Message-ID: <gh-pgjdbc-pgjdbc-3606-c2824090276@github.com>
From: "cgm-aw (@cgm-aw)" <noreply+cgm-aw@github.com>
To: "pgjdbc/pgjdbc" <noreply+pgjdbc-pgjdbc@github.com>
Date: Wed, 23 Apr 2025 12:15:13 +0000
Subject: Re: [pgjdbc/pgjdbc] PR #3606: fix: add support for multiple schemas via PgConnection.setSchema
In-Reply-To: <gh-pgjdbc-pgjdbc-3606@github.com>
References: <gh-pgjdbc-pgjdbc-3606@github.com>
List-Id: <gh-pgjdbc-pgjdbc.github.com>
X-GitHub-Author-Login: cgm-aw
X-GitHub-Comment-Id: 2824090276
X-GitHub-Comment-Type: issue_comment
X-GitHub-Issue: 3606
X-GitHub-Repo: pgjdbc/pgjdbc
X-GitHub-Type: comment
X-GitHub-Url: https://github.com/pgjdbc/pgjdbc/pull/3606#issuecomment-2824090276
Content-Type: text/plain; charset=utf-8

Ok, I changed `getSchema` to call `SHOW search_path`.
I noticed this in a test case:
Before (current_schema()):
```
execute("SET search_path TO \"$user\",public,schema2");
assertEquals(TestUtil.getUser(), conn.getSchema()); // Now fails
```

After (show search_path):
```
execute("SET search_path TO \"$user\",public,schema2");
assertEquals("\"$user\", public, schema2", conn.getSchema()); // OK
```

So this means that `current_schema()` resolves the schema `$user` to the respective user name, while `SHOW search_path` keeps the variable in its output. Since you would use the variable in the SET command, I guess this is fine.

Let me know if the PR is ok now, then I will look forward to the release 😃