jeramyRR

jeramyRR

Does anyone have any experience decoding json into structs, with Poison, where the struct has an improper list?

I’m trying to decode json from binance api that looks like the following:

 {
  "timezone": "UTC",
  "serverTime": 1508631584636,
  "rateLimits": [{
      "rateLimitType": "REQUESTS_WEIGHT",
      "interval": "MINUTE",
      "limit": 1200
    },
    {
      "rateLimitType": "ORDERS",
      "interval": "SECOND",
      "limit": 10
    },
    {
      "rateLimitType": "ORDERS",
      "interval": "DAY",
      "limit": 100000
    }
  ],
  "exchangeFilters": [],
  "symbols": [{
    "symbol": "ETHBTC",
    "status": "TRADING",
    "baseAsset": "ETH",
    "baseAssetPrecision": 8,
    "quoteAsset": "BTC",
    "quotePrecision": 8,
    "orderTypes": ["LIMIT", "MARKET"],
    "icebergAllowed": false,
    "filters": [{
      "filterType": "PRICE_FILTER",
      "minPrice": "0.00000100",
      "maxPrice": "100000.00000000",
      "tickSize": "0.00000100"
    }, {
      "filterType": "LOT_SIZE",
      "minQty": "0.00100000",
      "maxQty": "100000.00000000",
      "stepSize": "0.00100000"
    }, {
      "filterType": "MIN_NOTIONAL",
      "minNotional": "0.00100000"
    }]
  }]
}

I have structs defined for each key value pair, and Poison is decoding it like a champ all the way up to the filters inside the symbols list. I believe it is because that data falls into the improper list category. Below is my call to Poison.decode:

@spec decode(json: String.t()) :: %ExchangeInfo{}
def decode(json) do
        json
        |> Poison.decode!(
          as: %ExchangeInfo{
            rateLimits: [%RateLimit{}],
            exchangeFilters: [%ExchangeFilter{}],
            symbols: [
              %Symbol{
                filters: [
                  %IcebergParts{},
                  %LotSize{},
                  %Price{},
                  %MaxNumOrders{},
                  %MaxNumAlgoOrders{},
                  %MinNotional{}
                ]
              }
            ]
          }
        )
end

And here is how I have the Symbol struct defined:

@derive [Poison.Encoder]
  defstruct [
    :symbol,
    :status,
    :baseAsset,
    :baseAssetPrecision,
    :quoteAsset,
    :quotePrecision,
    :orderTypes,
    :icebergAllowed,
    filters: []
  ]

The part that I’m jacking up is the filters section of that decode expression. I’ll spare ya’ll the other structs, but they are pretty simple.

Any ideas on how I can get Poison to decode the filters section into the proper struct?

Showing Posts 1 to 4

ConnorRigby

ConnorRigby

Nerves Core Team

This is actually one of my favorite use cases for macros in Elixir. I’ll show how i’d do it without macros first and if there is interest i can show how i’d build a macro later.

Anyway i really don’t like Poison’s as: %Data{} feature. I won’t go into much details, other than you are now locked into using Poison for this feature that other json libraries might not support. Long story short: you should separate you’re JSON decoding from structuring and restructuring. Elixir != JSON and it shouldn’t be treated as if it is. Anyway here’s my decoder:

defmodule ExchangeInfo do
  defmodule RateLimit do
    defstruct [
      :rateLimitType,
      :interval,
      :limit
    ]

    @keys ~w(rateLimitType interval limit)
    def decode(%{} = map) do
      map
      |> Map.take(@keys)
      |> Enum.map(fn({k, v}) -> {String.to_existing_atom(k), v} end)
      |> Enum.map(&decode/1)
      |> fn(data) -> struct(__MODULE__, data) end.()
    end

    def decode({k, v}), do: {k, v}
  end

  defmodule ExchangeFilter do
    @keys ~w()
    defstruct []

    def decode(%{} = map) do
      map
      |> Map.take(@keys)
      |> Enum.map(fn({k, v}) -> {String.to_existing_atom(k), v} end)
      |> Enum.map(&decode/1)
      |> fn(data) -> struct(__MODULE__, data) end.()
    end

    def decode({k, v}), do: {k, v}
  end

  defmodule Symbol do
    defmodule PriceFilter do
      @keys ~w(maxPrice minPrice tickSize)
      defstruct [:maxPrice, :minPrice, :tickSize]
      def decode(%{} = map) do
        map
        |> Map.take(@keys)
        |> Enum.map(fn({k, v}) -> {String.to_existing_atom(k), v} end)
        |> Enum.map(&decode/1)
        |> fn(data) -> struct(__MODULE__, data) end.()
      end

      def decode({k, v}), do: {k, v}
    end

    defmodule LotSize do
      @keys ~w(maxPrice minPrice tickSize)
      defstruct [:maxPrice, :minPrice, :tickSize]
      def decode(%{} = map) do
        map
        |> Map.take(@keys)
        |> Enum.map(fn({k, v}) -> {String.to_existing_atom(k), v} end)
        |> Enum.map(&decode/1)
        |> fn(data) -> struct(__MODULE__, data) end.()
      end

      def decode({k, v}), do: {k, v}
    end

    defmodule MinNotional do
      @keys ~w(minNotional)
      defstruct [:minNotional]
      def decode(%{} = map) do
        map
        |> Map.take(@keys)
        |> Enum.map(fn({k, v}) -> {String.to_existing_atom(k), v} end)
        |> Enum.map(&decode/1)
        |> fn(data) -> struct(__MODULE__, data) end.()
      end

      def decode({k, v}), do: {k, v}
    end

    @keys ~w(symbol status baseAsset baseAssetPrecision quoteAsset quotePrecision orderTypes icebergAllowed filters)
    defstruct [
      :symbol,
      :status,
      :baseAsset,
      :baseAssetPrecision,
      :quoteAsset,
      :quotePrecision,
      :orderTypes,
      :icebergAllowed,
      :filters,
    ]

    def decode(%{} = map) do
      map
      |> Map.take(@keys)
      |> Enum.map(fn({k, v}) -> {String.to_existing_atom(k), v} end)
      |> Enum.map(&decode/1)
      |> fn(data) -> struct(__MODULE__, data) end.()
    end

    def decode({:filters, filters}) when is_list(filters) do
      {:filters, Enum.map(filters, &decode_filters/1)}
    end

    def decode({k, v}), do: {k, v}

    defp decode_filters(%{"filterType" => "PRICE_FILTER"} = data), do: PriceFilter.decode(data)
    defp decode_filters(%{"filterType" => "MIN_NOTIONAL"} = data), do: MinNotional.decode(data)
    defp decode_filters(%{"filterType" => "LOT_SIZE"} = data), do: LotSize.decode(data)
  end

  defstruct [
    :rateLimits,
    :exchangeFilters,
    :symbols
  ]

  @keys ~w(rateLimits exchangeFilters symbols)
  def decode(%{} = map) do
    map
    |> Map.take(@keys)
    |> Enum.map(fn({k, v}) -> {String.to_existing_atom(k), v} end)
    |> Enum.map(&decode/1)
    |> fn(data) -> struct(__MODULE__, data) end.()
  end

  def decode({:rateLimits, limits}) when is_list(limits) do
    {:rateLimits, Enum.map(limits, &RateLimit.decode/1)}
  end

  def decode({:exchangeFilters, filters}) do
    {:exchangeFilters, Enum.map(filters, &ExchangeFilter.decode/1)}
  end

  def decode({:symbols, symbols}) when is_list(symbols) do
    {:symbols, Enum.map(symbols, &Symbol.decode/1)}
  end
end
jeramyRR

jeramyRR OP

Thanks for taking the time out to put that together. You’ve got me really curious about the macro now!

ConnorRigby

ConnorRigby

Nerves Core Team

So you may have noticed a lot of duplication in that decoder. Duplication can be a great use case for a macro. Each and every module defined above has a decode/1 function. Lets extract that out. Here’s a refactored module:

defmodule ExchangeInfo do
  import Decoder

  defmodule RateLimit do
    mdecode ~w(rateLimitType interval limit)
  end

  defmodule ExchangeFilter do
    mdecode ~w()
  end

  defmodule Symbol do
    defmodule PriceFilter do
      mdecode ~w(maxPrice minPrice tickSize)
    end

    defmodule LotSize do
      mdecode ~w(maxPrice minPrice tickSize)
    end

    defmodule MinNotional do
      mdecode ~w(minNotional)
    end

    mdecode ~w(symbol status baseAsset baseAssetPrecision quoteAsset quotePrecision orderTypes icebergAllowed filters) do
      mlist :filters, fn(data) ->
        case data do
          %{"filterType" => "PRICE_FILTER"} -> PriceFilter.decode(data)
          %{"filterType" => "MIN_NOTIONAL"} -> MinNotional.decode(data)
          %{"filterType" => "LOT_SIZE"} -> LotSize.decode(data)
        end
      end
    end

  end

  mdecode ~w(rateLimits exchangeFilters symbols) do
    mlist :rateLimits, RateLimit
    mlist :exchangeFilters, ExchangeFilter
    mlist :symbols, Symbol
  end
end

a bit cleaner in my opinion. Here’s the Decoder module implementation:

defmodule Decoder do
  @doc "Defines a struct and decode/1 function."
  defmacro mdecode(str_fields, block \\ [do: nil])

  defmacro mdecode(str_fields, [do: _] = block) do
    quote do
      defstruct Enum.map(unquote(str_fields), &String.to_atom/1)

      def decode(%{} = map) do
        map
        |> Map.take(unquote(str_fields))
        |> Enum.map(fn({k, v}) -> {String.to_existing_atom(k), v} end)
        |> Enum.map(&decode/1)
        |> fn(data) -> struct(__MODULE__, data) end.()
      end

      unquote(block)

      def decode({k, v}), do: {k, v}
    end
  end

  @doc "Defines a decode/1 function that does an `Enum.map` with a function or module."
  defmacro mlist(field, args) do
    quote do
      cond do
        is_atom(unquote(args)) ->
          def decode({unquote(field), arg}) when is_list(arg) do
            {unquote(field), Enum.map(arg, fn(data) -> unquote(args).decode(data) end)}
          end
        is_function(unquote(args)) ->
          def decode({unquote(field), arg}) when is_list(arg) do
            {unquote(field), Enum.map(arg, fn(data) -> unquote(args).(data) end)}
          end
      end

    end
  end
end

Pretty simple.
mdecode defines a struct and the head decode/1 function from the original implementation.
that Enum.map(&decode/1) will call decode({key, value}) for every one of our expected keys after converting them to atoms. the unquote(block) is a bit of a hack to make the syntax look beteter imo. You could do this without that pretty easily.

the mlist/2 macro just defines that function decode({key, value}) and maps over value by either calling unquote(module).decode/1 or if supplied a function, it calls that instead which is what we do for :filters

After writing that all up i realized your original question wasn’t really answered and this may have been an overload of information depending on how comfortable with Elixir/Metaprogramming you are.

The main point i guess i want to get across is a subjective one: don’t just take my word for it -
Don’t rely on features of a dep outside of the scope of it’s domain. Poison is a json decoder, not a domain data structurer/destructurer. It’s main concern should be taking binary data and parsing it as Elixir data, NOT taking binary data and turning it into youre domain specific data. It doesn’t have enough information to perform that task.

jeramyRR

jeramyRR OP

Thanks again for spending so much time to go through this.

— All posts loaded —

Where Next? Top

Trending in Questions Top

stjefim
Hello! Suppose you are building workflow (order / task / payment) processing system with the following requirements: Each workflow con...
New
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
jaybe78
Hello, I’m developing a online persistent chat system (what’s app) like using elixir/dynamodb/aws for a mobile app(flutter). The diffic...
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
widianto
I think I’ve found a small improvement I could contribute to <%= web_namespace %>.CoreComponents (installer/templates/phx_web/compo...
New

Other Trending Topics Top

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 & 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
webofbits
Aludel - LLM Evaluation Workbench Aludel is an embeddable Phoenix LiveView dashboard for evaluating and comparing LLM prompts across mult...
New

We're in Beta

About us Mission Statement

Options

Thread Display Mode




Thread Preview

Skip Thread Previews