def json_text():
"""Returns the parser for Json formatted data"""
# Taken from https://github.com/vlasovskikh/funcparserlib/blob/master/funcparserlib/tests/json.py
# and modified slightly
unwrap = lambda x: x.value
null = (n('null') | n('Null')) >> const(None) >> unwrap
value = forward_decl()
member = (string >> unwrap) + op_(u':') + value >> tuple
object = (
op_(u'{') +
maybe(member + many(op_(u',') + member) + maybe(op_(','))) +
op_(u'}')
>> make_object)
array = (
op_(u'[') +
maybe(value + many(op_(u',') + value) + maybe(op_(','))) +
op_(u']')
>> make_array)
value.define(
null
| (true >> unwrap)
| (false >> unwrap)
| object
| array
| (number >> unwrap)
| (string >> unwrap))
json_text = object | array
return json_text
def parse(seq):
'Sequence(Token) -> object'
unarg = lambda f: lambda args: f(*args)
tokval = lambda x: x.value
flatten = lambda list: sum(list, [])
value_flatten = lambda l: sum([[l[0]]] + list(l[1:]), [])
n = lambda s: a(Token('Name', s)) >> tokval
op = lambda s: a(Token('Op', s)) >> tokval
op_ = lambda s: skip(op(s))
id = some(lambda t:
t.type in ['Name', 'Number', 'Color', 'String']).named('id') >> tokval
make_chart_attr = lambda args: DefAttrs(u'chart', [Attr(*args)])
node_id = id # + maybe(port)
pair = (
op_('(') + id + skip(maybe(op(','))) + id + op_(')')
>> tuple)
value = (id | pair)
value_list = (
value +
many(op_(',') + value)
>> value_flatten)
a_list = (
id +
maybe(op_('=') + id) +
skip(maybe(op(',')))
>> unarg(Attr))
attr_list = (
many(op_('[') + many(a_list) + op_(']'))
>> flatten)
chart_attr = id + (op_('=') | op_(':')) + value_list >> make_chart_attr
node_stmt = node_id + attr_list >> unarg(Node)
stmt = (
chart_attr
| node_stmt
)
stmt_list = many(stmt + skip(maybe(op(';'))))
chart_type = (
n('p') | n('pie') | n('piechart')
| n('p3') | n('pie3d') | n('piechart_3d')
| n('lc') | n('line') | n('linechart')
| n('lxy') | n('linechartxy')
| n('bhs') | n('holizontal_barchart')
| n('bvs') | n('vertical_barchart')
| n('bhg') | n('holizontal_bargraph')
| n('bvg') | n('vertical_bargraph')
| n('v') | n('venn') | n('venndiagram')
| n('s') | n('plot') | n('plotchart')
)
chart = (
chart_type +
maybe(id) +
op_('{') +
stmt_list +
op_('}')
>> unarg(Chart))
dotfile = chart + skip(finished)
return dotfile.parse(seq)
def test_non_halting_left_recursive():
h1 = fwd()
h1.define(x + h1)
ok_(not non_halting(h1))
h2 = fwd()
h2.define(x + (h2 | x))
ok_(not non_halting(h2))
nh1 = fwd()
nh1.define(nh1 + x)
ok_(non_halting(nh1))
nh2 = fwd()
nh2.define(x | nh2)
ok_(non_halting(nh2))
nh3_fwd = fwd()
nh3_fwd.define(nh3_fwd)
nh3 = x + nh3_fwd + x
ok_(non_halting(nh3))
nh4 = fwd()
nh4.define(maybe(x) + nh4 + x)
ok_(non_halting(nh4))
nh5 = fwd()
nh5.define(many(x) + maybe(x) + nh5 + x)
ok_(non_halting(nh5))
h3 = fwd()
h3.define(maybe(x) + many(x) + x + h3)
ok_(not non_halting(h3))
def parse(seq):
'Sequence(Token) -> object'
unarg = lambda f: lambda args: f(*args)
tokval = lambda x: x.value
flatten = lambda list: sum(list, [])
n = lambda s: a(Token('Name', s)) >> tokval
op = lambda s: a(Token('Op', s)) >> tokval
op_ = lambda s: skip(op(s))
date = some(lambda s: a(Token('Date', s))).named('date') >> tokval
id = some(lambda t: t.type in ['Name', 'Number', 'String']).named(
'id') >> tokval
make_chart_attr = lambda args: DefAttrs(u'chart', [Attr(*args)])
node_id = id # + maybe(port)
term = date + op_('-') + date
value = (id | term | date)
a_list = (id + maybe(op_('=') + id) + skip(maybe(op(','))) >> unarg(Attr))
attr_list = (many(op_('[') + many(a_list) + op_(']')) >> flatten)
chart_attr = id + (op_('=') | op_(':')) + value >> make_chart_attr
node_stmt = node_id + attr_list >> unarg(Node)
stmt = (chart_attr | node_stmt)
stmt_list = many(stmt + skip(maybe(op(';'))))
chart = (maybe(n('diagram')) + maybe(id) + op_('{') + stmt_list + op_('}')
>> unarg(Chart))
dotfile = chart + skip(finished)
return dotfile.parse(seq)
def test_non_halting_left_recursive():
h1 = fwd()
h1.define(x + h1)
ok_(not non_halting(h1))
h2 = fwd()
h2.define(x + (h2 | x))
ok_(not non_halting(h2))
nh1 = fwd()
nh1.define(nh1 + x)
ok_(non_halting(nh1))
nh2 = fwd()
nh2.define(x | nh2)
ok_(non_halting(nh2))
nh3_fwd = fwd()
nh3_fwd.define(nh3_fwd)
nh3 = x + nh3_fwd + x
ok_(non_halting(nh3))
nh4 = fwd()
nh4.define(maybe(x) + nh4 + x)
ok_(non_halting(nh4))
nh5 = fwd()
nh5.define(many(x) + maybe(x) + nh5 + x)
ok_(non_halting(nh5))
h3 = fwd()
h3.define(maybe(x) + many(x) + x + h3)
ok_(not non_halting(h3))
def when_equation(tokens, state):
kw = keyword
parser = (kw("when") + expression + kw("then") +
maybe(many(equation + op(";"))) + maybe(
many(
kw("elsewhen") + expression + kw("then") +
maybe(many(equation + op(";"))))) + kw("end") +
kw("when")) >> WhenEquation
return parser.run(tokens, state)
def when_statement(tokens, state):
kw = keyword
parser = (kw("when") + expression + kw("then") +
maybe(many(statement + op(";"))) + maybe(
many(
kw("elsewhen") + expression + kw("then") +
maybe(many(statement + op(";"))))) + kw("end") +
kw("when")) >> WhenStatement
return parser.run(tokens, state)
def if_equation(tokens, state):
kw = keyword
parser = (kw("if") + expression + kw("then") +
maybe(many(equation + op(";"))) + maybe(
many(
kw("elseif") + expression + kw("then") +
maybe(many(equation + op(";"))))) +
maybe(kw("else") + maybe(many(equation + op(";")))) + kw("end") +
kw("if")) >> IfEquation
return parser.run(tokens, state)
def if_statement(tokens, state):
kw = keyword
parser = (kw("if") + expression + kw("then") +
maybe(many(statement + op(";"))) + maybe(
many(
kw("elseif") + expression + kw("then") +
maybe(many(statement + op(";"))))) +
maybe(kw("else") + maybe(many(statement + op(";")))) +
kw("end") + kw("if")) >> IfStatement
return parser.run(tokens, state)
def short_class_definition(tokens, state):
# circular import!
from modparc.syntax.class_definition import (class_prefixes, enum_list,
base_prefix)
parser = (class_prefixes + token_type("ident") + op("=")
+ (base_prefix + name + maybe(array_subscript)
+ maybe(class_modification) + comment
| keyword('enumeration') + op('(') +
(maybe(enum_list) | op(":")) + op(')') + comment))
return (parser >> ShortClassDefinition).run(tokens, state)
def field_query_parser():
"""Return a parser for numeric queries.
Example queries: '1900-1995' or '>= 1998'
"""
number = token_type('number')
field_name = token_type('name')
lt = token_type('lt')
le = token_type('le')
gt = token_type('gt')
ge = token_type('ge')
eq = token_type('equals')
approx = token_type('approx')
# Simple comparisons
# NOTE: We put le before lt to parse both
comparison = parser.maybe(token_type('not'))\
+ field_name\
+ (le | lt | ge | gt)\
+ number
# Values can be given as intervals ('1990-2000')
interval = parser.maybe(token_type('not'))\
+ field_name\
+ skip('equals')\
+ number\
+ skip('dash')\
+ number
# Values can be given as ranges ('1990<=year<=2000')
# NOTE: We put le before lt to parse both
range_ = parser.maybe(token_type('not'))\
+ number\
+ (le | lt)\
+ field_name\
+ (le | lt)\
+ number
# Field value queries ('year=2000' or 'author~Augustus')
field_value = parser.maybe(token_type('not'))\
+ field_name\
+ (eq | approx)\
+ (token_type('name') | token_type('number') | token_type('any'))
# Field occurrence ('publisher' or '^publisher')
field_occurrence = parser.maybe(token_type('not')) + field_name
return (interval >> make_query_result('interval')
| comparison >> make_query_result('comparison')
| range_ >> make_query_result('range')
| field_value >> make_query_result('value')
| field_occurrence >> make_query_result('occurrence'))\
+ parser.skip(parser.finished)
def date_parser():
"""Return a parser for biblatex dates."""
dash = skip('dash')
forward_slash = token_type('slash')
year = if_token_type('number', lambda v: len(v) == 4) >> token_value
month = if_token_type('number', lambda v: len(v) == 2) >> token_value
day = month
date = year + parser.maybe(dash + month) + parser.maybe(dash + day)
return date\
+ parser.maybe((forward_slash + date) | forward_slash)\
+ parser.skip(parser.finished)\
>> make_date
def test_first_maybe():
eq_(first(maybe(a('x'))), ['x', _EPSYLON])
eq_(first(maybe(a('x')) + a('y')), ['x', 'y'])
eq_(first(maybe(a('x')) + maybe(a('y'))), ['x', 'y', _EPSYLON])
eq_(first(maybe(a('x')) | a('y')), ['x', _EPSYLON, 'y'])
eq_(first(a('x') | maybe(a('y'))), ['x', 'y', _EPSYLON])
eq_(first(maybe(a('x')) | maybe(a('y'))), ['x', _EPSYLON, 'y', _EPSYLON])
def test_first_maybe():
eq_(first(maybe(a('x'))), ['x', _EPSYLON])
eq_(first(maybe(a('x')) + a('y')), ['x', 'y'])
eq_(first(maybe(a('x')) + maybe(a('y'))), ['x', 'y', _EPSYLON])
eq_(first(maybe(a('x')) | a('y')), ['x', _EPSYLON, 'y'])
eq_(first(a('x') | maybe(a('y'))), ['x', 'y', _EPSYLON])
eq_(first(maybe(a('x')) | maybe(a('y'))), ['x', _EPSYLON, 'y', _EPSYLON])
def parse(seq):
'Sequence(Token) -> object'
tokval = lambda x: x.value
op = lambda s: a(Token('Op', s)) >> tokval
op_ = lambda s: skip(op(s))
id = some(lambda t: t.type in ['Name', 'Number', 'Color', 'String']).named(
'id') >> tokval
date = some(lambda t: t.type == 'Date').named('date') >> tokval
make_node = lambda args: Node(*args)
node_stmt = id + op_(':') + date + maybe(op_('-') + date) >> make_node
chart = (many(node_stmt + skip(maybe(op(';')))) >> Chart)
dotfile = chart + skip(finished)
return dotfile.parse(seq)
def expression(tokens, state):
kw = keyword
parser = (simple_expression
| kw("if") + expression + kw("then") + expression + maybe(
many(kw("elseif") + expression + kw("then") + expression)) +
kw("else") + expression) >> Expression
return parser.run(tokens, state)
def parse(tokens):
'''Parses an SQL date range.
Parses a list of Token object to see if it's a valid SQL
clause meeting the following conditions:
An optional sequence of ANDed simple conditions ANDed with
an optional sequence of ORed complex condtions.
Where a simple condition is a date unit, a sign, and a date value.
And a complex condition is any legal SQL combination of simple
conditions ANDed or ORed together.
Date unit: YYYY, MM, DD, HH, MIN
Sign: <, <=, =, >=, >
Date value: any integer value, with an optional leading zero
Returns:
True if the tokens reprsent a valid SQL date range, False otherwise.
'''
try:
left_paren = some(lambda t: t.value in '(')
right_paren = some(lambda t: t.value in ')')
oper = some(lambda t: t.value in SIGNS)
unit = some(lambda t: t.value in UNITS)
padded_num = some(lambda t: t.code == 2) + some(
lambda t: t.code == 2) # hmmm, better way???
raw_num = some(lambda t: t.code == 2)
num = padded_num | raw_num
cond = unit + oper + num
endmark = a(Token(token.ENDMARKER, ''))
end = skip(endmark + finished)
ands = maybe(cond + maybe(many(a(Token(token.NAME, 'AND')) + cond)))
or_ands = left_paren + ands + right_paren
ors_without_ands = or_ands + maybe(
many(a(Token(token.NAME, 'OR')) + or_ands))
ors_with_ands = (a(Token(token.NAME, 'AND')) + left_paren + or_ands +
maybe(many(a(Token(token.NAME, 'OR')) + or_ands)) +
right_paren)
ors = maybe(ors_without_ands | ors_with_ands)
full = left_paren + ands + ors + right_paren + end
full.parse(tokens)
except NoParseError:
return False
except TokenError:
return False
return True
def element(tokens, state):
kw = keyword
parser = (import_clause | extends_clause
| km('redeclare') + km('final') + km('inner') + km('outer') +
((class_definition | component_clause) | kw('replaceable') +
(class_definition | component_clause) +
maybe(constraining_clause + comment)))
return (parser >> Element).run(tokens, state)
def class_specifier(tokens, state):
normal = (token_type('ident') + string_comment + composition +
keyword('end') + token_type('ident'))
derived = (token_type('ident') + op("=") + base_prefix + name +
maybe(array_subscript) + maybe(class_modification) + comment)
enum_def = (token_type('ident') + op("=") + keyword('enumeration') +
op("(") + (maybe(enum_list) | op(":")) + op(")") + comment)
derivative = (token_type('ident') + op("=") + keyword('der') + op("(") +
name + op(",") + token_type('ident') +
maybe(many(op(",") + token_type('ident'))) + op(")") +
comment)
extended = (keyword('extends') + token_type('ident') +
maybe(class_modification) + string_comment + composition +
keyword("end") + token_type('ident'))
parser = (normal | derived | enum_def | derivative | extended)
return (parser >> ClassSpecifier).run(tokens, state)
def parser(last_error=None):
last_error = LastError() if last_error is None else last_error
def apl(f):
return lambda x: f(*x)
def delim(t):
return skip(_tok(t))
symbol = _tok(Token.SYMBOL) >> _gen(Symbol)
string = _tok(Token.STRING) >> _gen(String)
placeholder = _tok(Token.PLACEHOLDER) >> _gen(Placeholder)
keyword = _tok(Token.KEYWORD) >> _gen(Keyword)
# Note: tokenizer guarantee that value consists of dots and digits
# TODO: convert exceptions
number = _tok(Token.NUMBER) >> _gen(Number, literal_eval)
expr = forward_decl()
implicit_tuple = forward_decl()
list_ = ((_tok(Token.OPEN_BRACKET) + many(expr | keyword) +
_tok(Token.CLOSE_BRACKET)) >> apl(_list))
dict_ = (error_ctx(
_tok(Token.OPEN_BRACE) + many(keyword + expr) +
_tok(Token.CLOSE_BRACE), last_error, DICT_ERROR) >> apl(_dict))
inline_args = many(expr | keyword)
explicit_tuple = (error_ctx(
_tok(Token.OPEN_PAREN) + symbol + inline_args +
_tok(Token.CLOSE_PAREN), last_error, EXPLICIT_TUPLE_ERROR) >>
apl(_tuple))
indented_arg = (
oneplus(implicit_tuple | expr + delim(Token.NEWLINE)) >> _maybe_join)
indented_kwarg = (((keyword + expr + delim(Token.NEWLINE)) |
(keyword + delim(Token.NEWLINE) + delim(Token.INDENT) +
indented_arg + delim(Token.DEDENT))))
indented_args_kwargs = (
(many(indented_kwarg) + many(indented_arg)) >>
apl(lambda pairs, args: list(chain(*(pairs + [args])))))
implicit_tuple.define(
error_ctx(
symbol + inline_args + delim(Token.NEWLINE) + maybe(
delim(Token.INDENT) + indented_args_kwargs +
delim(Token.DEDENT)), last_error, IMPLICIT_TUPLE_ERROR) >> apl(
_implicit_tuple))
expr.define(symbol | string | number | explicit_tuple | list_ | dict_
| placeholder)
body = ((many(implicit_tuple) + _tok(Token.EOF)) >> apl(_module))
return body
def get_marginal_parser():
"""Return parser for tokens describing marginals."""
solution_type = parser.skip(parser.a(Token(token.NAME, 'MAR')))
minus = parser.a(Token(token.OP, '-'))
begin = parser.skip(
parser.maybe(minus + parser.a(Token(token.NAME, 'BEGIN')) + minus))
marginal_parser = (solution_type + parser.many(number_parser + begin) +
end_parser)
return marginal_parser
def between(parser, left=op_('('), right=op_(')'), empty=False):
"""Return a parser that parses an occurrence of parser between left and
right.
Can omit parser if empty is True."""
if empty:
return left + maybe(parser) + right
else:
return left + parser + right
def between(parser, left=op_('('), right=op_(')'), empty=False):
"""Return a parser that parses an occurrence of parser between left and
right.
Can omit parser if empty is True."""
if empty:
return left + maybe(parser) + right
else:
return left + parser + right
def parse(tokens):
## building blocks
kw_priority = some(toktype("kw_priority"))
kw_probability = some(toktype("kw_probability"))
kw_reaction = some(toktype("kw_reaction"))
kw_exists = some(toktype("kw_exists"))
kw_as = some(toktype("kw_as"))
op_tilde = some(toktype("op_tilde"))
op_priority_maximal = some(toktype("op_priority_maximal"))
op_production = some(toktype("op_production"))
atom = some(toktype("name"))
number = some(toktype("number"))
dissolve = some(toktype("op_dissolve"))
osmose = some(toktype("op_osmose"))
osmose_location = some(toktype("op_osmose_location"))
env_open = some(toktype("env_open"))
env_close = some(toktype("env_close"))
membrane_open = some(toktype("membrane_open"))
membrane_close = some(toktype("membrane_close"))
## grammar from the bottom up
name = atom | number
symbol = atom | (dissolve + maybe(name)) | (osmose + name + maybe(osmose_location + name))
priority = kw_priority + op_tilde + name + op_priority_maximal + name
reaction = (kw_reaction + maybe(kw_as + name) + op_tilde +
oneplus(name) + op_production + many(symbol))
exists = kw_exists + op_tilde + oneplus(name)
expr = (exists | reaction | priority)
statement = with_forward_decls(lambda: membrane | expr) >> Statement
body = maybe(name) + many(statement)
membrane = (skip(membrane_open) + body + skip(membrane_close)) >> Membrane
env = (skip(env_open) + body + skip(env_close)) >> Environment
program = many(env) + skip(finished) >> Program
return program.parse(tokens)
def function_arguments(tokens, state):
from modparc.syntax.equations import for_indices # circular dependency
# Since funcparserlib doesn't have full backtracking
# the `named_arguments` parser is matched first to avoid problems
parser = (
named_arguments
| function_argument +
maybe(op(",") + function_arguments
| keyword('for') + for_indices))
return (parser >> FunctionArguments).run(tokens, state)
def parse( tokenSequence ): """Sequence(Token) -> object""" # Top-level Parser expression = scanCode_expression | usbCode_expression | variable_expression | capability_expression | define_expression kll_text = many( expression ) kll_file = maybe( kll_text ) + skip( finished ) return kll_file.parse( tokenSequence )
def parse(tokenSequence):
"""Sequence(Token) -> object"""
# Top-level Parser
expression = ignore_expression | scanCode_expression | usbCode_expression | variable_expression | capability_expression | define_expression
kll_text = many(expression)
kll_file = maybe(kll_text) + skip(finished)
return kll_file.parse(tokenSequence)
def parse(tokens):
## building blocks
kw_priority = some(toktype("kw_priority"))
kw_probability = some(toktype("kw_probability"))
kw_reaction = some(toktype("kw_reaction"))
kw_exists = some(toktype("kw_exists"))
kw_as = some(toktype("kw_as"))
op_tilde = some(toktype("op_tilde"))
op_priority_maximal = some(toktype("op_priority_maximal"))
op_production = some(toktype("op_production"))
atom = some(toktype("name"))
number = some(toktype("number"))
dissolve = some(toktype("op_dissolve"))
osmose = some(toktype("op_osmose"))
osmose_location = some(toktype("op_osmose_location"))
env_open = some(toktype("env_open"))
env_close = some(toktype("env_close"))
membrane_open = some(toktype("membrane_open"))
membrane_close = some(toktype("membrane_close"))
## grammar from the bottom up
name = atom | number
symbol = atom | (dissolve + maybe(name)) | (osmose + name + maybe(osmose_location + name))
priority = kw_priority + op_tilde + name + op_priority_maximal + name
reaction = kw_reaction + maybe(kw_as + name) + op_tilde + oneplus(name) + op_production + many(symbol)
exists = kw_exists + op_tilde + oneplus(name)
expr = exists | reaction | priority
statement = with_forward_decls(lambda: membrane | expr) >> Statement
body = maybe(name) + many(statement)
membrane = (skip(membrane_open) + body + skip(membrane_close)) >> Membrane
env = (skip(env_open) + body + skip(env_close)) >> Environment
program = many(env) + skip(finished) >> Program
return program.parse(tokens)
def parse(input):
period = sometok("period")
string = p.oneplus(sometok("string")) >> (lambda x: " ".join(x))
number = sometok("number")
title = string + p.skip(period) >> RecipeTitle
ingredients_start = sometok("ingredients_start") + p.skip(period) >> IngredientStart
dry_measure = p.maybe(sometok("measure_type")) + sometok("dry_measure")
liquid_measure = sometok("liquid_measure")
mix_measure = sometok("mix_measure")
# is this valid ? 'g of butter', unit w/o initial_value
ingredient = p.maybe(number) + p.maybe(dry_measure | liquid_measure | mix_measure) + string >> unarg(Ingredient)
ingredients = p.many(ingredient)
cooking_time = p.skip(sometok("cooking_time")) + (number >> unarg(CookingTime)) + p.skip(sometok("period"))
oven_temp = p.skip(sometok("oven")) + p.many(number) + p.skip(sometok("oven_temp")) >> unarg(Oven)
method_start = sometok("method_start") + p.skip(period)
comment = p.skip(p.many(string | period))
header = title + p.maybe(comment)
instruction = (string + p.skip(period)) >> parse_instruction
instructions = p.many(instruction)
program = (method_start + instructions) >> unarg(MethodStart)
serves = (sometok("serve") + number >> (lambda x: Serve("serve", x[1]))) + p.skip(period)
ingredients_section = (ingredients_start + ingredients) >> unarg(IngredientSection)
recipe = (
header
+ p.maybe(ingredients_section)
+ p.maybe(cooking_time)
+ p.maybe(oven_temp)
+ p.maybe(program)
+ p.maybe(serves)
) >> RecipeNode
main_parser = p.oneplus(recipe)
return main_parser.parse(tokenize(input))
def _next(y, prefix=_prefix_nl):
if isinstance(y, _OptionalBlock):
if y.avoid:
# stop parsing if we see the distractor
distractor = prefix(y.avoid)
p_next = _not_followed_by(distractor) + prefix(y.p)
else:
p_next = prefix(y.p)
return fp.maybe(p_next)
else:
return prefix(y)
def class_prefixes(tokens, state):
kw = keyword
function_prefix = (maybe(kw("pure") | kw("impure")) + km("operator") +
kw("function"))
parser = (km("partial") +
((kw("class") | kw("model") | km("operator") + kw("record")
| kw("block") |
(km("expandable") + kw("connector")) | kw("type")
| kw("package") | function_prefix | kw("operator"))))
return (parser >> ClassPrefixes).run(tokens, state)
def _parse_rule(seq):
tokval = lambda x: x.value
toktype = lambda t: some(lambda x: x.type == t) >> tokval
sep = lambda s: a(Token(u'Sep', s)) >> tokval
s_sep = lambda s: skip(sep(s))
level = toktype(u'Level')
comparator = toktype(u'Comparator') >> COMPARATORS.get
number = toktype(u'Number') >> float
historical = toktype(u'Historical')
unit = toktype(u'Unit')
operator = toktype(u'Operator')
logical_operator = toktype(u'LogicalOperator') >> LOGICAL_OPERATORS.get
exp = comparator + (
(number + maybe(unit)) | historical) + maybe(operator + number)
rule = (level + s_sep(':') + exp + many(logical_operator + exp))
overall = rule + skip(finished)
return overall.parse(seq)
def _next(y, prefix=_prefix_nl):
if isinstance(y, _OptionalBlock):
if y.avoid:
# stop parsing if we see the distractor
distractor = prefix(y.avoid)
p_next = _not_followed_by(distractor) + prefix(y.p)
else:
p_next = prefix(y.p)
return fp.maybe(p_next)
else:
return prefix(y)
def _create_type_rules():
comma = _token_type("comma")
colon = _token_type("colon")
question_mark = _token_type("question-mark")
bar = _token_type("bar")
equals = _token_type("equals")
attr_name = type_name = arg_name = _token_type("name") >> _make_name
primary_type = forward_decl()
union_type = _one_or_more_with_separator(primary_type, bar) >> _make_union_type
type_ = union_type
type_ref = type_name >> _make_type_ref
applied_type = (
type_ref +
skip(_token_type("open")) +
_one_or_more_with_separator(type_, comma) +
skip(_token_type("close"))
) >> _make_apply
arg = (maybe(question_mark) + maybe(arg_name + skip(colon)) + type_) >> _make_arg
generic_params = maybe(type_name + _token_type("fat-arrow")) >> _make_params
args = _zero_or_more_with_separator(arg, comma)
signature = (generic_params + args + _token_type("arrow") + type_) >> _make_signature
sub_signature = (_token_type("paren-open") + signature + _token_type("paren-close")) >> (lambda result: result[1])
primary_type.define(sub_signature | applied_type | type_ref)
explicit_type = signature | type_
type_definition = (type_name + skip(equals) + type_ + skip(finished)) >> _make_type_definition
structural_type_attr = (attr_name + skip(colon) + explicit_type) >> tuple
structural_type_attrs = many(structural_type_attr)
structural_type_definition = (type_name + skip(colon) + structural_type_attrs + skip(finished)) >> _make_structural_type_definition
generic = (_one_or_more_with_separator(type_name, comma) + skip(finished)) >> _make_generic
return explicit_type + skip(finished), type_definition, structural_type_definition, generic
def create_grammar():
tokval = lambda x: x.value
toktype = lambda t: some(lambda x: x.type == t) >> tokval
op = lambda s: a(Token('Op', s)) >> tokval
op_ = lambda s: skip(op(s))
n = lambda s: a(Token('Name', s)) >> tokval
null = n('null')
true = n('true')
false = n('false')
number = toktype('Number')
string = toktype('String')
value = forward_decl()
member = string + op_(':') + value
object_ = (op_('{') + maybe(member + many(op_(',') + member)) + op_('}'))
array = (op_('[') + maybe(value + many(op_(',') + value)) + op_(']'))
value.define(null | true | false | object_ | array | number | string)
json_text = object_ | array
json_file = json_text + skip(finished)
return json_file
def _parse_rule(seq):
tokval = lambda x: x.value
toktype = lambda t: some(lambda x: x.type == t) >> tokval
sep = lambda s: a(Token(u'Sep', s)) >> tokval
s_sep = lambda s: skip(sep(s))
level = toktype(u'Level')
comparator = toktype(u'Comparator') >> COMPARATORS.get
number = toktype(u'Number') >> float
historical = toktype(u'Historical')
unit = toktype(u'Unit')
operator = toktype(u'Operator')
logical_operator = toktype(u'LogicalOperator') >> LOGICAL_OPERATORS.get
exp = comparator + ((number + maybe(unit)) | historical) + maybe(operator + number)
rule = (
level + s_sep(':') + exp + many(logical_operator + exp)
)
overall = rule + skip(finished)
return overall.parse(seq)
def _parse(seq):
const = lambda x: lambda _: x
tokval = lambda x: x.value
toktype = lambda t: some(lambda x: x.type == t) >> tokval
op = lambda s: a(Token(u'Op', s)) >> tokval
op_ = lambda s: skip(op(s))
def make_string(args):
context, value = args
if not context: context = 'any:'
return String(unescape_str(value[1:-1]), context[:-1])
def make_regex(args):
context, value = args
value, modifiers = value.rsplit('/', 1)
value = value[1:]
if not context: context = 'any:'
return Regex(unescape_regex(value), modifiers, context[:-1])
def make_or(args):
return Or(*args)
def make_and(args):
return And(*args)
def make_not(x):
return Not(x)
context = maybe(toktype(u'Prefix'))
string = (context + toktype(u'String')) >> make_string
regex = (context + toktype(u'Regex')) >> make_regex
par_term = forward_decl()
simple_term = forward_decl()
term = forward_decl()
not_term = forward_decl()
and_term = forward_decl()
or_term = forward_decl()
par_term.define(op_(u'(') + term + op_(u')'))
simple_term.define(par_term | string | regex)
not_term.define(op_('not') + not_term >> make_not | simple_term)
and_term.define(not_term + op_('and') + and_term >> make_and | not_term)
or_term.define(and_term + op_('or') + or_term >> make_or | and_term)
term.define(or_term)
eof = skip(toktype(u'EOF'))
filter_expr = (term + eof) | (eof >> const(Any()))
return filter_expr.parse(seq)
def primary(tokens, state):
kw = keyword
parser = (token_type("number")
| token_type("string")
| kw("false")
| kw("true")
| (name | kw("der") | kw("initial")) + function_call_args
| component_reference
| op("(") + output_expression_list + op(")")
| (op("[") + expression_list +
maybe(many(op(";") + expression_list)) + op("]"))
| op("{") + function_arguments + op("}")
| kw("end"))
return (parser >> Primary).run(tokens, state)
def parse(seq):
"""
Parses the list of tokens and generates an AST.
"""
def eval_expr(z, list):
return reduce(lambda s, (f, x): f(s, x), list, z)
unarg = lambda f: lambda x: f(*x)
tokval = lambda x: x.value # returns the value of a token
toktype = lambda t: some(lambda x: x.type == t) >> tokval # checks type of token
const = lambda x: lambda _: x # like ^^^ in Scala
op = lambda s: a(Token('Op', s)) >> tokval # return the value if token is Op
op_ = lambda s: skip(op(s)) # checks if token is Op and ignores it
lst = lambda x: [x[0],] + x[1]
tup = lambda x: (x[0], x[1])
makeop = lambda s, f: op(s) >> const(f)
add = makeop('+', Add)
sub = makeop('-', Sub)
mul = makeop('*', Mul)
div = makeop('/', Div)
lt = makeop('<', Lt)
gt = makeop('>', Gt)
eq = makeop('=', Eq)
operation = add | sub | mul | div | lt | gt | eq
decl = with_forward_decls(lambda:toktype('Var') + op_('=') + (exp | fun) >> tup)
decls = decl + many(skip(toktype('Semicolon')) + decl) >> lst
variable = toktype('Var') >> Variable
variables = variable + many(skip(toktype('Comma')) + variable) >> lst
fun = with_forward_decls(lambda: skip(toktype('Fun')) + variables + skip(toktype('Arrow')) + exp + skip(toktype('End'))) >> unarg(Fun)
parameters = with_forward_decls(lambda: exp + many(skip(toktype('Comma')) + exp) >> lst)
call = skip(toktype('Call')) + (fun | variable) + skip(toktype('Lp')) + parameters + skip(toktype('Rp')) >> unarg(Call)
ex = with_forward_decls(lambda:variable | toktype('Number') >> (lambda x: Const(int(x))) |\
toktype('True') >> (lambda x: Const(True)) | toktype('False') >> (lambda x: Const(False)) |\
skip(toktype('Let')) + decls + skip(toktype('In')) + exp + skip(toktype('End')) >> unarg(Let) |\
skip(toktype('If')) + exp + skip(toktype('Then')) + exp + maybe(skip(toktype('Else')) + exp) + skip(toktype('Fi')) >> unarg(If) |\
fun | call)
exp = ex + many(operation + ex) >> unarg(eval_expr)
prog = skip(toktype('Prog')) + exp >> Prog
return prog.parse(seq)
def parse(source):
task = Task()
get_value = lambda x: x.value
value_of = lambda t: some(lambda x: x.type == t) >> get_value
keyword = lambda s: skip(value_of(s))
make_rule = lambda x: task.add_rule(Rule(**{x[0]: x[1][1:-1]}))
set_root = lambda value: task.set_root_dir(value[1:-1])
set_mask = lambda value: task.set_mask(value[1:-1])
root = keyword('In') + value_of('Value') >> set_root
mask = keyword('With') + value_of('Value') >> set_mask
rule = keyword('Set') + \
value_of('Attribute') + \
keyword('Equals') + \
value_of('Value') \
>> make_rule
parser = maybe(mask) + root + many(rule)
parser.parse(source)
return task
def parse(seq):
"""Sequence(Token) -> object"""
unarg = lambda f: lambda args: f(*args)
tokval = lambda x: x.value
flatten = lambda list: sum(list, [])
n = lambda s: a(Token(u'Name', s)) >> tokval
op = lambda s: a(Token(u'Op', s)) >> tokval
op_ = lambda s: skip(op(s))
id_types = [u'Name', u'Number', u'String']
id = some(lambda t: t.type in id_types).named(u'id') >> tokval
make_graph_attr = lambda args: DefAttrs(u'graph', [Attr(*args)])
make_edge = lambda x, xs, attrs: Edge([x] + xs, attrs)
node_id = id # + maybe(port)
a_list = (
id + maybe(op_(u'=') + id) + skip(maybe(op(u','))) >> unarg(Attr))
attr_list = (many(op_(u'[') + many(a_list) + op_(u']')) >> flatten)
attr_stmt = (
(n(u'graph') | n(u'node') | n(u'edge')) + attr_list >> unarg(DefAttrs))
graph_attr = id + op_(u'=') + id >> make_graph_attr
node_stmt = node_id + attr_list >> unarg(Node)
# We use a forward_decl becaue of circular definitions like (stmt_list ->
# stmt -> subgraph -> stmt_list)
subgraph = forward_decl()
edge_rhs = skip(op(u'->') | op(u'--')) + (subgraph | node_id)
edge_stmt = ((subgraph | node_id) + oneplus(edge_rhs) + attr_list >>
unarg(make_edge))
stmt = (attr_stmt | edge_stmt | subgraph | graph_attr | node_stmt)
stmt_list = many(stmt + skip(maybe(op(u';'))))
subgraph.define(
skip(n(u'subgraph')) + maybe(id) + op_(u'{') + stmt_list +
op_(u'}') >> unarg(SubGraph))
graph = (maybe(n(u'strict')) + maybe(n(u'graph') | n(u'digraph')) +
maybe(id) + op_(u'{') + stmt_list + op_(u'}') >> unarg(Graph))
dotfile = graph + skip(finished)
return dotfile.parse(seq)
lambda: lparen + expr + rparen
)
# *Mark here are not really required, but if you are going to do
# anything complex that requires that you discern between different
# parsing paths, marks are often give you least hassle.
expr = with_forward_decls(
lambda:
(number + pure(NumberMark) + expr_rest |
paren_expr + pure(ParenMark) + expr_rest) >> make_expr)
# This one allows us to add more complex expressions like function
# application and ternary operators to the above definition with ease.
# Otherwise terms such as `apply = expr lparen many(expr) rpanen`
# would be impossible to add, always leading to infinite left recursion.
expr_rest = maybe(op + expr)
toplev = expr + skip(eof)
@py.test.mark.parametrize("given, parser, expected", [
("1", number, Number("1")),
("+", op, "+"),
("-", op, "-"),
("*", op, "*"),
("^", op, "^"),
])
def test_parse_primitives(given, parser, expected):
data = parser.parse(list(tokenize(given))[:-1])
assert data == expected
return (n[0], n[1])
def make_object(n):
return PLObject(n)
def make_name(n):
assert len(n) == 2
return PLName(n[0], n[1])
def make_array(n):
return PLArray(n)
number = sometok('number') >> make_number
string = sometok('string') >> make_string
comment = sometok('comment') >> make_comment
name = (sometok('name') | sometok('string')) + maybe(comment) >> make_name
value = fwd()
member = name + op_('=') + value >> make_member
section = maybe(comment) + oneplus(member + op_(';')) + maybe(comment) >> make_section
object = (
op_('{') +
many(section) +
op_('}')) >> make_object
array = (
op_('(') +
many(value + op_(',')) +
op_(')')) >> make_array
value.define(
object
| number
def parse(seq):
"""Sequence(Token) -> object"""
const = lambda x: lambda _: x
tokval = lambda x: x.value
toktype = lambda t: some(lambda x: x.type == t) >> tokval
op = lambda s: a(Token('Op', s)) >> tokval
op_ = lambda s: skip(op(s))
n = lambda s: a(Token('Name', s)) >> tokval
def make_array(n):
if n is None:
return []
else:
return [n[0]] + n[1]
def make_object(n):
return dict(make_array(n))
def make_number(n):
try:
return int(n)
except ValueError:
return float(n)
def unescape(s):
std = {
'"': '"', '\\': '\\', '/': '/', 'b': '\b', 'f': '\f',
'n': '\n', 'r': '\r', 't': '\t',
}
def sub(m):
if m.group('standard') is not None:
return std[m.group('standard')]
else:
return chr(int(m.group('unicode'), 16))
return re_esc.sub(sub, s)
def make_string(n):
return unescape(n[1:-1])
null = n('null') >> const(None)
true = n('true') >> const(True)
false = n('false') >> const(False)
number = toktype('Number') >> make_number
string = toktype('String') >> make_string
value = forward_decl()
member = string + op_(':') + value >> tuple
object = (
op_('{') +
maybe(member + many(op_(',') + member)) +
op_('}')
>> make_object)
array = (
op_('[') +
maybe(value + many(op_(',') + value)) +
op_(']')
>> make_array)
value.define(
null
| true
| false
| object
| array
| number
| string)
json_text = object | array
json_file = json_text + skip(finished)
return json_file.parse(seq)
raise p.NoParseError("parsing failed", s)
_not_parser.name = "!{}".format(parser.name)
return _not_parser
a = lambda value: p.some(lambda tok: tok.value == value)
string = lambda s: p.some(lambda tok: tok.value.lower() == s.lower()).named(s)
some_type = lambda t: p.some(lambda tok: tok.type == t).named(t)
not_type = lambda t: p.some(lambda tok: tok.type != t).named("!{}".format(t))
any_type = p.some(lambda _: True).named("Any")
eof = p.finished.named("")
some = p.some
maybe = lambda parser: p.maybe(parser).named("[{}]".format(parser.name))
many = lambda parser: p.many(parser).named("[{}]...".format(parser.name))
skip = lambda parser: p.skip(parser).named("")
oneplus = lambda parser: p.oneplus(parser).named("{},[{}]...".format(parser.name, parser.name))
sparsed = lambda parser: (skip(many(not_parser(parser))) + parser).named("_{}".format(parser.name))
integer = (some_type(TokenType.Integer) >> to_i).named("I")
number = ((some_type(TokenType.Integer) | some_type(TokenType.Float)) >> to_f).named("N")
word = (some_type(TokenType.Word) >> to_s).named("W")
mention = (some_type(TokenType.Mention) >> extract_mention_id).named("M")
channel = (some_type(TokenType.Channel) >> extract_channel_id).named("C")
emoji = (some_type(TokenType.Emoji) >> extract_emoji_id).named("E")
snippet = (some_type(TokenType.Snippet) >> extract_snippet).named("S")
# High level helpers
on_off_switch = (string("on") >> const(True)) | (string("off") >> const(False))
def evaluate(expression, environment):
"""Evaluate an expression in the specified variable environment."""
# Well known functions
const = lambda x: lambda _: x
unarg = lambda f: lambda args: f(*args)
# Semantic actions and auxiliary functions
tokval = lambda tok: tok.value
makeop = lambda s, f: op(s) >> const(f)
sometok = lambda type: some(lambda tok: tok.type == type)
def eval_name(s):
try:
return environment[s] # Case-sensitive
except KeyError:
raise ValueError('unbound variable: %s' % s)
def make_number(s):
try:
return int(s)
except ValueError:
return float(s)
def eval_expr(expr, op_expr_pairs):
result = expr
for op, expr in op_expr_pairs:
result = op(result, expr)
return result
def eval_call(func_name, maybe_expr_and_exprs):
if maybe_expr_and_exprs:
expr, exprs = maybe_expr_and_exprs
args = [expr] + exprs
else:
args = []
f = eval_name(func_name)
if not callable(f):
raise TypeError('variable is not callable: %s' % func_name)
argcount = len(args)
f_argcount = f.func_code.co_argcount
if f_argcount != argcount:
raise TypeError('%s takes %d arguments (%d given)' %
(func_name, f_argcount, argcount))
return f(*args)
# Primitives
number = (
sometok('number')
>> tokval
>> make_number)
raw_name = sometok('name') >> tokval
name = raw_name >> eval_name
op = lambda s: a(Token('op', s)) >> tokval
op_ = lambda s: skip(op(s))
add = makeop('+', operator.add)
sub = makeop('-', operator.sub)
mul = makeop('*', operator.mul)
div = makeop('/', operator.div)
mul_op = mul | div
add_op = add | sub
# Means of composition
expr = forward_decl()
call = (
raw_name + op_('(') + maybe(expr + many(op_(',') + expr)) + op_(')')
>> unarg(eval_call))
primary = (
number
| call
| name
| op_('(') + expr + op_(')'))
term = (
primary + many(mul_op + primary)
>> unarg(eval_expr))
expr.define(
term + many(add_op + term)
>> unarg(eval_expr))
# Toplevel parsers
toplevel = maybe(expr) + skip(finished)
return toplevel.parse(tokenize(expression))
def parse(seq):
"""Sequence(Token) -> object"""
unarg = lambda f: lambda args: f(*args)
tokval = lambda x: x.value
flatten = lambda list: sum(list, [])
n = lambda s: a(Token('Name', s)) >> tokval
op = lambda s: a(Token('Op', s)) >> tokval
op_ = lambda s: skip(op(s))
id_types = ['Name', 'Number', 'String']
id = some(lambda t: t.type in id_types).named('id') >> tokval
make_graph_attr = lambda args: DefAttrs('graph', [Attr(*args)])
make_edge = lambda x, xs, attrs: Edge([x] + xs, attrs)
node_id = id # + maybe(port)
a_list = (
id +
maybe(op_('=') + id) +
skip(maybe(op(',')))
>> unarg(Attr))
attr_list = (
many(op_('[') + many(a_list) + op_(']'))
>> flatten)
attr_stmt = (
(n('graph') | n('node') | n('edge')) +
attr_list
>> unarg(DefAttrs))
graph_attr = id + op_('=') + id >> make_graph_attr
node_stmt = node_id + attr_list >> unarg(Node)
# We use a forward_decl becaue of circular definitions like (stmt_list ->
# stmt -> subgraph -> stmt_list)
subgraph = forward_decl()
edge_rhs = skip(op('->') | op('--')) + (subgraph | node_id)
edge_stmt = (
(subgraph | node_id) +
oneplus(edge_rhs) +
attr_list
>> unarg(make_edge))
stmt = (
attr_stmt
| edge_stmt
| subgraph
| graph_attr
| node_stmt
)
stmt_list = many(stmt + skip(maybe(op(';'))))
subgraph.define(
skip(n('subgraph')) +
maybe(id) +
op_('{') +
stmt_list +
op_('}')
>> unarg(SubGraph))
graph = (
maybe(n('strict')) +
maybe(n('graph') | n('digraph')) +
maybe(id) +
op_('{') +
stmt_list +
op_('}')
>> unarg(Graph))
dotfile = graph + skip(finished)
return dotfile.parse(seq)
Return a parser that tries to parse p, and raises a CustomParseError
when it fails.
"""
@Parser
def _try_p(tokens, s):
try:
return p.run(tokens, s)
except NoParseError as err:
raise CustomParseError(msg, err)
return _try_p
# Grammar starts here
test = with_forward_decls(lambda: choice([
or_test + maybe(IF + or_test + ELSE + test),
lambdef]))
not_test = with_forward_decls(lambda: (NOT + not_test) | comparison)
and_test = sep_by(not_test, AND)
or_test = sep_by(and_test, OR)
test_nocond = with_forward_decls(lambda: or_test | lambdef_nocond)
testlist = with_forward_decls(lambda: sep_by(test))
atom = with_forward_decls(lambda: choice([
between(yield_expr | testlist_comp, empty=True),
between(testlist_comp, op_('['), op_(']'), empty=True),
between(dictorsetmaker, op_('{'), op_('}'), empty=True),
IDENTIFIER,
NUMBER,
STRING + many(STRING),
def sep_by(parser, sep=COMMA, trailing=True):
"""Parses one or more instances of p separated by sep."""
p = parser + many(sep + parser)
if trailing:
p += maybe(sep)
return p
unseqString = tokenType('SequenceString') >> Make.unseqString # For use with variables
colRowOperator = lambda s: a( Token( 'ColRowOperator', s ) )
relCROperator = lambda s: a( Token( 'RelCROperator', s ) )
pixelOperator = tokenType('PixelOperator')
# Code variants
code_begin = tokenType('CodeBegin')
code_end = tokenType('CodeEnd')
# Specifier
specifier_basic = ( timing >> Make.specifierTiming ) | ( name >> Make.specifierState )
specifier_complex = ( name + skip( operator(':') ) + timing ) >> unarg( Make.specifierState )
specifier_state = specifier_complex | specifier_basic
specifier_analog = number >> Make.specifierAnalog
specifier_list = skip( parenthesis('(') ) + many( ( specifier_state | specifier_analog ) + skip( maybe( comma ) ) ) + skip( parenthesis(')') )
# Scan Codes
scanCode_start = tokenType('ScanCodeStart')
scanCode_range = number + skip( dash ) + number >> Make.scanCode_range
scanCode_listElem = number >> Make.scanCode
scanCode_specifier = ( scanCode_range | scanCode_listElem ) + maybe( specifier_list ) >> unarg( Make.specifierUnroll )
scanCode_innerList = many( scanCode_specifier + skip( maybe( comma ) ) ) >> flatten
scanCode_expanded = skip( scanCode_start ) + scanCode_innerList + skip( code_end ) + maybe( specifier_list ) >> unarg( Make.specifierUnroll )
scanCode_elem = scanCode + maybe( specifier_list ) >> unarg( Make.specifierUnroll )
scanCode_combo = oneplus( ( scanCode_expanded | scanCode_elem ) + skip( maybe( plus ) ) )
scanCode_sequence = oneplus( scanCode_combo + skip( maybe( comma ) ) )
scanCode_single = ( skip( scanCode_start ) + scanCode_listElem + skip( code_end ) ) | scanCode
scanCode_il_nospec = oneplus( ( scanCode_range | scanCode_listElem ) + skip( maybe( comma ) ) )
scanCode_nospecifier = skip( scanCode_start ) + scanCode_il_nospec + skip( code_end )
def parser(last_error=None):
last_error = LastError() if last_error is None else last_error
def apl(f):
return lambda x: f(*x)
def delim(t):
return skip(_tok(t))
symbol = _tok(Token.SYMBOL) >> _gen(Symbol)
string = _tok(Token.STRING) >> _gen(String)
placeholder = _tok(Token.PLACEHOLDER) >> _gen(Placeholder)
keyword = _tok(Token.KEYWORD) >> _gen(Keyword)
# Note: tokenizer guarantee that value consists of dots and digits
# TODO: convert exceptions
number = _tok(Token.NUMBER) >> _gen(Number, literal_eval)
expr = forward_decl()
implicit_tuple = forward_decl()
list_ = (
(_tok(Token.OPEN_BRACKET) +
many(expr | keyword) +
_tok(Token.CLOSE_BRACKET)) >>
apl(_list)
)
dict_ = (
error_ctx(_tok(Token.OPEN_BRACE) +
many(keyword + expr) +
_tok(Token.CLOSE_BRACE),
last_error,
DICT_ERROR) >>
apl(_dict)
)
inline_args = many(expr | keyword)
explicit_tuple = (
error_ctx(_tok(Token.OPEN_PAREN) +
symbol + inline_args +
_tok(Token.CLOSE_PAREN),
last_error,
EXPLICIT_TUPLE_ERROR) >>
apl(_tuple)
)
indented_arg = (
oneplus(implicit_tuple | expr + delim(Token.NEWLINE)) >>
_maybe_join
)
indented_kwarg = (
((keyword + expr + delim(Token.NEWLINE)) |
(keyword + delim(Token.NEWLINE) +
delim(Token.INDENT) + indented_arg + delim(Token.DEDENT)))
)
indented_args_kwargs = (
(many(indented_kwarg) + many(indented_arg)) >>
apl(lambda pairs, args: list(chain(*(pairs + [args]))))
)
implicit_tuple.define(
error_ctx(symbol + inline_args + delim(Token.NEWLINE) +
maybe(delim(Token.INDENT) +
indented_args_kwargs +
delim(Token.DEDENT)),
last_error,
IMPLICIT_TUPLE_ERROR) >>
apl(_implicit_tuple)
)
expr.define(symbol | string | number | explicit_tuple | list_ | dict_ |
placeholder)
body = (
(many(implicit_tuple) + _tok(Token.EOF)) >>
apl(_module)
)
return body
application = forward_decl()
renaming = forward_decl()
set_context = forward_decl()
# string expression
strexpr = string + many(add_op + string) >> u(concatenate)
# numerical expression
numeric = application | lambda_ | number | name | (op_('(') + nexpr + op_(')'))
factor = numeric + many(power + numeric) >> u(make_expression)
term = factor + many(mul_op + factor) >> u(make_expression)
nexpr.define( term + many(add_op + term) >> u(make_expression) )
# boolean expression
proposition = (sexpr | strexpr | nexpr)
pred = proposition + maybe(cmp_op + proposition) >> u(make_predicate)
formula = true | false | pred | (op_('(') + bexpr + op_(')'))
conjunction = formula + many(and_ + formula) >> u(make_expression)
disjunction = conjunction + many(or_ + conjunction) >> u(make_expression)
bexpr.define( maybe(not_) + disjunction >> make_boolean )
# set expression
enumeration = op_('{') + maybe(expr + many(op_(',') + expr)) + op_('}') >> make_enum
range_ = op_('{') + nexpr + op_(':') + nexpr + op_('}') >> u(make_range)
set_ = op_('{') + expr + maybe(kw_('for') + set_context) + op_('}') >> u(make_set)
sexpr.define( enumeration | range_ | set_ )
# anonymous function
lambda_.define( op_('[') + maybe(name + many(op_(',') + name)) + op_(':') + expr + op_(']')
>> make_lambda )
# =========== Expressions parser
# FIXME: it should be rewritten using full Lua 5.2 grammar.
BINARY_OPS = set("+-*/^%><") | {"..", "==", "~=", ">=", "<=", "and", "or"}
UNARY_OPS = {"not", "-", "#"}
binary_op = p.some(lambda t: t.value in BINARY_OPS) >> token_value
unary_op = p.some(lambda t: t.value in UNARY_OPS) >> token_value
# expressions with binary and unary ops + parenthesis
@p.with_forward_decls
def value():
single_value = table | tok_number | tok_string | tok_constant | iden
return single_value | (close_rnd_brace + expr + open_rnd_brace)
_term = value + p.skip(p.maybe(unary_op))
expr = _term + p.many(binary_op + _term) >> flat
# [expression]
_index_lookup = p.skip(close_sq_brace) + expr + p.skip(open_sq_brace)
# foo=expr
# [foo]=expr
_key = iden | _index_lookup
_keyvalue = expr + token("=") + _key
# foo=expr, ["bar"]=42,
_table_sep = token(",") | token(";")
table_parameters = (
p.maybe(_table_sep) + # allow trailing comma/semicolon
(_keyvalue | expr) +
dash = tokenType("Dash")
plus = tokenType("Plus")
content = tokenType("VariableContents")
string = tokenType("String") >> make_string
unString = tokenType("String") # When the double quotes are still needed for internal processing
seqString = tokenType("SequenceString") >> make_seqString
unseqString = tokenType("SequenceString") >> make_unseqString # For use with variables
# Code variants
code_end = tokenType("CodeEnd")
# Scan Codes
scanCode_start = tokenType("ScanCodeStart")
scanCode_range = number + skip(dash) + number >> make_scanCode_range
scanCode_listElem = number >> listElem
scanCode_innerList = oneplus((scanCode_range | scanCode_listElem) + skip(maybe(comma))) >> flatten
scanCode_expanded = skip(scanCode_start) + scanCode_innerList + skip(code_end)
scanCode_elem = scanCode >> listElem
scanCode_combo = oneplus((scanCode_expanded | scanCode_elem) + skip(maybe(plus)))
scanCode_sequence = oneplus(scanCode_combo + skip(maybe(comma)))
# USB Codes
usbCode_start = tokenType("USBCodeStart")
usbCode_number = number >> make_usbCode_number
usbCode_range = (usbCode_number | unString) + skip(dash) + (number | unString) >> make_usbCode_range
usbCode_listElemTag = unString >> make_usbCode
usbCode_listElem = (usbCode_number | usbCode_listElemTag) >> listElem
usbCode_innerList = oneplus((usbCode_range | usbCode_listElem) + skip(maybe(comma))) >> flatten
usbCode_expanded = skip(usbCode_start) + usbCode_innerList + skip(code_end)
usbCode_elem = usbCode >> listElem
usbCode_combo = oneplus((usbCode_expanded | usbCode_elem) + skip(maybe(plus))) >> listElem
from __future__ import absolute_import
import funcparserlib.parser as p
string = lambda x: x or ''
cat = ''.join
negative = p.maybe(p.a('-')) >> string
digits = p.oneplus(p.some(lambda char: char.isdigit())) >> cat
decimal_part = (p.maybe(p.a('.') + digits)) >> string >> cat
number = (negative + digits + decimal_part) >> cat >> float
addition = number + p.skip(p.a('+')) + number >> sum
expression = addition | number
expression = expression + p.finished
def calculate(text):
return expression.parse(text)[0]
rest = parsed[1]
rest.insert(0, first)
return rest
return ( parser + many(by + parser) ) >> append
"""
E.g.
"An individual carrying two gain-of-function alleles or one functional allele and one gain-of-function allele"
=> [('gain-of-function', 'gain-of-function'), ('functional', 'gain-of-function')]
"""
_phenotype_genotype = (
skip( _ipattern(r'an', r'individual', r'carrying') ) +
_separated(
( _two_alleles >> (lambda allele: ( allele, allele )) ) |
( _one_allele + skip(_ipattern(r'and')) + _one_allele ),
skip(maybe(_ipattern(r'or')))
)
)
_parsers = {
'phenotype_genotype': _phenotype_genotype,
}
class ParserError(Exception):
pass
def parse(parser_name, string):
"""
Invoke the parser identified by ``parser_name`` on the provided string.
Parsers:
* **phenotype_genotype:**
return s
element = forward_decl()
elements = forward_decl()
skip_of = skip_str('of')
any_class = some(lambda x: re.match('[a-zA-Z0-9_]+', x.string)) >> to_simple_type
set_parser = (a_str('set') + skip_of + element) >> compile_set_type
list_parser = (a_str('list') + skip_of + element) >> compile_list_type
dict_parser = (a_str('dict') + skip_of + skip_str('{') + element + skip_str(',') + element + skip_str('}')) >> \
compile_dict_type
tuple_parser = (a_str('tuple') + skip_of + skip_str('(') + elements + skip_str(')')) >> compile_tuple_type
element.define(set_parser | list_parser | dict_parser | tuple_parser | any_class)
elements.define((many(element + skip_str(',')) + element) >>
(lambda x: x[0] + [x[1]]))
type_contract_parser = skip_str('(') + maybe(elements) + skip_str(')') + skip_str('->') + element
docstring_description = many(some(lambda token: '>>>' not in token.line)) >> (lambda tokens: ' '.join(token.string for token in tokens))
docstring_doctest = many(some(lambda token: True)) >> (lambda tokens: ' '.join(token.string for token in tokens))
entire_docstring = maybe(type_contract_parser) + docstring_description +\
docstring_doctest + finished
def parse_csc108_docstring(docstring):
"""Reads a docstring in the CSC108 format and extracts the argument types.
@param str docstring: The docstring to read.
@return: A parsed output of the docstring.
"""
output = list(generate_tokens(StringIO(docstring.strip()).readline))
def parse_instruction(spec):
string = p.oneplus(sometok("string")) >> (lambda x: " ".join(x))
ordinal = sometok("ordinal")
bowl = sometok("bowl")
the = sometok("the")
dish = sometok("dish")
to = sometok("to")
into = sometok("into")
concat = lambda list: " ".join(list)
take_i = sometok("take") + (p.oneplus(string) >> concat) + sometok("from") + sometok("refrigerator")
put_i = (
sometok("put")
+ p.skip(p.maybe(the))
+ (p.oneplus(string) >> concat)
+ p.skip(into)
+ p.maybe(ordinal | the)
+ bowl
)
liquefy_1 = sometok("liquefy") + sometok("contents") + p.maybe(ordinal) + bowl
liquefy_2 = sometok("liquefy") + (p.oneplus(string) >> concat)
liquefy_i = liquefy_1 | liquefy_2
pour_i = (
sometok("pour")
+ sometok("contents")
+ p.maybe(ordinal)
+ bowl
+ sometok("into")
+ the
+ p.maybe(ordinal)
+ dish
)
fold_i = (
sometok("fold") + p.skip(p.maybe(the)) + (p.oneplus(string) >> concat) + into + p.maybe(ordinal | the) + bowl
)
# cleanup repitition
add_i = sometok("add") + (p.oneplus(string) >> concat) + p.maybe(to + p.maybe(ordinal | the) + bowl)
remove_i = (
sometok("remove") + (p.oneplus(string) >> concat) + p.maybe(sometok("from") + p.maybe(ordinal | the) + bowl)
)
combine_i = sometok("combine") + (p.oneplus(string) >> concat) + p.maybe(into + p.maybe(ordinal | the) + bowl)
divide_i = sometok("divide") + (p.oneplus(string) >> concat) + p.maybe(into + p.maybe(ordinal | the) + bowl)
add_dry_i = sometok("add_dry") + p.maybe(to + p.maybe(ordinal | the) + bowl)
stir_1 = (
sometok("stir")
+ p.maybe(the + p.maybe(ordinal | the) + bowl)
+ sometok("for")
+ sometok("number")
+ (sometok("minute") | sometok("minutes"))
)
stir_2 = sometok("stir") + (p.oneplus(string) >> concat) + into + the + p.maybe(ordinal) + bowl
stir_i = stir_1 | stir_2
mix_i = sometok("mix") + p.maybe(the + p.maybe(ordinal) + bowl) + sometok("well")
clean_i = sometok("clean") + p.maybe(ordinal | the) + bowl
loop_start_i = (sometok("string") + p.maybe(the) + (p.oneplus(string) >> concat)) >> (lambda x: ("loop_start", x))
loop_end_i = (
sometok("string") + p.maybe(p.maybe(the) + (p.oneplus(string) >> concat)) + sometok("until") + string
) >> (lambda x: ("loop_end", x))
set_aside_i = sometok("set") >> (lambda x: (x, None))
serve_with_i = sometok("serve_with") + (p.oneplus(string) >> concat)
refrigerate_i = sometok("refrigerate") + p.maybe(
sometok("for") + sometok("number") + (sometok("hour") | sometok("hours"))
)
instruction = (
take_i
| put_i
| liquefy_i
| pour_i
| add_i
| fold_i
| remove_i
| combine_i
| divide_i
| add_dry_i
| stir_i
| mix_i
| clean_i
| loop_end_i # -| ORDER matters
| loop_start_i # -|
| set_aside_i
| serve_with_i
| refrigerate_i
) >> (lambda x: Instruction(x[0].lower().replace(" ", "_"), x[1:]))
return instruction.parse(tokenize_instruction(spec))
def parse(seq):
"""Sequence(Token) -> object"""
id_tokens = ['Name', 'IPAddr', 'Number', 'String']
tokval = lambda x: x.value
op = lambda s: a(Token('Op', s)) >> tokval
op_ = lambda s: skip(op(s))
_id = some(lambda t: t.type in id_tokens) >> tokval
keyword = lambda s: a(Token('Name', s)) >> tokval
def make_peer(first, edge_type, second, followers, attrs):
edges = [Edge(first, edge_type, second, attrs)]
from_node = second
for edge_type, to_node in followers:
edges.append(Edge(from_node, edge_type, to_node, attrs))
from_node = to_node
return Peer(edges)
def make_route(first, edge_type, second, followers, attrs):
edges = [Edge(first, edge_type, second, attrs)]
from_node = second
for edge_type, to_node in followers:
edges.append(Edge(from_node, edge_type, to_node, attrs))
from_node = to_node
return Route(edges)
#
# parts of syntax
#
node_list = (
_id +
many(op_(',') + _id)
>> create_mapper(oneplus_to_list)
)
option_stmt = (
_id +
maybe(op_('=') + _id)
>> create_mapper(Attr)
)
option_list = (
maybe(op_('[') + option_stmt + many(op_(',') + option_stmt) + op_(']'))
>> create_mapper(oneplus_to_list, default_value=[])
)
# node statement::
# A;
# B [attr = value, attr = value];
#
node_stmt = (
_id + option_list
>> create_mapper(Node)
)
# peer network statement::
# A -- B;
#
edge_stmt = (
_id +
op('--') +
_id +
many(op('--') + _id) +
option_list
>> create_mapper(make_peer)
)
# attributes statement::
# default_shape = box;
# default_fontsize = 16;
#
attribute_stmt = (
_id + op_('=') + _id
>> create_mapper(Attr)
)
# extension statement (class, plugin)::
# class red [color = red];
# plugin attributes [name = Name];
#
extension_stmt = (
(keyword('class') | keyword('plugin')) +
_id +
option_list
>> create_mapper(Extension)
)
# group statement::
# group {
# A;
# }
#
group_inline_stmt = (
attribute_stmt |
node_stmt
)
group_inline_stmt_list = (
many(group_inline_stmt + skip(maybe(op(';'))))
)
group_stmt = (
skip(keyword('group')) +
maybe(_id) +
op_('{') +
group_inline_stmt_list +
op_('}')
>> create_mapper(Group)
)
# network statement::
# network {
# A;
# }
#
network_inline_stmt = (
attribute_stmt |
group_stmt |
node_stmt
)
network_inline_stmt_list = (
many(network_inline_stmt + skip(maybe(op(';'))))
)
network_stmt = (
skip(keyword('network')) +
maybe(_id) +
op_('{') +
network_inline_stmt_list +
op_('}')
>> create_mapper(Network)
)
# route statement::
# route {
# A -> B -> C;
# }
#
route_inline_stmt = (
_id +
op_('->') +
_id +
many(op_('->') + _id) +
option_list
>> create_mapper(make_route)
)
route_stmt = (
skip(keyword('route')) +
maybe(_id) +
op_('{') +
network_inline_stmt_list +
op_('}')
>> create_mapper(Network)
)
#
# diagram statement::
# nwdiag {
# A;
# }
#
diagram_id = (
(keyword('diagram') | keyword('nwdiag')) +
maybe(_id)
>> list
)
diagram_inline_stmt = (
extension_stmt |
network_stmt |
group_stmt |
attribute_stmt |
route_stmt |
edge_stmt |
node_stmt
)
diagram_inline_stmt_list = (
many(diagram_inline_stmt + skip(maybe(op(';'))))
)
diagram = (
maybe(diagram_id) +
op_('{') +
diagram_inline_stmt_list +
op_('}')
>> create_mapper(Diagram)
)
dotfile = diagram + skip(finished)
return dotfile.parse(seq)
