You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I've been looking at the possibility of implementing a custom type parser for PostGIS's geometry type, using the wkx package to parse the (E)WKB it serializes as. Currently, I have a query working that gets the type OID from the database, and I have the type parser working correctly using require('pg').types.setTypeParser(); it's returning a Buffer. However, there's a big hurdle to using this in my project (a web frontend for postgres, web-pgq) - since the OIDs for types introduced by extensions may change between databases (or even on the same database over time), I can't connect to 2 databases and have the type parser work as expected on both.
I can partially work around this by registering and unregistering the type parser before and after each query, but then we get into race conditions if 2 connections are being queried simultaneously... so that's not really a tenable solution.
Could support be added for setting custom type parsers on a per-connection basis?
The text was updated successfully, but these errors were encountered:
I've been looking at the possibility of implementing a custom type parser for PostGIS's geometry type, using the wkx package to parse the (E)WKB it serializes as. Currently, I have a query working that gets the type OID from the database, and I have the type parser working correctly using
require('pg').types.setTypeParser()
; it's returning a Buffer. However, there's a big hurdle to using this in my project (a web frontend for postgres, web-pgq) - since the OIDs for types introduced by extensions may change between databases (or even on the same database over time), I can't connect to 2 databases and have the type parser work as expected on both.I can partially work around this by registering and unregistering the type parser before and after each query, but then we get into race conditions if 2 connections are being queried simultaneously... so that's not really a tenable solution.
Could support be added for setting custom type parsers on a per-connection basis?
The text was updated successfully, but these errors were encountered: