blender
/
blender
130
398
Fork 572
Code Issues 6.2k Pull Requests 747 Projects 19 Wiki Activity

Refactor: fcurve bounds functions #105177

Merged
Christoph Lendenfeld merged 16 commits from ChrisLend/blender:refactor_fcurve_bounds into main 2023-03-10 11:33:21 +01:00
Conversation 0 Commits 16 Files Changed 9 +394 -237
Christoph Lendenfeld commented 2023-02-24 15:49:45 +01:00
Member
Copy Link

Refactor the following functions and cover with tests

  • BKE_fcurve_calc_bounds - used to get the rctf bounding box of an fcurve
  • BKE_fcurve_calc_range - used to get only the x-range, potentially faster when not needing y extents

get_fcurve_end_keyframes has been replaced with get_bounding_bezt_indices - dealing with indices allows to iterate over that range later

BKE_fcurve_calc_bounds

  • pass in an rctf instead of float pointers
  • extract logic to get bounds into separate functions

BKE_fcurve_calc_range

  • removed the parameter do_min_length it was always false, and this function shouldn't concern itself with clamping.
    Calling code can do that if the return bool is false
  • use function to get x bounds separated from BKE_fcurve_calc_bounds
Refactor the following functions and cover with tests * `BKE_fcurve_calc_bounds` - used to get the `rctf` bounding box of an fcurve * `BKE_fcurve_calc_range` - used to get only the x-range, potentially faster when not needing y extents `get_fcurve_end_keyframes` has been replaced with `get_bounding_bezt_indices` - dealing with indices allows to iterate over that range later ### BKE_fcurve_calc_bounds * pass in an `rctf` instead of float pointers * extract logic to get bounds into separate functions ### BKE_fcurve_calc_range * removed the parameter `do_min_length` it was always false, and this function shouldn't concern itself with clamping. Calling code can do that if the return bool is false * use function to get x bounds separated from `BKE_fcurve_calc_bounds`
Christoph Lendenfeld added 3 commits 2023-02-24 15:49:56 +01:00
Christoph Lendenfeld added the
Module
Animation & Rigging
 label 2023-02-24 15:50:07 +01:00
Christoph Lendenfeld requested review from Sybren A. Stüvel 2023-02-24 15:50:13 +01:00
Christoph Lendenfeld added 1 commit 2023-02-24 15:56:52 +01:00
a2b16f370c remove warnings 
shouldn't be part of this patch
Sybren A. Stüvel requested changes 2023-02-27 15:57:29 +01:00
Sybren A. Stüvel left a comment
Member
Copy Link

Nice work, thanks!
Just a few inline comments.

Nice work, thanks! Just a few inline comments.
source/blender/blenkernel/BKE_fcurve.h
@ -368,3 +368,3 @@
/**
* Calculate the extents of F-Curve's keyframes.
* Calculate the x range of the given F-Curve's data.
*/
Sybren A. Stüvel commented 2023-02-27 15:07:47 +01:00
Member
Copy Link

Document what the return value means.

Document what the return value means.
ChrisLend marked this conversation as resolved
source/blender/blenkernel/BKE_fcurve.h
@ -370,3 +370,2 @@
*/
bool BKE_fcurve_calc_range(
const struct FCurve *fcu, float *min, float *max, bool do_sel_only, bool do_min_length);
bool BKE_fcurve_calc_range(const struct FCurve *fcu, float *min, float *max, bool do_sel_only);
Sybren A. Stüvel commented 2023-02-27 15:07:22 +01:00
Member
Copy Link

Return parameters should be prefixed with r_. I think this is an excellent opportunity to tackle that too.

And it's weird, French is not my favourite language, I'm not even that good at it. But still when I read do_sel_only I'm thinking "only salt?". No strong opinion here, but at the risk of doing even more in the same patch, I wouldn't mind if that parameter were renamed to use_selected_keys_only. I'll leave the final decision on whether to include that change or not to you though.

Return parameters should be prefixed with `r_`. I think this is an excellent opportunity to tackle that too. And it's weird, French is not my favourite language, I'm not even that good at it. But still when I read `do_sel_only` I'm thinking "only salt?". No strong opinion here, but at the risk of doing even more in the same patch, I wouldn't mind if that parameter were renamed to `use_selected_keys_only`. I'll leave the final decision on whether to include that change or not to you though.
ChrisLend marked this conversation as resolved
source/blender/blenkernel/intern/fcurve.c
@ -640,2 +632,2 @@
const bool use_range = range != NULL;
*r_min = bezt_array[index_range[0]].vec[1][0];
*r_max = bezt_array[index_range[0]].vec[1][0];
Sybren A. Stüvel commented 2023-02-27 15:18:12 +01:00
Member
Copy Link

I think this should be using index_range[1].

I think this should be using `index_range[1]`.
ChrisLend marked this conversation as resolved
source/blender/blenkernel/intern/fcurve.c
@ -663,0 +635,4 @@
if (include_handles) {
/* Need to check all handles because they might extend beyond their neighboring keys. */
for (int i = index_range[0]; i <= index_range[1]; i++) {
BezTriple bezt = bezt_array[i];
Sybren A. Stüvel commented 2023-02-27 15:26:12 +01:00
Member
Copy Link

bezt can be const, same in the function below.

`bezt` can be `const`, same in the function below.
Hans Goudey commented 2023-02-27 16:22:37 +01:00
Member
Copy Link

Better to not make a copy at all here. BezTriple is a rather large struct. In C the best option is const BezTriple *bezt = &bezt_array[i]; The compiler probably optimizes that but it's a good way to show intentions too.

Better to not make a copy at all here. `BezTriple` is a rather large struct. In C the best option is `const BezTriple *bezt = &bezt_array[i];` The compiler probably optimizes that but it's a good way to show intentions too.
Christoph Lendenfeld commented 2023-03-02 12:51:28 +01:00
Author
Member
Copy Link

I've made it a Beztriple * now

I've made it a `Beztriple *` now
dr.sybren marked this conversation as resolved
source/blender/blenkernel/intern/fcurve.c
@ -663,0 +640,4 @@
*r_max = max_fff(*r_max, bezt.vec[1][0], bezt.vec[2][0]);
}
}
else {
Sybren A. Stüvel commented 2023-02-27 15:18:39 +01:00
Member
Copy Link

If I'm right about the initial assignment to *r_max using index_range[1], this entire else block can be removed.

If I'm right about the initial assignment to `*r_max` using `index_range[1]`, this entire `else` block can be removed.
ChrisLend marked this conversation as resolved
source/blender/blenkernel/intern/fcurve.c
@ -717,0 +669,4 @@
if (include_handles) {
/* Left handle - only if applicable.
* NOTE: for the very first keyframe,
* the left handle actually has no bearings on anything. */
Sybren A. Stüvel commented 2023-02-27 15:33:06 +01:00
Member
Copy Link

This is a tricky one. If the idea is to zoom the FCurve so that all the keys and handles are on screen, it doesn't matter that the left handle has no meaning; an animator may still want to interact with it in order to edit the animation. And with, for example, "Aligned" handle type, manipulating the left handle will move the right one, and thus still has an effect.

This is a tricky one. If the idea is to zoom the FCurve so that all the keys and handles are on screen, it doesn't matter that the left handle has no meaning; an animator may still want to interact with it in order to edit the animation. And with, for example, "Aligned" handle type, manipulating the left handle will move the right one, and thus still has an effect.
Christoph Lendenfeld commented 2023-03-02 13:03:14 +01:00
Author
Member
Copy Link

I was hesitant to change that because I wasn't quite sure of the purpose
I removed that logic completely now. IMO this function shouldn't concern itself with the visibility/functionality of handles.
Controversial though :)

I was hesitant to change that because I wasn't quite sure of the purpose I removed that logic completely now. IMO this function shouldn't concern itself with the visibility/functionality of handles. Controversial though :)
dr.sybren marked this conversation as resolved
source/blender/blenkernel/intern/fcurve.c
@ -734,3 +697,1 @@
}
if (ymax) {
*ymax = ymaxv;
if (fcu->bezt) {
Sybren A. Stüvel commented 2023-02-27 15:35:55 +01:00
Member
Copy Link

I think this nesting of conditionals is too deep. Better to split up the function into one for Bezier curves and one for baked points.

I think this nesting of conditionals is too deep. Better to split up the function into one for Bezier curves and one for baked points.
ChrisLend marked this conversation as resolved
source/blender/blenkernel/intern/fcurve.c
@ -746,0 +711,4 @@
if (frame_range != NULL) {
bool found_indices = false;
for (int i = 0; i < fcu->totvert; i++) {
if (fcu->fpt[i].vec[0] < frame_range[0] || fcu->fpt[i].vec[0] > frame_range[1]) {
Sybren A. Stüvel commented 2023-02-27 15:39:28 +01:00
Member
Copy Link

Since there are no handles, you can use the ordering of the points and do a binary search for the start of the range. You can then immediately handle the xmin, and when the last point in the range is found, update the xmax. Per point you then only need to consider the y-coordinates, cutting the number of operations in half.

Since there are no handles, you can use the ordering of the points and do a binary search for the start of the range. You can then immediately handle the `xmin`, and when the last point in the range is found, update the `xmax`. Per point you then only need to consider the y-coordinates, cutting the number of operations in half.
Christoph Lendenfeld commented 2023-03-02 12:55:15 +01:00
Author
Member
Copy Link

instead of doing a binary search i calculate the index based on the assumption that a baked fcurve has a sample on every full frame.

instead of doing a binary search i calculate the index based on the assumption that a baked fcurve has a sample on every full frame.
dr.sybren marked this conversation as resolved
source/blender/blenkernel/intern/fcurve.c
@ -746,0 +714,4 @@
if (fcu->fpt[i].vec[0] < frame_range[0] || fcu->fpt[i].vec[0] > frame_range[1]) {
continue;
}
if (!found_indices) {
Sybren A. Stüvel commented 2023-02-27 15:41:49 +01:00
Member
Copy Link

Initialize r_bounds to ±∞ (so all minimum values to ∞ and all maximum values to -∞). Those are valid IEEE 754 floating point values, and work well with the min/max functions. That'll remove a boolean, a condition, and half of the lines of code.

Initialize `r_bounds` to `±∞` (so all minimum values to `∞` and all maximum values to `-∞`). Those are valid IEEE 754 floating point values, and work well with the min/max functions. That'll remove a boolean, a condition, and half of the lines of code.
ChrisLend marked this conversation as resolved
source/blender/blenkernel/intern/fcurve.c
@ -746,0 +728,4 @@
r_bounds->ymax = max_ff(r_bounds->ymax, fcu->fpt[i].vec[1]);
}
}
return found_indices;
Sybren A. Stüvel commented 2023-02-27 15:42:42 +01:00
Member
Copy Link

Here you can return r_bounds->xmin <= r_bounds->xmax.

Here you can return `r_bounds->xmin <= r_bounds->xmax`.
ChrisLend marked this conversation as resolved
source/blender/blenkernel/intern/fcurve.c
@ -747,2 +733,2 @@
if (xmax) {
*xmax = use_range ? range[1] : 1.0f;
else {
/* X range can be directly calculated from end verts. */
Sybren A. Stüvel commented 2023-02-27 15:46:13 +01:00
Member
Copy Link

A different structure of the function could work well here:

  1. Find the start/end index (either 0 resp. totvert - 1, or the result of a binary search)
  2. Loop over the points in the index range to find the min/max y.
A different structure of the function could work well here: 1. Find the start/end index (either 0 resp. `totvert - 1`, or the result of a binary search) 2. Loop over the points in the index range to find the min/max y.
Christoph Lendenfeld commented 2023-03-02 12:56:10 +01:00
Author
Member
Copy Link

can you clarify that, is that in regards to finding the start index when using a frame range?

can you clarify that, is that in regards to finding the start index when using a frame range?
Sybren A. Stüvel commented 2023-03-03 14:16:58 +01:00
Member
Copy Link

I was thinking about code like this:

static bool calculate_fpt_bounds(const FCurve *fcu, const float frame_range[2], rctf *r_bounds)
{
  r_bounds->xmin = INFINITY;
  r_bounds->xmax = -INFINITY;
  r_bounds->ymin = INFINITY;
  r_bounds->ymax = -INFINITY;

  const int first_index = 0;
  const int last_index = fcu->totvert - 1;
  int start_index = first_index;
  int end_index = last_index;

  if (frame_range != NULL) {
    /* Start index can be calculated because fpt has a key on every full frame. */
    const float start_index_f = frame_range[0] - fcu->fpt[0].vec[0];
    const float end_index_f = start_index_f + frame_range[1] - frame_range[0];

    if (start_index_f > last_index || end_index_f < first_index) {
      /* Range is outside of keyframe samples. */
      return false;
    }

    /* Range might be partially covering keyframe samples. */
    start_index = clamp_i(start_index_f, first_index, last_index);
    end_index = clamp_i(end_index_f, first_index, last_index);
  }

  /* X range can be directly calculated from end verts. */
  r_bounds->xmin = fcu->fpt[start_index].vec[0];
  r_bounds->xmax = fcu->fpt[end_index].vec[0];

  for (int i = start_index; i <= end_index; i++) {
    r_bounds->ymin = min_ff(r_bounds->ymin, fcu->fpt[i].vec[1]);
    r_bounds->ymax = max_ff(r_bounds->ymax, fcu->fpt[i].vec[1]);
  }

  return BLI_rctf_is_valid(r_bounds);
}

This uses INFINITY instead of FLT_MAX so that valid and invalid values can always be distinguished. And it separates "determine the start/endices" from "finding the bounds".

I'm not 100% about the name of start_index_f, maybe range_start_index would be better.

Update: using BLI_rctf_is_valid() instead of writing all of that out.

I was thinking about code like this: ```c static bool calculate_fpt_bounds(const FCurve *fcu, const float frame_range[2], rctf *r_bounds) { r_bounds->xmin = INFINITY; r_bounds->xmax = -INFINITY; r_bounds->ymin = INFINITY; r_bounds->ymax = -INFINITY; const int first_index = 0; const int last_index = fcu->totvert - 1; int start_index = first_index; int end_index = last_index; if (frame_range != NULL) { /* Start index can be calculated because fpt has a key on every full frame. */ const float start_index_f = frame_range[0] - fcu->fpt[0].vec[0]; const float end_index_f = start_index_f + frame_range[1] - frame_range[0]; if (start_index_f > last_index || end_index_f < first_index) { /* Range is outside of keyframe samples. */ return false; } /* Range might be partially covering keyframe samples. */ start_index = clamp_i(start_index_f, first_index, last_index); end_index = clamp_i(end_index_f, first_index, last_index); } /* X range can be directly calculated from end verts. */ r_bounds->xmin = fcu->fpt[start_index].vec[0]; r_bounds->xmax = fcu->fpt[end_index].vec[0]; for (int i = start_index; i <= end_index; i++) { r_bounds->ymin = min_ff(r_bounds->ymin, fcu->fpt[i].vec[1]); r_bounds->ymax = max_ff(r_bounds->ymax, fcu->fpt[i].vec[1]); } return BLI_rctf_is_valid(r_bounds); } ``` This uses `INFINITY` instead of `FLT_MAX` so that valid and invalid values can always be distinguished. And it separates "determine the start/endices" from "finding the bounds". I'm not 100% about the name of `start_index_f`, maybe `range_start_index` would be better. **Update:** using `BLI_rctf_is_valid()` instead of writing all of that out.
Christoph Lendenfeld commented 2023-03-09 12:21:21 +01:00
Author
Member
Copy Link

thanks got it
and changed to that code

thanks got it and changed to that code
source/blender/blenkernel/intern/fcurve_test.cc
@ -349,0 +367,4 @@
/* All keys. */
success = BKE_fcurve_calc_range(fcu, &min, &max, false);
EXPECT_TRUE(success);
Sybren A. Stüvel commented 2023-02-27 15:54:35 +01:00
Member
Copy Link

Avoid plain EXPECT_TRUE/EXPECT_FALSE. You can add an explanation with stream operators, like this:

EXPECT_TRUE(success) << "A non-empty FCurve should have a range";
Avoid plain `EXPECT_TRUE`/`EXPECT_FALSE`. You can add an explanation with stream operators, like this: ```cpp EXPECT_TRUE(success) << "A non-empty FCurve should have a range"; ```
ChrisLend marked this conversation as resolved
Christoph Lendenfeld added 1 commit 2023-03-02 12:17:27 +01:00
Christoph Lendenfeld added 3 commits 2023-03-02 12:50:46 +01:00
Christoph Lendenfeld added 1 commit 2023-03-02 13:01:12 +01:00
Christoph Lendenfeld requested review from Sybren A. Stüvel 2023-03-02 13:03:30 +01:00
Christoph Lendenfeld added 2 commits 2023-03-03 10:43:02 +01:00
Christoph Lendenfeld added 2 commits 2023-03-09 12:07:26 +01:00
Christoph Lendenfeld added 1 commit 2023-03-09 12:20:41 +01:00
Sybren A. Stüvel approved these changes 2023-03-09 15:23:28 +01:00
Sybren A. Stüvel left a comment
Member
Copy Link

Just one thing that Clang-Tidy will trip over, rest LGTM! You can resolve this before merging the PR.

Just one thing that Clang-Tidy will trip over, rest LGTM! You can resolve this before merging the PR.
source/blender/blenkernel/intern/fcurve.c
@ -774,3 +749,2 @@
if (bezt_first) {
BLI_assert(bezt_last != NULL);
else if (fcu->fpt) {
Sybren A. Stüvel commented 2023-03-09 15:22:19 +01:00
Member
Copy Link

No else after return; clang-tidy will complain about this.

No `else` after `return`; `clang-tidy` will complain about this.
Christoph Lendenfeld added 2 commits 2023-03-10 11:30:09 +01:00
Christoph Lendenfeld merged commit 2e03352492 into main 2023-03-10 11:33:21 +01:00
Christoph Lendenfeld deleted branch refactor_fcurve_bounds 2023-03-10 11:33:22 +01:00
Sign in to join this conversation.
Reviewers
No reviewers
Sybren A. Stüvel
Labels
Clear labels   
Interest
Alembic

  
Interest
Animation & Rigging

  
Interest
Asset Browser

  
Interest
Asset Browser Project Overview

  
Interest
Audio

  
Interest
Automated Testing

  
Interest
Blender Asset Bundle

  
Interest
BlendFile

  
Interest
Collada

  
Interest
Compatibility
This issue affects/is about backward or forward compatibility

  
Interest
Compositing

  
Interest
Core

  
Interest
Cycles

  
Interest
Dependency Graph

  
Interest
Development Management

  
Interest
EEVEE

  
Interest
EEVEE & Viewport

  
Interest
Freestyle

  
Interest
Geometry Nodes

  
Interest
Grease Pencil

  
Interest
ID Management

  
Interest
Images & Movies

  
Interest
Import Export

  
Interest
Line Art

  
Interest
Masking

  
Interest
Metal

  
Interest
Modeling

  
Interest
Modifiers

  
Interest
Motion Tracking

  
Interest
Nodes & Physics

  
Interest
OpenGL

  
Interest
Overlay

  
Interest
Overrides

  
Interest
Performance

  
Interest
Physics

  
Interest
Pipeline, Assets & IO

  
Interest
Platforms, Builds & Tests

  
Interest
Python API

  
Interest
Render & Cycles

  
Interest
Render Pipeline

  
Interest
Sculpt, Paint & Texture

  
Interest
Text Editor

  
Interest
Translations

  
Interest
Triaging

  
Interest
Undo

  
Interest
USD

  
Interest
User Interface

  
Interest
UV Editing

  
Interest
VFX & Video

  
Interest
Video Sequencer

  
Interest
Virtual Reality

  
Interest
Vulkan

  
Interest
Wayland
Issues relating to Wayland windowing support.

  
Interest
Workbench

  
Interest: X11

Issues relating to Xorg/X11 support.

  
Legacy
Blender 2.8 Project

  
Legacy
Milestone 1: Basic, Local Asset Browser

  
Legacy
OpenGL Error

  
Meta
Good First Issue

  
Meta
Papercut

  
Meta
Retrospective

  
Meta
Security
Issues relating to security: https://wiki.blender.org/wiki/Process/Vulnerability_Reports

  
Module
Animation & Rigging

  
Module
Core

  
Module
Development Management

  
Module
EEVEE & Viewport

  
Module
Grease Pencil

  
Module
Modeling

  
Module
Nodes & Physics

  
Module
Pipeline, Assets & IO

  
Module
Platforms, Builds & Tests

  
Module
Python API

  
Module
Render & Cycles

  
Module
Sculpt, Paint & Texture

  
Module
Triaging

  
Module
User Interface

  
Module
VFX & Video

  
Platform
FreeBSD

  
Platform
Linux

  
Platform
macOS

  
Platform
Windows

  
Priority
High

  
Priority
Low

  
Priority
Normal

  
Priority
Unbreak Now!

  
Status
Archived

  
Status
Confirmed

  
Status
Duplicate

  
Status
Needs Info from Developers

  
Status
Needs Information from User

  
Status
Needs Triage

  
Status
Resolved

  
Type
Bug

  
Type
Design

  
Type
Known Issue

  
Type
Patch

  
Type
Report

  
Type
To Do

No Label
Interest
Alembic
Interest
Animation & Rigging
Interest
Asset Browser
Interest
Asset Browser Project Overview
Interest
Audio
Interest
Automated Testing
Interest
Blender Asset Bundle
Interest
BlendFile
Interest
Collada
Interest
Compatibility
Interest
Compositing
Interest
Core
Interest
Cycles
Interest
Dependency Graph
Interest
Development Management
Interest
EEVEE
Interest
EEVEE & Viewport
Interest
Freestyle
Interest
Geometry Nodes
Interest
Grease Pencil
Interest
ID Management
Interest
Images & Movies
Interest
Import Export
Interest
Line Art
Interest
Masking
Interest
Metal
Interest
Modeling
Interest
Modifiers
Interest
Motion Tracking
Interest
Nodes & Physics
Interest
OpenGL
Interest
Overlay
Interest
Overrides
Interest
Performance
Interest
Physics
Interest
Pipeline, Assets & IO
Interest
Platforms, Builds & Tests
Interest
Python API
Interest
Render & Cycles
Interest
Render Pipeline
Interest
Sculpt, Paint & Texture
Interest
Text Editor
Interest
Translations
Interest
Triaging
Interest
Undo
Interest
USD
Interest
User Interface
Interest
UV Editing
Interest
VFX & Video
Interest
Video Sequencer
Interest
Virtual Reality
Interest
Vulkan
Interest
Wayland
Interest
Workbench
Interest: X11
Legacy
Blender 2.8 Project
Legacy
Milestone 1: Basic, Local Asset Browser
Legacy
OpenGL Error
Meta
Good First Issue
Meta
Papercut
Meta
Retrospective
Meta
Security
Module
Animation & Rigging
Module
Core
Module
Development Management
Module
EEVEE & Viewport
Module
Grease Pencil
Module
Modeling
Module
Nodes & Physics
Module
Pipeline, Assets & IO
Module
Platforms, Builds & Tests
Module
Python API
Module
Render & Cycles
Module
Sculpt, Paint & Texture
Module
Triaging
Module
User Interface
Module
VFX & Video
Platform
FreeBSD
Platform
Linux
Platform
macOS
Platform
Windows
Priority
High
Priority
Low
Priority
Normal
Priority
Unbreak Now!
Status
Archived
Status
Confirmed
Status
Duplicate
Status
Needs Info from Developers
Status
Needs Information from User
Status
Needs Triage
Status
Resolved
Type
Bug
Type
Design
Type
Known Issue
Type
Patch
Type
Report
Type
To Do
Milestone
Clear milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
No Assignees
3 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: blender/blender#105177
No description provided.
Delete Branch "ChrisLend/blender:refactor_fcurve_bounds"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?