string.translate() with unicode data in python

Question

I have 3 API's that return json data to 3 dictionary variables. I am taking some of the values from the dictionary to process them. I read the specific values that I want to the list valuelist. One of the steps is to remove the punctuation from them. I normally use string.translate(None, string.punctuation) for this process but because the dictionary data is unicode I get the error:

 wordlist = [s.translate(None, string.punctuation)for s in valuelist]
TypeError: translate() takes exactly one argument (2 given)

Is there a way around this? Either by encoding the unicode or a replacement for string.translate?

• Suzana: s.encode('utf-8').translate(None, string.punctuation) worked for me.
• asmaier: @Suzana_K Thank you! This was the simplest solution for me.
• jfs: related: Remove punctuation from Unicode formatted strings

Answer 1

The translate method work differently on Unicode objects than on byte-string objects:

>>> help(unicode.translate)
S.translate(table) -> unicode
Return a copy of the string S, where all characters have been mapped
through the given translation table, which must be a mapping of
Unicode ordinals to Unicode ordinals, Unicode strings or None.
Unmapped characters are left untouched. Characters mapped to None
are deleted.

So your example would become:

remove_punctuation_map = dict((ord(char), None) for char in string.punctuation)
word_list = [s.translate(remove_punctuation_map) for s in value_list]

Note however that string.punctuation only contains ASCII punctuation. Full Unicode has many more punctuation characters, but it all depends on your use case.

• jfs: dict.fromkeys(map(ord, string.punctuation))
• patrick: this is by far the best of all the answers on here. thanks.
• Mike Honey: thanks! just to note, I think this implies import string

Answer 2

I noticed that string.translate is deprecated. Since you are removing punctuation, not actually translating characters, you can use the re.sub function.

 >>> import re
 >>> s1="this.is a.string, with; (punctuation)."
 >>> s1
 'this.is a.string, with; (punctuation).'
 >>> re.sub("[\.\t,円\:;\(\)\.]", "", s1, 0, 0)
 'thisis astring with punctuation'
 >>>

• adohertyd: the translate function works great in python 2.7 and is computationally faster than REGEX. I may have no other option though. Thanks
• bheklilr: The module function string.translate is deprecated in favor of the method str.translate, the translate method (which OP is using) is still usable.

Answer 3

In this version you can relatively make one's letters to other

def trans(to_translate):
 tabin = u'привет'
 tabout = u'тевирп'
 tabin = [ord(char) for char in tabin]
 translate_table = dict(zip(tabin, tabout))
 return to_translate.translate(translate_table)

Answer 4

Python re module allows to use a function as a replacement argument, which should take a Match object and return a suitable replacement. We may use this function to build a custom character translation function:

import re
def mk_replacer(oldchars, newchars):
 """A function to build a replacement function"""
 mapping = dict(zip(oldchars, newchars))
 def replacer(match):
 """A replacement function to pass to re.sub()"""
 return mapping.get(match.group(0), "")
 return replacer

An example. Match all lower-case letters ([a-z]), translate 'h' and 'i' to 'H' and 'I' respectively, delete other matches:

>>> re.sub("[a-z]", mk_replacer("hi", "HI"), "hail")
'HI'

As you can see, it may be used with short (incomplete) replacement sets, and it may be used to delete some characters.

A Unicode example:

>>> re.sub("[\W]", mk_replacer(u'\u0435\u0438\u043f\u0440\u0442\u0432', u"EIPRTV"), u'\u043f\u0440\u0438\u0432\u0435\u0442')
u'PRIVET'

Answer 5

As I stumbled upon the same problem and Simon's answer was the one that helped me to solve my case, I thought of showing an easier example just for clarification:

from collections import defaultdict

And then for the translation, say you'd like to remove '@' and '\r' characters:

remove_chars_map = defaultdict()
remove_chars_map['@'] = None
remove_chars_map['\r'] = None
new_string = old_string.translate(remove_chars_map)

And an example:

old_string = "word1@\r word2@\r word3@\r"

new_string = "word1 word2 word3"

'@' and '\r' removed