pPEGpy 0.3.18

pPEG -- portable PEG in Python

pPEGpy

This is an implementation of a portable PEG parser in Python.

For documentation see pPEG, the portable PEG project.

The pPEGpy package can be installed from PyPi with: Woops, lost my PyPi access credentials, so can't update PyPi yet... See Notes below for how to clone and install this repo.

> pip install pPEGpy                  # version 3.18+  but not yet!

Note the spelling of pPEGpy, there are unrelated packages with similar names.

For other ways to use the pPEGpy grammar-parser see the Package Notes below.

Examples

from pPEGpy import peg

print("Hello world!")

greet = peg.compile("""
    greet = _ hail _ whom _
    hail  = 'Hello' / 'Hi'
    whom  = 'you' / 'world!'
    _     = [ \t\n\r]*
    """,
    transforms = {"greet": dict}
)

ok, words = greet.read("Hello world!")

print(words) # => {'hail': 'Hello', 'whom': 'world!'}

Run this example, then edit to see what works or fails.

Comment out the transforms on line 11 to see the parse tree printed out as JSON.

More examples:

from pPEGpy import peg

# Equivalent to the regular expression for
# well-formed URI's in RFC 3986.

uri = peg.compile("""
    URI     = (scheme ':')? ('//' auth)? 
               path ('?' query)? ('#' frag)?
    scheme  = ~[:/?#]+
    auth    = ~[/?#]*
    path    = ~[?#]*
    query   = ~'#'*
    frag    = ~[ \t\n\r]*
    """,
    transforms = {'URI': dict}
)

test = "http://www.ics.uci.edu/pub/ietf/uri/#Related"

ok, data = uri.read(test)

print(data)

# => {'scheme': 'http', 'auth': 'www.ics.uci.edu',
#     'path': '/pub/ietf/uri/', 'frag': 'Related'}

import pPEGpy as peg

print("CSV example....")

csv = peg.compile("""
    CSV     = Row+
    Row     = field (',' field)* _nl
    field   = _string / _text

    _text   = ~[,\n\r]*
    _string = '"' (~["] / '""')* '"'
    _nl     = '\n' / '\r' '\n'?
    """,
    transforms = {
        'CSV':list, 'Row':list, 'field':str
    }
)

test = """A,B,C
a1,b1,c1
a2,"b,2",c2
a3,b3,c3
"""

ok, data = csv.read(test)

print(data)

# [['A', 'B', 'C'],
#  ['a1', 'b1', 'c1'],
#  ['a2', '"b,2"', 'c2'],
#  ['a3', 'b3', 'c3']]

# -- parse tree --------

p = csv.parse(test);

print(p)

# CSV
# │ Row
# │ │ field 'A'
# │ │ field 'B'
# │ │ field 'C'
# │ Row
# │ │ field 'a1'
# │ │ field 'b1'
# │ │ field 'c1'
# │ Row
# │ │ field 'a2'
# │ │ field '"b,2"'
# │ │ field 'c2'
# │ Row
# │ │ field 'a3'
# │ │ field 'b3'
# │ │ field 'c3'

Package Notes

To experiment you can clone the GitHub repository pPEGpy.

These command lines can be used to build a local package:

> cd <your pPEGpy directory>
> uv init --lib

> uv build
> pip install -e .

> python3 examples/date.py
> ...

The -e option allows local editing of the local files.

The repo includes an examples/ folder, try running the date.py for example.

Bare File

The peg.py file in: pPEGpy/src/pPEGy/peg.py is the only file you really need.

If you put a copy of this file into a folder together with your own programs you can import the grammar-parser directly with import peg. Very simple and easy.

But that does not work across directories, to import the bare peg.py file from another directory requires a hack like this:

import sys
sys.path.insert(1, <path to your copy of peg.py>)
import peg

To avoid that (at the cost of all the Python packaging complications!) you can build a package pPEGpy and install it with pip, as above.

Release Notes version 3.18

• transforms in now in peg.compile(..), and not parse.transform(..) as before

• Transforms use {'x': fx, 'y:': fy, ...}

• parse.tree always constructed (pruned with faults ignored)

• PEGpy-29.py 2026-05-29 from fossil/mycode/

TODO

• extras.py need to be re-written to use trace directly
• add extras, and transforms API

Future plans

• source code map
• binary (not Unicode) and/or compile(..., ASCII=True)??

---