Skip to content

Conversation

@baierd
Copy link
Contributor

@baierd baierd commented Oct 28, 2025

This MR adds CVC5 multi-dimensional array model evaluation based on @daniel-raffler s implementation in commit 0d463c83410a332ed81ff4f7628d25ae729b3782, with minor modifications.

It also adds 2 tests for multi-dimensional arrays not based on SMTLIB2 parsing (we did not have any) as well.

@baierd baierd requested a review from kfriedberger October 28, 2025 15:57
@baierd baierd self-assigned this Oct 28, 2025
@baierd baierd added this to the Release 6.0.0 milestone Oct 28, 2025
…, but throw them until a common method.

This simplifies some nesting in the code.
We should avoid extensive recursion if possible.
…visitor usage.

This solution removes the ability to list UF assignments for bound variables as part of the model.
However, the model API is even documented in such a way that such values could be missing,
and the user should directly evaluate formulas in the model.
kfriedberger
kfriedberger previously approved these changes Nov 14, 2025
Copy link
Member

@kfriedberger kfriedberger left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ok for me

It seems as if Z3 produces n^2 additional variables in the model,
if arrays are represented.
Let's reduce this number and only list the relevant variables.
…g arrays and UFs) towards CVC4.

This change replaces the recursive implementation with faster and simpler visitor usage.
Additionally, we get proper array-handling in the CVC4 model.

This solution removes the ability to list UF assignments for bound variables as part of the model.
However, the model API is even documented in such a way that such values could be missing,
and the user should directly evaluate formulas in the model.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Development

Successfully merging this pull request may close these issues.

2 participants