Jürg Lehni
d85b4f0c80
Fix issues with wrong sorting of CurveLocation in Curve.filterIntersections()
2015-08-26 16:56:28 +02:00
Jürg Lehni
815991d556
Some minor adjustments and fixes.
2015-08-26 11:43:39 +02:00
Jürg Lehni
2c4dcb96c2
Streamline addOverlap() code a bit.
2015-08-26 11:26:55 +02:00
Jürg Lehni
3946bcb04e
Remove unnecessary segments between two collinear curves in Path#reduce()
2015-08-24 12:49:33 +02:00
Jürg Lehni
75f669d839
Implement proper method to determine which segment to choose when encountering overlaps.
2015-08-24 04:36:49 +02:00
Jürg Lehni
85d21c84b8
Start implementing support for touching and overlapping shapes in boolean operations.
...
Relates to #449 , #450 , #648 , #719
2015-08-23 21:19:19 +02:00
Jürg Lehni
9bd399b5b8
Introduce Curve#isStraight() and use it in splitPath() and divide() to keep the result of splitting straight curves straight.
...
Do not use Curve#isLinear(), as that would include curves with collinear handles, and we don't want to set these straight.
2015-08-22 22:06:42 +02:00
Jürg Lehni
bddff95fe3
Remove workarounds in splitPath() thanks to improved precision of Numerical.solveCubic().
2015-08-22 14:24:31 +02:00
Jürg Lehni
716d21e369
Always use fat-line clipping since fallback doesn't behave the same way.
2015-08-21 16:39:41 +02:00
Jürg Lehni
0436b2749a
Fix some comments.
2015-08-19 17:28:41 +02:00
Jürg Lehni
84b202fd79
Expose and reuse Curve evaluation method names.
2015-08-19 17:26:52 +02:00
Jürg Lehni
9fe93d1434
Move Curve evaluate() method to private scope.
2015-08-19 17:19:42 +02:00
Jürg Lehni
da82116501
Rework handling of weighted and normalized curve tangents and normals.
...
Relates to #563
2015-08-19 17:15:41 +02:00
Jürg Lehni
59eec1f27b
Implement Curve#isLinear(), #isCollinear() & co without depending on the segments being part of a path.
2015-08-19 11:43:37 +02:00
Jürg Lehni
e476672748
Use better epsilon values in Numerical.solveQuadratic() and solveCubic()
...
To finally satisfy both #541 and #708 .
With this change in place, https://github.com/paperjs/boolean-test is also finally back to run with 0 errors. Woop!
2015-08-18 23:47:28 +02:00
Jürg Lehni
4ee68e5782
Various changes to #isCollinear(), #isOrthogonal(), #isArc()
...
- Rename #isArc() - > #isOrthogonalArc()
- Implement versions for Curve
- Write proper documentation
2015-08-17 14:33:48 +02:00
Jürg Lehni
f8314f927e
Various changes on #isLinear(), #hasHandles() & co
...
Relates to #652
- Implement #hasHandles() on Path, Segment and Curve
- Remove Path#isPolygon()
- Define #isLinear() consistently across Path, Segment and Curve.
- Introduce new Segment#isStraight()
2015-08-17 14:31:23 +02:00
Jürg Lehni
19bcd19b49
Implement path versioning for segment changes.
...
And use it to keep CurveLocation objects up to date. Closes #682
2015-06-16 19:36:22 +02:00
Jürg Lehni
c1485e7068
Use < instead of <= when comparing against TOLERANCE
2015-06-16 19:33:21 +02:00
Jürg Lehni
b541088c6f
A whole lot of documentation clean-up.
2015-06-16 17:50:37 +02:00
Jürg Lehni
e3f04478d9
Clean up Curve.getParameterOf() code.
2015-06-16 12:29:27 +02:00
Jürg Lehni
7749641e66
Handle out-of-bounds offsets/parameters correctly.
...
Closes #660
2015-04-04 17:05:39 +02:00
Jürg Lehni
22be259167
Deal with edge cases in Curve.solveCubic() where small cubics are very close to lines.
...
Also reduce Numerical.EPSILON again, as required by some rare edge cases.
Relates to #541 .
2015-01-10 22:05:42 +01:00
Jürg Lehni
1c758a415a
Switch to new and improved Line.getSignedDistance() in fat-line clipping code as well.
2015-01-05 16:20:33 +01:00
Jürg Lehni
1239eb55ec
Minor clean up in Curve.filterIntersections().
2015-01-04 17:47:41 +01:00
Jürg Lehni
afc14b9634
Reduce Numerical.TOLERANCE to 1e-6, and use the same value across the boolean code.
...
Yields very good results now, across all tests.
2015-01-04 15:54:50 +01:00
Jürg Lehni
3564dafa93
Do not keep iterating if the curve was reduced to a point.
2015-01-04 13:23:59 +01:00
Jürg Lehni
d33d65fcc2
Revert change in recursion threshold as the new code reaches the desired precision earlier.
2015-01-04 13:21:12 +01:00
hkrish
3781aedb22
Small but critical fix on getSignedDistance method
...
when the line is perfectly vertical, return the difference between x values depending on the side. The earlier "l1y" is a silly mistake.
2015-01-03 23:41:47 +01:00
Jürg Lehni
b6a4815d03
Remove Curve#isHorizontal() as its implementation and precision is specific to boolean operations.
2015-01-03 21:02:12 +01:00
Jürg Lehni
a854c55914
Split off code from Path.Clockwise() into Curve.getEdgeSum()
2015-01-03 01:46:22 +01:00
Jürg Lehni
538eac6dc7
Use smaller tolerances in winding code, to address edge cases.
...
Closes #559
2015-01-02 23:47:26 +01:00
Jürg Lehni
477527f8ec
Implement more intuitive version of clipConvexHull(), outlined in #570 by @iconexperience
...
Code further shortened using a sub-routine.
2015-01-02 22:36:04 +01:00
Jürg Lehni
232739fae9
Add comments to fat-line clipping code linking to issues #568 and #571 .
2015-01-02 22:34:09 +01:00
Jürg Lehni
4ed9ef54f2
Fix bug in Curve.filterIntersections() and clean up code a bit.
2015-01-02 21:19:18 +01:00
Jürg Lehni
b2261fd512
Introduce Curve.filterIntersections() and use it both in PathItems#getIntersections() and Curve#getIntersections()
2015-01-02 21:00:05 +01:00
Jürg Lehni
3302a7dff5
Increase recursion threshold to 4 when checking for fatline convergence.
2015-01-02 20:49:09 +01:00
Jürg Lehni
fbd026ea0d
Bring back code removed by @hkrish in commit 5fb75593c2
...
Relates to issue #570 , still required to figure out why it is still needed.
2015-01-02 20:02:45 +01:00
Jürg Lehni
e3e3d13fa5
Add comment regarding issue #565
2015-01-02 16:33:15 +01:00
Jürg Lehni
b75def3f3a
Implement optimal values in fat line clipping code for maximum recursion and curve time difference, as determined by @iconexperience.
...
Closes #565
2015-01-02 16:29:12 +01:00
Jürg Lehni
45c86a3035
Switch to suggested new implementation of Line.getSignedDistance() by @iconexperience
...
Closes #554
2015-01-02 16:17:19 +01:00
Jürg Lehni
318209e9e9
Replace compareNumbers() with equals() and improve numerical tolerance handling in test helpers.
2014-11-30 11:27:14 -08:00
Jürg Lehni
ebdc63fbfa
Use correct formula in Curve#getTangentAt() for end-points when there is no handle.
...
Closes #553 .
2014-11-08 17:16:22 -08:00
Jürg Lehni
e5714584ff
Fix return type of #getCurvatureAt() methods.
...
Closes #535 properly this time.
2014-10-15 12:02:21 +02:00
Jürg Lehni
b71e2c69d6
Fix documentation of Curve#getCurvatureAt(), as well as Path#getCurvatureAt() & co.
...
Closes #535
2014-10-14 13:52:57 +02:00
Jürg Lehni
6a11532322
Clean up and simplify code a bit further.
2014-09-28 11:49:46 +02:00
Jürg Lehni
7462fe0b70
Fix Curve.getParameterAt() for negative values of offset, and implement unit test.
...
Closes #524 .
2014-09-28 11:44:38 +02:00
Jürg Lehni
ee27fe820f
Check for valid parameter range in Curve#getLocationAt() and delegate to it in #getLocationOf()
2014-09-20 11:09:09 +02:00
Jürg Lehni
c94fb3038f
Use valid parameter range [0,1] in calls to Curve.solveCubic().
2014-09-20 11:08:20 +02:00
Jürg Lehni
c2465b60fa
Transform the correct amount of points in Curve.getValues() when a matrix is provided.
2014-09-20 11:06:25 +02:00