read_csv blank lines

[hbeale]

If there are blank lines before a header, the header is recognized as such but also included as a row. I came across this in read_tsv, but the min example was easier in read_csv.

e.g.

read_csv("\n\nfileName,variableName,fileDescription\nmy.tsv,awesomeStuff,that awesome stuff i made")
# A tibble: 2 × 3
fileName variableName fileDescription
<chr> <chr> <chr>
1 fileName variableName fileDescription
2 my.tsv awesomeStuff that awesome stuff i made

It may not be an additional problem, but in my case I'm also using comments:
read_csv("# this is a comment\n\nfileName,variableName,fileDescription\nmy.tsv,awesomeStuff,that awesome stuff i made", comment="#")

[slonik-az]

I ran into this problem today. Very annoying behavior. read.csv has blank.lines.skip parameter. read_csv should have something similar.

[jimhester]

If you know the number of blank / comment lines use the skip parameter.

[jimhester]

Closed by 6517708

[slonik-az]

• It is very dangerous and bug-prone to rely on knowing number of blank lines. I would not let such code through a code review.
• Empty lines can happen in the middle of the cvs file. In this case skip lines parameter is useless.

I still cannot get it -- why is it so difficult to skip over blank lines. What am I missing?

--Leo

[lock]

This old issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with reprex) and link to this issue. https://reprex.tidyverse.org/