defmodule Ecto.Adapters.SQL do @moduledoc ~S""" This application provides functionality for working with SQL databases in `Ecto`. ## Built-in adapters By default, we support the following adapters: * `Ecto.Adapters.Postgres` for Postgres * `Ecto.Adapters.MyXQL` for MySQL * `Ecto.Adapters.Tds` for SQLServer ## Additional functions If your `Ecto.Repo` is backed by any of the SQL adapters above, this module will inject additional functions into your repository: * `disconnect_all(interval, options \\ [])` - shortcut for `Ecto.Adapters.SQL.disconnect_all/3` * `explain(type, query, options \\ [])` - shortcut for `Ecto.Adapters.SQL.explain/4` * `query(sql, params, options \\ [])` - shortcut for `Ecto.Adapters.SQL.query/4` * `query!(sql, params, options \\ [])` - shortcut for `Ecto.Adapters.SQL.query!/4` * `query_many(sql, params, options \\ [])` - shortcut for `Ecto.Adapters.SQL.query_many/4` * `query_many!(sql, params, options \\ [])` - shortcut for `Ecto.Adapters.SQL.query_many!/4` * `to_sql(type, query)` - shortcut for `Ecto.Adapters.SQL.to_sql/3` Generally speaking, you must invoke those functions directly from your repository, for example: `MyApp.Repo.query("SELECT true")`. You can also invoke them directly from `Ecto.Adapters.SQL`, but keep in mind that in such cases features such as "dynamic repositories" won't be available. ## Migrations `ecto_sql` supports database migrations. You can generate a migration with: $ mix ecto.gen.migration create_posts This will create a new file inside `priv/repo/migrations` with the `change` function. Check `Ecto.Migration` for more information. To interface with migrations, developers typically use mix tasks: * `mix ecto.migrations` - lists all available migrations and their status * `mix ecto.migrate` - runs a migration * `mix ecto.rollback` - rolls back a previously run migration If you want to run migrations programmatically, see `Ecto.Migrator`. ## SQL sandbox `ecto_sql` provides a sandbox for testing. The sandbox wraps each test in a transaction, making sure the tests are isolated and can run concurrently. See `Ecto.Adapters.SQL.Sandbox` for more information. ## Structure load and dumping If you have an existing database, you may want to dump its existing structure and make it reproducible from within Ecto. This can be achieved with two Mix tasks: * `mix ecto.load` - loads an existing structure into the database * `mix ecto.dump` - dumps the existing database structure to the filesystem For creating and dropping databases, see `mix ecto.create` and `mix ecto.drop` that are included as part of Ecto. ## Custom adapters Developers can implement their own SQL adapters by using `Ecto.Adapters.SQL` and by implementing the callbacks required by `Ecto.Adapters.SQL.Connection` for handling connections and performing queries. The connection handling and pooling for SQL adapters should be built using the `DBConnection` library. When using `Ecto.Adapters.SQL`, the following options are required: * `:driver` (required) - the database driver library. For example: `:postgrex` """ require Logger @type query_result :: %{ :rows => nil | [[term] | binary], :num_rows => non_neg_integer, optional(atom) => any } @doc false defmacro __using__(opts) do quote do @behaviour Ecto.Adapter @behaviour Ecto.Adapter.Migration @behaviour Ecto.Adapter.Queryable @behaviour Ecto.Adapter.Schema @behaviour Ecto.Adapter.Transaction opts = unquote(opts) @conn __MODULE__.Connection @driver Keyword.fetch!(opts, :driver) @impl true defmacro __before_compile__(env) do Ecto.Adapters.SQL.__before_compile__(@driver, env) end @impl true def ensure_all_started(config, type) do Ecto.Adapters.SQL.ensure_all_started(@driver, config, type) end @impl true def init(config) do Ecto.Adapters.SQL.init(@conn, @driver, config) end @impl true def checkout(meta, opts, fun) do Ecto.Adapters.SQL.checkout(meta, opts, fun) end @impl true def checked_out?(meta) do Ecto.Adapters.SQL.checked_out?(meta) end @impl true def loaders({:map, _}, type), do: [&Ecto.Type.embedded_load(type, &1, :json)] def loaders(:binary_id, type), do: [Ecto.UUID, type] def loaders(_, type), do: [type] @impl true def dumpers({:map, _}, type), do: [&Ecto.Type.embedded_dump(type, &1, :json)] def dumpers(:binary_id, type), do: [type, Ecto.UUID] def dumpers(_, type), do: [type] ## Query @impl true def prepare(:all, query) do {:cache, {System.unique_integer([:positive]), IO.iodata_to_binary(@conn.all(query))}} end def prepare(:update_all, query) do {:cache, {System.unique_integer([:positive]), IO.iodata_to_binary(@conn.update_all(query))}} end def prepare(:delete_all, query) do {:cache, {System.unique_integer([:positive]), IO.iodata_to_binary(@conn.delete_all(query))}} end @impl true def execute(adapter_meta, query_meta, query, params, opts) do Ecto.Adapters.SQL.execute(:named, adapter_meta, query_meta, query, params, opts) end @impl true def stream(adapter_meta, query_meta, query, params, opts) do Ecto.Adapters.SQL.stream(adapter_meta, query_meta, query, params, opts) end ## Schema @impl true def autogenerate(:id), do: nil def autogenerate(:embed_id), do: Ecto.UUID.generate() def autogenerate(:binary_id), do: Ecto.UUID.bingenerate() @impl true def insert_all( adapter_meta, schema_meta, header, rows, on_conflict, returning, placeholders, opts ) do Ecto.Adapters.SQL.insert_all( adapter_meta, schema_meta, @conn, header, rows, on_conflict, returning, placeholders, opts ) end @impl true def insert(adapter_meta, schema_meta, params, on_conflict, returning, opts) do %{source: source, prefix: prefix} = schema_meta {kind, conflict_params, _} = on_conflict {fields, values} = :lists.unzip(params) sql = @conn.insert(prefix, source, fields, [fields], on_conflict, returning, []) Ecto.Adapters.SQL.struct( adapter_meta, @conn, sql, :insert, source, [], values ++ conflict_params, kind, returning, opts ) end @impl true def update(adapter_meta, schema_meta, fields, params, returning, opts) do %{source: source, prefix: prefix} = schema_meta {fields, field_values} = :lists.unzip(fields) filter_values = Keyword.values(params) sql = @conn.update(prefix, source, fields, params, returning) Ecto.Adapters.SQL.struct( adapter_meta, @conn, sql, :update, source, params, field_values ++ filter_values, :raise, returning, opts ) end @impl true def delete(adapter_meta, schema_meta, params, returning, opts) do %{source: source, prefix: prefix} = schema_meta filter_values = Keyword.values(params) sql = @conn.delete(prefix, source, params, returning) Ecto.Adapters.SQL.struct( adapter_meta, @conn, sql, :delete, source, params, filter_values, :raise, returning, opts ) end ## Transaction @impl true def transaction(meta, opts, fun) do Ecto.Adapters.SQL.transaction(meta, opts, fun) end @impl true def in_transaction?(meta) do Ecto.Adapters.SQL.in_transaction?(meta) end @impl true def rollback(meta, value) do Ecto.Adapters.SQL.rollback(meta, value) end ## Migration @impl true def execute_ddl(meta, definition, opts) do Ecto.Adapters.SQL.execute_ddl(meta, @conn, definition, opts) end defoverridable prepare: 2, execute: 5, insert: 6, update: 6, delete: 5, insert_all: 8, execute_ddl: 3, loaders: 2, dumpers: 2, autogenerate: 1, checkout: 3, ensure_all_started: 2, __before_compile__: 1 end end @timeout 15_000 @doc """ Converts the given query to SQL according to its kind and the adapter in the given repository. ## Examples The examples below are meant for reference. Each adapter will return a different result: iex> Ecto.Adapters.SQL.to_sql(:all, Repo, Post) {"SELECT p.id, p.title, p.inserted_at, p.created_at FROM posts as p", []} iex> Ecto.Adapters.SQL.to_sql(:update_all, Repo, from(p in Post, update: [set: [title: ^"hello"]])) {"UPDATE posts AS p SET title = $1", ["hello"]} This function is also available under the repository with name `to_sql`: iex> Repo.to_sql(:all, Post) {"SELECT p.id, p.title, p.inserted_at, p.created_at FROM posts as p", []} """ @spec to_sql(:all | :update_all | :delete_all, Ecto.Repo.t(), Ecto.Queryable.t()) :: {String.t(), [term]} def to_sql(kind, repo, queryable) do case Ecto.Adapter.Queryable.prepare_query(kind, repo, queryable) do {{:cached, _update, _reset, {_id, cached}}, params} -> {String.Chars.to_string(cached), params} {{:cache, _update, {_id, prepared}}, params} -> {prepared, params} {{:nocache, {_id, prepared}}, params} -> {prepared, params} end end @doc """ Executes an EXPLAIN statement or similar for the given query according to its kind and the adapter in the given repository. ## Examples # Postgres iex> Ecto.Adapters.SQL.explain(Repo, :all, Post) "Seq Scan on posts p0 (cost=0.00..12.12 rows=1 width=443)" # MySQL iex> Ecto.Adapters.SQL.explain(Repo, :all, from(p in Post, where: p.title == "title")) |> IO.puts() +----+-------------+-------+------------+------+---------------+------+---------+------+------+----------+-------------+ | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra | +----+-------------+-------+------------+------+---------------+------+---------+------+------+----------+-------------+ | 1 | SIMPLE | p0 | NULL | ALL | NULL | NULL | NULL | NULL | 1 | 100.0 | Using where | +----+-------------+-------+------------+------+---------------+------+---------+------+------+----------+-------------+ # Shared opts iex> Ecto.Adapters.SQL.explain(Repo, :all, Post, analyze: true, timeout: 20_000) "Seq Scan on posts p0 (cost=0.00..11.70 rows=170 width=443) (actual time=0.013..0.013 rows=0 loops=1)\\nPlanning Time: 0.031 ms\\nExecution Time: 0.021 ms" It's safe to execute it for updates and deletes, no data change will be committed: iex> Ecto.Adapters.SQL.explain(Repo, :update_all, from(p in Post, update: [set: [title: "new title"]])) "Update on posts p0 (cost=0.00..11.70 rows=170 width=449)\\n -> Seq Scan on posts p0 (cost=0.00..11.70 rows=170 width=449)" This function is also available under the repository with name `explain`: iex> Repo.explain(:all, from(p in Post, where: p.title == "title")) "Seq Scan on posts p0 (cost=0.00..12.12 rows=1 width=443)\\n Filter: ((title)::text = 'title'::text)" ### Options Built-in adapters support passing `opts` to the EXPLAIN statement according to the following: Adapter | Supported opts ---------------- | -------------- Postgrex | `analyze`, `verbose`, `costs`, `settings`, `buffers`, `timing`, `summary`, `format` MyXQL | `format` All options except `format` are boolean valued and default to `false`. The allowed `format` values are `:map`, `:yaml`, and `:text`: * `:map` is the deserialized JSON encoding. * `:yaml` and `:text` return the result as a string. The built-in adapters support the following formats: * Postgrex: `:map`, `:yaml` and `:text` * MyXQL: `:map` and `:text` Any other value passed to `opts` will be forwarded to the underlying adapter query function, including shared Repo options such as `:timeout`. Non built-in adapters may have specific behaviour and you should consult their documentation for more details. For version compatiblity, please check your database's documentation: * _Postgrex_: [PostgreSQL doc](https://www.postgresql.org/docs/current/sql-explain.html). * _MyXQL_: [MySQL doc](https://dev.mysql.com/doc/refman/8.0/en/explain.html). """ @spec explain( pid() | Ecto.Repo.t() | Ecto.Adapter.adapter_meta(), :all | :update_all | :delete_all, Ecto.Queryable.t(), opts :: Keyword.t() ) :: String.t() | Exception.t() | list(map) def explain(repo, operation, queryable, opts \\ []) def explain(repo, operation, queryable, opts) when is_atom(repo) or is_pid(repo) do explain(Ecto.Adapter.lookup_meta(repo), operation, queryable, opts) end def explain(%{repo: repo} = adapter_meta, operation, queryable, opts) do Ecto.Multi.new() |> Ecto.Multi.run(:explain, fn _, _ -> {prepared, prepared_params} = to_sql(operation, repo, queryable) sql_call(adapter_meta, :explain_query, [prepared], prepared_params, opts) end) |> Ecto.Multi.run(:rollback, fn _, _ -> {:error, :forced_rollback} end) |> repo.transaction(opts) |> case do {:error, :rollback, :forced_rollback, %{explain: result}} -> result {:error, :explain, error, _} -> raise error _ -> raise "unable to execute explain" end end @doc """ Forces all connections in the repo pool to disconnect within the given interval. Once this function is called, the pool will disconnect all of its connections as they are checked in or as they are pinged. Checked in connections will be randomly disconnected within the given time interval. Pinged connections are immediately disconnected - as they are idle (according to `:idle_interval`). If the connection has a backoff configured (which is the case by default), disconnecting means an attempt at a new connection will be done immediately after, without starting a new process for each connection. However, if backoff has been disabled, the connection process will terminate. In such cases, disconnecting all connections may cause the pool supervisor to restart depending on the max_restarts/max_seconds configuration of the pool, so you will want to set those carefully. For convenience, this function is also available in the repository: iex> MyRepo.disconnect_all(60_000) :ok """ @spec disconnect_all( pid | Ecto.Repo.t() | Ecto.Adapter.adapter_meta(), non_neg_integer, opts :: Keyword.t() ) :: :ok def disconnect_all(repo, interval, opts \\ []) def disconnect_all(repo, interval, opts) when is_atom(repo) or is_pid(repo) do disconnect_all(Ecto.Adapter.lookup_meta(repo), interval, opts) end def disconnect_all(%{pid: pid} = _adapter_meta, interval, opts) do DBConnection.disconnect_all(pid, interval, opts) end @doc """ Returns a stream that runs a custom SQL query on given repo when reduced. In case of success it is a enumerable containing maps with at least two keys: * `:num_rows` - the number of rows affected * `:rows` - the result set as a list. `nil` may be returned instead of the list if the command does not yield any row as result (but still yields the number of affected rows, like a `delete` command without returning would) In case of failure it raises an exception. If the adapter supports a collectable stream, the stream may also be used as the collectable in `Enum.into/3`. Behaviour depends on the adapter. ## Options * `:log` - When false, does not log the query * `:max_rows` - The number of rows to load from the database as we stream ## Examples iex> Ecto.Adapters.SQL.stream(MyRepo, "SELECT $1::integer + $2", [40, 2]) |> Enum.to_list() [%{rows: [[42]], num_rows: 1}] """ @spec stream(Ecto.Repo.t(), String.t(), [term], Keyword.t()) :: Enum.t() def stream(repo, sql, params \\ [], opts \\ []) do repo |> Ecto.Adapter.lookup_meta() |> Ecto.Adapters.SQL.Stream.build(sql, params, opts) end @doc """ Same as `query/4` but raises on invalid queries. """ @spec query!(pid() | Ecto.Repo.t() | Ecto.Adapter.adapter_meta(), iodata, [term], Keyword.t()) :: query_result def query!(repo, sql, params \\ [], opts \\ []) do case query(repo, sql, params, opts) do {:ok, result} -> result {:error, err} -> raise_sql_call_error(err) end end @doc """ Runs a custom SQL query on the given repo. In case of success, it must return an `:ok` tuple containing a map with at least two keys: * `:num_rows` - the number of rows affected * `:rows` - the result set as a list. `nil` may be returned instead of the list if the command does not yield any row as result (but still yields the number of affected rows, like a `delete` command without returning would) ## Options * `:log` - When false, does not log the query * `:timeout` - Execute request timeout, accepts: `:infinity` (default: `#{@timeout}`); ## Examples iex> Ecto.Adapters.SQL.query(MyRepo, "SELECT $1::integer + $2", [40, 2]) {:ok, %{rows: [[42]], num_rows: 1}} For convenience, this function is also available under the repository: iex> MyRepo.query("SELECT $1::integer + $2", [40, 2]) {:ok, %{rows: [[42]], num_rows: 1}} """ @spec query(pid() | Ecto.Repo.t() | Ecto.Adapter.adapter_meta(), iodata, [term], Keyword.t()) :: {:ok, query_result} | {:error, Exception.t()} def query(repo, sql, params \\ [], opts \\ []) def query(repo, sql, params, opts) when is_atom(repo) or is_pid(repo) do query(Ecto.Adapter.lookup_meta(repo), sql, params, opts) end def query(adapter_meta, sql, params, opts) do sql_call(adapter_meta, :query, [sql], params, opts) end @doc """ Same as `query_many/4` but raises on invalid queries. """ @spec query_many!(Ecto.Repo.t() | Ecto.Adapter.adapter_meta(), iodata, [term], Keyword.t()) :: [query_result] def query_many!(repo, sql, params \\ [], opts \\ []) do case query_many(repo, sql, params, opts) do {:ok, result} -> result {:error, err} -> raise_sql_call_error(err) end end @doc """ Runs a custom SQL query that returns multiple results on the given repo. In case of success, it must return an `:ok` tuple containing a list of maps with at least two keys: * `:num_rows` - the number of rows affected * `:rows` - the result set as a list. `nil` may be returned instead of the list if the command does not yield any row as result (but still yields the number of affected rows, like a `delete` command without returning would) ## Options * `:log` - When false, does not log the query * `:timeout` - Execute request timeout, accepts: `:infinity` (default: `#{@timeout}`); ## Examples iex> Ecto.Adapters.SQL.query_many(MyRepo, "SELECT $1; SELECT $2;", [40, 2]) {:ok, [%{rows: [[40]], num_rows: 1}, %{rows: [[2]], num_rows: 1}]} For convenience, this function is also available under the repository: iex> MyRepo.query_many("SELECT $1; SELECT $2;", [40, 2]) {:ok, [%{rows: [[40]], num_rows: 1}, %{rows: [[2]], num_rows: 1}]} """ @spec query_many( pid() | Ecto.Repo.t() | Ecto.Adapter.adapter_meta(), iodata, [term], Keyword.t() ) :: {:ok, [query_result]} | {:error, Exception.t()} def query_many(repo, sql, params \\ [], opts \\ []) def query_many(repo, sql, params, opts) when is_atom(repo) or is_pid(repo) do query_many(Ecto.Adapter.lookup_meta(repo), sql, params, opts) end def query_many(adapter_meta, sql, params, opts) do sql_call(adapter_meta, :query_many, [sql], params, opts) end defp sql_call(adapter_meta, callback, args, params, opts) do %{pid: pool, telemetry: telemetry, sql: sql, opts: default_opts} = adapter_meta conn = get_conn_or_pool(pool) opts = with_log(telemetry, params, opts ++ default_opts) args = args ++ [params, opts] apply(sql, callback, [conn | args]) end defp put_source(opts, %{sources: sources}) when is_binary(elem(elem(sources, 0), 0)) do {source, _, _} = elem(sources, 0) [source: source] ++ opts end defp put_source(opts, _) do opts end @doc """ Check if the given `table` exists. Returns `true` if the `table` exists in the `repo`, otherwise `false`. The table is checked against the current database/schema in the connection. """ @spec table_exists?(Ecto.Repo.t(), table :: String.t(), opts :: Keyword.t()) :: boolean def table_exists?(repo, table, opts \\ []) when is_atom(repo) do %{sql: sql} = adapter_meta = Ecto.Adapter.lookup_meta(repo) {query, params} = sql.table_exists_query(table) query!(adapter_meta, query, params, opts).num_rows != 0 end # Returns a formatted table for a given query `result`. # # ## Examples # # iex> Ecto.Adapters.SQL.format_table(query) |> IO.puts() # +---------------+---------+--------+ # | title | counter | public | # +---------------+---------+--------+ # | My Post Title | 1 | NULL | # +---------------+---------+--------+ @doc false @spec format_table(%{ :columns => [String.t()] | nil, :rows => [term()] | nil, optional(atom) => any() }) :: String.t() def format_table(result) def format_table(nil), do: "" def format_table(%{columns: nil}), do: "" def format_table(%{columns: []}), do: "" def format_table(%{columns: columns, rows: nil}), do: format_table(%{columns: columns, rows: []}) def format_table(%{columns: columns, rows: rows}) do column_widths = [columns | rows] |> List.zip() |> Enum.map(&Tuple.to_list/1) |> Enum.map(fn column_with_rows -> column_with_rows |> Enum.map(&binary_length/1) |> Enum.max() end) [ separator(column_widths), "\n", cells(columns, column_widths), "\n", separator(column_widths), "\n", Enum.map(rows, &(cells(&1, column_widths) ++ ["\n"])), separator(column_widths) ] |> IO.iodata_to_binary() end # NULL defp binary_length(nil), do: 4 defp binary_length(binary) when is_binary(binary), do: String.length(binary) defp binary_length(other), do: other |> inspect() |> String.length() defp separator(widths) do Enum.map(widths, &[?+, ?-, String.duplicate("-", &1), ?-]) ++ [?+] end defp cells(items, widths) do cell = [items, widths] |> List.zip() |> Enum.map(fn {item, width} -> [?|, " ", format_item(item, width), " "] end) [cell | [?|]] end defp format_item(nil, width), do: String.pad_trailing("NULL", width) defp format_item(item, width) when is_binary(item), do: String.pad_trailing(item, width) defp format_item(item, width) when is_number(item), do: item |> inspect() |> String.pad_leading(width) defp format_item(item, width), do: item |> inspect() |> String.pad_trailing(width) ## Callbacks @doc false def __before_compile__(_driver, _env) do quote do @doc """ A convenience function for SQL-based repositories that executes the given query. See `Ecto.Adapters.SQL.query/4` for more information. """ def query(sql, params \\ [], opts \\ []) do Ecto.Adapters.SQL.query(get_dynamic_repo(), sql, params, opts) end @doc """ A convenience function for SQL-based repositories that executes the given query. See `Ecto.Adapters.SQL.query!/4` for more information. """ def query!(sql, params \\ [], opts \\ []) do Ecto.Adapters.SQL.query!(get_dynamic_repo(), sql, params, opts) end @doc """ A convenience function for SQL-based repositories that executes the given multi-result query. See `Ecto.Adapters.SQL.query_many/4` for more information. """ def query_many(sql, params \\ [], opts \\ []) do Ecto.Adapters.SQL.query_many(get_dynamic_repo(), sql, params, opts) end @doc """ A convenience function for SQL-based repositories that executes the given multi-result query. See `Ecto.Adapters.SQL.query_many!/4` for more information. """ def query_many!(sql, params \\ [], opts \\ []) do Ecto.Adapters.SQL.query_many!(get_dynamic_repo(), sql, params, opts) end @doc """ A convenience function for SQL-based repositories that translates the given query to SQL. See `Ecto.Adapters.SQL.to_sql/3` for more information. """ def to_sql(operation, queryable) do Ecto.Adapters.SQL.to_sql(operation, get_dynamic_repo(), queryable) end @doc """ A convenience function for SQL-based repositories that executes an EXPLAIN statement or similar depending on the adapter to obtain statistics for the given query. See `Ecto.Adapters.SQL.explain/4` for more information. """ def explain(operation, queryable, opts \\ []) do Ecto.Adapters.SQL.explain(get_dynamic_repo(), operation, queryable, opts) end @doc """ A convenience function for SQL-based repositories that forces all connections in the pool to disconnect within the given interval. See `Ecto.Adapters.SQL.disconnect_all/3` for more information. """ def disconnect_all(interval, opts \\ []) do Ecto.Adapters.SQL.disconnect_all(get_dynamic_repo(), interval, opts) end end end @doc false def ensure_all_started(driver, _config, type) do Application.ensure_all_started(driver, type) end @pool_opts [:timeout, :pool, :pool_size] ++ [:queue_target, :queue_interval, :ownership_timeout, :repo] @doc false def init(connection, driver, config) do unless Code.ensure_loaded?(connection) do raise """ could not find #{inspect(connection)}. Please verify you have added #{inspect(driver)} as a dependency: {#{inspect(driver)}, ">= 0.0.0"} And remember to recompile Ecto afterwards by cleaning the current build: mix deps.clean --build ecto """ end log = Keyword.get(config, :log, :debug) valid_log_levels = [ false, :debug, :info, :notice, :warning, :error, :critical, :alert, :emergency ] if log not in valid_log_levels do raise """ invalid value for :log option in Repo config The accepted values for the :log option are: #{Enum.map_join(valid_log_levels, ", ", &inspect/1)} See https://hexdocs.pm/ecto/Ecto.Repo.html for more information. """ end stacktrace = Keyword.get(config, :stacktrace, nil) telemetry_prefix = Keyword.fetch!(config, :telemetry_prefix) telemetry = {config[:repo], log, telemetry_prefix ++ [:query]} config = adapter_config(config) opts = Keyword.take(config, @pool_opts) meta = %{telemetry: telemetry, sql: connection, stacktrace: stacktrace, opts: opts} {:ok, connection.child_spec(config), meta} end defp adapter_config(config) do if Keyword.has_key?(config, :pool_timeout) do message = """ :pool_timeout option no longer has an effect and has been replaced with an improved queuing system. See \"Queue config\" in DBConnection.start_link/2 documentation for more information. """ IO.warn(message) end config |> Keyword.delete(:name) |> Keyword.update(:pool, DBConnection.ConnectionPool, &normalize_pool/1) end defp normalize_pool(pool) do if Code.ensure_loaded?(pool) && function_exported?(pool, :unboxed_run, 2) do DBConnection.Ownership else pool end end @doc false def checkout(adapter_meta, opts, callback) do checkout_or_transaction(:run, adapter_meta, opts, callback) end @doc false def checked_out?(adapter_meta) do %{pid: pool} = adapter_meta get_conn(pool) != nil end ## Query @doc false def insert_all( adapter_meta, schema_meta, conn, header, rows, on_conflict, returning, placeholders, opts ) do %{source: source, prefix: prefix} = schema_meta {_, conflict_params, _} = on_conflict {rows, params} = case rows do {%Ecto.Query{} = query, params} -> {query, Enum.reverse(params)} rows -> unzip_inserts(header, rows) end sql = conn.insert(prefix, source, header, rows, on_conflict, returning, placeholders) opts = if is_nil(Keyword.get(opts, :cache_statement)) do [{:cache_statement, "ecto_insert_all_#{source}"} | opts] else opts end all_params = placeholders ++ Enum.reverse(params, conflict_params) %{num_rows: num, rows: rows} = query!(adapter_meta, sql, all_params, [source: source] ++ opts) {num, rows} end defp unzip_inserts(header, rows) do Enum.map_reduce(rows, [], fn fields, params -> Enum.map_reduce(header, params, fn key, acc -> case :lists.keyfind(key, 1, fields) do {^key, {%Ecto.Query{} = query, query_params}} -> {{query, length(query_params)}, Enum.reverse(query_params, acc)} {^key, {:placeholder, placeholder_index}} -> {{:placeholder, Integer.to_string(placeholder_index)}, acc} {^key, value} -> {key, [value | acc]} false -> {nil, acc} end end) end) end @doc false def execute(prepare, adapter_meta, query_meta, prepared, params, opts) do %{num_rows: num, rows: rows} = execute!(prepare, adapter_meta, prepared, params, put_source(opts, query_meta)) {num, rows} end defp execute!(prepare, adapter_meta, {:cache, update, {id, prepared}}, params, opts) do name = prepare_name(prepare, id) case sql_call(adapter_meta, :prepare_execute, [name, prepared], params, opts) do {:ok, query, result} -> maybe_update_cache(prepare, update, {id, query}) result {:error, err} -> raise_sql_call_error(err) end end defp execute!( :unnamed = prepare, adapter_meta, {:cached, _update, _reset, {id, cached}}, params, opts ) do name = prepare_name(prepare, id) prepared = String.Chars.to_string(cached) case sql_call(adapter_meta, :prepare_execute, [name, prepared], params, opts) do {:ok, _query, result} -> result {:error, err} -> raise_sql_call_error(err) end end defp execute!( :named = _prepare, adapter_meta, {:cached, update, reset, {id, cached}}, params, opts ) do case sql_call(adapter_meta, :execute, [cached], params, opts) do {:ok, query, result} -> update.({id, query}) result {:ok, result} -> result {:error, err} -> raise_sql_call_error(err) {:reset, err} -> reset.({id, String.Chars.to_string(cached)}) raise_sql_call_error(err) end end defp execute!(_prepare, adapter_meta, {:nocache, {_id, prepared}}, params, opts) do case sql_call(adapter_meta, :query, [prepared], params, opts) do {:ok, res} -> res {:error, err} -> raise_sql_call_error(err) end end defp prepare_name(:named, id), do: "ecto_" <> Integer.to_string(id) defp prepare_name(:unnamed, _id), do: "" defp maybe_update_cache(:named = _prepare, update, value), do: update.(value) defp maybe_update_cache(:unnamed = _prepare, _update, _value), do: :noop @doc false def stream(adapter_meta, query_meta, prepared, params, opts) do do_stream(adapter_meta, prepared, params, put_source(opts, query_meta)) end defp do_stream(adapter_meta, {:cache, _, {_, prepared}}, params, opts) do prepare_stream(adapter_meta, prepared, params, opts) end defp do_stream(adapter_meta, {:cached, _, _, {_, cached}}, params, opts) do prepare_stream(adapter_meta, String.Chars.to_string(cached), params, opts) end defp do_stream(adapter_meta, {:nocache, {_id, prepared}}, params, opts) do prepare_stream(adapter_meta, prepared, params, opts) end defp prepare_stream(adapter_meta, prepared, params, opts) do adapter_meta |> Ecto.Adapters.SQL.Stream.build(prepared, params, opts) |> Stream.map(fn %{num_rows: nrows, rows: rows} -> {nrows, rows} end) end defp raise_sql_call_error(%DBConnection.OwnershipError{} = err) do message = err.message <> "\nSee Ecto.Adapters.SQL.Sandbox docs for more information." raise %{err | message: message} end defp raise_sql_call_error(err), do: raise(err) @doc false def reduce(adapter_meta, statement, params, opts, acc, fun) do %{pid: pool, telemetry: telemetry, sql: sql, opts: default_opts} = adapter_meta opts = with_log(telemetry, params, opts ++ default_opts) case get_conn(pool) do %DBConnection{conn_mode: :transaction} = conn -> sql |> apply(:stream, [conn, statement, params, opts]) |> Enumerable.reduce(acc, fun) _ -> raise "cannot reduce stream outside of transaction" end end @doc false def into(adapter_meta, statement, params, opts) do %{pid: pool, telemetry: telemetry, sql: sql, opts: default_opts} = adapter_meta opts = with_log(telemetry, params, opts ++ default_opts) case get_conn(pool) do %DBConnection{conn_mode: :transaction} = conn -> sql |> apply(:stream, [conn, statement, params, opts]) |> Collectable.into() _ -> raise "cannot collect into stream outside of transaction" end end @doc false def struct( adapter_meta, conn, sql, operation, source, params, values, on_conflict, returning, opts ) do opts = if is_nil(Keyword.get(opts, :cache_statement)) do [{:cache_statement, "ecto_#{operation}_#{source}_#{length(params)}"} | opts] else opts end case query(adapter_meta, sql, values, [source: source] ++ opts) do {:ok, %{rows: nil, num_rows: 1}} -> {:ok, []} {:ok, %{rows: [values], num_rows: 1}} -> {:ok, Enum.zip(returning, values)} {:ok, %{num_rows: 0}} -> if on_conflict == :nothing, do: {:ok, []}, else: {:error, :stale} {:ok, %{num_rows: num_rows}} when num_rows > 1 -> raise Ecto.MultiplePrimaryKeyError, source: source, params: params, count: num_rows, operation: operation {:error, err} -> case conn.to_constraints(err, source: source) do [] -> raise_sql_call_error(err) constraints -> {:invalid, constraints} end end end ## Transactions @doc false def transaction(adapter_meta, opts, callback) do checkout_or_transaction(:transaction, adapter_meta, opts, callback) end @doc false def in_transaction?(%{pid: pool}) do match?(%DBConnection{conn_mode: :transaction}, get_conn(pool)) end @doc false def rollback(%{pid: pool}, value) do case get_conn(pool) do %DBConnection{conn_mode: :transaction} = conn -> DBConnection.rollback(conn, value) _ -> raise "cannot call rollback outside of transaction" end end ## Migrations @doc false def execute_ddl(meta, conn, definition, opts) do ddl_logs = definition |> conn.execute_ddl() |> List.wrap() |> Enum.map(&query!(meta, &1, [], opts)) |> Enum.flat_map(&conn.ddl_logs/1) {:ok, ddl_logs} end @doc false def raise_migration_pool_size_error do raise Ecto.MigrationError, """ Migrations failed to run because the connection pool size is less than 2. Ecto requires a pool size of at least 2 to support concurrent migrators. When migrations run, Ecto uses one connection to maintain a lock and another to run migrations. If you are running migrations with Mix, you can increase the number of connections via the pool size option: mix ecto.migrate --pool-size 2 If you are running the Ecto.Migrator programmatically, you can configure the pool size via your application config: config :my_app, Repo, ..., pool_size: 2 # at least """ end ## Log defp with_log(telemetry, params, opts) do [log: &log(telemetry, params, &1, opts)] ++ opts end defp log({repo, log, event_name}, params, entry, opts) do %{ connection_time: query_time, decode_time: decode_time, pool_time: queue_time, idle_time: idle_time, result: result, query: query } = entry source = Keyword.get(opts, :source) query = String.Chars.to_string(query) result = with {:ok, _query, res} <- result, do: {:ok, res} stacktrace = Keyword.get(opts, :stacktrace) log_params = opts[:cast_params] || params acc = if idle_time, do: [idle_time: idle_time], else: [] measurements = log_measurements( [query_time: query_time, decode_time: decode_time, queue_time: queue_time], 0, acc ) metadata = %{ type: :ecto_sql_query, repo: repo, result: result, params: params, cast_params: opts[:cast_params], query: query, source: source, stacktrace: stacktrace, options: Keyword.get(opts, :telemetry_options, []) } if event_name = Keyword.get(opts, :telemetry_event, event_name) do :telemetry.execute(event_name, measurements, metadata) end case {opts[:log], log} do {false, _level} -> :ok {opts_level, false} when opts_level in [nil, true] -> :ok {true, level} -> Logger.log( level, fn -> log_iodata(measurements, repo, source, query, log_params, result, stacktrace) end, ansi_color: sql_color(query) ) {opts_level, args_level} -> Logger.log( opts_level || args_level, fn -> log_iodata(measurements, repo, source, query, log_params, result, stacktrace) end, ansi_color: sql_color(query) ) end :ok end defp log_measurements([{_, nil} | rest], total, acc), do: log_measurements(rest, total, acc) defp log_measurements([{key, value} | rest], total, acc), do: log_measurements(rest, total + value, [{key, value} | acc]) defp log_measurements([], total, acc), do: Map.new([total_time: total] ++ acc) defp log_iodata(measurements, repo, source, query, params, result, stacktrace) do [ "QUERY", ?\s, log_ok_error(result), log_ok_source(source), log_time("db", measurements, :query_time, true), log_time("decode", measurements, :decode_time, false), log_time("queue", measurements, :queue_time, false), log_time("idle", measurements, :idle_time, true), ?\n, query, ?\s, inspect(params, charlists: false), log_stacktrace(stacktrace, repo) ] end defp log_ok_error({:ok, _res}), do: "OK" defp log_ok_error({:error, _err}), do: "ERROR" defp log_ok_source(nil), do: "" defp log_ok_source(source), do: " source=#{inspect(source)}" defp log_time(label, measurements, key, force) do case measurements do %{^key => time} -> us = System.convert_time_unit(time, :native, :microsecond) ms = div(us, 100) / 10 if force or ms > 0 do [?\s, label, ?=, :io_lib_format.fwrite_g(ms), ?m, ?s] else [] end %{} -> [] end end defp log_stacktrace(stacktrace, repo) do with [_ | _] <- stacktrace, {module, function, arity, info} <- last_non_ecto(Enum.reverse(stacktrace), repo, nil) do [ ?\n, IO.ANSI.light_black(), "↳ ", Exception.format_mfa(module, function, arity), log_stacktrace_info(info), IO.ANSI.reset() ] else _ -> [] end end defp log_stacktrace_info([file: file, line: line] ++ _) do [", at: ", file, ?:, Integer.to_string(line)] end defp log_stacktrace_info(_) do [] end @repo_modules [Ecto.Repo.Queryable, Ecto.Repo.Schema, Ecto.Repo.Transaction] defp last_non_ecto([{mod, _, _, _} | _stacktrace], repo, last) when mod == repo or mod in @repo_modules, do: last defp last_non_ecto([last | stacktrace], repo, _last), do: last_non_ecto(stacktrace, repo, last) defp last_non_ecto([], _repo, last), do: last ## Connection helpers defp checkout_or_transaction(fun, adapter_meta, opts, callback) do %{pid: pool, telemetry: telemetry, opts: default_opts} = adapter_meta opts = with_log(telemetry, [], opts ++ default_opts) callback = fn conn -> previous_conn = put_conn(pool, conn) try do callback.() after reset_conn(pool, previous_conn) end end apply(DBConnection, fun, [get_conn_or_pool(pool), callback, opts]) end defp get_conn_or_pool(pool) do Process.get(key(pool), pool) end defp get_conn(pool) do Process.get(key(pool)) end defp put_conn(pool, conn) do Process.put(key(pool), conn) end defp reset_conn(pool, conn) do if conn do put_conn(pool, conn) else Process.delete(key(pool)) end end defp key(pool), do: {__MODULE__, pool} defp sql_color("SELECT" <> _), do: :cyan defp sql_color("ROLLBACK" <> _), do: :red defp sql_color("LOCK" <> _), do: :white defp sql_color("INSERT" <> _), do: :green defp sql_color("UPDATE" <> _), do: :yellow defp sql_color("DELETE" <> _), do: :red defp sql_color("begin" <> _), do: :magenta defp sql_color("commit" <> _), do: :magenta defp sql_color(_), do: nil end