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, [])
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 comparison_():
"""Returns the parse for a compound compare statement"""
ops = op('==') | op('<') | op('>') | op('<=') | op('>=') | op('!=')
op_vals = (boolean | number | timestamp_or_string)
comp_op = string + ops + op_vals >> make(ASTCompOp)
def multi(func):
"""For x + many(x) lists, call func only when there are multiple xs"""
def multi_(args):
x, xs = args
if len(xs) == 0:
return x
return func(args)
return multi_
comp_stmt = forward_decl()
comp_base = forward_decl()
comp_base.define((op_('(') + comp_stmt + op_(')')) | comp_op
| ((n('not') + comp_base) >> make(ASTCompNot)))
comp_and = comp_base + many(n_('and') + comp_base) >> multi(
make(ASTCompAnd))
comp_or = comp_and + many(n_('or') + comp_and) >> multi(make(ASTCompOr))
comp_stmt.define(comp_or)
return comp_stmt
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):
"""Returns the AST of the given token sequence."""
def eval_expr(z, list):
return reduce(lambda s, (f, x): f(s, x), list, z)
unarg = lambda f: lambda x: f(*x)
const = lambda x: lambda _: x # like ^^^ in Scala
tokval = lambda x: x.value # returns the value of a token
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
toktype = lambda t: some(lambda x: x.type == t) >> tokval # checks type of token
def lst(h,t):
return [h,] + t
makeop = lambda s, f: op(s) >> const(f)
or_op = makeop('|', Or)
char = with_forward_decls(lambda:
toktype('Char') >> Char | op_('(') + exp + op_(')'))
star = char + op_('*') >> Star | char
lst2_exp = star + many(star) >> unarg(lst)
lst_exp = lst2_exp >> Lst
exp = lst_exp + many(or_op + lst_exp) >> unarg(eval_expr)
return exp.parse(seq)
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 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(tokens):
t = lambda s: some(lambda tok: tok.type == s)
inttype = t('Int')
chartype = t('Char')
unsignedtype = t('Unsigned')
name = t('Name')
star = t('Star')
void = t('Void')
lpar = skip(t('LPar'))
rpar = skip(t('RPar'))
comma = skip(t('Comma'))
semicolon = skip(t('SemiColon'))
def collapse(x):
bp()
if len(x[1]) > 0:
# TODO: handle multiple stars
return Token("UserTypePointer", x[0].value + " " + x[1][0].value)
else:
return Token("UserType", x[0].value)
def make_func(x):
return Token('Function', x.value)
def make_type(x):
if len(x) == 3:
return Token("UnsignedTypePointer", x)
elif len(x) == 2:
if x[0].type == "Unsigned":
return Token("UnsignedType", x)
else:
return Token("TypePointer", x)
else:
return Token("Type", x)
udt = name + many(star) >> collapse
prim = (inttype | chartype | unsignedtype + inttype | unsignedtype + chartype ) + many(star) >> make_type
voidptr = void + star + many(star)
func = name >> make_func
accepted_types = voidptr | prim | udt
# Return Type
rettype = void | accepted_types
# Argument List
decl = accepted_types + name
decl_list = decl + many(comma + decl)
arg_list = void | decl_list
func_decl = rettype + func + lpar + arg_list + rpar + semicolon
return func_decl.parse(tokens)
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 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 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 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 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 separated(parser, by):
def append(parsed):
first = parsed[0]
rest = parsed[1]
rest.insert(0, first)
return rest
return ( parser + many(by + parser) ) >> append
def one_plus(p):
"""Parser(a, b) -> Parser(a, [b])
Return a parser that applies the parser p one or more times.
"""
q = p + many(p)
return q.named('(%s , { %s })' % (p.name, p.name))
def parse(seq):
"""Returns the AST of the given token sequence."""
global depth
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
paren = lambda s: a(Token('Parentheses', s)) >> tokval # return the value if token is Op
paren_ = lambda s: skip(paren(s)) # checks if token is Op and ignores it
def application(z, list):
return reduce(lambda s, x: Application(s, x), list, z)
depth = 0
variable = lambda x: Variable(str(x)+":"+str(depth))
def abstraction(x):
global depth
abst = Abstraction(str(x[0])+":"+str(depth), x[1])
depth += 1
return abst
variable = toktype('Name') >> variable
term = variable | with_forward_decls(lambda: paren_('(') + exp + paren_(')')) | \
with_forward_decls(lambda: skip(toktype('Lambda')) + toktype('Name') + \
skip(toktype('Dot')) + exp >> abstraction)
exp = term + many(term) >> unarg(application)
return exp.parse(seq)
def one_plus(p):
"""Parser(a, b) -> Parser(a, [b])
Return a parser that applies the parser p one or more times.
"""
q = p + many(p)
return q.named('(%s , { %s })' % (p.name, p.name))
def parse(s):
def value(t):
return t.value
def token(type):
return some(lambda t: t.type == type)
def instruction(name):
return some(lambda t: t.type == 'instruction' and t.value == name)
def unarg(f):
return lambda args: f(*args)
newline = token('newline')
variable = token('variable') >> value
number = token('number') >> value >> int
increment = skip(instruction('inc')) + variable >> Inc
decrement = skip(instruction('dec')) + variable >> Dec
zero = (skip(instruction('zero')) + variable + number +
skip(instruction('else')) + number >> unarg(Zero))
stop = instruction('stop') >> (lambda x: Stop())
instruction = increment | decrement | zero | stop
top_level = many(instruction + skip(newline)) + skip(finished)
tokens = tokenize(s)
return top_level.parse(tokens)
def parse_identifier(line: str):
"""Parses just the identifer (first element) of the write"""
tokval = lambda t: t.value
joinval = "".join
someToken = lambda type: some(lambda t: t.type == type)
char = someToken('Char') >> tokval
space = someToken('Space') >> tokval
comma = someToken('Comma') >> tokval
quote = someToken('Quote') >> tokval
escape = someToken('Escape') >> tokval
equal = someToken('Equal') >> tokval
escape_space = skip(escape) + space >> joinval
escape_comma = skip(escape) + comma >> joinval
escape_equal = skip(escape) + equal >> joinval
escape_escape = skip(escape) + escape >> joinval
plain_int_text = someToken('Int') >> tokval
plain_float_text = someToken('Float') >> tokval
identifier = many(char | plain_float_text | plain_int_text |
escape_space | escape_comma | escape_equal |
escape_escape | plain_int_text | quote) >> joinval
toplevel = identifier >> (lambda x: x)
parsed = toplevel.parse(LineTokenizer.tokenize(line))
if len(parsed) == 0:
raise NoParseError('parsed nothing')
else:
return parsed
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 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 parse(constraints):
"""Using funcparserlib turn constraints into a mongo query
NOTE: this uses functors, see:
http://spb-archlinux.ru/2009/funcparserlib/Tutorial
"""
tokval = lambda tok: tok.value
char = lambda tok: tok.code == 'CHAR'
chars = some(char) >> tokval
operator = lambda s: a(Token('OP', s)) >> tokval
const = lambda x: lambda _: x
makeop = lambda s: operator(s) >> const(s)
item = many(chars) >> (lambda x: ''.join(x))
test1 = item.parse(tokenize('hello123'))
assert test1 == 'hello123'
test1b = item.parse(tokenize('42a'))
assert test1b == '42a'
test1c = item.parse(tokenize('cam-oeprod-123299-master'))
assert test1c == 'cam-oeprod-123299-master'
test1d = item.parse(tokenize('Hello world'))
assert test1d == 'Hello world'
equals = makeop('=')
assert equals.parse(tokenize('=')) == '='
slash = makeop('/')
value = item >> possible_int
term = (item + equals + value) >> (lambda x: (x[0], x[2]))
test2 = term.parse(tokenize('dut=catgut'))
assert test2 == ('dut', 'catgut')
endmark = a(Token('END', ''))
seq = (many(((slash + term) >> (lambda x: x[1]))) >> dict)
top = (seq + endmark) >> (lambda x: x[0])
test3 = seq.parse(
tokenize('/dut=catgut/foo=bar/n=30/bet=a42a/message=Hello World'))
assert test3 == {
'dut': 'catgut',
'foo': 'bar',
'n': 30,
'message': 'Hello World',
'bet': 'a42a'
}
test4 = seq.parse(tokenize('/suppress=,bar'))
assert test4 == {'suppress': ',bar'}
lexemes = tokenize(constraints)
return top.parse(lexemes)
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 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 __init__(self):
# TODO: Make your changes in this section #############################
value = number | string
#######################################################################
attribute = rawname + op_("=") + value + semicolon >> unarg(Attribute)
attributes = many(attribute)
# For chicken-and-egg problems, forward_decl will be your friend
widgets = forward_decl()
widget = rawname + opencurlyparen + attributes + widgets + closecurlyparen >> unarg(Widget)
widgets.define(many(widget))
# For the toplevel, we allow only one widget, not multiple widgets
self.toplevel = widget + end
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 __init__(self):
self.toplevel = None
makeop = lambda s: op(s) >> const(lambda l, r: BinaryExpression(l, s, r))
kw_while = kw('while')
kw_if = kw('if')
kw_else = kw('else')
kw_put = kw('put')
add = makeop('+')
sub = makeop('-')
mul = makeop('*')
div = makeop('/')
comparison_operator = makeop('<') | makeop('<=') | makeop('>') | \
makeop('>=') | makeop('==') | makeop('!=')
equ = op_('=')
sign = op('+') | op('-')
int_const = maybe(sign) + number >> unarg(IntConst)
variable = rawname >> Variable
atom = int_const | variable
expr1 = atom + many((mul | div) + atom) >> eval_expr
expr2 = expr1 + many((add | sub) + expr1) >> eval_expr
comparison_expr = expr2 + many(comparison_operator + expr2) >> eval_expr
put_call = kw_put + inparens(comparison_expr) >> PutCall
assignment = variable + equ + expr2 >> unarg(Assignment)
condition = openparen + comparison_expr + closeparen
stmt_block = forward_decl()
while_stmt = kw_while + condition + stmt_block >> unarg(WhileStatement)
if_stmt = kw_if + condition + stmt_block + maybe(kw_else + stmt_block) >> unarg(IfStatement)
stmt = assignment | put_call | while_stmt | if_stmt
stmt_block.define(opencurlyparen + many(stmt) + closecurlyparen)
program = many(stmt) >> Program
self.toplevel = program + end
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(constraints):
"""Using funcparserlib turn constraints into a mongo query
NOTE: this uses functors, see:
http://spb-archlinux.ru/2009/funcparserlib/Tutorial
"""
tokval = lambda tok: tok.value
char = lambda tok: tok.code == 'CHAR'
chars = some( char )>>tokval
operator = lambda s: a(Token('OP',s)) >> tokval
const = lambda x : lambda _: x
makeop = lambda s: operator(s) >> const(s)
item = many (chars) >> (
lambda x: ''.join(x))
test1 = item.parse(tokenize('hello123'))
assert test1 == 'hello123'
test1b = item.parse(tokenize('42a'))
assert test1b == '42a'
test1c = item.parse(tokenize('cam-oeprod-123299-master'))
assert test1c == 'cam-oeprod-123299-master'
test1d = item.parse(tokenize('Hello world'))
assert test1d == 'Hello world'
equals = makeop('=')
assert equals.parse(tokenize('=')) == '='
slash = makeop('/')
value = item >> possible_int
term = (item + equals + value) >> (lambda x: (x[0], x[2]))
test2 = term.parse(tokenize('dut=catgut'))
assert test2 == ('dut','catgut')
endmark = a(Token('END', ''))
seq = (many( ((slash + term) >> (lambda x: x[1])) ) >> dict)
top = (seq + endmark) >> (lambda x: x[0])
test3 = seq.parse(tokenize(
'/dut=catgut/foo=bar/n=30/bet=a42a/message=Hello World'))
assert test3 == {'dut': 'catgut', 'foo': 'bar', 'n': 30,
'message':'Hello World', 'bet': 'a42a'}
test4 = seq.parse(tokenize('/suppress=,bar'))
assert test4 == {'suppress': ',bar'}
lexemes = tokenize(constraints)
return top.parse(lexemes)
def parse(tokens):
makeop = lambda s: op(s) >> const(lambda l, r: BinaryExpression(l, s, r))
num = number >> Number
var = rawname >> Variable
add, sub, mul, div = map(makeop, ['+', '-', '*', '/'])
lt, lte, gt, gte = map(makeop, ['<', '<=', '>', '>='])
method_call = forward_decl()
atom = num | method_call | var
expr1 = atom + many((mul | div) + atom) >> eval_expr
expr2 = expr1 + many((add | sub) + expr1) >> eval_expr
expr3 = expr2 + many((lt | lte | gt | gte) + expr2) >> eval_expr
method_call.define(rawname + inparens(maybe_empty_listof(expr2)) >> unarg(MethodCall))
defn = expr3 + end
return defn.parse(tokens)
def parse(tokens):
var = some(toktype("name")) | some(toktype("number"))
open_form = some(toktype("form_open"))
close_form = some(toktype("form_close"))
op_lambda = some(toktype("op_lambda"))
op_map = some(toktype("op_map"))
prim_bind = some(toktype("kw_bind"))
prim_halt = some(toktype("kw_halt"))
exp = with_forward_decls(lambda: lam | var | prim_exp | exprn) >> Expression
lam = open_form + op_lambda + many(var) + op_map + oneplus(exp) + close_form >> Lambda
bind_exp = open_form + prim_bind + var + lam + close_form
halt_exp = open_form + prim_halt + exp + close_form
prim_exp = bind_exp | halt_exp
exprn = open_form + oneplus(exp) + close_form >> Form
prog = many(exp) + skip(finished) >> Program
return prog.parse(tokens)
def __init__(self):
# TODO: Make your changes in this section #############################
nr = number >> Number
plus_operator = op('+') >> Operator
calculation = nr + plus_operator + nr >> unarg(Calculation)
string_list = inbrackets(listof(string))
value = calculation | number | string | string_list
#######################################################################
attribute = rawname + op_("=") + value + semicolon >> unarg(Attribute)
attributes = many(attribute)
widgets = forward_decl()
widget = rawname + opencurlyparen + attributes + widgets + closecurlyparen >> unarg(Widget)
widgets.define(many(widget))
self.toplevel = widget + end
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 __init__(self):
self.toplevel = None
makeop = lambda s: op(s) >> const(lambda l, r: BinaryExpression(l, s, r))
kw_class = kw('class')
kw_new = kw('new')
kw_if = kw('if')
kw_else = kw('else')
kw_while = kw('while')
kw_return = kw('return')
add = makeop('+')
sub = makeop('-')
mul = makeop('*')
div = makeop('/')
equ = op_('=')
int_const = number >> IntConst
variable = rawname >> Variable
type_ = rawname >> Type
var_decl = type_ + rawname >> unarg(VarDeclaration)
var_decls = many(var_decl + semicolon)
method_call = forward_decl()
atom = int_const | method_call | variable
expr1 = atom + many((mul | div) + atom) >> eval_expr
expr2 = expr1 + many((add | sub) + expr1) >> eval_expr
method_call.define(rawname + inparens(maybe_empty_listof(expr2)) >> unarg(MethodCall))
new_obj_expr = kw_new + type_ + openparen + closeparen >> NewObject
assignment = variable + equ + (new_obj_expr | expr2) >> unarg(Assignment)
return_stmt = kw_return + expr2 >> ReturnStatement
simple_stmt = (assignment | method_call | return_stmt) + semicolon
stmt_block = forward_decl()
condition = openparen + expr2 + closeparen
if_stmt = kw_if + condition + stmt_block + kw_else + stmt_block >> unarg(IfStatement)
while_stmt = kw_while + condition + stmt_block >> unarg(WhileStatement)
stmt = simple_stmt | if_stmt | while_stmt | return_stmt
stmt_block.define(opencurlyparen + many(stmt) + closecurlyparen)
method_decl = type_ + rawname + inparens(maybe_empty_listof(var_decl)) + \
opencurlyparen + var_decls + many(stmt) + \
closecurlyparen >> unarg(MethodDeclaration)
method_decls = many(method_decl)
class_decl = kw_class + rawname + opencurlyparen + var_decls + \
method_decls + closecurlyparen >> unarg(ClassDeclaration)
program = many(class_decl) >> Program
self.toplevel = program + end
def create_parser():
# operator: '~=' | '>=' | '<=' | '<' | '>' | '='
operator = some(lambda tok: tok.type == 'CMP') >> choose_class
# value: STRING | WORD
word = some(lambda tok: tok.type == 'WORD') >> Text
string = some(lambda tok: tok.type == 'STRING') >> QuotedText
value = string | word
# function: WORD '(' ')'
open_brace = skip(a(Token('BR', '(')))
close_brace = skip(a(Token('BR', ')')))
function = word + open_brace + close_brace >> Function
# field_expr: WORD operator value
fieldexpr = (word + operator + (function | value)) >> (lambda x: x[1]
([x[0], x[2]]))
OR = a(Token('OP', 'OR')) >> choose_class
AND = a(Token('OP', 'AND')) >> choose_class
def eval(data):
arg1, lst = data
for f, arg2 in lst:
arg1 = f([arg1, arg2])
return arg1
def eval(data):
lft, args = data
return reduce(lambda arg1, (f, arg2): f([arg1, arg2]), args, lft)
expr = forward_decl()
basexpr = open_brace + expr + close_brace | fieldexpr
andexpr = (basexpr + many(AND + basexpr)) >> eval
orexpr = (andexpr + many(OR + andexpr)) >> eval
expr.define(orexpr)
return expr
def parse(seq):
"""Returns the AST of the given token sequence."""
def eval_expr(z, list):
return reduce(lambda s, (f, x): f(s, x), list, z)
unarg = lambda f: lambda x: f(*x)
const = lambda x: lambda _: x # like ^^^ in Scala
tokval = lambda x: x.value # returns the value of a token
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
toktype = lambda t: some(lambda x: x.type == t) >> tokval # checks type of token
def lst(h,t):
return [h,] + t
call = lambda x: Call(x[0], x[1])
makeop = lambda s, f: op(s) >> const(f)
add = makeop('+', Plus)
sub = makeop('-', Minus)
mul = makeop('*', Times)
div = makeop('/', Div)
def make_const(i):
return const(int(i))
number = toktype('Number') >> Const
mul_op = mul | div
add_op = add | sub
factor = with_forward_decls(lambda:
number | op_('(') + exp + op_(')') | call)
term = factor + many(mul_op + factor) >> unarg(eval_expr)
exp = term + many(add_op + term) >> unarg(eval_expr)
exp_lst = with_forward_decls(lambda:
exp + many(op_(',') + exp) >> unarg(lst))
call = toktype('Name') + op_('(') + exp_lst + op_(')') >> call
return exp.parse(seq)
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 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)
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):
'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 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 test_error_info(self):
tokenize = make_tokenizer([
('keyword', (r'(is|end)',)),
('id', (r'[a-z]+',)),
('space', (r'[ \t]+',)),
('nl', (r'[\n\r]+',)),
])
try:
list(tokenize('f is ф'))
except LexerError as e:
self.assertEqual(six.text_type(e),
'cannot tokenize data: 1,6: "f is \u0444"')
else:
self.fail('must raise LexerError')
sometok = lambda type: some(lambda t: t.type == type)
keyword = lambda s: a(Token('keyword', s))
id = sometok('id')
is_ = keyword('is')
end = keyword('end')
nl = sometok('nl')
equality = id + skip(is_) + id >> tuple
expr = equality + skip(nl)
file = many(expr) + end
msg = """\
spam is eggs
eggs isnt spam
end"""
toks = [x for x in tokenize(msg) if x.type != 'space']
try:
file.parse(toks)
except NoParseError as e:
self.assertEqual(e.msg,
"got unexpected token: 2,11-2,14: id 'spam'")
self.assertEqual(e.state.pos, 4)
self.assertEqual(e.state.max, 7)
# May raise KeyError
t = toks[e.state.max]
self.assertEqual(t, Token('id', 'spam'))
self.assertEqual((t.start, t.end), ((2, 11), (2, 14)))
else:
self.fail('must raise NoParseError')
def test_error_info(self):
tokenize = make_tokenizer([
('keyword', (r'(is|end)',)),
('id', (r'[a-z]+',)),
('space', (r'[ \t]+',)),
('nl', (r'[\n\r]+',)),
])
try:
list(tokenize('f is ф'))
except LexerError as e:
self.assertEqual(str(e),
'cannot tokenize data: 1,6: "f is \u0444"')
else:
self.fail('must raise LexerError')
sometok = lambda type: some(lambda t: t.type == type)
keyword = lambda s: a(Token('keyword', s))
id = sometok('id')
is_ = keyword('is')
end = keyword('end')
nl = sometok('nl')
equality = id + skip(is_) + id >> tuple
expr = equality + skip(nl)
file = many(expr) + end
msg = """\
spam is eggs
eggs isnt spam
end"""
toks = [x for x in tokenize(msg) if x.type != 'space']
try:
file.parse(toks)
except NoParseError as e:
self.assertEqual(e.msg,
"got unexpected token: 2,11-2,14: id 'spam'")
self.assertEqual(e.state.pos, 4)
self.assertEqual(e.state.max, 7)
# May raise KeyError
t = toks[e.state.max]
self.assertEqual(t, Token('id', 'spam'))
self.assertEqual((t.start, t.end), ((2, 11), (2, 14)))
else:
self.fail('must raise NoParseError')
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 parse1(tokens):
"""
Experimenting with collapsing part of the parse tree
to make it easier to work with.
"""
t = lambda s: some(lambda tok: tok.type == s)
name = t('Name')
star = t('Star')
def collapse(x):
if len(x[1]) > 0:
# TODO: handle multiple stars
return Token("UserTypePointer", x[0].value + " " + x[1].value)
else:
return Token("UserType", x[0].value)
udt = name + many(star) >> collapse
return udt.parse(tokens)
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 test_error_info():
tokenize = make_tokenizer([
Spec('keyword', r'(is|end)'),
Spec('id', r'[a-z]+'),
Spec('space', r'[ \t]+'),
Spec('nl', r'[\n\r]+'),
])
try:
list(tokenize('f is ф'))
except LexerError as e:
pass
else:
ok_(False, 'must raise LexerError')
keyword = lambda s: tok('keyword', s)
id = tok('id')
is_ = keyword('is')
end = keyword('end')
nl = tok('nl')
equality = id + skip(is_) + id >> tuple
expr = equality + skip(nl)
file = many(expr) + end
msg = """\
rake is eggs
eggs isnt spam
end"""
toks = [x for x in tokenize(msg) if x.type != 'space']
try:
file.parse(toks)
except ParserError as e:
msg, pos, i = e.args
eq_(msg, "got unexpected token: id 'spam'")
eq_(pos, ((2, 11), (2, 14)))
# May raise KeyError
t = toks[i]
eq_(t, Token('id', 'spam'))
else:
ok_(False, 'must raise ParserError')
def test_error_info():
tokenize = make_tokenizer([
Spec('keyword', r'(is|end)'),
Spec('id', r'[a-z]+'),
Spec('space', r'[ \t]+'),
Spec('nl', r'[\n\r]+'),
])
try:
list(tokenize('f is ф'))
except LexerError as e:
pass
else:
ok_(False, 'must raise LexerError')
keyword = lambda s: tok('keyword', s)
id = tok('id')
is_ = keyword('is')
end = keyword('end')
nl = tok('nl')
equality = id + skip(is_) + id >> tuple
expr = equality + skip(nl)
file = many(expr) + end
msg = """\
rake is eggs
eggs isnt spam
end"""
toks = [x for x in tokenize(msg) if x.type != 'space']
try:
file.parse(toks)
except ParserError as e:
msg, pos, i = e.args
eq_(msg, "got unexpected token: id 'spam'")
eq_(pos, ((2, 11), (2, 14)))
# May raise KeyError
t = toks[i]
eq_(t, Token('id', 'spam'))
else:
ok_(False, 'must raise ParserError')
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(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 dolike(head):
"Parse a `do`-like form."
return pexpr(sym(head), many(FORM))
posnumber = (some(lambda tok: tok.type == 'NUMBER') >> tokval >> make_number)
add = makeop('+', operator.add)
sub = makeop('-', operator.sub)
mul = makeop('*', operator.mul)
div = makeop('/', operator.div)
pow = makeop('**', operator.pow)
negnumber = (sub + posnumber) >> negate
number = posnumber | negnumber
mul_op = mul | div
add_op = add | sub
primary = with_forward_decls(lambda: number | (op_('(') + expr + op_(')')))
factor = primary + many(pow + primary) >> f
term = factor + many(mul_op + factor) >> f
expr = term + many(add_op + term) >> f
endmark = a(Token(token.ENDMARKER, ''))
end = skip(endmark + finished)
toplevel = maybe(expr) + end
#@+node:peckj.20140124085532.4012: *3* parse
def parse(tokens):
return toplevel.parse(tokens)
parse_and_run = lambda x: parse(tokenize(x))
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 dolike(head):
"Parse a `do`-like form."
return pexpr(sym(head), many(FORM))
# =========== 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) +
p.many(_table_sep + (_keyvalue | expr))
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)