Message-ID: <gh-pgjdbc-pgjdbc-3543-c2694444364@github.com>
From: "MaxMello (@MaxMello)" <noreply+MaxMello@github.com>
To: "pgjdbc/pgjdbc" <noreply+pgjdbc-pgjdbc@github.com>
Date: Mon, 03 Mar 2025 13:43:35 +0000
Subject: Re: [pgjdbc/pgjdbc] issue #3543: Table meta data is broken if catalog name not equal database name
In-Reply-To: <gh-pgjdbc-pgjdbc-3543@github.com>
References: <gh-pgjdbc-pgjdbc-3543@github.com>
List-Id: <gh-pgjdbc-pgjdbc.github.com>
X-GitHub-Author-Login: MaxMello
X-GitHub-Comment-Id: 2694444364
X-GitHub-Comment-Type: issue_comment
X-GitHub-Edited-At: 2025-03-03T13:48:59Z
X-GitHub-Issue: 3543
X-GitHub-Repo: pgjdbc/pgjdbc
X-GitHub-Type: comment
X-GitHub-Url: https://github.com/pgjdbc/pgjdbc/issues/3543#issuecomment-2694444364
Content-Type: text/plain; charset=utf-8

One additonal info, calling `connection.getCatalog()` will return "read" or "write", but `metaData.getCatalogs()` will not return those values, instead only the proper dbname ("postgres"). 

I traced back how `getCatalog()` is implemented and if I understand the code correctly, it will at one point call `PGProperty.PG_DBNAME.getOrDefault(info);` which is "just" the path part of the connection url? 

Compared to that, `getCatalogs()` "properly" queries the database. Is it maybe requried to use something like `SELECT current_database();` to get the database / catalog name, instead of relying on the connection properties? At least in this use case with pgbouncer?