I am trying to create an interpreter that take my script as input. I have some problem with writing regular expressions. One of the defined token is considering all strings as it's tokens.
import ply.lex as lex
import ply.yacc as yacc
tokens = (
'STAIRCASE',
'STAIRCASE_END',
'STAIR',
'STAIR_END',
'TAG',
'COLON_SYM',
'LINE_START_SYM',
'NONE',
'USER_DEFINED',
'ARRAY',
'IS',
)
assignments = {}
t_STAIRCASE = r'staircase'
t_TAG = r'\(([a-zA-Z0-9\ ])*\)'
t_COLON_SYM = r' :'
t_LINE_START_SYM = r'-'
t_STAIRCASE_END = 'EOSC'
t_ignore = ' \t\n'
t_STAIR = 'stair'
t_STAIR_END = 'EOS'
t_NONE = 'EOP'
t_USER_DEFINED = r'[a-zA-Z0-9]+'
t_IS = 'is'
def t_error(t):
print 'Illegal character "%s"' % t.value[0]
t.lexer.skip(1)
lex.lex()
NONE, STAIRCASE, TAG, STAIRCASE_DESCRIPTION = range(4)
states = ['NONE', 'STAIRCASE','STAIRCASE_DESCRIPTION']
current_state = NONE
def x():
print "Hi How you doing"
def p_staircase_def(t):
"""STAIRCASE_DEF : STAIRCASE TAG COLON_SYM STAIRCASE_DESCRIPTION
"""
print t[0:]
help(t)
def p_staircase_description(t):
"""STAIRCASE_DESCRIPTION : LINE_START_SYM DICTONARY STAIRCASE_DESCRIPTION
| STAIRCASE_END STAIR_DEF
"""
print t[0:]
def p_dictonary(t):
"""
DICTONARY : USER_DEFINED IS USER_DEFINED
"""
temp = { t[1] : t[2] }
print assignments.update( temp )
def p_stair_def(t):
"""STAIR_DEF : STAIR TAG COLON_SYM STAIR_DESCRIPTION
"""
print t[0:]
def p_stair_description(t):
"""STAIR_DESCRIPTION : LINE_START_SYM DICTONARY STAIR_DESCRIPTION
| STAIR_END STAIR_DEF
| STAIR_END
"""
print t[0:]
def p_error(t):
print 'Syntax error at "%s"' % t.value if t else 'NULL'
global current_state
current_state = NONE
yacc.yacc()
file_input = open("x.staircase","r")
yacc.parse(file_input.read())
This is a sample input that need to be accepted by my interpreter "x.staircase"
staircase(XXXX XXX XXX):
- abc is 23183 # which need to {'abc' : '23183'}
- bcf is fda
- deh is szsC
EOSC
stair(XXXX XXX XXX):
- lkm is 35
- raa is 233
EOS
stair(XXXX XXX XXX):
- faa is zxhfb
- faa is 1
EOS
Syntax error at "staircase"
[Finished in 0.1s]
import ply.lex as lex
import ply.yacc as yacc
tokens = (
'STAIRCASE',
'STAIRCASE_END',
'STAIR',
'STAIR_END',
'TAG',
'COLON_SYM',
'LINE_START_SYM',
'NONE',
'USER_DEFINED',
'ARRAY',
'IS',
)
assignments = {}
t_STAIRCASE = r'staircase'
t_TAG = r'\(([a-zA-Z0-9\ ])*\)'
t_COLON_SYM = r' :'
t_LINE_START_SYM = r'-'
t_STAIRCASE_END = 'EOSC'
t_ignore = ' \t\n'
t_STAIR = 'stair'
t_STAIR_END = 'EOS'
t_NONE = 'EOP'
##########################################
Here is the issue with this regular exprission
It worked, If I Use this
t_USER_DEFINED = r'a'
Instead of this
#t_USER_DEFINED = r'[a-zA-Z0-9]+'
But, when it comes to my input file it only accept one variable called 'a'
##########################################
Code continues
t_IS = 'is'
def t_error(t):
print 'Illegal character "%s"' % t.value[0]
t.lexer.skip(1)
lex.lex()
NONE, STAIRCASE, TAG, STAIRCASE_DESCRIPTION = range(4)
states = ['NONE', 'STAIRCASE','STAIRCASE_DESCRIPTION']
current_state = NONE
def x():
print "Hi How you doing"
def p_staircase_def(t):
"""STAIRCASE_DEF : STAIRCASE TAG COLON_SYM STAIRCASE_DESCRIPTION
"""
print t[0:]
help(t)
def p_staircase_description(t):
"""STAIRCASE_DESCRIPTION : LINE_START_SYM DICTONARY STAIRCASE_DESCRIPTION
| STAIRCASE_END STAIR_DEF
"""
print t[0:]
def p_dictonary(t):
"""
DICTONARY : USER_DEFINED IS USER_DEFINED
"""
HERE is my assignment operation, actually it create a dictionary of variables
temp = { t[1] : t[2] }
print assignments.update( temp )
def p_stair_def(t):
"""STAIR_DEF : STAIR TAG COLON_SYM STAIR_DESCRIPTION
"""
print t[0:]
def p_stair_description(t):
"""STAIR_DESCRIPTION : LINE_START_SYM DICTONARY STAIR_DESCRIPTION
| STAIR_END STAIR_DEF
| STAIR_END
"""
print t[0:]
def p_error(t):
print 'Syntax error at "%s"' % t.value if t else 'NULL'
global current_state
current_state = NONE
yacc.yacc()
file_input = open("x.staircase","r")
yacc.parse(file_input.read())
staircase(XXXX XXX XXX):
- a is a # which need to {'abc' : '23183'}
- a is a
- a is a
EOSC
stair(XXXX XXX XXX):
- a is a
- a is a
EOS
stair(XXXX XXX XXX):
- a is a
- a is a 1
EOS
Please (re-)read the description of how Ply's lexer recognises tokens in the Ply manual. Pay particular attention to the ordering rules; since pattern variables are sorted longest to shortest, the t_USER_DEFINED pattern is tried before any keyword pattern (such as staircase) so none of the keywords will be recognised. (This is why shortening t_USER_DEFINED to a single character changes lexical behaviour.)
There's a good clue that this is the problem and not the assignment production: the error message is triggered at the token staircase, long before the assignment is encountered. You would have gotten another clue by printing t.type as well as t.value in your p_error function. (Or, of course, by testing the tokeniser before trying to parse anything.)
If you read through to the end of the section of the Ply manual I linked, you'll find a suggestion for how to handle keyword tokens, using a scanner function and auxiliary dictionary of keywords. I strongly suggest you use that as your model.
Also note that you require a colon to be preceded by a space character:
t_COLON_SYM = r' :'
But your sample input does not have spaces before colons.