Skip to content

README.md

Latest commit

 

History

History
64 lines (45 loc) · 2.48 KB

README.md

File metadata and controls

64 lines (45 loc) · 2.48 KB

py2blocks

A Python module, to be run in a worker using Pyodide (CPython) that takes arbitrary Python code and converts it to a block-based representation expressed as JSON conforming to the Blockly format.

We use the built-in Python AST module to parse the Python code and then traverse the resulting AST to generate the Blockly JSON.

For more information about the Blockly JSON serialization format, see:

https://developers.google.com/blockly/guides/configure/web/serialization

Developer Setup

  1. Create a virtual environment and activate it.
  2. Install local packages with: pip install -r requirements.txt
  3. Run the test suite via: make serve and visit localhost:8000
  4. While the local server is running, changes to the source code are automatically added to the archive used by the test framework. Simply refresh the browser to run the updated code.

That's it!

We expect folks to contribute and collaborate in the spirit of our Care of Community statement.

Available blocks

In the JSON representation of block, each block has a type whose name relates to the matching Python ast node type. Depending on the type of block, they may contain other attributes. The available blocks can be found in the src/blocks folder of this repository. Please see the README.md therein for more information.

Pre-defined vs User-defined functions

Sometimes, for ease of use, guidance and simplicity we will provide pre-defined blocks that represent built-in functionality. For example, the builtin print function in Python may have a simple, and equivalent, print block.

In addition, users may define their own functions via blocks, and will then have to call these functions via a special function_call block.

For this to work, when py2blocks encounters a function call the following algorithm is used:

  • Is the name of the function in the builtin collection of pre-defined blocks?
  • If True, use the provided template for that pre-defined block, to render the JSON output that constitutes the blockly object model.
  • If False, just create a function_call block with the relevant name and arguments slotted into place.
  • That's it!

We will provide mechanism for third parties to be able to pre-define blocks, in addition to the blocks we provide automatically.

Open Source Acknowledgements

The drag strategy plugin is derived from the Apache 2.0 licenced Makecode, a project of Microsoft.