PostGIS Java2D geometry implementation of points

[laborgi]

The PostGIS Java2D implementation of points as GeneralPath accordingly to the Javadoc is: "Points thus are translated into MoveTo() followed by a closePath."
I've tested various methods.The point is drawn when the GeneralPath has a lineTo() to the same coordinates instead closePath() (as described below: "Maybe we need a LineTo() to the same coordinages instead.")
Perhaps someone can correct it :-)
Thanks

[phillipross]

You mean to say that the documentation is incorrect. This implementation is actually quite old... referring to java 1.5. The note about adjacent MoveTo commands being thrown out is no longer applicable?

[laborgi]

Hallo Phillip,

I'm sorry, I do not understand exactly the content of your post,
especially what you mean with "The note about adjacent MoveTo commands
being thrown out is no longer applicable?".

At the moment it is not possible to draw geometry containing only
points. So, my concern was to solve this problem.
This can be done by changing the line 158 in the class
ShapeBinaryParser.java (method parsePoint()) from "path.closePath(...)"
to "path.lineTo(...)" with the coordinates of the starting point (from
the line above) as parameter.
If this is changed, also points are drawn correctly by Java2D.

Best regards.
Ines

Am 16.07.2015 um 18:05 schrieb Phillip Ross:

You mean to say that the documentation is incorrect. This
implementation is actually quite old... referring to java 1.5. The
note about adjacent MoveTo commands being thrown out is no longer
applicable?

—
Reply to this email directly or view it on GitHub
#27 (comment).

[phillipross]

Sorry Ines, I think I understand what you mean now. At first I thought you were saying that the documentation PGShapeGeometry was incorrect/inaccurate, and I was saying that it might be inaccurate because it was written when there might have been java2d rendering bugs in java 1.5.

But I guess what you are actually saying is that drawing points can now be done with moveTo()+lineTo() instead of moveTo()+closePath(). Is this correct?

Ideally we'd have some test code to illustrate this. If you could provide some test code we'd be very grateful. I'll see if I can produce a test case on my own that illustrates the problem. It sounds like your fix is a simple code change, but we should have some test code in place that tests geometries that are only points along with geometries that are not exclusively points.

Thanks!

[phillipross]

Hi @laborgi... would you be able to review the pull request I just submitted which provides a patch to the point parsing code? Thanks!

[laborgi]

Hi Phillip,

I think the changed code works. Your test class and my own tests ran
through without any problems.

Many Thanks!
Ines

Am 18.07.2015 um 23:27 schrieb Phillip Ross:

Hi @laborgi https://github.com/laborgi... would you be able to
review the pull request
#28 I just submitted
which provides a patch to the point parsing code? Thanks!

—
Reply to this email directly or view it on GitHub
#27 (comment).

[phillipross]

Thanks Ines. I merged the pull request into the master branch. I'll do one more run of manual testing on the java2d test (I'm not sure how to properly automate testing of this if it's even possible) and then close out this issue if everything is fine.
Thanks again!

[phillipross]

This issue has been verified via manual testing on master 518a9d7