Pycket is a Racket/Scheme implementation that is automatically generated using the RPython framework. Given an interpreter written in RPython (in our case a CEK machine interpreter for Racket), RPython framework produces a fast binary for it. It can also add a tracing JIT.

There are currently two different modes of evaluation that we refer as OLD and NEW. The NEW Pycket uses linklets and bootstraps the Racket using the expander linklet exported by Racket (version 7+). The OLD Pycket, on the other hand, uses Racket's binary to fully expand the program and generates json asts and evaluates them.

Note that both versions require an unmodified Racket installation. The OLD Pycket requires a Racket binary, and while the NEW Pycket doesn't require a Racket binary, it still requires the Racket packages and libraries to bootstrap.

See the Makefile targets section about how to build both versions.

• #pycket Slack channel under Racket
• #pycket channel on IRC freenode
• Pycket's benchmarks are available at this repository, along with instructions for running them.

In order to do anything with Pycket, you need to check out PyPy:

$ hg clone https://bitbucket.org/pypy/pypy

The below instructions assume that you do this checkout in Pycket's directory.

Additionally, it helps to have the build dependencies of PyPy installed. On a Debian or Ubuntu system:

$ sudo apt-get build-dep pypy

To produce an executable, run:

$ ./pypy/rpython/bin/rpython -Ojit targetpycket.py

This expects that a binary named pypy is in your path. (Note that a hand-compiled PyPy by running make produces pypy-c, not pypy).

If you don't have a PyPy binary, you can also translate with the CPython:

$ python ./pypy/rpython/bin/rpython -Ojit targetpycket.py

This will take upwards of 10 minutes.

You can also use make for any of the above,

• make setup to setup and update the pypy checkout and install pycket-lang to Racket
• make expander to generate the expander linklet (it assumes an unmodified Racket install and PLTHOME environment variable -- see the Environment Variables section below)
• make pycket-c to translate OLD Pycket with JIT
• make pycket-c-linklets to translate NEW Pycket with JIT
• make pycket-c-nojit to translate OLD Pycket without JIT (which may be a lot faster to translate but runs a lot lot slower)
• make pycket-c-linklets-nojit to translate NEW Pycket without JIT (which may be a lot faster to translate but runs a lot lot slower)
• make test to run the unit tests on OLD Pycket.
• make test-new-with-expander to run the unit tests on NEW Pycket using the Racket's expander linklet.
• make test-new-no-expander to run the unit tests on NEW Pycket without using the expander.

Now that Pycket has two different modes with options, we run the unit tests on each of those settings using the following targets:

• make test : to run the OLD Pycket on the tests
• make test-new-no-expander : to run NEW Pycket without using the expander linklet
• make test-new-with-expander : to run NEW Pycket using the expander linklet

For the NEW Pycket, using the expander linklet means that for each test expression string we use the read and eval functions in that linklet to read and evaluate the test. If we're not using the expander, on the other hand, then we manually create a linklet containing the expression and instantiate it directly (mostly with an empty target) to get the result.

The NEW Pycket is able to generate and use its own .zo files. For now both the generation and the use are manual.

To generate a .zo file for a .rkt source file, use make compile-file:

$ make compile-file FILE=$(PLTHOME)/racket/collects/racket/private/qq-and-or.rkt

The parameter that enables Racket expander to use compiled code is use-compiled-file-paths, which defaults to pycket-compiled in Pycket. Whenever a module is required, the expander will use the compiled code if it exists, otherwise it will use the source code of the module (read, expand, etc.).

pycket-repl> (#%require racket/private/qq-and-or)

Note that pycket-compiled is a folder that make compile-file is going to generate by itself.

Currently we have two make targets for working with compiled files:

$ make compile-racket-modules

will create .zo files for the predefined list of Racket modules (see compile-file-pycket.rkt).

$ make clean-compiled-files

will remove all the .zo files under the pycket-compiled directories in Racket libraries.

This is a work in progress. We plan to have a make target that compiles all the Racket modules automatically by following the module dependencies (as opposed to using a predefined list of modules with the respective paths).

Running the interpreter on CPython or PyPy (i.e. running the targetpycket.py) requires a PYTHONPATH that includes both RPython (that should be the pypy directory cloned above) and pycket (that should be this directory).

Also there are a couple of variables need to be set for the NEW Pycket to interact with the Racket, since it bootstraps Racket by reading and evaluating Racket's main collection by loading and using the bootstrap linklets (currently only the expander.rktl.linklet) exported by Racket. So the NEW Pycket needs to be able to locate various different parts of the Racket installation. The OLD Pycket is lucky to use the Racket's own binary.

Naturally, it varies on the way in which the Racket is installed:

Then all the NEW Pycket needs is a PLTHOME environment variable to point to the surrounding directory. For example it will assume the collects directory is at:

$PLTHOME/racket/collects

Then NEW Pycket needs to know the locations of various directories. In particular, it needs:

• PLTEXECFILE to point to the location of the racket binary * PLTCOLLECTS to point to the collects directory for the main collections
• PLTCONFIGDIR to point to Racket's etc directory that contains config.rktd
• (optional) PLTADDONDIR to point to a directory for user-specific Racket configuration, packages, and extensions. It defaults to .racket in USERHOME.
• (optional) PLTUSERHOME to point to the home directory of the user. It's optional since Pycket will also look at other environment variables to figure out the home directory (e.g. $HOME).

You can also use the command line options to provide these paths, e.g. -X, -G etc.. Run it with -h to see all the commands and options.

$ ./pycket-c-linklets -h

Also, the Makefile reacts to some variables:

• PYPYPATH for when your pypy checkout is not in this directory. Defaults to pypy.
• PYTEST for when you don’t want to use pypy’s version of pytest. Defaults to $(PYPYPATH)/pytest.py.
• RPYTHON for when you want to use something other than the default rpython script, but you probably would not want that. Defaults to $(PYPYPATH)/rpython/bin/rpython --batch.

Pycket currently defaults to the OLD Pycket. To use the NEW version with the linklets, run it with:

$ ./pycket-c-linklets <arguments>

You can run with the -h option to see the different command line options for each versions:

$ ./pycket-c -h

$ ./pycket-c-linklets -h

You can run pycket-c like the racket binary:

$ ./pycket-c program.rkt

You can also run pycket under plain python (or pypy if its available), like this:

$ ./pycket-slow.sh program

One of the beautiful perks of bootstrapping Racket is to be able to run on Pykcet some interesting programs implemented in Racket (as long as we have enough runtime support for it in Pycket -- i.e. having the runtime primitives implemented in RPython).

The NEW Pycket now features an original Racket REPL that's implemented in Racket. Try it out!

$ ./pycket-c-linklets

You can make it more verbose with the --verbose flag if you're curious about what's going on in the background.

$ ./pycket-c-linklets --verbose

Also increase the verbosity level (defaults to 0).

$ ./pycket-c-linklets --verbose 1

Or even this, when this directory is in your PYTHONPATH:

$ python -mpycket program

You can edit the shell script to make it use pypy, if desired.

You can generate a coverage report with pytest:

$ pypy/pytest.py --cov .

or via

$ make coverage

which also generates an HTML report in pycket/test/coverage_report. You need these Python packages for that to work:

• pytest-cov (provided with the pypy checkout)
• cov-core and coverage