zen-xu / xpattern

The Pattern Matching for Python with XObject.

Xpattern: Pattern Matching with XObject.

xpattern is inspired by hask, pipetools, pampy, and with xpattern your code will be more readable and graceful!

Let's play now

Each pattern matchline evaluated in the order they appear.

from xpattern import caseof
from xpattern import m


~(caseof(v)
    | m(pattern_1) >> action_1
    | m(pattern_2) >> action_2
)

caseof is lazy, so you need to add ~ operator to run it!

m is a matcher, if you don't like circular brackets style, you can alse use square brackets style

from xpattern import caseof
from xpattern import m


~(caseof(v)
    | m[pattern_1] >> action_1
    | m[pattern_2] >> action_2
)

Write a Fibonacci

The operator _ means "match everything"

from xpattern import _
from xpattern import caseof
from xpattern import m


def fibonacci(n):
    return ~(caseof(n)
        | m(1) >> 1
        | m(2) >> 1
        | _ >> (lambda x: fibonacci(x - 1) + fibonacci(x - 2))
    )

Write a Lisp calculator

from functools import reduce

from xpattern import _
from xpattern import REST
from xpattern import caseof
from xpattern import m


def lisp(exp):
    return ~(caseof(exp)
        | m(int)            >> (lambda x: x)
        | m(callable)       >> (lambda x: x)
        | m(callable, REST) >> (lambda f, rest: f(*map(lisp, rest)))
        | m(tuple)          >> (lambda t: list(map(lisp, t)))
    )

plus = lambda a, b: a + b
minus = lambda a, b: a - b
lisp((plus, 1, 2))                  # => 3
lisp((plus, 1 (minus, 4, 2)))       # => 3
lisp((reduce, plus, (range, 10)))   # => 45

You can match so many things!

~(caseof(x)
    | m(3)         >> "this matches the number 3"
    | m(int)       >> "matches any integer"
    | m(str, int)  >> (lambda a, b: "a tuple (a, b) you can use in a function")
    | m(1, 2, _)   >> "any list of 3 elements that begins with [1, 2]"
    | m({"x", _})  >> "any dict with a key 'x' and any value associated"
    | _            >> "anything else"
)

Match dataclass

from dataclasses import dataclass

from xpattern import _
from xpattern import caseof
from xpattern import m


@dataclass
class Point:
    x: int
    y: int

@dataclass
class Point2:
    x: int
    y: int

@dataclass
class Line:
    p1: Point
    p2: Point

@dataclass
class Rect:
    l1: Line
    l2: Line

~(caseof(Rect(Point(1, 2), Point(3, 4)))
    | m(Rect(Point(_, str), Point(_, 4))) >> "first"
    | m(Rect(Point(_, int), Point2(_, 4))) >> "second"
    | m(Rect(Point(_, int), Point(_, 4))) >> (lambda x, y, z: (x, y, z))
)  # => (1, 2, 3)

Match [HEAD, TAIL]

from xpattern import _
from xpattern import HEAD
from xpattern import TAIL
from xpattern import caseof
from xpattern import m


x = [1, 2, 3]

~(caseof(x)
    | m(1, TAIL) >> (lambda t: t)             # => [2, 3]
)

~(caseof(x)
    | m(HEAD, TAIL) >> (lambda h, t: (h, t))  # => [1, [2, 3]]
)

Chain match

from xpattern import _
from xpattern import caseof
from xpattern import m


pet = {"type": "dog", "details": {"age": 3}}
~(caseof(pet)
    | m({_: {"age": _}}) >> ~(caseof(X)
        | m(int, int) >> (lambda x, y: x + y)
        | m(str, int) >> (lambda x, y: y)
    )
)  # => 3

With Bitwise Operators

from xpattern import _
from xpattern import caseof
from xpattern import m


~(caseof(1)
    | (m[2] | m[1]) >> 1
    | _ >> "nothing"
) # => 1

~(caseof(11)
    | (m[lambda x: x > 1] & m[lambda x: x < 10]) >> "1 < x < 10"
    | (m[lambda x: x > 1] & m[lambda x: x < 15]) >> "1 < x < 15"
) # => "1 < x < 15"

~(caseof(6)
    | ~m[lambda x: x > 5]  >> "x <=5"
    | ~m[lambda x: x > 10] >> "x <= 10"
) # => "x <= 10"

More pattern cases

Your can visit repo pampy get more pattern cases, xpattern is Syntactic Sugar of pampy

Why name Xpattern, what's X mean?

X means XObject !!!

XObject is a Syntactic Sugar of lambda function

from xpattern import X
from xpattern import caseof
from xpattern import m


~(caseof(1)
    | m(1) >> X + 1       # => 2
)

~(caseof("apple")
    | m(str) >> X.upper()  # => "APPLE"
)

~(caseof([1, 2, 3])
    | m(1, 2, 3) >> X[2]   # => 3
)

~(caseof([1, 2, 3])
    | m(1, 2, 3) >> X + [4, 5, 6]              # => [1, 2, 3, 4, 5, 6]
)

~(caseof(9)
    | m(int) >> X + X ** (X << 2) % 2 / 3 - X  # => 0.333333333
)

~(caseof(1)
    | m(int) >> X._in_([1, 2, 3])   # => True
)

~(caseof(lambda x, y: x + y)
    | m(callable) >> X(1, 2)        # => 3
)

Operation	Syntax
Addition	X + 1
Call	X(a, b, c)
Concatenation	X + [1, 2, 3]
Containment Test	X._in_( [1, 2, 3] )
Contains	X._contains_(1)
Division	X / 2 or X // 2
Bitwise And	X & 2
Bitwise Exclusive Or	X ^ 2
Bitwise Inversion	~X
Bitwise Or	X | 2
Exponentiation	X ** 2
Identity	X._is_(2)
Indexing	X[k]
Left Shift	X << 2
Modulo	X % 2
Multiplication	X * 2
Matrix Multiplication	X @ matrix
Negation (Arithmetic)	-X
Negation (Logical)	X._not()
Positive	+X
Right Shift	X >> 2
Subtraction	X - 2
Ordering	X < 2 or X <= 2 or X > 2 or X >= 2
Equality	X == 2
Difference	X != 2

match pattern with XObject

from xpattern import X
from xpattern import caseof
from xpattern import m


~(caseof((1, 2, 3))
    | m(X[2] == 3) >> X[2] + 4  # => 7
)

# same as
~(caseof((1, 2, 3))
    | m(lambda x: x[2] == 3) >> (lambda x: x[2] + 4)  # => 7
)

xfunction

XObject only can represent as function like X(1, 2), but can not be as args in function func(X, X + 2)

Now you can try to use xfunction realizing it!

from xpattern import X
from xpattern import _
from xpattern import caseof
from xpattern import m
from xpattern import xfunction


@xfunction
def add(a, b):
    return a + b

# in actions
~(caseof(1)
    | m(int) >> add(X, X)                             # => 2
)

# recursion xfunction
~(caseof(2)
    | m(int) >> add(add(X + 1, 4), b=add(X * 9, 7))   # => 32
)


@xfunction
def greater_than_4(x):
    return x > 4


# in patterns
~(caseof(1)
    | m(greater_than_4(X + 5)) >> "greater than 4"
    | _ >> "equal or lesser than 4"
)  # => "greater than 4"

~(caseof(1)
    | m(greater_than_4(X)) >> "greater than 4"
    | _ >> "equal or lesser than 4"
)  # => "equal or lesser than 4"