defmodule Mix.Tasks.Ecto.Migrate do use Mix.Task import Mix.Ecto import Mix.EctoSQL @shortdoc "Runs the repository migrations" @aliases [ n: :step, r: :repo ] @switches [ all: :boolean, step: :integer, to: :integer, quiet: :boolean, prefix: :string, pool_size: :integer, log_sql: :boolean, strict_version_order: :boolean, repo: [:keep, :string], no_compile: :boolean, no_deps_check: :boolean, migrations_path: :keep ] @moduledoc """ Runs the pending migrations for the given repository. Migrations are expected at "priv/YOUR_REPO/migrations" directory of the current application, where "YOUR_REPO" is the last segment in your repository name. For example, the repository `MyApp.Repo` will use "priv/repo/migrations". The repository `Whatever.MyRepo` will use "priv/my_repo/migrations". You can configure a repository to use another directory by specifying the `:priv` key under the repository configuration. The "migrations" part will be automatically appended to it. For instance, to use "priv/custom_repo/migrations": config :my_app, MyApp.Repo, priv: "priv/custom_repo" This task runs all pending migrations by default. To migrate up to a specific version number, supply `--to version_number`. To migrate a specific number of times, use `--step n`. The repositories to migrate are the ones specified under the `:ecto_repos` option in the current app configuration. However, if the `-r` option is given, it replaces the `:ecto_repos` config. Since Ecto tasks can only be executed once, if you need to migrate multiple repositories, set `:ecto_repos` accordingly or pass the `-r` flag multiple times. If a repository has not yet been started, one will be started outside your application supervision tree and shutdown afterwards. ## Examples mix ecto.migrate mix ecto.migrate -r Custom.Repo mix ecto.migrate -n 3 mix ecto.migrate --step 3 mix ecto.migrate --to 20080906120000 ## Command line options * `-r`, `--repo` - the repo to migrate * `--all` - run all pending migrations * `--step` / `-n` - run n number of pending migrations * `--to` - run all migrations up to and including version * `--quiet` - do not log migration commands * `--prefix` - the prefix to run migrations on * `--pool-size` - the pool size if the repository is started only for the task (defaults to 2) * `--log-sql` - log the raw sql migrations are running * `--strict-version-order` - abort when applying a migration with old timestamp * `--no-compile` - does not compile applications before migrating * `--no-deps-check` - does not check depedendencies before migrating * `--migrations-path` - the path to load the migrations from, defaults to `"priv/repo/migrations"`. This option may be given multiple times in which case the migrations are loaded from all the given directories and sorted as if they were in the same one. Note, if you have migrations paths e.g. `a/` and `b/`, and run `mix ecto.migrate --migrations-path a/`, the latest migrations from `a/` will be run (even if `b/` contains the overall latest migrations.) """ @impl true def run(args, migrator \\ &Ecto.Migrator.run/4) do repos = parse_repo(args) {opts, _} = OptionParser.parse! args, strict: @switches, aliases: @aliases opts = if opts[:to] || opts[:step] || opts[:all], do: opts, else: Keyword.put(opts, :all, true) opts = if opts[:quiet], do: Keyword.merge(opts, [log: false, log_sql: false]), else: opts # Start ecto_sql explicitly before as we don't need # to restart those apps if migrated. {:ok, _} = Application.ensure_all_started(:ecto_sql) for repo <- repos do ensure_repo(repo, args) paths = ensure_migrations_paths(repo, opts) pool = repo.config[:pool] fun = if Code.ensure_loaded?(pool) and function_exported?(pool, :unboxed_run, 2) do &pool.unboxed_run(&1, fn -> migrator.(&1, paths, :up, opts) end) else &migrator.(&1, paths, :up, opts) end case Ecto.Migrator.with_repo(repo, fun, [mode: :temporary] ++ opts) do {:ok, _migrated, _apps} -> :ok {:error, error} -> Mix.raise "Could not start repo #{inspect repo}, error: #{inspect error}" end end :ok end end