sean

sean

TL;DR The problem I’m trying to solve is analogous to building an Ecto.Query, say, for posts that includes n number of latest associated comments for each.

I realize this has been discussed several times already, but I still have not found a clear solution:

As an example, given the following:

defmodule Comment do
  schema "comments" do
    belongs_to(:post, Post)
  end
end

defmodule Post do
  schema "posts" do
    has_many(:comments, Comment)    
  end

  defmacrop latest_comments(post_id, limit \\ 5) do
    quote do
      fragment(
        "select * from comments
        where comments.post_id = ?
        order by comments.inserted_at desc nulls last
        limit ?",
        unquote(post_id),
        unquote(limit)
      )
    end
  end

  def with_latest_comments(queryable) do
    queryable
    |> join(:left_lateral, [p], c in latest_comments(p.id))
    |> select_merge([p, c], %{comments: c})
  end
end

I think this is so close to a solution, but it raises this error:

PostgreSQL does not support selecting all fields from without a schema. Please specify a schema or specify exactly which fields you want to select in query

If I wrap c with map or struct e.g.:

|> select_merge([p, c], %{comments: struct(c, [:id, :post_id])})

The following error is raised:

it is not possible to return a map/struct subset of a fragment, you must explicitly return the desired individual fields in query

I see this was discussed here #1592 Ecto.QueryError) PostgreSQL requires a schema module when using selector “u0” but none was given.

The following suggestions were mentioned, either of which I think would work great!

We could also support struct(u, [:foo, :bar, :baz], MyModule)

or even struct(u, MyModule) if there are all the columns since the module contains the column names

But alas, that issue is from 2016 and has long been closed. :confused:

Any other ideas? Am I missing something obvious?

Thanks!

Showing Posts 1 to 2

peerreynders

peerreynders

Edit: Full example of a working lateral join.

iex(1)> alias MusicDB.{Repo,Album,Track}
[MusicDB.Repo, MusicDB.Album, MusicDB.Track]
iex(2)> import Ecto.Query
Ecto.Query
iex(3)> compact_rest = fn
...(3)>   {nil,nil,nil,[]} ->
...(3)>     []
...(3)>   {_, album, tracks, rest} = init ->
...(3)>     [Map.put(album, :tracks, :lists.reverse(tracks)) | rest]
...(3)> end
#Function<6.127694169/1 in :erl_eval.expr/5>
iex(4)> next_acc = fn
...(4)>   (%{id: id} = a, t, {id, album, tracks, rest}) ->
...(4)>     {id, album, [t|tracks], rest}
...(4)>   (%{id: id} = a, t, acc) ->
...(4)>     {id, a, [t], compact_rest.(acc)}
...(4)> end
#Function<18.127694169/3 in :erl_eval.expr/5>
iex(5)> row_to_acc = fn (%{album: album, track: track}, acc) -> next_acc.(album, track, acc) end
#Function<12.127694169/2 in :erl_eval.expr/5>
iex(6)> make_result = fn rows ->
...(6)>   rows
...(6)>   |> List.foldl({nil,nil,nil,[]}, row_to_acc)
...(6)>   |> compact_rest.()
...(6)>   |> :lists.reverse()
...(6)> end
#Function<6.127694169/1 in :erl_eval.expr/5>
iex(7)> album_ids = [1,2]
[1, 2]
iex(8)> album_query = from(a in Album, [where: a.id in ^album_ids])
#Ecto.Query<from a in MusicDB.Album, where: a.id in ^[1, 2]>
iex(9)> query = from(q in subquery(album_query), [
...(9)>   left_lateral_join: t in fragment("SELECT * FROM tracks AS t WHERE t.album_id = ? ORDER BY t.inserted_at DESC NULLS LAST LIMIT ?", q.id,3), on: true,
...(9)>   select: %{album: %{id: q.id, artist_id: q.artist_id, title: q.title}},
...(9)>   select_merge: %{track: %{id: t.id, index: t.index, title: t.title, duration: t.duration}}
...(9)> ])
#Ecto.Query<from a in subquery(from a in MusicDB.Album,
  where: a.id in ^[1, 2]),
 left_join_lateral: f in fragment("SELECT * FROM tracks AS t WHERE t.album_id = ? ORDER BY t.inserted_at DESC NULLS LAST LIMIT ?", a.id, 3),
 on: true,
 select: %{album: %{id: a.id, artist_id: a.artist_id, title: a.title}, track: %{id: f.id, index: f.index, title: f.title, duration: f.duration}}>
iex(10)> results = Repo.all(query)

18:31:36.947 [debug] QUERY OK db=2.5ms
SELECT s0."id", s0."artist_id", s0."title", f1."id", f1."index", f1."title", f1."duration" FROM (SELECT a0."id" AS "id", a0."title" AS "title", a0."inserted_at" AS "inserted_at", a0."updated_at" AS "updated_at", a0."artist_id" AS "artist_id" FROM "albums" AS a0 WHERE (a0."id" = ANY($1))) AS s0 LEFT OUTER JOIN LATERAL (SELECT * FROM tracks AS t WHERE t.album_id = s0."id" ORDER BY t.inserted_at DESC NULLS LAST LIMIT 3) AS f1 ON TRUE [[1, 2]]
[
  %{
    album: %{artist_id: 1, id: 1, title: "Kind Of Blue"},
    track: %{duration: 481, id: 5, index: 5, title: "Flamenco Sketches"}
  },
  %{
    album: %{artist_id: 1, id: 1, title: "Kind Of Blue"},
    track: %{duration: 693, id: 4, index: 4, title: "All Blues"}
  },
  %{
    album: %{artist_id: 1, id: 1, title: "Kind Of Blue"},
    track: %{duration: 327, id: 3, index: 3, title: "Blue In Green"}
  },
  %{
    album: %{artist_id: 1, id: 2, title: "Cookin' At The Plugged Nickel"},
    track: %{duration: 1061, id: 10, index: 5, title: "No Blues"}
  },
  %{
    album: %{artist_id: 1, id: 2, title: "Cookin' At The Plugged Nickel"},
    track: %{duration: 754, id: 9, index: 4, title: "Miles"}
  },
  %{
    album: %{artist_id: 1, id: 2, title: "Cookin' At The Plugged Nickel"},
    track: %{duration: 896, id: 8, index: 3, title: "Walkin'"}
  }
]
iex(11)> result = make_result.(results)
[
  %{
    artist_id: 1,
    id: 1,
    title: "Kind Of Blue",
    tracks: [
      %{duration: 481, id: 5, index: 5, title: "Flamenco Sketches"},
      %{duration: 693, id: 4, index: 4, title: "All Blues"},
      %{duration: 327, id: 3, index: 3, title: "Blue In Green"}
    ]
  },
  %{ 
    artist_id: 1,
    id: 2,
    title: "Cookin' At The Plugged Nickel",
    tracks: [
      %{duration: 1061, id: 10, index: 5, title: "No Blues"},
      %{duration: 754, id: 9, index: 4, title: "Miles"},
      %{duration: 896, id: 8, index: 3, title: "Walkin'"}
    ]
  }
]
iex(12)>

Raw SQL that served as the template:

SELECT a1.id AS album_id, a1.artist_id, a1.title AS album_title, t1.id AS track_id, t1.index, t1.title, t1.duration FROM
  (SELECT id, artist_id, title FROM albums WHERE id in (1,2)) a1
  LEFT JOIN LATERAL
  (SELECT id, index, title, duration FROM tracks WHERE album_id = a1.id ORDER BY inserted_at DESC NULLS LAST LIMIT 3) t1
  ON true;
peerreynders

peerreynders

Correction: Caveat

The issue is that the conditions apply to all the records returned by that single preload query - not just each single association.


preload query.

Example:

iex(1)> alias MusicDB.{Repo,Album,Track}
[MusicDB.Repo, MusicDB.Album, MusicDB.Track]
iex(2)> import Ecto.Query
Ecto.Query
iex(3)> album_id = 2
2
iex(4)> tracks_query = from(t in Track, [
...(4)>    order_by: fragment("? DESC NULLS LAST", t.inserted_at),
...(4)>    limit: 3
...(4)> ])
#Ecto.Query<from t in MusicDB.Track,
 order_by: [asc: fragment("? DESC NULLS LAST", t.inserted_at)], limit: 3>
iex(5)> query = from(a in Album, [
...(5)>   preload: [tracks: ^tracks_query],
...(5)>   where: a.id == ^album_id
...(5)> ])
#Ecto.Query<from a in MusicDB.Album, where: a.id == ^2,
 preload: [tracks: #Ecto.Query<from t in MusicDB.Track, order_by: [asc: fragment("? DESC NULLS LAST", t.inserted_at)], limit: 3>]>
iex(6)> Repo.all(query)

12:43:52.217 [debug] QUERY OK source="albums" db=2.2ms decode=1.7ms
SELECT a0."id", a0."title", a0."inserted_at", a0."updated_at", a0."artist_id" FROM "albums" AS a0 WHERE (a0."id" = $1) [2]
 
12:43:52.223 [debug] QUERY OK source="tracks" db=2.5ms
SELECT t0."id", t0."title", t0."duration", t0."index", t0."number_of_plays", t0."inserted_at", t0."updated_at", t0."album_id", t0."album_id" FROM "tracks" AS t0 WHERE (t0."album_id" = $1) ORDER BY t0."album_id", t0."inserted_at" DESC NULLS LAST LIMIT 3 [2]
[
  %MusicDB.Album{
    __meta__: #Ecto.Schema.Metadata<:loaded, "albums">,
    artist: #Ecto.Association.NotLoaded<association :artist is not loaded>,
    artist_id: 1,
    genres: #Ecto.Association.NotLoaded<association :genres is not loaded>,
    id: 2,
    inserted_at: ~N[2018-06-16 20:29:41.476625],
    title: "Cookin' At The Plugged Nickel",
    tracks: [
      %MusicDB.Track{
        __meta__: #Ecto.Schema.Metadata<:loaded, "tracks">,
        album: #Ecto.Association.NotLoaded<association :album is not loaded>,
        album_id: 2,
        duration: 1061,
        duration_string: nil,
        id: 10,
        index: 5,
        inserted_at: ~N[2018-06-16 20:29:41.480612],
        number_of_plays: 0,
        title: "No Blues",
        updated_at: ~N[2018-06-16 20:29:41.480618]
      },
      %MusicDB.Track{
        __meta__: #Ecto.Schema.Metadata<:loaded, "tracks">,
        album: #Ecto.Association.NotLoaded<association :album is not loaded>,
        album_id: 2,
        duration: 754,
        duration_string: nil,
        id: 9,
        index: 4,
        inserted_at: ~N[2018-06-16 20:29:41.480045],
        number_of_plays: 0,
        title: "Miles",
        updated_at: ~N[2018-06-16 20:29:41.480051]
      },
      %MusicDB.Track{
        __meta__: #Ecto.Schema.Metadata<:loaded, "tracks">,
        album: #Ecto.Association.NotLoaded<association :album is not loaded>,
        album_id: 2,
        duration: 896,
        duration_string: nil,
        id: 8,
        index: 3,
        inserted_at: ~N[2018-06-16 20:29:41.479460],
        number_of_plays: 0,
        title: "Walkin'",
        updated_at: ~N[2018-06-16 20:29:41.479465]
      }
    ],
    updated_at: ~N[2018-06-16 20:29:41.476631]
  }
]
iex(7)> 
— All posts loaded —

Where Next? Top

Trending in Questions Top

Blokh
Hey guys, I’ve got a huge CSV ( around 10 GB ) that needs to be processed hourly Do you guys have any suggestions what is the best prac...
New
kszambelanczyk
Hello! Could someone please give me a help/sample code, how to delete a file from s3 using waffle/waffle_ecto from Phoenix app. I creat...
New
Onor.io
I have what I’ve heard referred to as a “lookup table” in my database. This is a way of assigning codes to common values. One common lo...
New
Trolleger
What approach to take when sending live updates to “random” users Hi! I have a question, I have a little chat app, and when I create a DM...
New
matt-savvy
Anyone here using Honeybadger? My Honeybadger account is being overwhelmed with noise from some bots. Seeing a lot of Bandit.HTTPError...
New
RemyXRenard
I’m seeing that a list inside a Kino.DataTable will be interpreted as a charlist, even if the Kino.configure() is set to charlists: :as_l...
New
samoloth
Hi, I’ve just set up an application with ash_authentication. There is only magic link strategy for now, so there is no confirmation add o...
New

Other Trending Topics Top

mudasobwa
I am happy to introduce the very α version of the new programming language compiled to BEAM. Welcome Cure. It has literally three kille...
New
garrison
Hobbes is a low-level distributed database for the Elixir programming language. Hobbes provides a simple, safe, and scalable storage lay...
New
mcass19
ExRatatui lets you cook up rich terminal UIs in Elixir, powered by Rust’s ratatui via Rustler NIFs. Build interactive terminal applicatio...
New
Damirados
Hello everyone. After busy few months I am happy to announce v0.1.0 of Emerge &amp; Solve. They are GUI (Emerge) and State management (S...
New
netoum
Corex is an accessible, unstyled UI component library for Phoenix that integrates Zag.js state machines using Vanilla JavaScript and Live...
New
wintermeyer
There are three potential reasons for members of this forum to have a look at https://vutuv.de You are tired or annoyed of LinkedIn. Yo...
New

We're in Beta

About us Mission Statement

Options

Thread Display Mode




Thread Preview

Skip Thread Previews