read_csv fails when a column ends in a quote.

[jvangael]

I use pandas to read csv files that I get from mysql. Using pandas 0.9 and 0.10 beta, the following files give unexpected results when using read_csv

"A","B"
"row_1","foo""
"row_2","bar"

output is

row_1 foo"\nrow_2" bar

When I add a space after the first quote after foo,

"A","B"
"row_1","foo" "
"row_2","bar"

we do get the correct number of rows back but output is still unexpected

0 row_1 foo\ "
1 row_2 bar

[wesm]

Hey Jurgen, long time no see.

I think you'll get what you want if you specify \ as the escape character:

In [1]: read_csv('tmp.csv', sep=',', escapechar='\\')
Out[1]: 
       A     B
0  row_1  foo"
1  row_2   bar

[jvangael]

Hey Wes, thanks for the quick response. Terribly sorry for the noise - I hadn't realized escapechar is the new kid on the block in 0.10. Very clever work on the new release - I look forward to upgrade!