Parse food recipes from Python with pylangparser

Link to project page

from pylangparser import *

QTY = Symbols(r'\d+')
QTY_HALF = Symbols(r'(\d+ ){0,1}1/2')
QTYS = QTY_HALF & QTY

ONION = Symbols(r'onion')
ROAST = Symbols(r'roast')
SOY_SAUCE = Symbols(r'soy sauce')
STALK_CELERY = Symbols(r'stalk celery')
WATER = Symbols(r'water')
YOGURT = Symbols(r'yogurt')

FAT = Symbols(r'\d+%')

INGREDIENT = \
    ONION & \
    ROAST & \
    SOY_SAUCE & \
    STALK_CELERY & \
    WATER & \
    YOGURT

RANDOM_WORD = Symbols(r'[a-zA-Z]+')



LBS = Symbols(r'lbs')
TABLESPOON = Symbols(r'tablespoon')
TABLESPOONS = Symbols(r'tablespoons')
TEASPOON = Symbols(r'teaspoon')
TEASPOONS = Symbols(r'teaspoons')
QUART = Symbols(r'quart')
QUARTS = Symbols(r'quarts')
DASH = Symbols(r'dash')
DASHES = Symbols(r'dashes')
CUP = Symbols(r'cup')
CUPS = Symbols(r'cups')

METRICS = \
    LBS & \
    TABLESPOON & \
    TABLESPOONS & \
    TEASPOON & \
    TEASPOONS & \
    QUART & \
    QUARTS & \
    DASH & \
    DASHES & \
    CUP & \
    CUPS

IGNORE_CHARS = Ignore(r'[ \n,.-]+')

IGNORES = IGNORE_CHARS

# order is important, first token that matches will be considered
TOKENS = FAT & QTYS & METRICS & IGNORES & INGREDIENT & RANDOM_WORD



food_recipe = """

3 lbs chuck roast
1 quart water
1 1/2 quarts water
1 onion, chopped
1 stalk celery, chopped
2 tablespoons soy sauce
1/2 cup 2% Greek yogurt

"""

product = \
    SymbolsParser(ONION) | \
    SymbolsParser(ROAST) | \
    SymbolsParser(SOY_SAUCE) | \
    SymbolsParser(STALK_CELERY) | \
    SymbolsParser(WATER) | \
    SymbolsParser(YOGURT)

measure = \
    SymbolsParser(CUP) | \
    SymbolsParser(CUPS) | \
    SymbolsParser(LBS) | \
    SymbolsParser(QUART) | \
    SymbolsParser(QUARTS) | \
    SymbolsParser(TABLESPOON) | \
    SymbolsParser(TABLESPOONS)

qty = \
    SymbolsParser(QTY_HALF) | \
    SymbolsParser(QTY)

fat = SymbolsParser(FAT)

ingredient = qty << ZeroOrMore(measure) << \
    Repeat(fat | product | IgnoreResult(SymbolsParser(RANDOM_WORD)))

recipe = ZeroOrMore(ingredient)



# extract tokens
lexer = Lexer(TOKENS)
tokens = lexer.parseTokens(food_recipe, False)
print(tokens)

parser = AllTokensConsumed(recipe)

# generating AST
ast = parser(tokens, 0)
print("\nast:")
ast.pretty_print()

for group in ast:
    if group.check_parser(ingredient):
        print("\ningredient:")
        for sub_group in group:
            if sub_group.check_parser(qty):
                print("qty: %s" % sub_group.get_token())
            if sub_group.check_parser(measure):
                print("measure: %s" % sub_group.get_token())
            if sub_group.check_parser(product):
                print("product: %s" % sub_group.get_token())
            if sub_group.check_parser(fat):
                print("fat level: %s" % sub_group.get_token())







output:

-------



[(3, \d+), (lbs, lbs), (chuck, [a-zA-Z]+), (roast, roast), (1, \d+), (quart, quart), (water, water), (1 1/2, (\d+ ){0,1}1/2), (quarts, quarts), (water, water), (1, \d+), (onion, onion), (chopped, [a-zA-Z]+), (1, \d+), (stalk celery, stalk celery), (chopped, [a-zA-Z]+), (2, \d+), (tablespoons, tablespoons), (soy sauce, soy sauce), (1/2, (\d+ ){0,1}1/2), (cup, cup), (2%, \d+%), (Greek, [a-zA-Z]+), (yogurt, yogurt)]

ast:
[[['3'], ['lbs'], ['roast']],
 [['1'], ['quart'], ['water']],
 [['1 1/2'], ['quarts'], ['water']],
 [['1'], ['onion']],
 [['1'], ['stalk celery']],
 [['2'], ['tablespoons'], ['soy sauce']],
 [['1/2'], ['cup'], ['2%'], ['yogurt']]]

ingredient:
qty: 3
measure: lbs
product: roast

ingredient:
qty: 1
measure: quart
product: water

ingredient:
qty: 1 1/2
measure: quarts
product: water

ingredient:
qty: 1
product: onion

ingredient:
qty: 1
product: stalk celery

ingredient:
qty: 2
measure: tablespoons
product: soy sauce

ingredient:
qty: 1/2
measure: cup
fat level: 2%
product: yogurt