Feature Request: Support Dynamic File Loading with `url` function instead of Table Function

# Current Behavior
Currently, the `url` table function allows querying files hosted on an HTTP server as follows:

```SQL
SELECT * FROM url('http://localhost:80/file1', 'JSONEachRow')
```
This produces a result like:
```SQL
┌─result─────────────────────────────────┐
│ { "source": "file1", "key": "value1" } │
│ { "source": "file1", "key": "value2" } │
└────────────────────────────────────────┘
```
Although it works well for a single file, handling multiple files becomes very complicated.
For example, to query multiple files, I must use `UNION`:
```SQL
SELECT * FROM url('http://localhost/file1', 'JSONEachRow')
UNION
SELECT * FROM url('http://localhost/file2', 'JSONEachRow')
UNION
SELECT * FROM url('http://localhost/file3', 'JSONEachRow');
```
This approach becomes not feasible when dealing with a large or dynamic number of files.

# Desired Behavior
I would like to query multiple files dynamically by leveraging a table of file names.
For instance, given a table of files, I could construct a query like this:
```SQL
SELECT url('http://localhost/' || fileName, 'JSONEachRow') FROM files;
```
The expected result is:
```SQL
┌─result───────────────────────────────────────────────────────────────────────────┐
│ [{ "source": "file1", "key": "value1" }, { "source": "file1", "key": "value2" }] │
│ [{ "source": "file2", "key": "value1" }, { "source": "file2", "key": "value2" }] │
│ [{ "source": "file3", "key": "value1" }, { "source": "file3", "key": "value2" }] │
└──────────────────────────────────────────────────────────────────────────────────┘
```
Than I could use `arrayJoin` to unfold and flatten this data into a table format identical to the one produced using multiple `UNION` clauses.

This feature would be particularly helpful for cases where files are created dynamically, as I could use a materialized view to query new files automatically upon creation.

This feature might be also helpful for other table functions like `file` or `s3`

If there is already a way to achieve this behaviour efficiently, I would like to know!