Codegen c multi-dimensional locals

[rols121]

Brief description of what is fixed or changed

Sympy did not allow for codegen with multi-dimensional arrays as arguments that would require a local variable copy. This change just enforces that multi-dimensional arrays are converted to a single dimension array with the equivalent amount of elements. The de-referencing of the new local array already happened correctly.

Other comments

This could be improved upon even more by having a flag for "copy arguments", which, if set has the current behavior, and if not then just uses the argument instead of making a copy and then operates on the argument. It may not be the neatest solution but contiguous multi-dim arrays could just be treated as a single dim array if the user is willing to pass the pointer to the very first element. This change just is just a starting point to allow the use of sympy Matrix types with more that one dimension for c code generation.

Release Notes

• codegen
  • allowing for multi-dimensional arrays as arguments/locals in c code generation

[sympy-bot]

✅

Hi, I am the SymPy bot (v161). I'm here to help you write a release notes entry. Please read the guide on how to write release notes.

Your release notes are in good order.

Here is what the release notes will look like:

• codegen
  • allowing for multi-dimensional arrays as arguments/locals in c code generation (#21190 by @rols121)

This will be added to https://github.com/sympy/sympy/wiki/Release-Notes-for-1.8.

Click here to see the pull request description that was parsed.

<!-- Your title above should be a short description of what
was changed. Do not include the issue number in the title. -->

#### Brief description of what is fixed or changed
Sympy did not allow for codegen with multi-dimensional arrays as arguments that would require a local variable copy. This change just enforces that multi-dimensional arrays are converted to a single dimension array with the equivalent amount of elements. The de-referencing of the new local array already happened correctly.


#### Other comments
This could be improved upon even more by having a flag for "copy arguments", which, if set has the current behavior, and if not then just uses the argument instead of making a copy and then operates on the argument. It may not be the neatest solution but contiguous multi-dim arrays could just be treated as a single dim array if the user is willing to pass the pointer to the very first element. This change just is just a starting point to allow the use of sympy Matrix types with more that one dimension for c code generation.

#### Release Notes

<!-- Write the release notes for this release below between the BEGIN and END
statements. The basic format is a bulleted list with the name of the subpackage
and the release note for this PR. For example:

* solvers
  * Added a new solver for logarithmic equations.

* functions
  * Fixed a bug with log of integers.

or if no release note(s) should be included use:

NO ENTRY

See https://github.com/sympy/sympy/wiki/Writing-Release-Notes for more
information on how to write release notes. The bot will check your release
notes automatically to see if they are formatted correctly. -->

<!-- BEGIN RELEASE NOTES -->

* codegen
  * allowing for multi-dimensional arrays as arguments/locals in c code generation

<!-- END RELEASE NOTES -->

Update

The release notes on the wiki have been updated.

[bjodah]

Thank you for your contribution. This will also need a test.

[bjodah]

@gouarin I think you know the utilities.codegen module the best, would you mind taking a quick look here?

[gouarin]

@bjodah thanks. I will review this PR before the end of this week.

[rols121]

Hi @bjodah and @gouarin, is there an update on this PR? Comments are welcome as this is my first contribution, so still learning the ins and outs.

[gouarin]

Hi @rols121, few days ago I added a comment in your test case.
Do you see it ?

[rols121]

Hey @gouarin, no, unfortunately not. Where do I see comments? It still says "no reviews" at the top of this PR.

[gouarin]

Why do you use cse here? The example is more complicated to read, no?

[rols121]

I had to use cse so that the code reached the section where c locals are printed. If I did not use cse, then there where no locals and the "raise CodeGenError("Only column vectors are supported in local variabels. Local result {} has dimensions {}".format(result, dims))" line was not reached.

The test is a bit simple as it does not show cse being used but it does however force an array from a multi-dimensional input.

[gouarin]

of course! Thanks.

[gouarin]

Sorry, my bad.

[bjodah]

Thank you @rols121 for your contribution, and thank you @gouarin for the review!