Add epd7in5_v2

[asaaki]

When I ordered such 7.5" display and tried to use this library I was confused, as nothing worked. After some time I realized the very red note and thought »Oh, I must be a lucky one« since nobody besides Waveshare has support for V2 yet.

After spending hours in the python code and the spec PDF I came up with the following code, which works fine for my display. Not sure if anyone is willing to spend the money to confirm it. 😅

I also got rid of the deprecation warnings from embedded-hal, shouting at me to not use digital::v1 anymore. My workaround might look hacky, but should not be really worse than the logic before, though the let _ = … might be a bit unpleasing, as well as the .unwrap_or(false) — yes, I make probably a stupid assumption here to unlock the busy state on error, we could also negate and potentially lock forever; dunno what's best. Given that we ignored potentially failing state changes, I think it should be mostly fine.

---

Commit asaaki/epd-waveshare@19e3d51 was an attempt to fix the digital::v2 fallability, but I discovered that it would at least not work with linux-embedded-hal due to error type mismatches. And the examples step of CI is also pretty unhappy about it.

[codecov-io]

Codecov Report

Merging #36 into master will increase coverage by 1.28%.
The diff coverage is 43.31%.

@@            Coverage Diff             @@
##           master      #36      +/-   ##
==========================================
+ Coverage   35.74%   37.03%   +1.28%     
==========================================
  Files          18       21       +3     
  Lines         996     1145     +149     
==========================================
+ Hits          356      424      +68     
- Misses        640      721      +81

Impacted Files	Coverage Δ
src/epd1in54b/mod.rs	3.1% <ø> (ø)	⬆️
src/epd2in9/mod.rs	3.41% <ø> (ø)	⬆️
src/traits.rs	0% <ø> (ø)	⬆️
src/epd7in5/mod.rs	4.59% <ø> (ø)	⬆️
src/epd4in2/mod.rs	2.83% <ø> (ø)	⬆️
src/epd1in54/mod.rs	3.3% <ø> (ø)	⬆️
src/interface.rs	0% <0%> (ø)	⬆️
src/epd7in5_v2/graphics.rs	100% <100%> (ø)
src/epd7in5_v2/command.rs	100% <100%> (ø)
src/epd7in5_v2/mod.rs	4.76% <4.76%> (ø)
... and 3 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 636b314...d0e1d55. Read the comment docs.

[caemor]

Commit asaaki/epd-waveshare@19e3d51 was an attempt to fix the digital::v2 fallability, but I discovered that it would at least not work with linux-embedded-hal due to error type mismatches. And the examples step of CI is also pretty unhappy about it.

In my first attempt to update to digital::v2 the result also wasn't what I wanted so I am currently waiting for the digital::v3/other results of the discussion (rust-embedded/wg#393)

[caemor]

I'm a little bit unsure about calling this display Display7in5/EPD7in5 even though they are in the separate v2 module. But since every new newly bought 7in5 Display would be one of v2 it should be the better decision 👍

• Would it make sense to add your update_and_display_frame to the main trait?

[caemor]

Also thanks for your work :-) 👍

[asaaki]

Thanks for the pointer. Seems to be a very messy situation. Will undo all digital::* changes then.

[asaaki]

Would it make sense to add your update_and_display_frame to the main trait?

Could make sense, waveshare's python reference implementation uses a single display function for non-v2 as well, so providing a consolidated function in the main trait could be beneficial for everyone.
Will give it a try.

[caemor]

Will undo all digital::* changes then

I think we can leave your current v2 impl/changes and just keep in mind that we need to improve this situation later on (with v3) for which I made a new issue earlier.

[caemor]

I am gonna merge this for now and make the change to the main trait in a seperate pr

[asaaki]

Sorry, was occupied with other stuff. Thanks for taking it in. 🙇🏻‍♂️