Robinson Crusoe's parsing library
Python
Switch branches/tags
Nothing to show
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
examples
extras
.gitignore
CHANGES
LICENSE
MANIFEST.in
README.rst
peglet.py
setup.py

README.rst

Peglet

Peglet extends Python's regular expressions to handle recursive grammars. For example, to parse a tiny subset of HTML:

>>> from peglet import Parser
>>> a_little_html = Parser(r"""parts = part parts |
...                            part  = <(\w+)> parts </\w+> group
...                                  | ([^<]+)
...                        """, group=lambda *values: values)
>>> a_little_html("Hello. <p><em>Nesting</em> for <i>the win</i>.</p>")
('Hello. ', ('p', ('em', 'Nesting'), ' for ', ('i', 'the win'), '.'))

The goal was to make a parsing library

  1. pleasant enough to use;
  2. simple to adapt or rewrite from scratch if I'm faced with some new situation like a new programming language;
  3. with code easy enough to follow that it could introduce people to parsing.

So it came down to one page of clear code not using combinators. (And then ballooned to 200+ lines from documentation and a few extras.) Some bits that couldn't fit the latter two constraints went into a combinator library, parson.

For more, see the examples or the module doc in the code. Other programs using this include bicicleta.py and .js, a logic puzzle solver, and the lexer for unitcalc.

Installing it

pip install peglet, or else download then python setup.py install.