Fun with Pseudocode

Fun with Pseudocode (or fwp for short) is a Python utility to parse "pseudocode" with syntax based on the Cambridge International A & AS Level Computer Science Pseudocode Guide for Teachers.

This project is a fun exercise. It has no intended practical application. It is also largely unfinished as of now; not all language features have been implemented or rigorously tested.

Usage

$ python -m fwp --help

Requires Lark to be installed.

Sample pseudocode programs are provided under the examples directory.

The Process

                   +--------------+
                   | EBNF grammar |
                   +---|----------+
                       |
+-------------------+  V   +------------+     +----+
| Pseudocode source | ---> | Parse tree | --> | IR | ---> ???
+-------------------+      +------------+     +----+

1. Parsing: the pseudocode source is first run through the parser to generate a parse tree. Lark is used as the parser generator, and supplied with the EBNF definition of the pseudocode language (see fwp/grammar.lark). The parser performs lexical analysis (tokenization) and parsing of the pseudocode source, resulting in a parse tree. Pass the -P option to stop at the parsing stage and view the parse tree.

2. Intermediate representation (IR) generation: the IR is implemented as a hierarchy of Python objects that comprise a tree structure. It is essentially a parse tree with some semantics added. However, unlike most other IRs there is no actual text or binary format for which to store such generated IR, which exists purely as a Python object. To view the IR as a nicely formatted S-expression, pass the -S option.

3. Now all that's left is to do something useful with the IR. This has yet to be decided upon.

License

Fun with Pseudocode is licensed under the AGPLv3. See the LICENSE file for more information.