Multi-Index Search & Associated Library Drop-down in PyAnsys

[Revathyvenugopal162]

After migrating to the new [Fuse.js] static search, searching across multiple search.json files is no longer supported. Some developers have requested this functionality, particularly for projects with multiple libraries.

Proposed Solutions

Selection of the index when searching

• Introduce a dropdown near the search button that allows users to choose the desired index when performing a search.
• The dropdown will provide filtering options based on the associated libraries specified in conf.py.
• This ensures users can quickly refine searches within a specific library or dataset, improving usability and navigation.
• associated issue Restore multi-index search capabilities #573

example:

Caution

Styles may change

Dropdown for associated libraries

• Add dropdown buttons in the package home section of the PyAnsys documentation to associate different libraries.
• The associated libraries can be defined in conf.py, ensuring easy configuration.
• The search experience remains distinct for each library without additional complexity in maintaining search result weights.
• associated issue Dropdown buttons for associated libs in PyAnsys  #500

pinging @ansys/pyansys-core team for visibility and insights

[germa89]

I do not think both approaches are in conflict... aren't they?

From the perspective of two tighly integrated libraries (like PyAEDT and EDB API), I guess the first option (search dropdown) makes more sense.

From the perspective of two related libraries (for instance, PyDPF-Core and PyDPF-Post) it makes more sense the second approach (home dropdown).

I would consider having both, starting with the easiest one to have (maybe the home dropdown?). If I have to choose only one... I would go for the search dropdown as it adds "more" to the theme.

PyMAPDL and PyMAPDL-Reader could be a duo that falls within the second approach... too bad the reader is going to be removed from PyMAPDL (eventually).

[RobPasMue]

I agree with @germa89, they are not mutually exclusive - I would prioritize the first approach though, because that is what provides the actual searching capabilities in different datasets... which is what people where interested in.

For the second approach - people could do it themselves if they really wanted. But I agree that it is better if we provide it to them as part of the theme. However, this is not on the critical path. It's a nice-to-have.

[SMoraisAnsys]

I also agree with @germa89

[RobPasMue]

We all follow you @germa89

[germa89]

For your own sake... please do not