Update Current Page data pull by Scrivener

Phoenix Forum Questions / Help
, ,
1

i want to update only the current records being viewed by the user. but my current query is updating every record matching in the db.
how do i only apply this affect to only records that have been pulled by scrivener that the user is viewing
eg.
if my current table is only showing 10records on those 10 should change to there status to read
here is my query.

defmodule Messages do
 def mark_status_updated do
  MyApp.Messages.where(status: "PENDING" )
    |> case do
      [] -> []
      pending_sms ->
        Enum.map(pending_sms, fn smses ->
         MyApp.Messages.update(smses, status: "READ")
        end)
    end
  end
end
2

Well, you should have access to the current pagination parameters. You can then use them to fetch only those records and change only them.

3 
def update_status(ids) do
      Myschema
      |> where([a], a.id in ^ids and a.status == "PENDING")
      |> Repo.update_all(set: [status: "READ"])
    end

this returns an error of

value `10100` in `where` cannot be cast to type {:in, :id} in query:
4

Is the query being passed a single integer (10100) rather than the expected List of integers?

5

I would do this to try and understand what’s wrong:

      Myschema
      |> where([a], a.id in ^ids and a.status == "PENDING")
      |> Ecto.Adapters.SQL.to_sql()

And take it from there.

1 Like
6 
def slogs(search_params, page, size) do
    MySchema
    |> handle_report_filter(search_params)
    |> order_by(desc: :status)
    |> compose_report_select()
    |> Repo.paginate(page: page, page_size: size)
    |> update_status()
  end

it returns this error

 value `%Scrivener.Page{entries: [%{[mydata]}]  in `where` cannot be cast to type {:in, :id} in query:
7

it anonymously picks an id value places it in there depending on which function is use to call it.

 def slogs(search_params, page, size) do
    SmsLog
    |> handle_report_filter(search_params)
    |> order_by(desc: :status)
    |> compose_report_select()
    |> Repo.paginate(page: page, page_size: size)
     |> (fn x ->
       x |> Enum.map(& update_status(&1.id))
     end).()
  end

if i used the function above that is when it returns that

8

We will not go anywhere like that. Why don’t you try inspecting the SQL that Ecto is trying to send to the DB and show it to us?

9

It appears that in one example a Scrivener.Page struct is being passed to update_status and, in the other, single integers rather than the expected List of integers. But, to @dimitarvp 's point, inspecting the query with his example code will likely uncover the issue quickly. Additionally, IO.inspect can help a lot in debugging.

For instance, it could be added to update_status…

def update_status(ids) do
  IO.inspect(ids)

  Myschema
  |> where([a], a.id in ^ids and a.status == "PENDING")
  |> Repo.update_all(set: [status: "READ"])
end

It can also be added anywhere within a pipeline and multiple times…

def slogs(search_params, page, size) do
  MySchema
  |> handle_report_filter(search_params)
  |> order_by(desc: :status)
  |> compose_report_select()
  |> Repo.paginate(page: page, page_size: size)
  |> IO.inspect()
  |> update_status()
end

Exercise the code in your tests or even the running app and the inspected values will appear in the terminal.

1 Like
10

thanks found my error and fixed it

11

What was it? Share it for future reader’s reference?