ponder-lab / Hybridize-Functions-Refactoring / #2247
80%

DEFAULT BRANCH: main
Ran
Jobs 1
Files 0
Run time –
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
#2247

push

github

web-flow 
Pin the Python list-repetition tensor over-typing (wala/ML#653) (#707)

* Pin the Python list-repetition tensor over-typing (wala/ML#653)

`testListRepetitionTensorOverTyping` guards wala/ML#653: the tensor-type
analysis types a Python list-repetition result (`[0] * n`) as a tensor because
it models the `*` as tensor multiplication. The test asserts the current
(incorrect) behavior (`rep`'s `value`, fed `[0] * 3`, is over-typed) with a
control (`lit`'s `value`, fed the literal `[1, 2, 3]`, is correctly not typed),
to invert once wala/ML#653 is fixed.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FuhM8SsTRB6BraLt3Ph7ZC

* Add a scalar-`*` control to the list-repetition pin

`mul`'s `value` (fed `2 * 3`) confirms scalar multiplication is not over-typed,
isolating the trigger to a list operand of `*`, not `*` or the list alone. Use
direct autoboxing for the `lit`/`mul` controls to match the codebase
convention (`getHasTensorParameter()` is determinate here).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FuhM8SsTRB6BraLt3Ph7ZC

* Use issue URLs in the list-repetition pin comments

Reference https://github.com/wala/ML/issues/653 by URL rather than a bare
`wala/ML#653`, which does not link from a source comment.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FuhM8SsTRB6BraLt3Ph7ZC

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Jobs
ID Job ID Ran Files Coverage
1 #2247.1 32
Source Files on build #2247
Detailed source file information is not available for this build.