pyparsec

python parser combinators (parsec) library (under development)

Examples

In [2]: from pyparsec import *

In [3]: aparser = char("a")

In [4]: bparser = char("b")

In [5]: aparser("abc")
Out[5]: (Right ('a', 'bc'))

In [6]: bparser("abc")
Out[6]: <pyparsec.pyparsec.Left at 0x7f5fdd6746a0>

In [7]: bparser("bca")
Out[7]: (Right ('b', 'ca'))

so to create a parser for 1 character you use char and to run parser you call it with input string

In [8]: aparser.parse("abc")
Out[8]: (Right ('a', 'bc'))

In [9]: aparser("abc")
Out[9]: (Right ('a', 'bc'))

In [10]: run_parser(aparser, "abc")
Out[10]: (Right ('a', 'bc'))

using __call__ is the same as calling .parse or calling run_parser directly

you can combine parsers using >>

In [12]: abparser = aparser >> bparser

In [13]: abparser.parse("abc")
Out[13]: (Right (['a', 'b'], 'c'))

In [14]: abparser.parse("bca")
Out[14]: <pyparsec.pyparsec.Left at 0x7f5fdd58ab38>

aparser >> bparser is read aparser is followed by bparser

How about Or? can we do ors?

In [15]: aorb = aparser | bparser

In [16]: aorb.parse("adef")
Out[16]: (Right ('a', 'def'))

In [17]: aorb.parse("bdef")
Out[17]: (Right ('b', 'def'))

Parsing greetings

In [2]: from pyparsec import *

In [3]: sentp = letters >> char(",") >> many(whitespace) >> letters >> optionally(char("!"))

In [4]: sent = "Hello, World!"

In [5]: sentp.parse(sent)
Out[5]: (Right (['Hello', ',', ' ', 'World', '!'], ''))

transform!

In [8]: sentp = letters.map(lambda l: "".join(l).upper()) >> char(",") >> many(whitespace) >> letters >> optionally(char("!"))

In [9]: sentp.parse(sent)
Out[9]: (Right (['HELLO', ',', ' ', 'World', '!'], ''))

you can map a function to be applied on a list parsed by a parser.

suppress

you can also suppress certain parsed values

In [12]: sentp = letters.map(lambda l: "".join(l).upper()) >> char(",").suppress() >> many(whitespace).suppress() >> letters >> optionally(char("!")).suppress()

In [13]: sentp.parse(sent)
Out[13]: (Right (['HELLO', 'World'], ''))

In [17]: lowerparser = anyOf(string.ascii_lowercase)
In [18]: singlequote = char('"')

In [20]: quotedword = between(singlequote, many1(lowerparser), singlequote)
    ...: print(quotedword('"abc"'))

(Right (['"', 'abc', '"'], ''))

In [23]: digitparser = anyOf(string.digits)
    ...:

In [24]: adigitlist = sepBy(whitespace.suppress(), digitparser)
    ...:

In [25]: print(run_parser(adigitlist, '1 2 3 4 5'))
    ...:
(Right (['1', '2', '3', '4', '5'], ''))

forward declarations

In [4]: valp = forward(lambda:  digits | listp)
   ...: listp = char("[") >> sepBy(char(",").suppress(),many(valp)) >> char("]")
   ...:
   ...: print(run_parser(valp, "4"))
   ...: print(valp("[8,6,7]"))
   ...: print(valp("[1,2,[1,2]]"))
   ...:
(Right ('4', ''))
(Right (['[', '8', '6', '7', ']'], ''))
(Right (['[', '1', '2', '[', '1', '2', ']', ']'], ''))

builtins

In [1]: from pyparsec import *

In [2]: letter("abc")
Out[2]: (Right ('a', 'bc'))

In [3]: letters("abc")
Out[3]: (Right ('abc', ''))

In [4]: word("abc")
Out[4]: (Right ('abc', ''))

In [5]: word("123")
Out[5]: <pyparsec.pyparsec.Left at 0x7fea1904c048>

In [6]: digits("123")
Out[6]: (Right ('123', ''))

In [7]: digits_as_int("123")
Out[7]: (Right (123, ''))

In [8]: quotedword('"Hello, World"')
Out[8]: <pyparsec.pyparsec.Left at 0x7fea1904f7b8>

In [9]: quotedword('"Hello"')
Out[9]: (Right (['"', 'Hello', '"'], ''))

untils

In [2]: ifstmt = chars("IF") >> many1( word | whitespace) << until(char(":"))

In [4]: ifstmt.parse("IF x is ok:")
Out[4]: (Right (['IF', ' x is ok'], ':'))

In [5]: ifstmt.parse("IF x is ok:")
Out[5]: (Right (['IF', ' x is ok'], ':'))

In [6]: ifstmt = chars("IF") >> many1( word | whitespace) << until(char(":")) >> char(":")

In [7]: ifstmt.parse("IF x is ok:")
Out[7]: (Right (['IF', ' x is ok', ':'], ''))

parsing uuids

In [1]: from pyparsec import *

In [2]: hexstr = any_of(string.hexdigits)

In [3]: s = "db9674c4-72a9-4ab9-9ddd-1d641a37cde4"

In [4]: uuidparser =((hexstr*8) >> char("-") >> (hexstr*4) >> char("-") >>  (hexstr*4) >> char("-") >> (hexstr*4) >> char("-") >> (hexstr*12))

In [5]: uuidparser(s)
Out[5]: (Right (['db9674c4', '-', '72a9', '-', '4ab9', '-', '9ddd', '-', '1d641a37cde4'], ''))

In [7]: uuidparser(s[1:5])
Out[7]: (Left Failed at both)

It's not production ready, please use pyparsing instead or send me PRs to make it better :)