CODEWITHSUNDEEP
pythoncsvparsingpandas
Georg Heiler
Georg Heiler

Reputation: 17676

pandas parse csv with left and right quote chars

I am trying to read a file in pandas which is structured as follows

<first>$$><$$<second>$$><$$<first>$$>
<foo>$$><$$<bar>$$><$$<baz>$$>

using pd.read_csv('myflie.csv', encoding='utf8', sep='$$><$$', decimal=',') will fail to produce a meaningful result. All data is read into a single colum, and quotes are not extracted.

Upvotes: 2

Views: 436

Answers (1)

jezrael
jezrael

Reputation: 862681

You need escape $ by \, because it is read as regex (end of string):

(separators > 1 char and different from '\s+' are interpreted as regex)

import pandas as pd
from pandas.compat import StringIO


temp=u"""<first>$$><$$<second>$$><$$<first>$$>
<foo>$$><$$<bar>$$><$$<baz>$$>"""
#after testing replace 'StringIO(temp)' to 'filename.csv'
df = pd.read_csv(StringIO(temp), 
                 encoding='utf8', 
                 sep='\$\$><\$\$', 
                 decimal=',', 
                 header=None, 
                 engine='python')

print (df)
         0         1           2
0  <first>  <second>  <first>$$>
1    <foo>     <bar>    <baz>$$>

And then for remove $$> from last column is possible use replace (added & for end of string):

df.iloc[:, -1] = df.iloc[:, -1].str.replace('\$\$>$', '')
print (df)
         0         1        2
0  <first>  <second>  <first>
1    <foo>     <bar>    <baz>

And for remove quoting:

df = df.replace(['^<', '>$'], ['', ''], regex=True)
print (df)
       0       1      2
0  first  second  first
1    foo     bar    baz

Both replace together:

df = df.replace(['^<', '>$', '>\$\$'], ['', '', ''], regex=True)
print (df)
       0       1      2
0  first  second  first
1    foo     bar    baz

Upvotes: 3

Related Questions