Hi Akilan, I see how this can be useful. The problem is how would we integrate this with a database: 1. Read all the records from the db to memory and them apply the search. 2. Stream all records from the db to the search library. Both of these solutions require reading the complete table. This might work for some applications with hundreds of records, but it would slow things down when you start dealing with any significant amount of data. DB engines have perfected searching and record retrieval. Getting better results outside of that is engine is possible but costly. So what can we do? My original approach for fuzzy/smart search was to generate a bunch of alternative words and use the database engine to search for these terms. The sources of these "extra terms" are: - spelling correction (AI) - corpus built periodically from the database - using the custom corpus to find similar words in terms of some custom distance function based of standard spell checker algorithm and thesaurus for finding similar meaning. I know it's a huge undertaking, but it will keep uAdmin as fast as the database, which is one of my main goals.

…

On Tue, May 23, 2023, 5:04 PM Akilan Selvacoumar ***@***.***> wrote: Proposal The issue address implementing text indexing for uadmin. Implementation which could be used: (https://github.com/blevesearch/bleve) Interesting issue: blevesearch/bleve#1794 <blevesearch/bleve#1794> — Reply to this email directly, view it on GitHub <#139>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AA4AZ3TE7VMGUDXN76WJ2C3XHSYWZANCNFSM6AAAAAAYL3X4MU> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

Improved search uadmin #139

Description

Proposal

Metadata

Metadata

Assignees

Labels

Projects

Milestone

Relationships

Development

Issue actions