feat: add C implementation for blas/ext/base/ndarray/sfill - #14422
feat: add C implementation for blas/ext/base/ndarray/sfill#14422AryanSharma48 wants to merge 3 commits into
blas/ext/base/ndarray/sfill#14422Conversation
Signed-off-by: Aryan Sharma <aryansharma24106@gmail.com>
Coverage Report
The above coverage report was generated for the changes in this PR. |
Signed-off-by: Aryan Sharma <aryansharma24106@gmail.com>
|
Hey @0PrashantYadav0 ! I have a question regarding the architectural design for @stdlib/blas/ext/base/ndarray/sfill. Currently, I've implemented it as a pure JavaScript wrapper that delegates directly to the underlying strided C implementation via @stdlib/blas/ext/base/sfill (similar to how sfill-nan and dfill-nan are structured). However, I noticed that some packages like @stdlib/blas/ext/base/ndarray/dsum have their own dedicated C addon (src/addon.c, binding.gyp, etc.) and serialize the metadata directly to C. Should sfill remain a pure JS delegator to @stdlib/blas/ext/base/sfill, or do you prefer that I implement a dedicated C addon and compile configuration for it similar to dsum? |
Signed-off-by: Aryan Sharma <aryansharma24106@gmail.com>
Part of #14404.
Description
This pull request:
@stdlib/blas/ext/base/ndarray/sfill.@stdlib/blas/ext/base/sfill's.ndarraymethod by extracting inputs from anarraysarray-like object.dsumanddaxpbcanonical references.Related Issues
This pull request has the following related issues:
ndarrayinterface variants for missing modules in@stdlib/blas/ext/base#14404Questions
No.
Other
No.
Checklist
AI Assistance
If you answered "yes" above, how did you use AI assistance?
Disclosure
I used AI assistance to scaffold the TS tests, and documentation based on canonical reference packages (
dsum,daxpb), followed by manual review, local tape tests, and lint verification.@stdlib-js/reviewers