Iterative find/replace from a list of tuples in Python

Question

I have a list of tuples, each containing a find/replace value that I would like to apply to a string. What would be the most efficient way to do so? I will be applying this iteratively, so performance is my biggest concern.

More concretely, what would the innards of processThis() look like?

x = 'find1, find2, find3'
y = [('find1', 'replace1'), ('find2', 'replace2'), ('find3', 'replace3')]

def processThis(str,lst):
     # Do something here
     return something

>>> processThis(x,y)
'replace1, replace2, replace3'

Thanks, all!

Answer 1

x = 'find1, find2, find3'
y = [('find1', 'replace1'), ('find2', 'replace2'), ('find3', 'replace3')]

def processThis(str,lst):
    for find, replace in lst:
        str = str.replace(find, replace)

    return str

>>> processThis(x,y)
'replace1, replace2, replace3'

Answer 2

You could consider using re.sub:

import re
REPLACEMENTS = dict([('find1', 'replace1'),
                     ('find2', 'replace2'),
                     ('find3', 'replace3')])

def replacer(m):
    return REPLACEMENTS[m.group(0)]

x = 'find1, find2, find3'
r = re.compile('|'.join(REPLACEMENTS.keys()))
print r.sub(replacer, x)

Answer 3

s = reduce(lambda x, repl: str.replace(x, *repl), lst, s)

Answer 4

A couple notes:

1. The boilerplate argument about premature optimization, benchmarking, bottlenecks, 100 is small, etc.
2. There are cases where the different solutions will return different results. if y = [('one', 'two'), ('two', 'three')] and x = 'one' then mhawke's solution gives you 'two' and Unknown's gives 'three'.
3. Testing this out in a silly contrived example mhawke's solution was a tiny bit faster. It should be easy to try it with your data though.