[Rewriter] Add optimizer to fold Pad operators into Conv

[Johansmm]

Following (#2301), fuse_pad_into_conv rule set is introduced to reduce the following list of operators:

• Conv ∘ Pad -> Conv
• ConvInteger ∘ Pad -> ConvInteger

Additionally, NormalizePadFormat is introduced in order to change auto_pads Conv attribute in its explicit pads list (ref: https://onnx.ai/onnx/operators/onnx__Conv.html).

[codecov]

Codecov Report

❌ Patch coverage is 95.42254% with 13 lines in your changes missing coverage. Please review.
✅ Project coverage is 70.11%. Comparing base (38c4468) to head (8fc4326).

Files with missing lines	Patch %	Lines
onnxscript/rewriter/fuse_pad_into_conv_test.py	94.44%	4 Missing and 4 partials ⚠️
onnxscript/rewriter/fuse_pad_into_conv.py	96.42%	3 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2363      +/-   ##
==========================================
+ Coverage   69.81%   70.11%   +0.30%     
==========================================
  Files         209      211       +2     
  Lines       25323    25590     +267     
  Branches     2529     2565      +36     
==========================================
+ Hits        17678    17943     +265     
+ Misses       6770     6766       -4     
- Partials      875      881       +6     

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

[Johansmm]

Push force rebasing on main and fixing conflicts.

[Johansmm]

@justinchuby I forgot to explain in the previous message that the changes were not ready yet (just fixing rebase with main).
In the last commits I update the code with all the suggestions.
Now this work is ready to be reviewed.

[justinchuby]

Could you update the PR title and description? Thanks

[Johansmm]

@justinchuby is ok the new description and title ?

[Johansmm]

Last push force rebasing on main and addig @justinchuby suggestions

[Johansmm]

Last push force rebasing on upstream/main and with @gramalingam suggestions

[Copilot]

Pull Request Overview

This PR introduces a new optimization to fuse Pad operators into subsequent Conv operators to reduce the number of operations in ONNX models. It adds pattern matching rules to convert sequences like Conv(Pad(x)) and ConvInteger(Pad(x)) into optimized Conv(x) and ConvInteger(x) operations respectively.

Key changes include:

• Implementation of pattern matching and rewriting logic for Pad-Conv fusion
• Addition of padding format normalization to convert auto_pad attributes to explicit padding lists
• Integration of the new rule set into the default optimization pipeline

Reviewed Changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 3 comments.

File	Description
onnxscript/rewriter/fuse_pad_into_conv.py	Core implementation with pattern matching classes and rewrite logic for fusing Pad into Conv operations
onnxscript/rewriter/fuse_pad_into_conv_test.py	Comprehensive test suite covering fusion scenarios, edge cases, and normalization functionality
onnxscript/rewriter/init.py	Integration of the new rule set into the default optimization rules

[justinchuby]

What happens when input_shape/output_shape is None? Do we need to handle it?