Omega359 commented on issue #9102:
URL: 
https://github.com/apache/arrow-datafusion/issues/9102#issuecomment-1922265580

   > DataFusion also has `Expr::Like` but that implements the `%` SQL type 
matching
   > 
   > Also potentially interesting is that we have a version of this function in 
InfluxDB:
   > 
   > 
https://github.com/influxdata/influxdb/blob/ff567cd33f6f9c42a4e24e9c93d1799d1a8f4461/query_functions/src/regex.rs#L22-L121
   > 
   > (though that one is designed to match the behavior of the Go regexp 
library)
   
   That is definitely interesting. I spent some time today looking into the 
differences between the postgresql, Java and Rust implementations of regex. 
There is as expected a very large amount of overlap but some advanced features 
are only found in one implementation or another. Postgresql has an expanded 
version of one of the posix definitions, Java I think is more based on the Perl 
regex, and the Rust crate specifically calls out that is isn't posix based. 
   
   Essentially, I was thinking of just sticking with what is in use already and 
documenting the syntax via references to the rust crate's documentation and 
noting that anyone expecting to have 100% compatibility with either Java's, 
Perl's, Go's, Posix, etc is bound to be disappointed. Once we have proper 
separation between default, postgres and spark syntaxes then others can have a 
go at specific versions.
   
   Primarily I suspect this might possible impact the comet contribution being 
made - I haven't checked to see if they did any work in this area.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to