than "on".

- change the test driver to try all methods relevant to the element, checking they return true/false as appropriate. I decided that, rather than introducing a new tag saying which methods should match, it was clearer to use the 'expected' tag value as a description of how the element interacted with the polygons and generate the methods that should match from this and the non-matching from a list if all methods.

Ticker

_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Ticker, yes, probably the value is still too small, and I assume nobody would mind to use a halo of 0.08 m width. Small problem is that we don't use this halo with the point in shape tests. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 27. Februar 2020 17:23 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Gerd Yes, I think I see where it goes in with josm at maximum zoom. Shouldn't EPS be a bit bigger? My understanding was that a high -precision unit was ~37mm at the equator (40,075*1000/2^24/2^6)*1000. With your change from 10mm to 20mm all the tests pass, but wouldn't private static final double EPS = 0.080; be safer, to allow for 2 hp roundings away from each other. Ticker On Thu, 2020-02-27 at 14:42 +0000, Gerd Petermann wrote:

Hi Ticker,

I've looked at b14 again. The mp-cutting produces a point which is a small bit inside the real hole (~1.6 mm). I've attached the gpx files. e_hp is the tested segmet of the element b14, s_hp is the tested segment created by cutting, shape_hp is the complete shape produced by the mp-cutter. The southern end crosses e_hp. This small overlap is ignored when the merged shape is analyzed. Responsible for this is the method isOnOrCloseToEdgeOfShape(), I think I wrote it exactly for this problem but it is just a work-around for the problems produced by the polygon cutting.

Attached is a patch which might solve the problem, didn't test it much. It increases the allowed distance for "ON" from 1 to 2 mm and handles this special case.

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 27. Februar 2020 14:01 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

Changes:

- Added comment in style-manual.txt about tools needed and how to generate the pdf. These are asciidoc, fop, python-pygments and mkgmap -pygments.

- Adjusted the layout of "Table 4.2. Style functions" so that is_in(...) doesn't run into the flags and and slightly changed the highlighting again.

- Changed log.info to log.debug.

- Added default: throw... to a couple of switch statements. I disagree with SonarLint that this is always good practice; here it is just pointless clutter.

I'm not sure what the problem is with commented code blocks. I left @override value() {...} in as commented because it clarifies what needs to change between a CachedFunction and a StyleFunction.

doc/styles/main.txt should be deleted from svn; it has been superseded by style-manual.txt.

Reg. b14. It might be slight rounding errors, but the hole generated from the merge of the 2 halves of the outer retains the 2 cut-points and this does match the inner polygon, whereas, for the inner polygon, isLineInShape gives IN/ON/OUT against one of the halves of the outer. I'd have expected the problems to be the other way around.

Ticker

On Wed, 2020-02-26 at 08:46 +0000, Gerd Petermann wrote:

...

Hi Ticker,

no idea how the tool chain for the pdf works.

There are some commented code blocks and Eclipse and SonarLint warn about several missing default statements, e.g. Complete cases by adding the missing enum constants or add a default case to this switch. IsInFunction.java mkgmap/src/uk/me/parabola/mkgmap/osmstyle/function line 164 SonarLint On-The-Fly Issue

Reg. the TODO comment // problem with test b14 on the cut polygons and isLineInShape that goes away when merged. TODO: investigate sometime Maybe the reason is that we create a Coord instance at the place where the polygon is split. Due to the rounding errors this point can be a 1-2 mm inside or outside the original inner polygon. Merging should not change the result unless the added point is removed by the merge. In that case I would assume that there were no rounding errors.

Some log statements might be removed or changed to debug level? log.info("done", System.identityHashCode(this), hasIn, hasOn, hasOut);

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Dienstag, 25. Februar 2020 10:27 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

I think this is about ready for release.

There is a slight problem with the layout in the Style Manual in that the width of "is_in(tag,value,method)" causes it to run into the Node/Way/Relation flags. If there was a way to put the function args down the first column it would fix it, but I have no idea of the rules of the formatting language. What are the tools used to transform doc/styles/*.txt to the style-manual.pdf?

I'm not going to be able to do any complex line->polygon in/on/out debugging in the next few days and it seems to work well for most cases.

Ticker

On Mon, 2020-02-24 at 12:50 +0000, Ticker Berkin wrote:

...

Hi Gerd

Patch attached that:

- rewords the sentence is the Style Manual and changes the highlighting; I need to check the next build/download to see if this is clearer.

- fixes polygon 'any' method to also return true if exactly ON.

- merge polygons for 'any' so that line on shared boundary is "in" rather than "on".

- change the test driver to try all methods relevant to the element, checking they return true/false as appropriate. I decided that, rather than introducing a new tag saying which methods should match, it was clearer to use the 'expected' tag value as a description of how the element interacted with the polygons and generate the methods that should match from this and the non-matching from a list if all methods.

Ticker

_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Ticker, if you see a way to change the algo just do it. I've just noticed that I forgot to commit a change in Coord.makeBetweenPoint(). This routine should use Math.round. Will reduce the error by 50%, right? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 27. Februar 2020 20:24 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Gerd Looking at the distance calculations to compare with EPS (ie very small), wouldn't it be much simpler and clearer just to calculate it as highPrecisionSquared units, not bothering with degrees/radians, rhumb -line, metre conversion etc. Then EPS would be 4. Then, considering IsInUtil.insidePolgon and can it be changed to have some tolerance and maybe changing the interface to return IN/ON/OUT, it looks like it will stop, returning onBoundary, if there is a polygon segment that 'aims at' the point. Ticker On Thu, 2020-02-27 at 16:42 +0000, Gerd Petermann wrote:

Hi Ticker,

yes, probably the value is still too small, and I assume nobody would mind to use a halo of 0.08 m width. Small problem is that we don't use this halo with the point in shape tests.

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 27. Februar 2020 17:23 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

Yes, I think I see where it goes in with josm at maximum zoom.

Shouldn't EPS be a bit bigger? My understanding was that a high -precision unit was ~37mm at the equator (40,075*1000/2^24/2^6)*1000.

With your change from 10mm to 20mm all the tests pass, but wouldn't private static final double EPS = 0.080; be safer, to allow for 2 hp roundings away from each other.

Ticker

On Thu, 2020-02-27 at 14:42 +0000, Gerd Petermann wrote:

...

Hi Ticker,

I've looked at b14 again. The mp-cutting produces a point which is a small bit inside the real hole (~1.6 mm). I've attached the gpx files. e_hp is the tested segmet of the element b14, s_hp is the tested segment created by cutting, shape_hp is the complete shape produced by the mp-cutter. The southern end crosses e_hp. This small overlap is ignored when the merged shape is analyzed. Responsible for this is the method isOnOrCloseToEdgeOfShape(), I think I wrote it exactly for this problem but it is just a work-around for the problems produced by the polygon cutting.

Attached is a patch which might solve the problem, didn't test it much. It increases the allowed distance for "ON" from 1 to 2 mm and handles this special case.

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 27. Februar 2020 14:01 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

Changes:

- Added comment in style-manual.txt about tools needed and how to generate the pdf. These are asciidoc, fop, python-pygments and mkgmap -pygments.

- Adjusted the layout of "Table 4.2. Style functions" so that is_in(...) doesn't run into the flags and and slightly changed the highlighting again.

- Changed log.info to log.debug.

- Added default: throw... to a couple of switch statements. I disagree with SonarLint that this is always good practice; here it is just pointless clutter.

I'm not sure what the problem is with commented code blocks. I left @override value() {...} in as commented because it clarifies what needs to change between a CachedFunction and a StyleFunction.

doc/styles/main.txt should be deleted from svn; it has been superseded by style-manual.txt.

Reg. b14. It might be slight rounding errors, but the hole generated from the merge of the 2 halves of the outer retains the 2 cut -points and this does match the inner polygon, whereas, for the inner polygon, isLineInShape gives IN/ON/OUT against one of the halves of the outer. I'd have expected the problems to be the other way around.

Ticker

On Wed, 2020-02-26 at 08:46 +0000, Gerd Petermann wrote:

...

Hi Ticker,

no idea how the tool chain for the pdf works.

There are some commented code blocks and Eclipse and SonarLint warn about several missing default statements, e.g. Complete cases by adding the missing enum constants or add a default case to this switch. IsInFunction.java mkgmap/src/uk/me/parabola/mkgmap/osmstyle/function line 164 SonarLint On-The-Fly Issue

Reg. the TODO comment // problem with test b14 on the cut polygons and isLineInShape that goes away when merged. TODO: investigate sometime Maybe the reason is that we create a Coord instance at the place where the polygon is split. Due to the rounding errors this point can be a 1-2 mm inside or outside the original inner polygon. Merging should not change the result unless the added point is removed by the merge. In that case I would assume that there were no rounding errors.

Some log statements might be removed or changed to debug level? log.info("done", System.identityHashCode(this), hasIn, hasOn, hasOut);

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Dienstag, 25. Februar 2020 10:27 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

I think this is about ready for release.

There is a slight problem with the layout in the Style Manual in that the width of "is_in(tag,value,method)" causes it to run into the Node/Way/Relation flags. If there was a way to put the function args down the first column it would fix it, but I have no idea of the rules of the formatting language. What are the tools used to transform doc/styles/*.txt to the style-manual.pdf?

I'm not going to be able to do any complex line->polygon in/on/out debugging in the next few days and it seems to work well for most cases.

Ticker

On Mon, 2020-02-24 at 12:50 +0000, Ticker Berkin wrote:

...

Hi Gerd

Patch attached that:

- rewords the sentence is the Style Manual and changes the highlighting; I need to check the next build/download to see if this is clearer.

- fixes polygon 'any' method to also return true if exactly ON.

- merge polygons for 'any' so that line on shared boundary is "in" rather than "on".

- change the test driver to try all methods relevant to the element, checking they return true/false as appropriate. I decided that, rather than introducing a new tag saying which methods should match, it was clearer to use the 'expected' tag value as a description of how the element interacted with the polygons and generate the methods that should match from this and the non-matching from a list if all methods.

Ticker

_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

---

mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Ticker, nice :) Committed with r4461, with small changes. This even allows to change the test case for lamp4 from expected="?" to expected="2" (ON). reg. the distance calculations: I don't remember the details and maybe all is wrong. See svn log for r3333 : http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=3333 and search the archives for "Great Britain National Grid". The math behind projections and distance formulas is too complex for me. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Dienstag, 3. März 2020 14:24 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch Hi Gerd Patch attached: - inherits your is_in-function_v14.patch - adds Math.round to Coord.makeBetweenPoint(); Is this how you indented? - removes IsInUtil.insidePolygon() and changes callers to use isPointInShape() - adds ON tolerance to isPointInShape() - I couldn't work out how to do this with the winding algo, so changed it to crossing-point, which is fine for mkgmap polygons. - add some more tolerances to isLineInShape - isLineInShape had comment: // it is unlikely but not impossible that pTest is on boundary :( and it returned the wrong result if it was. This contributed to the difficulties with b14 (and b19). I've fixed it and, I think, improved and simplified the running status setting - fix spelling of rhumLine to rhumbLine There are still a couple of places that have complicated distance calculation and point insertion using, bearings, rhumbline, meter conversion but I don't think it is worth re-implementing them for this. I'm slightly surprised how much use there is of bearing/rhumbLine/Mercator projection calculations. I think real distance/metre calcs should be "great circle" and internal poly/line chopping, testing etc should be high-precision polar coords. Ticker On Thu, 2020-02-27 at 19:36 +0000, Gerd Petermann wrote:

Hi Ticker,

if you see a way to change the algo just do it. I've just noticed that I forgot to commit a change in Coord.makeBetweenPoint(). This routine should use Math.round. Will reduce the error by 50%, right?

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 27. Februar 2020 20:24 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

Looking at the distance calculations to compare with EPS (ie very small), wouldn't it be much simpler and clearer just to calculate it as highPrecisionSquared units, not bothering with degrees/radians, rhumb -line, metre conversion etc. Then EPS would be 4.

Then, considering IsInUtil.insidePolgon and can it be changed to have some tolerance and maybe changing the interface to return IN/ON/OUT, it looks like it will stop, returning onBoundary, if there is a polygon segment that 'aims at' the point.

Ticker

Hi Gerd Nothing else I want to change. I keep thinking there must be other ways to do the lineInShape logic, but I think it has taken enough time already. Ticker On Fri, 2020-03-06 at 08:16 +0000, Gerd Petermann wrote:

Hi Ticker,

My concerns regarding a merge to trunk are resolved. Anything else you want to change?

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Gerd Petermann <gpetermann_muenchen@hotmail.com> Gesendet: Dienstag, 3. März 2020 15:22 An: Ticker Berkin; Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Ticker,

nice :) Committed with r4461, with small changes.

This even allows to change the test case for lamp4 from expected="?" to expected="2" (ON).

reg. the distance calculations: I don't remember the details and maybe all is wrong.

See svn log for r3333 : http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=3333 and search the archives for "Great Britain National Grid".

The math behind projections and distance formulas is too complex for me.

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Dienstag, 3. März 2020 14:24 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

Patch attached:

- inherits your is_in-function_v14.patch

- adds Math.round to Coord.makeBetweenPoint(); Is this how you indented?

- removes IsInUtil.insidePolygon() and changes callers to use isPointInShape()

- adds ON tolerance to isPointInShape() - I couldn't work out how to do this with the winding algo, so changed it to crossing-point, which is fine for mkgmap polygons.

- add some more tolerances to isLineInShape

- isLineInShape had comment:

// it is unlikely but not impossible that pTest is on boundary :(

and it returned the wrong result if it was. This contributed to the difficulties with b14 (and b19). I've fixed it and, I think, improved and simplified the running status setting

- fix spelling of rhumLine to rhumbLine

There are still a couple of places that have complicated distance calculation and point insertion using, bearings, rhumbline, meter conversion but I don't think it is worth re-implementing them for this.

I'm slightly surprised how much use there is of bearing/rhumbLine/Mercator projection calculations. I think real distance/metre calcs should be "great circle" and internal poly/line chopping, testing etc should be high-precision polar coords.

Ticker

On Thu, 2020-02-27 at 19:36 +0000, Gerd Petermann wrote:

...

Hi Ticker,

if you see a way to change the algo just do it. I've just noticed that I forgot to commit a change in Coord.makeBetweenPoint(). This routine should use Math.round. Will reduce the error by 50%, right?

Gerd

________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 27. Februar 2020 20:24 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Work on is_in branch

Hi Gerd

Looking at the distance calculations to compare with EPS (ie very small), wouldn't it be much simpler and clearer just to calculate it as highPrecisionSquared units, not bothering with degrees/radians, rhumb -line, metre conversion etc. Then EPS would be 4.

Then, considering IsInUtil.insidePolgon and can it be changed to have some tolerance and maybe changing the interface to return IN/ON/OUT, it looks like it will stop, returning onBoundary, if there is a polygon segment that 'aims at' the point.

Ticker

---

mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev