class DiceParser(TextParsers):
nat = reg(r"[1-9]\d*") > int
constant = reg(r"\d+") > (lambda x: ValueConstant(int(x)))
roll = die_value << "d" & die_value > (lambda xs: ValueRoll(xs[0], xs[1]))
naked_roll = "d" >> die_value > (lambda x: ValueRoll(ValueConstant(1), x))
value = roll | naked_roll | constant | brackets
die_value = constant | brackets
op_add = value << "+" & expr > (
lambda xs: BinaryOperator(RollOperator.ADD, xs[0], xs[1]))
op_sub = value << "-" & expr > (
lambda xs: BinaryOperator(RollOperator.SUB, xs[0], xs[1]))
op_mul = value << "*" & expr > (
lambda xs: BinaryOperator(RollOperator.MUL, xs[0], xs[1]))
op_div = value << "/" & expr > (
lambda xs: BinaryOperator(RollOperator.DIV, xs[0], xs[1]))
op_pow = value << "^" & expr > (
lambda xs: BinaryOperator(RollOperator.POW, xs[0], xs[1]))
expr = op_add | op_sub | op_mul | op_div | op_pow | value
brackets = "(" >> expr << ")"
parse_all = expr
class DiscordParser(TextParsers):
"""Removes surrounding code blocks before the program can reach the main parser"""
main = (
"```" >> reg(r"(?s).*?(?=```)") << "```"
| "`" >> reg(r"(?s).*?(?=`)") << "`"
| reg(r"[^`](?s).*")
)
class JsonStringParsers(TextParsers, whitespace=None):
quote = lit(r'\"')
reverse_solidus = lit(r'\\')
solidus = lit(r'\/')
backspace = lit(r'\b')
form_feed = lit(r'\f')
line_feed = lit(r'\n')
carriage_return = lit(r'\r')
tab = lit(r'\t')
uni = reg(r'\\u([0-9a-fA-F]{4})')
escaped = (quote | reverse_solidus | solidus | backspace | form_feed |
line_feed | carriage_return | tab | uni)
unescaped = reg(r'[\u0020-\u0021\u0023-\u005B\u005D-\U0010FFFF]+')
string = '"' >> rep(escaped | unescaped) << '"' > ''.join
class FormatTextParsers(TextParsers, whitespace=None):
integer = reg(r'[0-9]+') > int
dense = lit('d') > constant(Mode.dense)
compressed = lit('s') > constant(Mode.compressed)
mode = dense | compressed
# Use eof to ensure each parser goes to end
format_without_orderings = rep(mode) << eof > (
lambda modes: Format(tuple(modes), tuple(range(len(modes)))))
format_with_orderings = rep(mode
& integer) << eof > make_format_with_orderings
format = format_without_orderings | format_with_orderings
class Homework(TextParsers):
number = reg(r"\d+") > int
plus = lit("+") > constant(add)
times = lit("*") > constant(mul)
operator = plus | times
# No precedence
base = "(" >> unprecedented << ")" | number
unprecedented = base & rep(operator & base) > reduce
# Addition first, then multiplication
base = "(" >> multiplication << ")" | number
addition = base & rep(plus & base) > reduce
multiplication = addition & rep(times & addition) > reduce
class UsawParser(TextParsers, whitespace=None):
floatP = reg(r'[+-]?[0-9]*\.?[0-9]+')
# actual fields
WeightClass = "Weight Class:|" >> reg(r'[0-9]+[-+]?') << ' Kg|'
Total = "Total:|" >> floatP << "|" > float
CompetitionWeight = "Competition Weight:|" >> floatP << '|' > float
BodyWeight = "Body Weight:|" >> floatP << '|' > float
Sn1 = "Snatch 1:|" >> opt(floatP << '|' > float)
Sn2 = "Snatch 2:|" >> opt(floatP << '|' > float)
Sn3 = "Snatch 3:|" >> opt(floatP << '|' > float)
BestSn = "Best Snatch:|" >> floatP << '|' > float
Cj1 = "CleanJerk 1:|" >> opt(floatP << '|' > float)
Cj2 = "CleanJerk 2:|" >> opt(floatP << '|' > float)
Cj3 = "CleanJerk 3:|" >> opt(floatP << '|' > float)
Cj1And = "Clean & Jerk 1:|" >> opt(floatP << '|' > float)
Cj2And = "Clean & Jerk 2:|" >> opt(floatP << '|' > float)
Cj3And = "Clean & Jerk 3:|" >> opt(floatP << '|' > float)
BestCj = "Best CleanJerk:|" >> floatP > float
BestCJAnd = "Best Clean & Jerk:|" >> floatP > float
value = WeightClass & Total & (
CompetitionWeight | BodyWeight) & Sn1 & Sn2 & Sn3 & BestSn & (
Cj1 | Cj1And) & (Cj2 | Cj2And) & (Cj3 | Cj3And) & (BestCj
| BestCJAnd)
class TensorExpressionParsers(TextParsers):
name = reg(r'[A-Za-z][A-Za-z0-9]*')
# taco does not support negatives or exponents
floating_point = reg(r'[0-9]+\.[0-9]+') > (lambda x: Float(float(x)))
integer = reg(r'[0-9]+') > (lambda x: Integer(int(x)))
number = floating_point | integer
# taco requires at least one index; scalar tensors are not parsed as `a()`
# taco also allows for `y_{i}` and `y_i` to mean `y(i)`, but that is not supported here
tensor = name & '(' >> rep1sep(name, ',') << ')' > splat(Tensor)
scalar = name > Scalar
variable = tensor | scalar
parentheses = '(' >> expression << ')' # noqa: F821
factor = variable | number | parentheses
term = rep1sep(factor, '*') > (lambda x: reduce(Multiply, x))
expression = term & rep(lit('+', '-') & term) > splat(make_expression)
simple_assignment = variable & '=' >> expression > splat(Assignment)
add_assignment = variable & '+=' >> expression > splat(lambda v, e: Assignment(v, Add(v, e)))
assignment = simple_assignment | add_assignment
class TreeParsers(TextParsers):
# Grammar rules using Parsita, with semantic routines
exam = opt(title) & rep(questionBox) # > make_binop
questionBox = question & opt(hint) & choiceBox
question = reg(r"Question(\s)*") & opt(hintType) & text
hintType = reg(
r"\[((\"|\')[a-zA-Z0-9]*((\"|\')(\s)*\,(\s)*)?)*(\"|\')[a-zA-Z0-9]*(\"|\')\]"
)
choiceBox = repsep(choice | choiceCorrect)
# choiceBox = choice
choice = reg(r"Choice[\s]+") & text
choiceCorrect = reg(r"Choice correct(\s)*") & text
hint = reg(r"Hint[\s]+") & text
title = reg(r"Exam[\s]+") & text
text = reg(r"[\w\W]*\n")
class JsonParsers(TextParsers, whitespace=r'[ \t\n\r]*'):
number = reg(r'-?(0|[1-9][0-9]*)(\.[0-9]+)?([eE][-+]?[0-9]+)?')
false = lit('false')
true = lit('true')
null = lit('null')
string = JsonStringParsers.string
array = '[' >> repsep(value, ',') << ']'
entry = string << ':' & value
obj = '{' >> repsep(entry, ',') << '}'
value = (number
| false
| true
| null
| string
| array
| obj)
class ProgramParser(TextParsers):
# Actual grammar
split1 = lambda item, separator: item & rep(separator & item)
split = lambda item, separator: opt(split1(item, separator))
identifier = reg(r"[a-zA-Z]\w*")
string = reg(r'".*?(?<!\\)(\\\\)*?"') | reg(r"'.*?(?<!\\)(\\\\)*?'") > (
lambda s: TokenString(s[1:-1])
)
num_int = reg(r"\d+") > int
num_float = reg(r"(\d*\.\d+|\d+\.\d*)") > float
num_positive = num_float | num_int
num_negative = "-" >> num > (lambda x: -x)
num = num_negative | num_positive
number = num > TokenNumber
op_eq = lit("==") > constant(Operator.EQ)
op_ne = lit("!=") > constant(Operator.NE)
op_ge = lit(">=") > constant(Operator.GE)
op_gt = lit(">") > constant(Operator.GT)
op_le = lit("<=") > constant(Operator.LE)
op_lt = lit("<") > constant(Operator.LT)
op_and = lit("&") > constant(Operator.AND)
op_or = lit("|") > constant(Operator.OR)
op_add = lit("+") > constant(Operator.ADD)
op_sub = lit("-") > constant(Operator.SUB)
op_mul = lit("*") > constant(Operator.MUL)
op_div = lit("/") > constant(Operator.DIV)
op_pow = lit("^") > constant(Operator.POW)
op_not = lit("!") > constant(Operator.NOT)
op_neg = lit("-") > constant(Operator.NEG)
equality_op = op_eq | op_ne
comparison_op = op_ge | op_gt | op_le | op_lt
logic_op = op_and | op_or
term_op = op_add | op_sub
factor_op = op_mul | op_div
power_op = op_pow
unary_op = op_neg | op_not
case_pair = expr << "->" & expr
assignment = identifier << "=" & expr
let_stmt = "^" >> rep1sep(assignment, ";") << "$" & expr > let
anon_func = (lit("\\") | lit("\\\\")) >> rep1(identifier) & "->" >> expr > anon
variable = identifier > TokenVariable
expr = rep1sep(equality, reg(r"\s*")) > maybe_application
equality = split1(comparison, equality_op) > bin_operator
comparison = split1(logic, comparison_op) > bin_operator
logic = split1(term, logic_op) > bin_operator
term = split1(factor, term_op) > bin_operator
factor = split1(power, factor_op) > bin_operator
power = split1(ternary, power_op) > bin_operator_right
ternary = case & opt("?" >> expr << ":" & expr) > maybe_ternary
case = dice & opt(lit("$") >> "(" >> rep1sep(case_pair, ";") << ")") > maybe_case
dice = unary & opt("d" >> unary) > maybe_dice
unary = unary_op & unary | primary > mon_operator
primary = number | string | bracketed | let_stmt | anon_func | variable
bracketed = "(" >> expr << ")"
func = identifier & "=" >> expr > function
program = repsep("@" >> func | expr, ";") << opt(";") > Program
main = program
class DiceParser(TextParsers):
split1 = lambda item, separator: item & rep(separator & item)
split = lambda item, separator: opt(split1(item, separator))
comment = "/*" >> expr << "*/" > constant(None)
string = reg(r'".*?(?<!\\)(\\\\)*?"') > (lambda s: ValueString(s[1:-1]))
num_int = reg(r"\d+") > int
num_float = reg(r"(\d*\.\d+|\d+\.\d*)") > float
num_positive = num_float | num_int
num_negative = "-" >> num > (lambda x: -x)
num = num_negative | num_positive
number = num > ValueNumber
# set = "(" >> repsep(expr, ",") << ")" > ValueSet
op_eq = lit("==") > constant(Operator.EQ)
op_ne = lit("!=") > constant(Operator.NE)
op_ge = lit(">=") > constant(Operator.GE)
op_gt = lit(">") > constant(Operator.GT)
op_le = lit("<=") > constant(Operator.LE)
op_lt = lit("<") > constant(Operator.LT)
op_and = lit("&") > constant(Operator.AND)
op_or = lit("|") > constant(Operator.OR)
op_add = lit("+") > constant(Operator.ADD)
op_sub = lit("-") > constant(Operator.SUB)
op_mul = lit("*") > constant(Operator.MUL)
op_div = lit("/") > constant(Operator.DIV)
op_pow = lit("^") > constant(Operator.POW)
op_not = lit("!") > constant(Operator.NOT)
op_neg = lit("-") > constant(Operator.NEG)
equality_op = op_eq | op_ne
comparison_op = op_ge | op_gt | op_le | op_lt
logic_op = op_and | op_or
term_op = op_add | op_sub
factor_op = op_mul | op_div
power_op = op_pow
unary_op = op_neg | op_not
case_pair = expr << "," & expr
program = repsep(expr, ";") > Program
expr = equality
equality = split1(comparison, equality_op) > bin_operator
comparison = split1(logic, comparison_op) > bin_operator
logic = split1(term, logic_op) > bin_operator
term = split1(factor, term_op) > bin_operator
factor = split1(power, factor_op) > bin_operator
power = split1(dice, power_op) > bin_operator_right
dice = opt(opt(ternary) << "d") & ternary > maybe_dice
ternary = case & opt("?" >> expr << ":" & expr) > maybe_ternary
case = unary & opt(
lit(":") >> "(" >> repsep(case_pair, ";") << ")") > maybe_case
unary = unary_op & unary | primary > mon_operator
primary = number | string | bracketed
bracketed = "(" >> expr << ")"
parse_all = program
class PlusParsers(TextParsers):
variable = positioned(reg('[A-Za-z][A-Za-z0-9_]*') > UnfinishedVariable)
plus = variable & '+' >> variable > splat(Plus)
class ModelParsers(TextParsers, whitespace=r'[ \t]*'):
number = reg(r'[+-]?\d+(\.\d+)?([Ee][+-]?\d+)?') > float
name = reg(r'[A-Za-z_][A-Za-z_0-9]*')
symbol = name > sy.Symbol
def make_function(x):
func_name, arguments = x
if func_name in sy.functions.__dict__:
func_handle = sy.__dict__[func_name]
else:
raise ValueError(f'Function "{func_name}" not found. Only functions in sympy.* may be used.')
return func_handle(*arguments)
function = name & '(' >> repsep(expression, ',') << ')' > make_function
factor = number | function | symbol | '(' >> expression << ')'
def make_exponent(x):
x = list(reversed(x)) # Exponentiation is right associative so reverse the list
value = x[0]
rest = x[1:]
for item in rest:
value = sy.Pow(item, value)
return value
exponent = rep1sep(factor, '^') > make_exponent
def make_term(x):
first, rest = x
value = first
for op, exponent in rest:
if op == '/':
exponent = sy.Pow(exponent, -1) # This is how sympy handles divide
value = sy.Mul(value, exponent)
return value
term = exponent & rep(lit('*', '/') & exponent) > make_term
def make_unary_term(x):
ops, value = x
for op in ops:
if op == '-':
value = -value
return value
unary_term = rep(lit('+', '-')) & term > make_unary_term
def make_expression(x):
first, rest = x
value = first
for op, term in rest:
if op == '-':
term = sy.Mul(-1, term) # This is how sympy handles minus
value = sy.Add(value, term)
return value
expression = unary_term & rep(lit('+', '-') & unary_term) > make_expression
def make_one_sided_time_constant(x):
direction, time = x
if direction == '<':
return -inf, time
else:
return time, inf
one_sided_time = '(' >> lit('t') >> lit('<', '>') & number << ')' > make_one_sided_time_constant
two_sided_time = '(' >> number << '<' << 't' << '<' & number << ')'
time_range = one_sided_time | two_sided_time
def make_constant(x):
name, op, value = x
if op == '+=':
additive = True
else:
additive = False
return name, Constant(name, value, additive)
constant = name & lit('=', '+=') & number > make_constant
def make_rule(x):
name, domain, op, expr = x
if not domain:
first = -inf
last = inf
else:
first, last = domain[0]
if op == '+=':
additive = True
else:
additive = False
return name, Rule(name, AnalyticSegment(first, last, expr), additive)
rule = name & opt(time_range) & lit('=', '+=') & expression > make_rule
def make_initial(x):
name, op, value = x
if op == '+=':
additive = True
else:
additive = False
return name, Initial(value, additive)
initial = name << '*' & lit('=', '+=') & expression > make_initial
def make_ode(x):
name, domain, op, expr = x
if not domain:
first = -inf
last = inf
else:
first, last = domain[0]
if op == '+=':
additive = True
else:
additive = False
return name, Ode(AnalyticSegment(first, last, expr), additive)
ode = name << "'" & opt(time_range) & lit('=', '+=') & expression > make_ode
def make_dose(x):
name, time, op, value = x
if op == '+=':
value += sy.Symbol(name)
return name, Dose(time, value)
dose = name << "(" & number << ")" & lit('=', '+=') & expression > make_dose
def make_effect(x):
name, op, value = x
if op == '+=':
value += sy.Symbol(name)
return Effect(name, value)
effect = name & lit('=', '+=') & expression > make_effect
def make_event(x):
left, direction, right, effects = x
trigger = left - right
if direction == '<':
effect_direction = EventDirection.down
else:
effect_direction = EventDirection.up
return Event(trigger, effect_direction, True, effects)
event = lit('@') >> '(' >> expression & lit('<', '>') & expression << ")" & rep1sep(effect, ',') > make_event
component = constant | rule | initial | ode | dose | event
eol = reg(r'(((#.*)?\n)+)|(((#.*)?\n)*(#.*)?\Z)')
options_section = '%' >> lit('options') >> eol >> rep(failure('not implemented') << eol)
components_section = '%' >> lit('components') >> eol >> rep(component << eol)
def make_model(x):
maybe_options, components = x
if maybe_options:
raise NotImplementedError
parts, events = collapse_components(components)
new_model = Model(parts, events)
return new_model
model = opt(options_section) & components_section > make_model
class KarmaParser(TextParsers):
anything = reg(r".") > constant(None)
word_topic = reg(r'[^"\s]+?(?=[+-]{2})')
string_topic = reg(r'".*?(?<!\\)(\\\\)*?"(?=[+-]{2})')
topic = (word_topic >
(lambda t: [t, False])) | (string_topic >
(lambda t: [t[1:-1], True]))
op_positive = reg(r"(?<![+-])\+\+(?![+-])") > constant(
KarmaOperation.POSITIVE)
op_neutral = (reg(r"(?<![+-])\+-(?![+-])")
| reg(r"(?<![+-])-\+(?![+-])")) > constant(
KarmaOperation.NEUTRAL)
op_negative = reg(r"(?<![+-])--(?![+-])") > constant(
KarmaOperation.NEGATIVE)
operator = op_positive | op_neutral | op_negative
bracket_reason = reg(r"\(.+?\)") > (lambda s: s[1:-1])
quote_reason = reg(r'".*?(?<!\\)(\\\\)*?"(?![+-]{2})') > (
lambda s: s[1:-1])
reason_words = reg(r"(?i)because") | reg(r"(?i)for")
text_reason = reason_words >> (reg(r'[^",]+') | quote_reason)
reason = bracket_reason | quote_reason | text_reason
karma = (topic & operator & opt(reason)) > make_karma
parse_all = rep(karma | anything) > filter_out_none
class TreeParsers(TextParsers):
exam = opt(title) & questions > formatString
questions = rep(actualQuestion)
actualQuestion = reg(r"Question ") & opt(listt) & text & opt(
hint) & mulChoices
mulChoices = rep(choiceCorrect | singleChoice)
singleChoice = reg(r"Choice ") & opt(
reg(r"\s*")) & text & reg(r"\s*") & opt(feedback)
choiceCorrect = reg(r"Choice ") & opt(
reg(r"\s*correct\s*")) & text & reg(r"\s*") & opt(feedback)
feedback = reg(r"Feedback ")
listt = reg(r"\[\".+\"(?:,.+)?\]")
hint = reg(r"Hint ") & text
title = reg(r"^Exam ") & text
text = reg(r"\".+\"(?:\n,\n)?")
class TreeParsers(TextParsers):
# Grammar rules using Parsita, with semantic routines
exam = opt(title) & questionBox # > make_binop
questionBox = rep(question)
question = reg(r"Question ") & opt(listt) & text & opt(hint) & choiceBox
choiceBox = rep(choiceCorrect | choice)
choice = reg(r"Choice ") & opt(reg(r"\s*")) & text & reg(r"\s*") & opt(feedback)
choiceCorrect = reg(r"Choice ") & opt(reg(r"\s*correct\s*")) & text & reg(r"\s*") & opt(feedback)
feedback = reg(r"Feedback ")
listt = reg(r"\[\".+\"(?:,.+)?\]")
hint = reg(r"Hint ") & text
title = reg(r"^Exam ") & text
text = reg(r"\".+\"(?:\n,\n)?")
