Adam Wulkiewicz
|
8e9a59bac3
|
[relate] helpers like segment_watcher, exit_watcher, etc. moved to a separate file, implemented preliminary version of L/A (commented out for now), found bug in L/L, added failing tests
|
2014-03-11 18:06:47 +01:00 |
|
Adam Wulkiewicz
|
fe0e34bff6
|
[relate][get_turns] fixed incorrect dispatching of static masks, tag_base<> replaced by group_tag<> and moved to namespace detail, added group_dim<>
|
2014-03-11 01:51:23 +01:00 |
|
Adam Wulkiewicz
|
4182cb916b
|
[relate] TODOs regarding predefined masks added
|
2014-03-10 17:25:11 +01:00 |
|
Adam Wulkiewicz
|
d91db3a2e0
|
[relate] added ready-to-use static masks for various spatial relations
|
2014-03-10 15:03:57 +01:00 |
|
Adam Wulkiewicz
|
d209b4fbd2
|
[relate] added functions check() and interrupt(), added parameter EnableInterrupt to static_interrupt for consistency
|
2014-03-10 12:08:16 +01:00 |
|
Adam Wulkiewicz
|
4987e8d13d
|
relate() implemented support for complex static masks as a MPL sequences
|
2014-03-10 03:02:11 +01:00 |
|
Adam Wulkiewicz
|
5f0ab7962f
|
relate() static_mask_handler::update() dispatching fixed
|
2014-03-09 22:29:27 +01:00 |
|
Adam Wulkiewicz
|
9ca11feb90
|
relate() static_mask and static_mask_handler implemented
|
2014-03-09 22:24:09 +01:00 |
|
Adam Wulkiewicz
|
09c41cb171
|
relate() - const MatrixOrMask may be passed to the function, the actual matrix is now stored in the result handler
|
2014-03-08 19:59:39 +01:00 |
|
Adam Wulkiewicz
|
1e6da76527
|
relate() implemented different interface, added support for complex masks passed in boost::tuple
|
2014-03-08 19:31:06 +01:00 |
|
Adam Wulkiewicz
|
b6ad82a0c2
|
relate() naming of results changed, added de4im matrix
|
2014-03-07 19:41:54 +01:00 |
|
Adam Wulkiewicz
|
9ff0624b41
|
relate(L,A) added preliminary implementation, for now working only for disjoint geometries
|
2014-03-04 17:37:12 +01:00 |
|
Adam Wulkiewicz
|
38147c2fc7
|
relate() for_each_disjoint_linestring_if utility now works for all geometries, renamed to for_each_disjoint_geometry_if
|
2014-03-04 16:55:18 +01:00 |
|
Adam Wulkiewicz
|
8ae3b591d3
|
relate(L,L) unused code commented out
|
2014-03-04 15:37:34 +01:00 |
|
Adam Wulkiewicz
|
02cdc305d3
|
get_turns(L,A) - fixed turns generation for endpoints of collinear front/back segments, added some tests
|
2014-03-03 19:21:12 +01:00 |
|
Adam Wulkiewicz
|
325b2e2bb0
|
get_turns(L,L) - TODO comment added
|
2014-03-03 18:19:37 +01:00 |
|
Adam Wulkiewicz
|
4fd21c2fa7
|
get_turns(L,A) - handled the back endpoint of the Linestring
|
2014-03-03 18:16:02 +01:00 |
|
Adam Wulkiewicz
|
ebaf26c0ff
|
get_turns(L,A) handled front endpoint
|
2014-03-03 13:36:23 +01:00 |
|
Adam Wulkiewicz
|
be1ac99a82
|
get_turns(L,A) added, the endpoints are NOT properly set yet!
|
2014-02-28 18:59:30 +01:00 |
|
Adam Wulkiewicz
|
ad201710c7
|
generation of turns for linestrings endpoints moved from get_turn_info_ll.hpp to separate file, some run-time parameters replaced by compile-time parameters
|
2014-02-28 16:51:57 +01:00 |
|
Adam Wulkiewicz
|
040f9ed6e9
|
some comments changed in relate/turns.hpp
|
2014-02-27 15:31:11 +01:00 |
|
Adam Wulkiewicz
|
e8b5e64f2b
|
added empty test for get_turns(L,A) and some small changes like naming, comments. to_svg() (test debug helper tool) is now using turn_info defined in relate::turns, also traversal info output commented out.
|
2014-02-27 15:20:51 +01:00 |
|
Adam Wulkiewicz
|
629d6b1246
|
relate/less.hpp copyrights fixed
|
2014-02-27 02:18:12 +01:00 |
|
Adam Wulkiewicz
|
9ad7ec98a8
|
detail::relate::less moved to separate file
|
2014-02-27 02:17:08 +01:00 |
|
Adam Wulkiewicz
|
9ed47574e1
|
relate(L,L) readability of some parts of the code improved
|
2014-02-25 18:11:11 +01:00 |
|
Adam Wulkiewicz
|
6f4094ec2a
|
relate(L,L) some comments added
|
2014-02-25 17:59:02 +01:00 |
|
Adam Wulkiewicz
|
37db735b38
|
relate(L,L) recently introduced turn position info used to optimize boundary points detection
|
2014-02-25 17:37:00 +01:00 |
|
Adam Wulkiewicz
|
736f26db4a
|
get_turns(L,L) turn_position now indicates the global position in linestring, not local in segment, for consistency - this was set only for the first and the last segment
|
2014-02-25 16:47:18 +01:00 |
|
Adam Wulkiewicz
|
261b35a323
|
get_turns(L,L) got rid of the duplicated part of the code
|
2014-02-25 16:38:06 +01:00 |
|
Adam Wulkiewicz
|
0a70310124
|
get_turns(L,L) added turn_operation_linear containing info about the position of the IP in the segment
|
2014-02-25 13:39:09 +01:00 |
|
Adam Wulkiewicz
|
00941b7d50
|
Implemented relate(MultiPoint, MultiPoint), added detail::relate::less comparing Point types with different coordinate_type, added some TODO notes regarding possible optimizations
|
2014-02-24 17:47:29 +01:00 |
|
Adam Wulkiewicz
|
39d0d3bd75
|
Merge branch 'feature/relate' of github.com:boostorg/geometry into feature/relate
|
2014-02-24 15:17:20 +01:00 |
|
Adam Wulkiewicz
|
7033337149
|
IP coordinates rounded to the nearest value instead of a cast in segments_intersection_points if coordinate_type is integer
|
2014-02-24 15:16:37 +01:00 |
|
Adam Wulkiewicz
|
684b29cbe3
|
relate() - fixed clang warnings
|
2014-02-22 12:38:04 +01:00 |
|
Adam Wulkiewicz
|
a68d9f5855
|
fixed GCC compilation errors
|
2014-02-22 04:29:11 +01:00 |
|
Adam Wulkiewicz
|
0fd875fb63
|
relate() added Point-MultiPoint, point_point moved to the separate file, Geometries sequence is reversed if needed to not duplicate the code - for this Transpose is also used as the parameter to set<>(result)
|
2014-02-21 19:25:27 +01:00 |
|
Adam Wulkiewicz
|
89e412f5b4
|
added test for relate(Pt,Pt), added a TODO in relate::linear_linear
|
2014-02-21 17:56:40 +01:00 |
|
Adam Wulkiewicz
|
6482716689
|
relate() Finished the implementation for Point/Geometry, added topology_check util, added tests
|
2014-02-21 17:51:37 +01:00 |
|
Adam Wulkiewicz
|
93b82b734d
|
relate() added interruption_enabled flag to policy which can be used to create optimized mask type not checking for interuption per each set<>()
|
2014-02-21 15:34:52 +01:00 |
|
Adam Wulkiewicz
|
a03f554e27
|
relate() Added the interruptable mask Result, handled arbitrary Result type in relate(), added tests for masks and interruption, fixed warning in get_turns_ll
|
2014-02-21 00:40:09 +01:00 |
|
Adam Wulkiewicz
|
47b958ec53
|
relate()/get_turn_info_ll added TODO and commented out unused code
|
2014-02-20 23:29:39 +01:00 |
|
Adam Wulkiewicz
|
d13ff170be
|
relate(L,L) code refactored, tests added
|
2014-02-20 23:10:16 +01:00 |
|
Adam Wulkiewicz
|
9b16fdcf23
|
relate(L,L) handled case when several BLOCK operations were generated for the same point, fixed an error for BLOCK generated for a geometry touching the other one from the "outside"
|
2014-02-20 22:47:23 +01:00 |
|
Adam Wulkiewicz
|
a125811d56
|
relate(L,L) the ref to perviously analysed turn and operation added to turn_analyser<>
|
2014-02-19 15:45:02 +01:00 |
|
Adam Wulkiewicz
|
c2e1331270
|
relate(L,L) handling of 1-point linestrings replaced by simple ignoring (for now) since those are not valid geometries, the code was not removed though in case we wanted to e.g. enable it with some macro check.
|
2014-02-19 14:42:05 +01:00 |
|
Adam Wulkiewicz
|
b03601fab8
|
get_turns(L,L) fixed 2x IPs generation for a special case of collinear, opposite last Ls segment
|
2014-02-19 02:26:15 +01:00 |
|
Adam Wulkiewicz
|
1e871f8ae6
|
relate() result refactored, result concept crystalized, prepared for compile-time interruptable result types
|
2014-02-17 17:37:46 +01:00 |
|
Adam Wulkiewicz
|
99610ca124
|
relate(L,L) handled the rest of cases with 1-point linestrings involved, added interrupt member to relate result - it may be used to break the calculation
|
2014-02-16 22:23:14 +01:00 |
|
Adam Wulkiewicz
|
9f54142896
|
relate(L,L) handled some cases with 1-point linestrings involved (if 1-PtLS is disjoint or there are other Ls in the other MLs)
|
2014-02-16 18:26:41 +01:00 |
|
Adam Wulkiewicz
|
4815e00048
|
relate(L,L) disjoin linestrings contained in multilinestring properly handled. In the case of disjoint linestrings almost all special cases are handled non-simple linear rings, 1-point linestrings. Linestrings containing >1 equal points are for now treated as normal linear rings, this should probably be changed.
|
2014-02-15 21:51:47 +01:00 |
|