Compare commits
| Author | SHA1 | Date | |
|---|---|---|---|
 Clifford Yapp
|
92f826ab30 | ||
|
Clifford Yapp
|
975e0417fb | ||
|
Clifford Yapp
|
b311f67947 | ||
|
Clifford Yapp
|
1eaf01d7b7 |
@@ -1,702 +1,17 @@
|
||||
commit 7a9b8958614d7af947e43a0e3bb18db6587d05d8
|
||||
commit 975e0417fbffe1fb7b2f68b39cbccaad9e04fcb8
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Fri Mar 13 09:09:29 2026 -0400
|
||||
Date: Thu Apr 16 11:50:42 2026 -0400
|
||||
|
||||
brledit is stand-alone intentionally - adjust
|
||||
Spotted by Keith - we weren't respecting the -S flag when redrawing a previously drawn comb
|
||||
|
||||
commit 0bc0943116408f7990cb95f5b106585e45706c47
|
||||
commit b311f6794789b647c191778bbc7b6382a0f46ab7
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Fri Mar 13 08:36:39 2026 -0400
|
||||
Date: Fri Sep 25 11:57:01 2026 -0400
|
||||
|
||||
Update ChangeLog and release date
|
||||
Update version number for 7.42.4 patch
|
||||
|
||||
commit fc905ca063dd18d98c94269351b1a0872a1c7299
|
||||
commit 1eaf01d7b708d73e85b288826c93e3025f15f489
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Thu Mar 12 16:49:14 2026 -0400
|
||||
Date: Wed Sep 23 15:20:33 2026 +0000
|
||||
|
||||
Move brledit to libbu - for bu_editor to work reliably, it has to be there for anything that might call bu_editor - which means it has to be there for libbu.
|
||||
|
||||
(cherry picked from commit dc2cca5853d24d9dea90f2cecb61529e04b42fb7)
|
||||
|
||||
commit 1a86cbd7acc40ec7affff2d046e6a0fa49ef0b21
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Thu Mar 12 15:16:57 2026 -0400
|
||||
|
||||
Move brledit to src/util, set up its license file - it needs to be installed as the last resort editor
|
||||
|
||||
(cherry picked from commit 433722851eb94d7cd0f1c24808191e32e3e742cd)
|
||||
|
||||
commit f2afb65a8e1a1fe4ee0962eb1986e227cec76184
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Thu Mar 12 12:55:48 2026 -0400
|
||||
|
||||
Update ChangeLog with commits being incorporated into rel-7-42-2
|
||||
|
||||
commit c1fc1ee1d8d70e8653c52122519c42ad666ac55b
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Thu Mar 12 12:42:58 2026 -0400
|
||||
|
||||
Update release numbers and date
|
||||
|
||||
commit 2a0af414df146fa4ee69cc8a51eea5dbe0e3a121
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Thu Aug 28 19:24:40 2025 -0400
|
||||
|
||||
Bump libpkg error msg buffer size
|
||||
|
||||
One of the GitHub CI runners had a long enough path that it hit the 128
|
||||
char limit. Need to pick some value, so going with the fallback default
|
||||
for MAXPATHLEN from common.h plus a little extra for message chars.
|
||||
|
||||
(cherry picked from commit ad91dec0282920feba3cff7dd176b898449786b7)
|
||||
|
||||
commit 88df8abbeb1fd56da360cfb45e02405fcc1f43a6
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Mon Mar 9 09:24:19 2026 -0400
|
||||
|
||||
document e46c6ddb6a and 1ac6bde653
|
||||
|
||||
commit 1ac6bde653b67d20cc6eb3c8b7949cccc49aeb5d
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Tue Dec 23 16:05:55 2025 -0500
|
||||
|
||||
another stab at better scaling gqa grid
|
||||
|
||||
We need to always work in mm. Otherwise we restrict our grid to 1 UNIT -
|
||||
ie if we're working in m, the grid can only go down to 1m = 1000mm (is a
|
||||
very loose grid).
|
||||
|
||||
Also, set the grid lower bound to be 1/10th the upper to see if we can
|
||||
get some cheap convergence and higher accuracy value without sacrificing
|
||||
too much speed.
|
||||
NOTE: this value is still largely arbitrary and would be nice to verify
|
||||
how much error this is introducing over a much slower precision run.
|
||||
|
||||
commit e46c6ddb6abaf3903b16315d356ae39aaee6f5a8
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Thu Sep 18 02:39:25 2025 -0400
|
||||
|
||||
need to have grid_step of atleast 1
|
||||
|
||||
commit e47ace67925c728acb527a15ea106ddf063f5d1e
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Fri Feb 20 09:58:20 2026 -0500
|
||||
|
||||
Bump patch number
|
||||
|
||||
commit b7ed363f3b10ee784c57cb99f81840156a46146b
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Thu Feb 19 12:12:44 2026 -0500
|
||||
|
||||
Note crash fix for invalid data input
|
||||
|
||||
commit 5982ac1303e56a5162916b63d3117ae1ab938c99
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Wed Feb 18 13:24:46 2026 -0500
|
||||
|
||||
cleanup unused var
|
||||
|
||||
commit b5440039ed920eabe92f47fca3cf76807ba284aa
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Wed Feb 18 12:51:13 2026 -0500
|
||||
|
||||
oops, inverted logic
|
||||
|
||||
commit 542354b6eb5140c56b89ae8e9e934d780190306c
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Wed Feb 18 11:05:06 2026 -0500
|
||||
|
||||
these should not be possible
|
||||
|
||||
since we've validated the point, vector math with real numbers should
|
||||
not produce NaN/INF's. Remove unnecessary if check, but leave an assert
|
||||
for completeness' sake
|
||||
|
||||
commit 75663b3f1842a81761d1d666f2cb8f08b0866265
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Wed Feb 18 11:03:27 2026 -0500
|
||||
|
||||
sanity asserts
|
||||
|
||||
commit 0fa200fb741d47f8f90977750f5718cbb3ce613e
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Wed Feb 18 10:03:20 2026 -0500
|
||||
|
||||
be a little more memory friendly
|
||||
|
||||
There's no reason to always allocate valid_faces array during prep.
|
||||
Instead take a two-loop approach: the first pass assumes all faces are
|
||||
valid and computes centroids + bounds directly. If an invalid face is
|
||||
encountered, we enter a 'heavier' path where we then create the valid
|
||||
face mapping to compact the array (skip the bad data).
|
||||
|
||||
This preserves crash resilience while reducing memory usage for the
|
||||
"correct" usage of bots with good data.
|
||||
|
||||
commit 56a6640cb9eb1443103011c8e8cc9718fb585061
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Thu Feb 12 12:51:06 2026 -0500
|
||||
|
||||
better safeguards on bot prep
|
||||
|
||||
Josh caught a bot that crashes when raytraced. So, add a safety function
|
||||
to protect prep from doing vector math on bad data (nan's / inf's).
|
||||
Might have gotten a bit carried away with the validation checks -
|
||||
performance impact is currently untested.
|
||||
|
||||
In the future, we could be smarter about stitching out the bad data, but
|
||||
this edit matches old (< 7.42.0) behavior of just ignoring the problem
|
||||
geometry and moving on to ray tracing.
|
||||
|
||||
commit e5044cc491ac7ae7ba34f9d6d0ca3af2990be1c5
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Tue Feb 17 11:27:38 2026 -0500
|
||||
|
||||
Right, the Qt6 approach won't work with Qt5
|
||||
|
||||
commit 02e8d3ec8e22fd10486413f5ce53a21a4d45d3c4
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Thu Feb 12 09:38:55 2026 -0500
|
||||
|
||||
Add stepcode dep update to NEWS items
|
||||
|
||||
commit b096fe7f7f0ce978e4fa0626a5b68452e743bc41
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Thu Feb 12 09:37:08 2026 -0500
|
||||
|
||||
Update sha1 to point to latest bext branch updates
|
||||
|
||||
commit 7eddd4769bb7159e857e73d9a95dba9b41a76080
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Wed Feb 11 10:16:18 2026 -0500
|
||||
|
||||
Updates for warnings and deprecations with newer Qt6
|
||||
|
||||
commit 3db6857feaff3afed1b5884951aef2d029120d3a
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Wed Sep 10 13:45:48 2025 -0400
|
||||
|
||||
Add brledit editor as a last-resort fallback
|
||||
|
||||
Originally based on https://github.com/troglobit/mg
|
||||
|
||||
We went with pico/nano style key bindings rather than emacs-style, since the
|
||||
latter is more complex to deal with for a casual user. Anyone wanting more
|
||||
powerful editing capabilities is also likely to have their preferred editor
|
||||
set up - the envisioned use case for this is constrained/last-resort cases
|
||||
where the only option is to use what is available.
|
||||
|
||||
The original code has been extensively reworked, simplified, etc. to produce a
|
||||
minimalist console editor that can work both on Linux/OSX/etc. type systems and
|
||||
the various Windows console environments. The latter in particular are very
|
||||
difficult to deal with, and there will likely be various corner cases we
|
||||
don't/can't handle well, but brledit has demonstrated the ability to edit a
|
||||
file in PowerShell, Command Prompt and git bash.
|
||||
|
||||
Typically, the only environment where we might not have access to at least a vi
|
||||
editor is a headless Windows server. If we're going to maintain console mode
|
||||
on a Windows desktop without 3rd party editors installed brledit also comes
|
||||
into play , but in that situation we can generally also fall back on local GUI
|
||||
editors to get the job done as well. In one sense yori's yedit, a Windows-only
|
||||
console editor and our previous last resort, covers the only truly problematic
|
||||
case. Running in git bash, for example, we will generally have nano available
|
||||
and so not need brledit. However, the switch is still worthwhile - brledit is
|
||||
both a much smaller code footprint than yori and also works on POSIX type
|
||||
systems in case some unforseen circumstance DOES end up making having our own
|
||||
fallback there desirable or necessary.
|
||||
|
||||
In addition to brledit itself we have added new logic to the libtermio.h header
|
||||
to deal with the requirements of its interace, which are much more extensive
|
||||
than any of the other BRL-CAD utilities still leveraging libtermio.h. Previous
|
||||
functionality was left in place, so the libtermio.h header update should
|
||||
introduce no behaviorial changes in other tools.
|
||||
|
||||
This commit also swaps brledit in for yedit in bu_editor logic and removes the
|
||||
yori directory we previously added to support compiling yedit as part of
|
||||
BRL-CAD.
|
||||
|
||||
(cherry picked from commit 400037e12d00e5dd1b7e5cade93390f43b5d5240)
|
||||
|
||||
commit 5d971c51d649c76fc4e411091f89358d3662ba2f
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Tue Feb 10 09:37:32 2026 -0500
|
||||
|
||||
Typo
|
||||
|
||||
commit 15029fae8b4db00e102113e2c4be354b909a0ca8
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Mon Feb 9 14:24:50 2026 -0500
|
||||
|
||||
Use full checkout rather than co alias
|
||||
|
||||
commit 7edeafaf48f7d23e117afb0eb23b2fabea2eafc8
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Mon Feb 9 12:57:03 2026 -0500
|
||||
|
||||
Match CMakeLists.txt sha1
|
||||
|
||||
commit a1ce5c10a3e4e6b8c9b642be848f6bb5850d6228
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Mon Feb 9 09:47:05 2026 -0500
|
||||
|
||||
As with bext itself, only target macos-14 for this release
|
||||
|
||||
commit ee5c0e12bcec13b6f890370771eb18ee5bce4555
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Mon Feb 9 09:45:29 2026 -0500
|
||||
|
||||
Point 7-42-2 test branch at the latest test 7-42-2 bext sha1
|
||||
|
||||
commit a5e37dd4472a6cbd8e55d2470bf7477f769c53bf
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Thu Feb 5 14:47:25 2026 -0500
|
||||
|
||||
Update libtiff to v4.7.1
|
||||
|
||||
commit 3862f3c94873548e15c98134c3823f574d257312
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Thu Feb 5 13:14:32 2026 -0500
|
||||
|
||||
A lot of dependencies with CVE updates - update NEWS with the list so far.
|
||||
|
||||
commit ecb84ddb38810d9feac3f58733ec7f632300ada5
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Wed Jan 7 10:22:32 2026 -0500
|
||||
|
||||
Update lc.xml
|
||||
|
||||
update notes for flags to reflect actual behavior (-d flag claims to only list 1 region per id but actually is designed to list only duplicate ids (regions that share ids)).
|
||||
|
||||
(cherry picked from commit a01688ff2d7750a8993b43e277574228310355dc)
|
||||
|
||||
commit cd28b5cfaadfe003459c025066fae87fa1ce140e
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Jan 5 10:05:29 2026 -0500
|
||||
|
||||
Update gqa.xml
|
||||
|
||||
add note on density file order
|
||||
|
||||
(cherry picked from commit e7fb6619591d342c295c6249d678a8645d9ee99f)
|
||||
|
||||
commit 6be84e7082c7e97e32c5c9f5cf4debc7abfd60ec
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Wed Dec 31 15:06:35 2025 -0500
|
||||
|
||||
document Josh's changes
|
||||
|
||||
(cherry picked from commit 6662c3318e3460720adcbab0f73c0c25b623763c)
|
||||
|
||||
commit 3ab24cabca2b4841a436d78893d903307cc80ca6
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Dec 29 14:34:39 2025 -0500
|
||||
|
||||
Update helplib.tcl
|
||||
|
||||
(cherry picked from commit b90147b26d61529d170874cc705d44a3462dba4f)
|
||||
|
||||
commit fe22bb795d232627b20d6f4aaa3b382cb9ac2628
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Dec 29 13:29:33 2025 -0500
|
||||
|
||||
Update view.xml
|
||||
|
||||
expand view save notes (with options listed out) and add example
|
||||
|
||||
(cherry picked from commit 27b23a1c1590a8a835f8134f5abf3f5ca43ed4eb)
|
||||
|
||||
commit 5e67a885b61515b16ce31400ec580e98ae8a8884
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Mon Dec 29 12:56:55 2025 -0500
|
||||
|
||||
want a literal ampersand here
|
||||
|
||||
(cherry picked from commit 0e17a6386d2ea0763a5d85e6c7f7cb6bbda9b929)
|
||||
|
||||
commit f9bb22f1ad999225375a707973cfe6895d7f8c9c
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Dec 29 12:49:20 2025 -0500
|
||||
|
||||
Update get.xml
|
||||
|
||||
reword name description to match syntax
|
||||
|
||||
(cherry picked from commit 44c5c2fbf7d2b84dca4e5a12aaad85274975ef55)
|
||||
|
||||
commit 5ae766bb16a71f383e5c62c3b723796aa23078e1
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Dec 29 12:29:52 2025 -0500
|
||||
|
||||
Update get.xml
|
||||
|
||||
add notes about adjust
|
||||
|
||||
(cherry picked from commit 908911eb6aeea1ba6a9695cfd0fdcd41e53c5ee9)
|
||||
|
||||
commit 4620183b2b538fb55b9713d86381f90079726992
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Dec 29 12:27:15 2025 -0500
|
||||
|
||||
Update CMakeLists.txt
|
||||
|
||||
add new adjust page
|
||||
|
||||
(cherry picked from commit b466ae0e7af483e97cc838cfb44e111251a51801)
|
||||
|
||||
commit f1993cb9148746780158302ea3a0fd158467dd45
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Dec 29 12:26:54 2025 -0500
|
||||
|
||||
Create adjust.xml
|
||||
|
||||
add new page for adjust
|
||||
|
||||
(cherry picked from commit f29054f9e133d177822a907ce54219bbf15728b3)
|
||||
|
||||
commit faf89c6fe216765049e6404f58d155adca02b058
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Dec 29 12:12:14 2025 -0500
|
||||
|
||||
Update helplib.tcl
|
||||
|
||||
update adjust to show that it works on solid parameters (not on metadata attributes). Make a note on get for the same (although it works on some basic attr as well (default region ones I think))
|
||||
|
||||
(cherry picked from commit a7cf1dc61237a73a637ad4ea8b865079d05327fd)
|
||||
|
||||
commit 768b4618344fec3f7637daca340f24c5ce1c881e
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Dec 29 11:52:52 2025 -0500
|
||||
|
||||
Update draw.xml
|
||||
|
||||
note on what -A -o does
|
||||
|
||||
(cherry picked from commit 561a8e50a79a69b771699f7814f8bf5e0e812f3a)
|
||||
|
||||
commit a400b360a4419bddaed893a90713ef248b434186
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Dec 29 11:46:38 2025 -0500
|
||||
|
||||
Update B.xml
|
||||
|
||||
Add flags to match draw command and quick note on -S and -m1.
|
||||
|
||||
(cherry picked from commit b3cccb00e93dc5f635c3b9219844f5f2768ce88a)
|
||||
|
||||
commit 6c7068a311669ae968f0aa90686498a337bda6ec
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Dec 29 11:39:07 2025 -0500
|
||||
|
||||
Update draw.xml
|
||||
|
||||
fix spacing issue after -A -o
|
||||
|
||||
(cherry picked from commit 091e4c881df57c95380d1452004cf22ed870e395)
|
||||
|
||||
commit 516f7c06c2e92409d13047865b352d25977949a1
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Dec 29 11:02:48 2025 -0500
|
||||
|
||||
Update voxelize.xml
|
||||
|
||||
adjust example to be 100 size and note defaults for each value.
|
||||
|
||||
(cherry picked from commit cbe6c1d6cb1c3811e2b4286bccd5ed30cff5edeb)
|
||||
|
||||
commit d9c2e82500435881d3c93ab7dbd7545f458b4f20
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Dec 29 10:29:56 2025 -0500
|
||||
|
||||
Update gqa.cpp
|
||||
|
||||
add p in options after -A (for display with -h)
|
||||
|
||||
(cherry picked from commit b570d5dc2e46b8d83e18f4e5a6d25776d8d44dba)
|
||||
|
||||
commit e90f4c46d24b3918b32e64df1a5c0a30c1b99270
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Dec 29 10:17:18 2025 -0500
|
||||
|
||||
Update center.xml
|
||||
|
||||
fix typo (extra > in example)
|
||||
|
||||
(cherry picked from commit 7cc0f24a162ee601cd824461760df67f572a88d1)
|
||||
|
||||
commit 07423da5bfa5a535c7c7f68c57e9bd5c84d0b4dc
|
||||
Author: joshbaker22 <141415098+joshbaker22@users.noreply.github.com>
|
||||
Date: Mon Aug 18 10:08:31 2025 -0400
|
||||
|
||||
Update gqa.xml - removed extra > after p (under analysis flags)
|
||||
|
||||
(cherry picked from commit 7c4455d3ea220dc1cac8bf8c86b17adbcb223620)
|
||||
|
||||
commit 947d7cf728a89b067ed67880167cda69ce6e0d36
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Wed Feb 4 10:45:24 2026 -0500
|
||||
|
||||
Document open cmd reporting fix
|
||||
|
||||
commit 09a092581ee3d125355ce8abe145ff26f44ec64b
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Fri Jan 23 14:13:12 2026 -0500
|
||||
|
||||
If we don't have a database open, report that.
|
||||
|
||||
(cherry picked from commit 399356125e8429125aad10ae7f2461a86ecb4843)
|
||||
|
||||
commit accdc0384278d81111f4f534a55ddb1506d31cdb
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Wed Jan 21 16:56:27 2026 -0500
|
||||
|
||||
While I'm at it, do the same check in g2asc
|
||||
|
||||
(cherry picked from commit db8e79d886dd12e417930deb9eee8d978aaf310b)
|
||||
|
||||
commit 633eec2a03c7abb431e1860d51e829efc3f9d785
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Wed Jan 21 16:52:18 2026 -0500
|
||||
|
||||
Update asc2g to not overwrite input file if specified as output
|
||||
|
||||
Ugh. Let's not do that again. Sanity check that the user didn't
|
||||
specify the input file as the output file. Previously that would
|
||||
dump output to the input file and corrupt it.
|
||||
|
||||
(cherry picked from commit 12ec562e6cdb1c9e7d910392e87f0831e100f7f1)
|
||||
|
||||
commit 1b4cb31ba6f88e634889c22306ca05861c27f928
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Mon Jan 12 15:11:40 2026 -0500
|
||||
|
||||
Fix default behavior of open cmd when creating non-extant files
|
||||
|
||||
Back in commit e0384c3abc29a6b4 in 2023, I added a -c option to open for
|
||||
creating files. In the process, I inadvertantly broke the default ged open
|
||||
behavior when supplied with a filename that didn't yet exist, and various uses
|
||||
of the opendb command in MGED and Archer depended on that.
|
||||
|
||||
This restores the "create by default" behavior. To allow for cases were we
|
||||
don't want that, we need to add an "o" option to disable it and only open
|
||||
existing files - the default behavior needs to remain create. We'll leave
|
||||
the -c option for compatibility, since it is harmless.
|
||||
|
||||
(cherry picked from commit e7c5ca378047573b00f806dc6c0f3177cd07b610)
|
||||
|
||||
commit 50c8016ffcead1d148b8f1bd93a17fd549339919
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Wed Nov 26 10:21:29 2025 -0500
|
||||
|
||||
Fix static initialization order issues in libged/libdm plugin registries
|
||||
|
||||
Global registry objects (using STL containers) in libged and libdm were defined
|
||||
as static variables at file scope. This risks "static initialization order
|
||||
fiasco" - static constructors from other translation units (TUs) may access the
|
||||
registries before they are initialized, causing crashes or undefined behavior,
|
||||
especially in static builds with plugins.
|
||||
|
||||
The solution is to convert global registry objects to function-local static
|
||||
"Meyers singletons". All registry accesses now use singleton getter functions,
|
||||
ensuring construction is ordered and thread-safe by C++11 guarantees.
|
||||
|
||||
With this approach function-local statics are initialized on first use,
|
||||
guaranteeing that any plugin constructor or TU macro accessing the registry
|
||||
always gets a fully constructed object, regardless of cross-TU initialization
|
||||
order. This makes static and dynamic plugin registration robust across all
|
||||
platforms and build/link scenarios.
|
||||
|
||||
(cherry picked from commit 48a87e7e134c84a71f46db95c59adbaf68f048be)
|
||||
|
||||
commit 7c1e3c303750ab046157dff9dfabfb6e01833eb0
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Tue Nov 25 16:28:46 2025 -0500
|
||||
|
||||
Fixed a crash with bot -h when we don't have a database open - no reason we can't still print help.
|
||||
|
||||
(cherry picked from commit 35bf1f3ca0ce283719037b2060cea7cd1fe2e032)
|
||||
|
||||
commit e8f8715fb1c638c7baf0e50e57ae86bcb90c3c09
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Fri Nov 21 16:33:56 2025 -0500
|
||||
|
||||
Fix libged (i.e. MGED's) handling of nirt -c option to disable ray point backout
|
||||
|
||||
(cherry picked from commit a788c096152fe0181402ab05b08f4744f9c4a250)
|
||||
|
||||
commit efe0496b07b87965cc22227f00043351f7f73b44
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Fri Sep 26 11:20:11 2025 -0400
|
||||
|
||||
Add workaround for broken OSX signal header
|
||||
|
||||
(cherry picked from commit 1a9c712da501e0bd92447e4aac4f3cf825792bb9)
|
||||
|
||||
commit bd2087f3d5877cb8e15bd2ced78cd47c2dbc275c
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Tue Sep 9 12:00:36 2025 -0400
|
||||
|
||||
Don't create an LoD cache unless we need it.
|
||||
|
||||
(cherry picked from commit bccb3733e69f38fed0a3ee3cc9fd1bc340ddef70)
|
||||
|
||||
commit 79949ffd43d0b4aef21621f770dc1c50d75c1bea
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Sat Aug 30 23:42:47 2025 -0400
|
||||
|
||||
Add workaround for broken Xcode signal.h header
|
||||
|
||||
Apparently Xcode 16.4 has a broken signal.h header that uses NSIG but
|
||||
doesn't define it.
|
||||
|
||||
(cherry picked from commit db6809643e4796ad13651301b357da28ec9bb7f4)
|
||||
|
||||
commit c7f2cbdab278bbb6ff0de9cd01cfb21cc10f81a1
|
||||
Author: ChangJoon Lee <changjoonlee@me.com>
|
||||
Date: Thu Aug 21 17:39:18 2025 +0900
|
||||
|
||||
Remove documentation error in clang
|
||||
|
||||
(cherry picked from commit 68ad20e42c7fb8ab18fff0a0ebeea4cad9f66ab3)
|
||||
|
||||
commit a3903fab55bc1eb7e6eac9bd6e246cb9ed45770f
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Sun Aug 17 00:13:30 2025 -0400
|
||||
|
||||
Add error case if all THREADLOCAL attempts fail
|
||||
|
||||
Per suggestion from Sean, make the failure to define THREADLOCAL
|
||||
explicit.
|
||||
|
||||
(cherry picked from commit 5e54a73797e415cbc8e794a6b2c94c78875e09c8)
|
||||
|
||||
commit 44782d88bdc7998c750fff305c007339d2e4f5f3
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Sat Aug 16 23:07:18 2025 -0400
|
||||
|
||||
Define THREADLOCAL to use for per-thread storage
|
||||
|
||||
How to properly identify per-thread storage varies somewhat with
|
||||
different standards and platforms - boil the various cases down to a
|
||||
THREADLOCAL encapsulation we can use in our code.
|
||||
|
||||
(cherry picked from commit fc7c51af95cc18e3adddcbf258429b77ef59c432)
|
||||
|
||||
commit c8a53047c910543be6a44cc9b1cece98d124dc72
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Tue Aug 12 16:30:43 2025 -0400
|
||||
|
||||
Stop bot remesh subcmd from clearing output
|
||||
|
||||
Was truncating the help message being built up for the bot subcommands.
|
||||
|
||||
(cherry picked from commit 3dd6914b578337e32861d6cff85ccd6dc1c3ef20)
|
||||
|
||||
commit 8482dc235a0a436a90995611b2ef8fae2b83cb21
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Wed Feb 4 10:00:05 2026 -0500
|
||||
|
||||
Correct BoT raytracing behavior when starting pt inside mesh
|
||||
|
||||
Josh found a test case (long small arb subtracted from narrow large arb) where
|
||||
the new raytracing code wasn't properly evaluating the subtraction.
|
||||
|
||||
Turned out that the new code wasn't backing a starting ray point that was
|
||||
inside a BoT mesh out of the mesh before intersecting with the BVH tree - the
|
||||
result was that only the triangles after the start of the ray were reported.
|
||||
That in turn resulted in single hit reports that produced no solid segments -
|
||||
with nothing for boolweave to process from the subtraction solid, we got no
|
||||
subtraction in the rendering result.
|
||||
|
||||
Fix is straightforward and consistent with how other primitives are handled -
|
||||
back the start point out along the ray for the solve so we properly get all
|
||||
triangles from a mesh intersection even when the originally specified ray pt is
|
||||
inside the mesh.
|
||||
|
||||
This evidently didn't show up in more normal testing because it is rare for a
|
||||
starting ray point to be set up inside geometry - when the same geometry
|
||||
configuration was set up as a union rather than a subtraction it worked,
|
||||
because the staring ray point calcluated for the scene incorprated both
|
||||
bounding boxes. The overall scene bounding calculation does not incorporate
|
||||
subtractions (since they do not contribute positive volume they're not a good
|
||||
basis for things like autoview sizing) and in most geometries that means the
|
||||
ray point is already backed up "far enough". The corner case of a subtraction
|
||||
volume outside the overall scene was what was needed to trigger this issue.
|
||||
|
||||
(cherry picked from commit 0a8a286e34a6f81d3f44f84b0f1f1bac863394e3)
|
||||
|
||||
commit a3091bed40140a40e49a83007511287d4fb0167b
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Wed Jan 21 18:11:35 2026 -0500
|
||||
|
||||
this is wrong
|
||||
|
||||
(cherry picked from commit fa85521fd564bca1adb033e1b38b5a6096216311)
|
||||
|
||||
commit 1972c0246677828ca5f1658780820e086c7b1ae3
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Wed Jan 21 18:06:55 2026 -0500
|
||||
|
||||
get rid of the magic number
|
||||
|
||||
since we're now ensuring the ray origin WILL be outside of the root
|
||||
node, we can ditch the -1 magic tolerance value here and use the
|
||||
standard slab rejection criteria.
|
||||
|
||||
Use SMALL_FASTF instead of straight up 0.0 to give us a little
|
||||
fuzz protection.
|
||||
|
||||
(cherry picked from commit 106f5fa6e553595b11b33b888841c4dad4ee91fa)
|
||||
|
||||
commit 1d759184924877f6ffc02678ddf63814dcb98907
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Wed Jan 21 18:00:10 2026 -0500
|
||||
|
||||
only backup our ray pt if we need to
|
||||
|
||||
the rp->r_min has already accounted for if the ray is inside the
|
||||
bounding sphere. We can us it directly to back up rather than blindly
|
||||
applying a bounding sphere diameter backout. This has the added
|
||||
benefit of keeping our bot_shot_hlbvh_flat() function signature clean
|
||||
of the soltab.
|
||||
|
||||
we can also do the backout in-line to avoid the tiny load/store overhead
|
||||
and improve formula readability.
|
||||
|
||||
(cherry picked from commit 0cf43403e8a3746b6312535e5b0ba884e0f0272e)
|
||||
|
||||
commit 7376385fd24f0e55dcc3525f5085bfffbe924bcd
|
||||
Author: Clifford Yapp <238416+starseeker@users.noreply.github.com>
|
||||
Date: Wed Jan 21 15:50:03 2026 -0500
|
||||
|
||||
For the initial fast culling test, back our r_pt out in the inverse ray direction.
|
||||
|
||||
Also try avoiding NaNs, but not sure this will make a major difference either
|
||||
way performance wise.
|
||||
|
||||
(cherry picked from commit 042338c49cbc3b72658ff5d4960f24de997a15fa)
|
||||
|
||||
commit e12863573f8d7809c7ab1b8792475cc2a63b0600
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Tue Jan 20 14:44:01 2026 -0500
|
||||
|
||||
improve readability
|
||||
|
||||
probably way overkill on the comments but oh well
|
||||
|
||||
(cherry picked from commit fb2757bee85730fe39eb3a8405300f573de512f7)
|
||||
|
||||
commit a7cf755c0365ae1174875e4ad94e06d8a13f18f7
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Tue Jan 20 13:27:54 2026 -0500
|
||||
|
||||
pretty sure we want don't want bitwise
|
||||
|
||||
(cherry picked from commit 42ce6f5d64c1ac1fb0a335d1fbc934ef6e0c7014)
|
||||
|
||||
commit 437eaaf3ae7c1871ad33548509b6a556eb3badfe
|
||||
Author: f4alt <christopherjmacgregor@gmail.com>
|
||||
Date: Tue Jan 20 13:26:42 2026 -0500
|
||||
|
||||
make sure we're not accidentally wiping out the ray
|
||||
|
||||
(cherry picked from commit 3eba161f7b0aa5e825b16b85bf8fcb5ec7a817b0)
|
||||
Backport commit de09372b73b fixing BoT raytracing interacting with bad r_min settings in multi-solid raytracing.
|
||||
|
||||
@@ -9,6 +9,17 @@ first and are grouped by release. Each release also includes optional
|
||||
descriptive text that emphasizes or further describes significant
|
||||
changes made. See document footer for additional details.
|
||||
|
||||
----------------------------------------------------------------------
|
||||
--- 2026-10-08 Release 7.42.4 ---
|
||||
----------------------------------------------------------------------
|
||||
|
||||
Follow-on patch release to 7.42.2 correcting a case where multiple
|
||||
object raytracing could result in rays whose origins start inside
|
||||
BoT primitives not backing out of the solid correctly.
|
||||
|
||||
* fixed ray origin pnt backout behavior in BoT primitives - Cliff Yapp
|
||||
* respect-S flag redrawing a previously drawn comb - Cliff Yapp
|
||||
|
||||
----------------------------------------------------------------------
|
||||
--- 2026-03-13 Release 7.42.2 ---
|
||||
----------------------------------------------------------------------
|
||||
|
||||
@@ -1,5 +1,5 @@
|
||||
BRL-CAD
|
||||
Release 7.42.2
|
||||
Release 7.42.4
|
||||
http://brlcad.org/
|
||||
|
||||
BRL-CAD is a powerful cross-platform open source combinatorial
|
||||
|
||||
+1
-1
@@ -1 +1 @@
|
||||
2
|
||||
4
|
||||
|
||||
@@ -1,3 +1,10 @@
|
||||
brlcad (7.42.4-0) unstable; urgency=low
|
||||
|
||||
* update brlcad version
|
||||
|
||||
-- Cliff Yapp <cyapp@brlcad.org> Thu, 08 Oct 2026 20:00:00 UTC
|
||||
|
||||
|
||||
brlcad (7.42.2-0) unstable; urgency=low
|
||||
|
||||
* update brlcad version
|
||||
|
||||
@@ -11,7 +11,7 @@
|
||||
\f0\i\b\fs72 \cf0 README
|
||||
\fs36 \
|
||||
|
||||
\f1\i0\b0\fs28 This is BRL-CAD 7.42.2 for Mac OS X. Please send any bugs, comments, questions, or suggestions to the project website.\
|
||||
\f1\i0\b0\fs28 This is BRL-CAD 7.42.4 for Mac OS X. Please send any bugs, comments, questions, or suggestions to the project website.\
|
||||
|
||||
\fs22 \
|
||||
|
||||
|
||||
@@ -4,7 +4,7 @@
|
||||
\margl1440\margr1440\vieww8020\viewh6280\viewkind0
|
||||
\pard\tx560\tx1120\tx1680\tx2240\tx2800\tx3360\tx3920\tx4480\tx5040\tx5600\tx6160\tx6720\ql\qnatural
|
||||
|
||||
\f0\fs22 \cf0 This installer will guide you through the installation of BRL-CAD 7.42.2 for Mac OS X and Darwin.\
|
||||
\f0\fs22 \cf0 This installer will guide you through the installation of BRL-CAD 7.42.4 for Mac OS X and Darwin.\
|
||||
\
|
||||
This distribution of BRL-CAD requires approximately 500 MB of available disk space and does not include source code.\
|
||||
\
|
||||
|
||||
@@ -2362,6 +2362,17 @@ BViewState::scene_obj(
|
||||
sp->s_soldash = (is_subtract) ? 1 : 0;
|
||||
}
|
||||
|
||||
// Align with vs draw_non_subtract_only settings
|
||||
if (vs->s_dmode == curr_mode) {
|
||||
if (sp->s_soldash && vs->draw_non_subtract_only) {
|
||||
if (sp->s_flag != DOWN)
|
||||
sp->s_flag = DOWN;
|
||||
} else {
|
||||
if (sp->s_flag != UP)
|
||||
sp->s_flag = UP;
|
||||
}
|
||||
}
|
||||
|
||||
// Set line width, if the user specified a non-default value
|
||||
if (vs && vs->s_line_width)
|
||||
sp->s_os->s_line_width = vs->s_line_width;
|
||||
|
||||
@@ -701,8 +701,46 @@ rt_bot_makesegs(hit_da *hits,
|
||||
struct rt_piecestate *psp);
|
||||
|
||||
|
||||
/* r_min may be the start of a spatial partition cell inside this solid.
|
||||
* Start BVH traversal just before the ray enters the BoT root box so its
|
||||
* entrance triangles are considered when assembling solid segments. */
|
||||
static void
|
||||
bot_backed_origin(point_t b_pt, const struct xray *rp, const struct bvh_flat_node *root,
|
||||
const vect_t inv_dir, const struct soltab *stp)
|
||||
{
|
||||
fastf_t backout = MAX_FASTF;
|
||||
int axis;
|
||||
|
||||
for (axis = X; axis <= Z; axis++) {
|
||||
if (rp->r_pt[axis] < root->bounds[axis] ||
|
||||
rp->r_pt[axis] > root->bounds[axis + 3]) {
|
||||
VMOVE(b_pt, rp->r_pt);
|
||||
return;
|
||||
}
|
||||
}
|
||||
|
||||
for (axis = X; axis <= Z; axis++) {
|
||||
fastf_t distance;
|
||||
if (rp->r_dir[axis] > 0.0)
|
||||
distance = (rp->r_pt[axis] - root->bounds[axis]) * inv_dir[axis];
|
||||
else if (rp->r_dir[axis] < 0.0)
|
||||
distance = (root->bounds[axis + 3] - rp->r_pt[axis]) * -inv_dir[axis];
|
||||
else
|
||||
continue;
|
||||
backout = FMIN(backout, distance);
|
||||
}
|
||||
|
||||
if (backout < MAX_FASTF) {
|
||||
fastf_t padding = stp->st_rtip ? stp->st_rtip->rti_tol.dist : BN_TOL_DIST;
|
||||
VJOIN1(b_pt, rp->r_pt, -(backout + padding), rp->r_dir);
|
||||
} else {
|
||||
VMOVE(b_pt, rp->r_pt);
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
void
|
||||
bot_shot_hlbvh_flat(struct bvh_flat_node *root, struct xray* rp, triangle_s *tris, size_t ntris, hit_da* hits, fastf_t toldist)
|
||||
bot_shot_hlbvh_flat(struct bvh_flat_node *root, struct xray* rp, triangle_s *tris, size_t ntris, hit_da* hits, fastf_t toldist, const struct soltab *stp)
|
||||
{
|
||||
struct bvh_flat_node *stack_node[HLBVH_STACK_SIZE];
|
||||
unsigned char stack_child_index[HLBVH_STACK_SIZE];
|
||||
@@ -718,25 +756,10 @@ bot_shot_hlbvh_flat(struct bvh_flat_node *root, struct xray* rp, triangle_s *tri
|
||||
inverse_r_dir[1] = RAYDIR_INV(rp->r_dir[1]);
|
||||
inverse_r_dir[2] = RAYDIR_INV(rp->r_dir[2]);
|
||||
|
||||
// Because we are doing solid shotlining, we need to intersect all
|
||||
// triangles on the line of the ray, even when the r_pt is inside the mesh.
|
||||
// This means we DON'T want to cull hlbvh leaves behind our r_pt when the
|
||||
// r_pt is inside the mesh. To avoid this, we back our r_pt up to the
|
||||
// bounding sphere if needed. This incurs a slight
|
||||
// performance penalty when r_pt is past the mesh but still within the
|
||||
// bounding sphere diameter distance, since backing up will cause false
|
||||
// bbox intersect matches in those cases. This is necessary to ensure we
|
||||
// get the intersections needed for solid segment creation - the subsequent
|
||||
// full intersection solves will still produce the correct results.
|
||||
//
|
||||
// NOTE: We DO, however, need to use the real rp->r_pt when doing the
|
||||
// actual intersection solve math so our segments end up in the right
|
||||
// place.
|
||||
fastf_t backout = FMAX(0.0, -rp->r_min);
|
||||
/* Only BVH culling uses the backed origin. Triangle intersections must
|
||||
* remain relative to the original ray origin. */
|
||||
point_t b_pt;
|
||||
b_pt[X] = rp->r_pt[X] - backout * rp->r_dir[X];
|
||||
b_pt[Y] = rp->r_pt[Y] - backout * rp->r_dir[Y];
|
||||
b_pt[Z] = rp->r_pt[Z] - backout * rp->r_dir[Z];
|
||||
bot_backed_origin(b_pt, rp, root, inverse_r_dir, stp);
|
||||
|
||||
while (stack_ind >= 0) {
|
||||
if (UNLIKELY(stack_ind >= HLBVH_STACK_SIZE)) {
|
||||
@@ -900,7 +923,7 @@ rt_bot_shot(struct soltab *stp, struct xray *rp, struct application *ap, struct
|
||||
toldist = (DBL_EPSILON * stp->st_aradius * 10);
|
||||
}
|
||||
|
||||
bot_shot_hlbvh_flat(sps->root, rp, sps->tris, bot->bot_ntri, &hits_per_cpu, toldist);
|
||||
bot_shot_hlbvh_flat(sps->root, rp, sps->tris, bot->bot_ntri, &hits_per_cpu, toldist, stp);
|
||||
|
||||
if (hits_per_cpu.count == 0) {
|
||||
return 0;
|
||||
|
||||
@@ -80,6 +80,10 @@ brlcad_addexec(rt_arb8 arb8_tests.c "librt;libwdb;libbg;libbu" TEST)
|
||||
|
||||
brlcad_addexec(bot_match bot_match.cpp "librt;libbg" TEST)
|
||||
|
||||
# BoT rays starting inside the mesh, including a spatially partitioned scene
|
||||
brlcad_addexec(rt_bot_inside bot_inside.c "librt;libwdb;libbu;${M_LIBRARY}" TEST)
|
||||
brlcad_add_test(NAME rt_bot_inside COMMAND rt_bot_inside)
|
||||
|
||||
# boolweave testing
|
||||
brlcad_addexec(rt_boolweave rt_boolweave.c "librt" TEST)
|
||||
|
||||
@@ -91,6 +95,7 @@ set(
|
||||
CMakeLists.txt
|
||||
arb_intersect.g
|
||||
binary_attribute.c
|
||||
bot_inside.c
|
||||
brep_boolean_tests.g
|
||||
cyclic_tests.g
|
||||
extreme_ssi_test.g
|
||||
|
||||
@@ -0,0 +1,327 @@
|
||||
/* B O T _ I N S I D E . C
|
||||
* BRL-CAD
|
||||
*
|
||||
* Copyright (c) 2026 United States Government as represented by
|
||||
* the U.S. Army Research Laboratory.
|
||||
*
|
||||
* This library is free software; you can redistribute it and/or
|
||||
* modify it under the terms of the GNU Lesser General Public License
|
||||
* version 2.1 as published by the Free Software Foundation.
|
||||
*
|
||||
* This library is distributed in the hope that it will be useful, but
|
||||
* WITHOUT ANY WARRANTY; without even the implied warranty of
|
||||
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
||||
* Lesser General Public License for more details.
|
||||
*
|
||||
* You should have received a copy of the GNU Lesser General Public
|
||||
* License along with this file; see the file named COPYING for more
|
||||
* information.
|
||||
*/
|
||||
|
||||
#include "common.h"
|
||||
|
||||
#include <math.h>
|
||||
#include <string.h>
|
||||
|
||||
#include "bu/app.h"
|
||||
#include "bu/log.h"
|
||||
#include "bu/malloc.h"
|
||||
#include "raytrace.h"
|
||||
#include "rt/geom.h"
|
||||
#include "wdb.h"
|
||||
|
||||
/* ARS prep also uses BoT raytracing, but its RPP bounds can refresh r_min.
|
||||
* Compare converted ARS and native BoT for direct and scene shots. */
|
||||
#define RING_SIDES 8
|
||||
#define RING_POINTS (RING_SIDES + 1)
|
||||
|
||||
static int
|
||||
check_shot(const struct rt_functab *ftp, struct soltab *stp, struct application *ap, struct xray *ray, const char *label)
|
||||
{
|
||||
struct seg seghead;
|
||||
struct seg *segp;
|
||||
struct resource *resp = ap->a_resource;
|
||||
int nhits;
|
||||
int failed = 0;
|
||||
|
||||
BU_LIST_INIT(&seghead.l);
|
||||
nhits = ftp->ft_shot(stp, ray, ap, &seghead);
|
||||
if (nhits <= 0 || BU_LIST_IS_EMPTY(&seghead.l)) {
|
||||
bu_log("%s: inside-origin ray missed\n", label);
|
||||
failed = 1;
|
||||
} else {
|
||||
segp = BU_LIST_FIRST(seg, &seghead.l);
|
||||
if (!NEAR_EQUAL(segp->seg_in.hit_dist, -9.0, SMALL_FASTF) ||
|
||||
!NEAR_EQUAL(segp->seg_out.hit_dist, 7.0, SMALL_FASTF)) {
|
||||
bu_log("%s: interval %g..%g, expected -9..7\n", label,
|
||||
segp->seg_in.hit_dist, segp->seg_out.hit_dist);
|
||||
failed = 1;
|
||||
}
|
||||
}
|
||||
|
||||
while (BU_LIST_WHILE(segp, seg, &seghead.l)) {
|
||||
BU_LIST_DEQUEUE(&segp->l);
|
||||
RT_FREE_SEG(segp, resp);
|
||||
}
|
||||
|
||||
return failed;
|
||||
}
|
||||
|
||||
struct scene_result {
|
||||
int subject_hits;
|
||||
fastf_t out_dist;
|
||||
};
|
||||
|
||||
static int
|
||||
scene_hit(struct application *ap, struct partition *part_head,
|
||||
struct seg *UNUSED(finished_segs))
|
||||
{
|
||||
struct scene_result *result = (struct scene_result *)ap->a_uptr;
|
||||
struct partition *part;
|
||||
|
||||
for (part = part_head->pt_forw; part != part_head; part = part->pt_forw) {
|
||||
if (strstr(part->pt_regionp->reg_name, "subject.r")) {
|
||||
result->subject_hits++;
|
||||
result->out_dist = part->pt_outhit->hit_dist;
|
||||
}
|
||||
}
|
||||
return result->subject_hits > 0;
|
||||
}
|
||||
|
||||
static int
|
||||
scene_miss(struct application *UNUSED(ap))
|
||||
{
|
||||
return 0;
|
||||
}
|
||||
|
||||
/* The support sphere makes the cut tree start the ray's first cell
|
||||
* inside the subject, while the subject itself spans the entire cell. */
|
||||
static int
|
||||
check_scene(int subject_id, const fastf_t *const curve_data[3],
|
||||
fastf_t *vertices, int *faces, const char *file)
|
||||
{
|
||||
struct db_i *dbip = NULL;
|
||||
struct rt_wdb *wdbp = NULL;
|
||||
struct rt_i *rtip = NULL;
|
||||
struct resource resp = RT_RESOURCE_INIT_ZERO;
|
||||
struct wmember members;
|
||||
struct application ap;
|
||||
struct scene_result result = {0};
|
||||
point_t support_center;
|
||||
int failed = 0;
|
||||
|
||||
if (file) {
|
||||
wdbp = wdb_fopen(file);
|
||||
if (wdbp)
|
||||
dbip = wdbp->dbip;
|
||||
} else {
|
||||
dbip = db_create_inmem();
|
||||
if (dbip)
|
||||
wdbp = wdb_dbopen(dbip, RT_WDB_TYPE_DB_INMEM);
|
||||
}
|
||||
if (!wdbp) {
|
||||
if (dbip)
|
||||
db_close(dbip);
|
||||
return 1;
|
||||
}
|
||||
|
||||
if (subject_id == ID_ARS) {
|
||||
fastf_t **curves = (fastf_t **)bu_malloc(3 * sizeof(fastf_t *), "ARS curves");
|
||||
size_t i;
|
||||
for (i = 0; i < 3; i++) {
|
||||
curves[i] = (fastf_t *)bu_malloc(RING_POINTS * ELEMENTS_PER_POINT *
|
||||
sizeof(fastf_t), "ARS curve");
|
||||
memcpy(curves[i], curve_data[i],
|
||||
RING_POINTS * ELEMENTS_PER_POINT * sizeof(fastf_t));
|
||||
}
|
||||
failed += mk_ars(wdbp, "subject.s", 3, RING_POINTS, curves) != 0;
|
||||
} else {
|
||||
failed += mk_bot(wdbp, "subject.s", RT_BOT_SOLID, RT_BOT_CCW, 0,
|
||||
RING_SIDES + 2, RING_SIDES * 2, vertices, faces,
|
||||
NULL, NULL) != 0;
|
||||
}
|
||||
BU_LIST_INIT(&members.l);
|
||||
if (!mk_addmember("subject.s", &members.l, NULL, WMOP_UNION))
|
||||
failed++;
|
||||
failed += mk_lcomb(wdbp, "subject.r", &members, 1, NULL, NULL, NULL, 0) != 0;
|
||||
|
||||
VSET(support_center, 1.5, 0, 4);
|
||||
failed += mk_sph(wdbp, "support.s", support_center, 1.0) != 0;
|
||||
BU_LIST_INIT(&members.l);
|
||||
if (!mk_addmember("support.s", &members.l, NULL, WMOP_UNION))
|
||||
failed++;
|
||||
failed += mk_lcomb(wdbp, "support.r", &members, 1, NULL, NULL, NULL, 0) != 0;
|
||||
if (failed)
|
||||
goto done;
|
||||
|
||||
rtip = rt_new_rti(dbip);
|
||||
if (!rtip || rt_gettree(rtip, "subject.r") ||
|
||||
rt_gettree(rtip, "support.r")) {
|
||||
failed = 1;
|
||||
goto done;
|
||||
}
|
||||
rt_prep(rtip);
|
||||
rt_init_resource(&resp, 0, rtip);
|
||||
|
||||
RT_APPLICATION_INIT(&ap);
|
||||
ap.a_rt_i = rtip;
|
||||
ap.a_resource = &resp;
|
||||
ap.a_hit = scene_hit;
|
||||
ap.a_miss = scene_miss;
|
||||
ap.a_uptr = &result;
|
||||
VSET(ap.a_ray.r_pt, 3, 0, 4);
|
||||
VSET(ap.a_ray.r_dir, 1, 0, 0);
|
||||
rt_shootray(&ap);
|
||||
if (result.subject_hits != 1 ||
|
||||
!NEAR_EQUAL(result.out_dist, 5.0, rtip->rti_tol.dist)) {
|
||||
bu_log("%s scene: %d subject partitions, exit %.17g; expected one exit at 5\n",
|
||||
subject_id == ID_ARS ? "ARS" : "BoT",
|
||||
result.subject_hits, result.out_dist);
|
||||
failed = 1;
|
||||
}
|
||||
|
||||
done:
|
||||
if (rtip)
|
||||
rt_free_rti(rtip);
|
||||
wdb_close(wdbp);
|
||||
return failed;
|
||||
}
|
||||
|
||||
int
|
||||
main(int argc, char **argv)
|
||||
{
|
||||
fastf_t curve_data[3][RING_POINTS * ELEMENTS_PER_POINT] = {{0}};
|
||||
fastf_t *curves[3] = {curve_data[0], curve_data[1], curve_data[2]};
|
||||
fastf_t bot_vertices[(RING_SIDES + 2) * ELEMENTS_PER_POINT] = {0};
|
||||
int bot_faces[RING_SIDES * 2 * 3];
|
||||
struct rt_bot_internal bot = {0};
|
||||
struct rt_ars_internal ars = {0};
|
||||
struct rt_db_internal intern;
|
||||
struct rt_db_internal bot_intern;
|
||||
struct soltab st = {0};
|
||||
struct soltab bot_st = {0};
|
||||
struct rt_i *rtip;
|
||||
struct resource *resp = &rt_uniresource;
|
||||
struct application ap;
|
||||
struct xray ray = {0};
|
||||
vect_t invdir;
|
||||
size_t i;
|
||||
int failed = 0;
|
||||
|
||||
(void)argc;
|
||||
bu_setprogname(argv[0]);
|
||||
|
||||
for (i = 0; i < RING_POINTS; i++) {
|
||||
fastf_t angle = 2.0 * M_PI * (i % RING_SIDES) / RING_SIDES;
|
||||
VSET(&curve_data[1][i * ELEMENTS_PER_POINT],
|
||||
10.0 * cos(angle), 10.0 * sin(angle), 5.0);
|
||||
if (i < RING_SIDES)
|
||||
VMOVE(&bot_vertices[(i + 1) * ELEMENTS_PER_POINT],
|
||||
&curve_data[1][i * ELEMENTS_PER_POINT]);
|
||||
curve_data[2][i * ELEMENTS_PER_POINT + Z] = 10.0;
|
||||
}
|
||||
bot_vertices[(RING_SIDES + 1) * ELEMENTS_PER_POINT + Z] = 10.0;
|
||||
for (i = 0; i < RING_SIDES; i++) {
|
||||
int current = (int)i + 1;
|
||||
int next = (int)((i + 1) % RING_SIDES) + 1;
|
||||
int *bottom = &bot_faces[i * 6];
|
||||
int *top = &bottom[3];
|
||||
bottom[0] = 0;
|
||||
bottom[1] = next;
|
||||
bottom[2] = current;
|
||||
top[0] = current;
|
||||
top[1] = next;
|
||||
top[2] = RING_SIDES + 1;
|
||||
}
|
||||
bot.magic = RT_BOT_INTERNAL_MAGIC;
|
||||
bot.mode = RT_BOT_SOLID;
|
||||
bot.orientation = RT_BOT_CCW;
|
||||
bot.num_vertices = RING_SIDES + 2;
|
||||
bot.vertices = bot_vertices;
|
||||
bot.num_faces = RING_SIDES * 2;
|
||||
bot.faces = bot_faces;
|
||||
ars.magic = RT_ARS_INTERNAL_MAGIC;
|
||||
ars.ncurves = 3;
|
||||
ars.pts_per_curve = RING_POINTS;
|
||||
ars.curves = curves;
|
||||
|
||||
RT_DB_INTERNAL_INIT(&intern);
|
||||
intern.idb_major_type = DB5_MAJORTYPE_BRLCAD;
|
||||
intern.idb_type = ID_ARS;
|
||||
intern.idb_meth = &OBJ[ID_ARS];
|
||||
intern.idb_ptr = &ars;
|
||||
RT_DB_INTERNAL_INIT(&bot_intern);
|
||||
bot_intern.idb_major_type = DB5_MAJORTYPE_BRLCAD;
|
||||
bot_intern.idb_type = ID_BOT;
|
||||
bot_intern.idb_meth = &OBJ[ID_BOT];
|
||||
bot_intern.idb_ptr = ⊥
|
||||
|
||||
rtip = rt_dirbuild_inmem(NULL, 0, NULL, 0);
|
||||
if (!rtip)
|
||||
return 1;
|
||||
rt_init_resource(resp, 0, rtip);
|
||||
|
||||
st.l.magic = RT_SOLTAB_MAGIC;
|
||||
st.l2.magic = RT_SOLTAB2_MAGIC;
|
||||
st.st_rtip = rtip;
|
||||
st.st_meth = &OBJ[ID_ARS];
|
||||
st.st_id = ID_ARS;
|
||||
if (OBJ[ID_ARS].ft_prep(&st, &intern, rtip)) {
|
||||
bu_log("ARS prep failed\n");
|
||||
return 1;
|
||||
}
|
||||
|
||||
bot_st.l.magic = RT_SOLTAB_MAGIC;
|
||||
bot_st.l2.magic = RT_SOLTAB2_MAGIC;
|
||||
bot_st.st_rtip = rtip;
|
||||
bot_st.st_meth = &OBJ[ID_BOT];
|
||||
bot_st.st_id = ID_BOT;
|
||||
if (OBJ[ID_BOT].ft_prep(&bot_st, &bot_intern, rtip)) {
|
||||
bu_log("BoT prep failed\n");
|
||||
return 1;
|
||||
}
|
||||
|
||||
RT_APPLICATION_INIT(&ap);
|
||||
ap.a_rt_i = rtip;
|
||||
ap.a_resource = resp;
|
||||
|
||||
ray.magic = RT_RAY_MAGIC;
|
||||
VSET(ray.r_pt, 1, 0, 4);
|
||||
VSET(ray.r_dir, 1, 0, 0);
|
||||
ray.r_min = -0.05; /* the current spatial cell starts inside the mesh */
|
||||
failed += check_shot(&OBJ[ID_ARS], &st, &ap, &ray, "ARS cell bound");
|
||||
failed += check_shot(&OBJ[ID_BOT], &bot_st, &ap, &ray, "BoT cell bound");
|
||||
|
||||
VSET(invdir, 1, INFINITY, INFINITY);
|
||||
if (!rt_in_rpp(&ray, invdir, st.st_min, st.st_max)) {
|
||||
bu_log("ARS bounding box missed the test ray\n");
|
||||
failed++;
|
||||
} else {
|
||||
failed += check_shot(&OBJ[ID_ARS], &st, &ap, &ray, "ARS solid bound");
|
||||
}
|
||||
|
||||
ray.r_min = -0.05;
|
||||
if (!rt_in_rpp(&ray, invdir, bot_st.st_min, bot_st.st_max)) {
|
||||
bu_log("BoT bounding box missed the test ray\n");
|
||||
failed++;
|
||||
} else {
|
||||
failed += check_shot(&OBJ[ID_BOT], &bot_st, &ap, &ray, "BoT solid bound");
|
||||
}
|
||||
|
||||
OBJ[ID_ARS].ft_free(&st);
|
||||
OBJ[ID_BOT].ft_free(&bot_st);
|
||||
rt_free_rti(rtip);
|
||||
|
||||
if (argc != 1 && argc != 3) {
|
||||
bu_log("Usage: %s [ars_scene.g bot_scene.g]\n", argv[0]);
|
||||
return 1;
|
||||
}
|
||||
{
|
||||
const fastf_t *scene_curves[3] = {curve_data[0], curve_data[1], curve_data[2]};
|
||||
failed += check_scene(ID_ARS, scene_curves, bot_vertices, bot_faces,
|
||||
argc == 3 ? argv[1] : NULL);
|
||||
failed += check_scene(ID_BOT, scene_curves, bot_vertices, bot_faces,
|
||||
argc == 3 ? argv[2] : NULL);
|
||||
}
|
||||
return failed ? 1 : 0;
|
||||
}
|
||||
Reference in New Issue
Block a user