Thanks for the suggestion. I don't think `readline` is a good suggestion, since the first line is frequently a header. Some headers don't need quoting, so you can't sniff quote characters from them. I see a magic number in an example as an invitation to use something better for my particular use :) On 20. 06. 23 10:42, Piergiovanni Bazzana wrote:

Hi, I want to submit an improvement, not really a bug but it is smth that may help people that need to deal with large csv files

In https://docs.python.org/3/library/csv.html <https://docs.python.org/3/library/csv.html>

I find in the example the line

|

|dialect = csv.Sniffer().sniff(csvfile.read(1024))|

while it would be probably better to use smth like: |

|dialect = csv.Sniffer().sniff(csvfile.readline())|

|

|because of no magic numbers and cleared rationale behind (I read it as: please try to infer the dialect from the header line of the csv file, which makes a lot of sense)|

|

-- Thank you, Best Regards

Piergiovanni Bazzana

_______________________________________________ docs mailing list -- docs@python.org To unsubscribe send an email to docs-leave@python.org https://mail.python.org/mailman3/lists/docs.python.org/ Member address: encukou@gmail.com