I have json files having more than 70 complex collection fields. I want to index these Json files using Azure Cognitive Search, but unfortunately the limit for complex collection fields is only 40 for every tier.
I suppose the limit should be increased from 40 to something more appropriate.

Azure Search should support higher numbers of scoring functions for higher pricing tiers, since they have more compute resources available.

Current limit is 8 function maximum regardless of pricing tier.

Accrording to doc Hit highlighting behavior updated on 15 July 2020,
but still I am getting same response as before,
I have recreate whole service on 16th July.

Can not get exact match results in hit highlight

Can we see support for indexing (or crawling) of Visio documents?

In the networking world, these are very important sources of data.

Visio documents are key information sources for design, intent and work planning when it comes to deploying and managing network appliances. It would be invaluable to engineers

All the other limits are per second - why such a low limit? We are indexing many documents at once and keep hitting it.

Azure Cognitive Search should be able to index Azure Files. I have no way to search in Azure Files directly, if I store files in Azure Files instead of Blobs.

DataSource Definition JSON always shows Connectionstring = null (By Design).

The recommendation is that, instead of null, says: "Hidden for security purposes".

When the search service is configured with a private link or in private mode, we are unable to manage the Search Service outside of doing it programmatically or through the API. We would like an option to be able to manage the service when private link is enabled.

Can we get solr pivot or sub aggregation equivalent functionality.

Additional mail formats such as .PST, EDRM, Mbox
AutoCAD
DICOM

We need the ability to set up a scoring profile based on the proximity to a numeric field.

i.e. All of our projects in our search index have a fee amount, we want to be able to score them based on similarly sized projects, i.e. proximity from the fee number, like you can do with geolocation.

Add search/filter text box to left CMDLETS, RUNBOOKS, & ASSETS panel to ease in selecting various items. Currently the full list of CMDLETS is very long to find the one you want.

When using Azure Search entity recognition with larger documents, there is no easier way to mark an offset for each page based on the position in the text. This makes it difficult to join the entities returned afterwards.

OCR Cognitive Skill is described in Docs as being compatible with ja (Japanese), but when actually used, Japanese words are output with each character separated by a space, so the index is properly Cannot be created. This is a fatal problem.

Now I'm creating a custom skill to add the process of removing the read space, but it's not perfect.

In Azure search we need to support "contains".

example:

If I search for apple
then it show be give apple,apple fruit ,green apple ,green apple fruit.

apple

Dear Team,

Please introduce a 'minimum should match' parameter in Azure Search REST API (Lucene docs: https://lucene.apache.org/solr/851/solr-core/org/apache/solr/util/doc-files/min-should-match.html and Elasticsearch docs: https://www.elastic.co/guide/en/elasticsearch/reference/current/query-dsl-minimum-should-match.html#query-dsl-minimum-should-match )

Currently there is no option to add comments to the existing fields in Azure search, can we make this feature available?

It would be great if you support more operators in collection matching lambda expressions.
For example, Members/any(m: m contains 'a').
Since there's no correlation between the query and the filters, this can be very useful.

When a document has a field which is a collection of a complex type, like:

Students-> Collection of complex(age, gender, name)

When we find a match in a name of a given student, the hit highlight returns the name. It would be great to get the entire Student complex type record in the highlight, so we'll have the context of the highlight and not just the name.

i recently tried the phonetic filter in custom analyser. But it wasnt complete. I also needed words which sounded similar but were also spelled wrong