⚠️ Note: This software is currently under active development. The API and interface should be considered unstable until a v1.0.0 release.
Tapioca is a library used to generate RBI (Ruby interface) files for use with Sorbet. RBI files provide the structure (classes, modules, methods, parameters) of the gem/library to Sorbet to assist with typechecking.
As yet, no gem exports type information in a consumable format and it would be a huge effort to manually maintain such an interface file for all the gems that your codebase depends on. Thus, there is a need for an automated way to generate the appropriate RBI file for a given gem. The tapioca
gem, developed at Shopify, is able to do exactly that to almost 99% accuracy. It can generate the definitions for all statically defined types and most of the runtime defined types exported from Ruby gems (non-Ruby gems are not handled yet).
When you run tapioca sync
in a project, tapioca
loads all the gems that are in your dependency list from the Gemfile into memory. It then performs runtime introspection on the loaded types to understand their structure and generates an appropriate RBI file for each gem with a versioned filename.
Tapioca helps simplify your setup too. Gems such as sorbet-typed
, sorbet-rails
are localized solutions that provide typing information for DSLs and gems. However, Tapioca aims to provide the complete tooling to generating RBIs for gems and DSLs. Tapioca does not require being combined with other gems, such as sorbet-rails
, which can be removed from your gemfile.
Please see the detailed answer on our wiki
If you are not yet using Sorbet then skip straight to step 2
The RBI files that Tapioca generates should not be added on top of those that Sorbet has already generated for you. Tapioca needs to run afresh. Tapioca also supercedes the need for gems like sorbet-rails
or sorbet-typed
. In order to start using Tapioca you will need to remove these gems and regenerate all of your automatically generated RBI files.
The easiest way to remove existing RBI files is by removing the entire sorbet/rbi
folder.
Note: if your application has shims (hand written RBI files), keep them saved somewhere in case you need to restore some of them. Chances are, fewer shims will be needed after transitioning to Tapioca. Additionally, don't forget to keep your sorbet/config
file too.
Remove sorbet-rails
or other such alternative solutions from the Gemfile
. As mentioned, Tapioca is meant to be used on its own.
If the previously used alternatives had their own configuration files, you should remove them as well.
Start here if you have a fresh sorbet installation.
group :development do
gem 'tapioca', require: false
end
Then run bundle install
# Generate sorbet/config and sorbet/tapioca/require.rb if required
# (they won't be touched if they already exist)
bundle exec tapioca init
# Generate RBI files for gems
# (Generate only for specific gems using `tapioca gem gem...]`)
bundle exec tapioca gem
# Generate the list of all unresolved constants
# This will generate the file `sorbet/rbi/todo.rbi`
# defining all unresolved constants as empty modules
bundle exec tapioca todo
The first step may result in some typechecking errors due to missing definitions. Let's pick them up and dela with them:
# Check whether any gems are missed by running typechecking
bundle exec srb tc
# If you notice missing definitions for gems, nudge Tapioca to
# look for them by running:
bin/tapioca require
# If this doesn't add the requisite gems, add them manually to:
# sorbet/tapioca/require.rb
# repeat as necessary
Still having problems with gems not being included? Check the section on manual gem requires.
Methods such as Rails' belongs_to
will dynamically generate a set of helper-methods
at runtime. Left to its own devices, Sorbet won't see them and will fall over. However Tapioca can figure
out and generate their RBIs:
# Generate runtime definitions
bin/tapioca dsl
While generating RBI files you might realize that some shims have become obsolete or incorrect. Make sure to edit or remove existing shims that are causing errors. You can read about DSL RBI generators supplied by tapioca
in the manual.
(Note - there are still a few rough patches around this WRT Rails - please see this interim fix)
(If there are still errors for missing definitions, you may want to try bringing back any shims you saved for later)
# Skip any remaining errors by generate a file with
# empty definitions (to be fixed later on)
bin/tapioca todo
After generating the RBIs, some files might actually be ready to move from typed: false
to typed: true
. If doing so produces no erorrs then Spoom can automatically bump files to true
.
# Automatically increase files to the highest level of type-strictness
# they will now support
bundle exec spoom bump
Here are some things that do not need to be run for the migration (or ever in some cases).
bin/tapioca sync
: this command is meant to be used after the initial setup is complete to update RBIs for gems that have been upgraded. It does not need to be used as part of the migration process, but should be used laterbundle exec srb rbi hidden-definitions
: not necessary at all, everbundle exec srb rbi suggest-typed
: not necessary at all, ever. Preferbundle exec spoom bump
Done! Your application should be all set and type checking should pass.
Do not forget to execute tapioca
using bundler
:
$ bundle exec tapioca
Commands:
tapioca --version, -v # show version
tapioca dsl [constant...] # generate RBIs for dynamic methods
tapioca generate [gem...] # generate RBIs from gems
tapioca help [COMMAND] # Describe available commands or one specific command
tapioca init # initializes folder structure
tapioca require # generate the list of files to be required by tapioca
tapioca sync # sync RBIs to Gemfile
tapioca todo # generate the list of unresolved constants
Options:
--pre, -b, [--prerequire=file] # A file to be required before Bundler.require is called
--post, -a, [--postrequire=file] # A file to be required after Bundler.require is called
--out, -o, [--outdir=directory] # The output directory for generated RBI files
--cmd, -c, [--generate-command=command] # The command to run to regenerate RBI files
-x, [--exclude=gem [gem ...]] # Excludes the given gem(s) from RBI generation
--typed, -t, [--typed-overrides=gem:level [gem:level ...]] # Overrides for typed sigils for generated gem RBIs
--prerequire [file]
: A file to be required beforeBundler.require
is called.--postrequire [file]
: A file to be required afterBundler.require
is called.--out [directory]
: The output directory for generated RBI files, default tosorbet/rbi/gems
.--generate-command [command]
: [DEPRECATED] The command to run to regenerate RBI files (used in header comment of the RBI files), defaults to the current command.--typed-overrides [gem:level]
: Overrides typed sigils for generated gem RBIs for gemgem
to levellevel
(level
can be one ofignore
,false
,true
,strict
, orstrong
, see the Sorbet docs for more details).
For gems that have a normal default require
and load all of their constants through such a require, everything works seamlessly. However, for gems that are marked as require: false
in the Gemfile, or for gems that export optionally loaded types via different requires, where a single require does not load the whole gem code into memory, tapioca
will not be able to load some of the types into memory and, thus, won't be able to generate complete RBIs for them. For this reason, we need to keep a small external file named sorbet/tapioca/require.rb
that is executed after all the gems in the Gemfile have been required and before generation of gem RBIs have started. This file is responsible for adding the requires for additional files from gems, which are not covered by the default require.
For example, suppose you are using the class BetterHtml::Parser
exported from the better_html
gem. Just doing a require "better_html"
(which is the default require) does not load that type:
$ bundle exec pry
[1] pry(main)> require 'better_html'
=> true
[2] pry(main)> BetterHtml
=> BetterHtml
[3] pry(main)> BetterHtml::Parser
NameError: uninitialized constant BetterHtml::Parser
from (pry):3:in `__pry__`
[4] pry(main)> require 'better_html/parser'
=> true
[5] pry(main)> BetterHtml::Parser
=> BetterHtml::Parser
In order to make sure that tapioca
can reflect on that type, we need to add the line require "better_html/parser"
to the sorbet/tapioca/require.rb
file. This will make sure BetterHtml::Parser
is loaded into memory and a type annotation is generated for it in the better_html.rbi
file. If this extra require
line is not added to sorbet/tapioca/require.rb
file, then the definition for that type will be missing from the RBI file.
If you ever run into a case, where you add a gem or update the version of a gem and run tapioca sync
but don't have some types you expect in the generated gem RBI files, you will need to make sure you have added the necessary requires to the sorbet/tapioca/require.rb
file.
You can use the command tapioca require
to auto-populate the sorbet/tapioca/require.rb
file with all the requires found
in your application. Once the file generated, you should review it, remove all unnecessary requires and commit it.
Bug reports and pull requests are welcome on GitHub at https://github.com/Shopify/tapioca. This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the Contributor Covenant code of conduct.
The gem is available as open source under the terms of the MIT License.