search

vermaseren / form

The FORM project for symbolic manipulation of very big expressions
GNU General Public License v3.0
1.15k stars 136 forks source link

Change EndSort type for transform statements #516

Closed jodavies closed 4 months ago

jodavies commented 4 months ago

Since the transform statements are sorting into the workspace, and might have to merge patches, make them use a par=1 EndSort. This commit makes this change for all transform statements, though only mularg is going to produce an explosion of terms here.

This fixes #230 and causes #183 to crash with a useful error (Term too complex during normalization).

jodavies commented 4 months ago

@vermaseren this one needs a sanity check, I am not 100% sure of the consequences of this change.

vermaseren commented 4 months ago

I think you are right. MulArg was added later. Before this was not a problem, and apparently when MulArg was added this problem was not considered. And of course there may be future options.

On 10 May 2024, at 11:22, jodavies @.***> wrote:

@vermaseren https://github.com/vermaseren this one needs a sanity check, I am not 100% sure of the consequences of this change.

— Reply to this email directly, view it on GitHub https://github.com/vermaseren/form/pull/516#issuecomment-2104263217, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABJPCERENQNUWK4LEOSMCVDZBSGUVAVCNFSM6AAAAABHQJJW5CVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCMBUGI3DGMRRG4. You are receiving this because you were mentioned.

coveralls commented 4 months ago

Coverage Status

coverage: 48.748% (+0.04%) from 48.712% when pulling 7831fd9e78b0b4c808afa8aad76f9d11955e76bc on jodavies:issue-230 into 0112aa6b21b48616575c6cd205ec00c7797dbc93 on vermaseren:master.

jodavies commented 4 months ago

It is probably worth mentioning that mularg can be slow compared to repeated id statements (which are sorting the arguments each repeat).

tueda commented 4 months ago

This is ready to merge, right?