Full unicode support about textx HOT 4 CLOSED

textX uses Python re module for regex matches. Whatever you can match using re module you can use in textX between slashes /..../.

Assuming you are trying to parse a file with multiple records like the one you provided you could do something like this:

# -*- coding: utf-8 -*-
from textx import metamodel_from_str

grammar = '''
Model: records*=Record;
Record: num=/\d+-\d+/ name=/\w+\s+\w+/;
'''
mm = metamodel_from_str(grammar)

input = '''
2785-599 São Domingo
2785-599 São Domingo
'''

model = mm.model_from_str(input)
print(model.records)
print(model.records[0].name)
print(model.records[0].num)

Note that input text should be interpreted as unicode. textX accepts unicode only. The example above is for Python 3.

from textx.

First of all, thanks for your quick response! I forgot to tell you that I am using Python 2.7. With this version, using:

• re.U
• re.UNICODE

makes the \w sequence dependent on Unicode

from textx.

For Python 2 you can do this:

# -*- coding: utf-8 -*-
from textx import metamodel_from_str

grammar = '''
Model: records*=Record;
Record: num=/\d+-\d+/ name=/(?u)\w+\s+\w+/;
'''
mm = metamodel_from_str(grammar)

input = u'''
2785-599 São Domingo
2785-599 São Domingo
'''

model = mm.model_from_str(input)
print(model.records[0].name)
print(model.records[0].num)

Notice (?u) at the beginning of name regex match. This turns on the re.UNICODE flag for that match.
There is also u at the beginning of input unicode string.

There was a slight problem with printing exceptions on syntax errors with unicode chars in Python 2. It should be fixed now on the master branch so use that version.

from textx.

Great! Great job! Thank you so much for your help!

from textx.