Add support for colorized rooms

[shmuelzon]

This PR adds support for coloring rooms/areas in the SVG map and it's only available for robots that support MapInfo.

Most of the elements remain the same as before. The main change is with the Room items which now need to be drawn twice. Once with the color of the room and another time with the striped pattern overlayed on top of it. I couldn't find a way to have both a background color and an overlay on the same path. The new styles and striped line definitions are only added when needed to keep the SVG small.

Examples:

Before	After

Note that the background image-based maps have changed as a result but their content is the same and only the defs element was moved to simplify the code.

[codecov]

Codecov Report

❌ Patch coverage is 99.28058% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 94.54%. Comparing base (ba38147) to head (904c41b).
⚠️ Report is 1 commits behind head on dev.
✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
src/map/style.rs	98.63%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##              dev    #1281      +/-   ##
==========================================
+ Coverage   94.51%   94.54%   +0.02%     
==========================================
  Files         148      148              
  Lines        5707     5775      +68     
  Branches      350      350              
==========================================
+ Hits         5394     5460      +66     
- Misses        251      253       +2     
  Partials       62       62              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[codspeed-hq]

CodSpeed Performance Report

Merging #1281 will not alter performance

_{Comparing shmuelzon:add-support-for-colored-maps (904c41b) with dev (ba38147)}

Summary

✅ 107 untouched

[edenhaus]

Why only 4 rooms? I'm expecting a normal flat has often more rooms

Can we have a different blue? It's nearly the same as the wall color and the other colors look a lot better. Maybe even don't have a blue et all

[shmuelzon]

I took the 4 colors that seem to repeat themselves from the app. The app also has the walls as gray (see #1177 (comment)) so that's another option but then it'll be different than what your app shows.

If we're good with deviating from the app, than I'd suggest adding more colors to reduce the risk of two rooms that have a shared edge having the same color. I couldn't understand yet the logic of the app to decide the color for each room but somehow (might be by chance) that doesn't happen even with "only" 4 colors.

Anyway, let me know what you prefer and I'll do that

[edenhaus]

I don't remember which command returns the connection between rooms, but maybe the app uses it to pick the correct color. Nevertheless, we should add more colors and make it better :)
We can try to go with the gray walls for all and see how it looks :)

Not sure if you want to add also the room colors for the png based map. If yes, you should only need to modify https://github.com/DeebotUniverse/client.py/blob/dev/src/map/background_image.rs#L70-L74

[shmuelzon]

I switched the walls to gray but, if you don't like it, #c6d5eb seems like a good replacement from the RoomColor1's blue. Also added two more colors:

ilt3k8	kr0277

Not sure if you want to add also the room colors for the png based map

I tried adding the following change:

diff --git a/src/map/background_image.rs b/src/map/background_image.rs
index 99a5101..7f29d6d 100644
--- a/src/map/background_image.rs
+++ b/src/map/background_image.rs
@@ -20,8 +20,14 @@ const MAP_IMAGE_PALETTE: &[u8] = &[
     0x1a, 0x81, 0xed, // Carpet
     0xde, 0xe9, 0xfb, // Not scanned space
     0xed, 0xf3, 0xfb, // Possible obstacle
+    0xa2, 0xbc, 0xe7, // Room color 1
+    0xec, 0xd0, 0x99, // Room color 2
+    0x9b, 0xd4, 0xda, // Room color 3
+    0xec, 0xc6, 0xc9, // Room color 4
+    0xd7, 0xbc, 0xe3, // Room color 5
+    0xc3, 0xe2, 0xb6, // Room color 6
 ];
-const MAP_IMAGE_PALETTE_LEN: u8 = 6;
+const MAP_IMAGE_PALETTE_LEN: u8 = 12;
 // 0 -> Transparent, 255 -> Fully opaque
 // (first entry is transparent, rest opaque)
 const MAP_IMAGE_PALETTE_TRANSPARENCY: &[u8] = &[0u8, 255, 255, 255, 255, 255];

But the maps in the test didn't change so I guess they don't have that added information in them (?). Not sure if we should add that without a map that we can see it working on

[edenhaus]

Why are you wrapping everything in a Box? I removed some of them already in your previous PRs

[shmuelzon]

Probably because I don't know any better and rely on compilation errors and LLMs 🤷‍♂️.
The structure of the hash map changed now because of the below request and it swears that it can't be removed anymore.

[edenhaus]

We don't need to have a list if we only have at most one element

[shmuelzon]

I thought there might be a case in the future but I guess we can complicate it in the future, it needed.

[edenhaus]

Thanks @shmuelzon 👍