Add target_table_info in tables.list_joinable's response

[Anish9901]

Fixes #3713

Checklist

• My pull request has a descriptive title (not a vague title like Update index.md).
• My pull request targets the develop branch of the repository
• My commit messages follow best practices.
• My code follows the established code style of the repository.
• I added tests for the changes I made (if applicable).
• I added or updated documentation (if applicable).
• I tried running the project locally and verified that there are no
  visible errors.

Developer Certificate of Origin

Developer Certificate of Origin

Developer Certificate of Origin
Version 1.1

Copyright (C) 2004, 2006 The Linux Foundation and its contributors.
1 Letterman Drive
Suite D4700
San Francisco, CA, 94129

Everyone is permitted to copy and distribute verbatim copies of this
license document, but changing it is not allowed.


Developer's Certificate of Origin 1.1

By making a contribution to this project, I certify that:

(a) The contribution was created in whole or in part by me and I
    have the right to submit it under the open source license
    indicated in the file; or

(b) The contribution is based upon previous work that, to the best
    of my knowledge, is covered under an appropriate open source
    license and I have the right under that license to submit that
    work with modifications, whether created in whole or in part
    by me, under the same open source license (unless I am
    permitted to submit under a different license), as indicated
    in the file; or

(c) The contribution was provided directly to me by some other
    person who certified (a), (b) or (c) and I have not modified
    it.

(d) I understand and agree that this project and the contribution
    are public and that a record of the contribution (including all
    personal information I submit with it, including my sign-off) is
    maintained indefinitely and may be redistributed consistent with
    this project or the open source license(s) involved.

[seancolsen]

Ahh — I see why you requested a different structure. Previously we were returning a JSON object with stringified Django column ids as keys. Now columns require two id values to be uniquely identified, so the old structure is problematic. Interesting!

I'm okay with nesting them.

But in running the code in this PR, I'm seeing some problems.

I'm running this request where 698589 is the oid of the "Books" table in our library management sample data.

{
  "jsonrpc": "2.0",
  "method": "tables.list_joinable",
  "id": 0,
  "params": {
    "database_id": 1,
    "table_oid": 698589,
    "max_depth": 1
  }
}

Here are some changes I'd like to see:

• Ensure that the joinable_tables and target_table_info arrays don't contain duplicates. There's something fishy going on somewhere. I have a hunch it's some kind of SQL join or aggregation problem. I'm expecting to see three results in each of these arrays. But instead I see 9.

• Restructure the target_table_info value. Currently it's an array of objects. An each object contains an entry which has a key of the table OID and a value of the table name. This would be very bad JSON schema design, so I don't imagine you did this intentionally. It looks like a bug. Instead of an array of objects, I'd like to see a single object where each entry has the table OID as the key and then an object as the value.

  You have:

  {
    "target_table_info": [
      {
        "698581": "Authors",
        "columns": []
      }
    ]
  }

  I want:

  {
    "target_table_info": {
      "698581": {
        "name": "Authors",
        "columns": []
      }
    }
  }

[Anish9901]

I think I've addressed your feedback @seancolsen, Here is the response for the "Books" table with depth=1.

Response

{
  "joinable_tables": [
    {
      "base": 2254321,
      "depth": 1,
      "target": 2254313,
      "fkey_path": [
        [2254389, false]
      ],
      "join_path": [
        [
          [2254321, 10],
          [2254313,  1]
        ]
      ],
      "multiple_results": false
    },
    {
      "base": 2254321,
      "depth": 1,
      "target": 2254358,
      "fkey_path": [
        [2254394, false]
      ],
      "join_path": [
        [
          [2254321, 11],
          [2254358,  1]
        ]
      ],
      "multiple_results": false
    },
    {
      "base": 2254321,
      "depth": 1,
      "target": 2254334,
      "fkey_path": [
        [2254399, true]
      ],
      "join_path": [
        [
          [2254321, 1],
          [2254334, 5]
        ]
      ],
      "multiple_results": true
    }
  ],
  "target_table_info": {
    "2254313": {
      "name": "Authors",
      "columns": [
        {"name": "id"        , "type": "integer", "attnum": 1},
        {"name": "Last Name" , "type": "text"   , "attnum": 3},
        {"name": "First Name", "type": "text"   , "attnum": 2}
      ]
    },
    "2254334": {
      "name": "Items",
      "columns": [
        {"name": "id"              , "type": "integer", "attnum": 1},
        {"name": "Book"            , "type": "integer", "attnum": 5},
        {"name": "Acquisition Date", "type": "date"   , "attnum": 3},
        {"name": "Barcode"         , "type": "text"   , "attnum": 2}
      ]
    },
    "2254358": {
      "name": "Publishers",
      "columns": [
        {"name": "id"  , "type": "integer", "attnum": 1},
        {"name": "Name", "type": "text"   , "attnum": 2}
      ]
    }
  }
}

Do you have an opinion on whether the columns should be a key-value pair of attnum and {name, type} or are you fine with the approach of the above response where I've nested the attnum instead?

[seancolsen]

Good question, @Anish9901. Can you please put the columns data in an object too? Stringified attnums as keys.

[Anish9901]

Can you please put the columns data in an object too? Stringified attnums as keys.

Done @seancolsen.

Here is the updated response:

{
  "joinable_tables": [
    {
      "base": 2254321,
      "depth": 1,
      "target": 2254313,
      "fkey_path": [
        [2254389, false]
      ],
      "join_path": [
        [
          [2254321, 10],
          [2254313,  1]
        ]
      ],
      "multiple_results": false
    },
    {
      "base": 2254321,
      "depth": 1,
      "target": 2254358,
      "fkey_path": [
        [2254394, false]
      ],
      "join_path": [
        [
          [2254321, 11],
          [2254358,  1]
        ]
      ],
      "multiple_results": false
    },
    {
      "base": 2254321,
      "depth": 1,
      "target": 2254334,
      "fkey_path": [
        [2254399, true]
      ],
      "join_path": [
        [
          [2254321, 1],
          [2254334, 5]
        ]
      ],
      "multiple_results": true
    }
  ],
  "target_table_info": {
    "2254313": {
      "name": "Authors",
      "columns": {
        "1": {"name": "id"        , "type": "integer"},
        "2": {"name": "First Name", "type": "text"   },
        "3": {"name": "Last Name" , "type": "text"   }
      }
    },
    "2254334": {
      "name": "Items",
      "columns": {
        "1": {"name": "id"              , "type": "integer"},
        "2": {"name": "Barcode"         , "type": "text"   },
        "3": {"name": "Acquisition Date", "type": "date"   },
        "5": {"name": "Book"            , "type": "integer"}
      }
    },
    "2254358": {
      "name": "Publishers",
      "columns": {
        "1": {"name": "id"  , "type": "integer"},
        "2": {"name": "Name", "type": "text"   }
      }
    }
  }
}

[mathemancer]

Overall, not bad. Given that we're waiting for feedback from Sean, I think it's a good idea to take another pass through the SQL function and tighten a couple of screws. See my specific comments.

[mathemancer]

Why make this a temp table rather than a proper CTE here? It's

• slightly less performant under some conditions (i.e., when it's small, and
• affects the volatility setting, meaning it's slower for callers

[mathemancer]

If you change the temp table to a CTE, you should mark this as STABLE.

[mathemancer]

LGTM. Nice!

[seancolsen]

This is great. Thanks!