 |
Forums |
Admin Discussion Forums: help Start New Thread
By: Thibaut Barrère
 RE: Row preprocessing ? [ reply ]
2009-09-04 19:28
|
well sure, it makes sense - I just tend to avoid writing code if a single library call makes the job.
thanks for the code, I will definitely use this - and thanks for replying so quickly :)
-- Thibaut
|
By: James Gray
RE: Row preprocessing ? [ reply ]
2009-09-04 19:22
|
Luckily, handling the headers is very easy in your case. For example:
data = <<END_DATA
a:b:c
1:2:3
4:5:6
END_DATA
headers = nil
data.each do |row|
row.strip!
if headers.nil?
headers = row.split(":")
else
fields = Hash[*headers.zip(row.split(":")).flatten]
p fields
end
end
I'm not trying to run you off, of course. It's just that you are feeding non-CSV data to a CSV parser. It's not too surprising this is problematic. Your data is very easy to read without help though.
I hope that makes sense.
|
By: Thibaut Barrère
RE: Row preprocessing ? [ reply ]
2009-09-04 19:09
|
Well the fact that FasterCSV handles the headers nicely and returns a hash for each record made me use it in this case, even if I don't really need a parser, you're right :)
I could rewrite the code to achieve the same thing - but I wanted to ensure there wasn't any quick way to do that still with FasterCSV.
thoughts ?
|
By: Thibaut Barrère
Row preprocessing ? [ reply ]
2009-09-04 18:52
|
Hi,
I was wondering: is there some built-in way (converters maybe ?) to preprocess the content of each row before it gets parsed by FasterCSV ?
My content hasn't got any quotes in it:
id;name
7;Bob David
8;John "Amazing" Bowie
Until today I had no lines with double quotes in there, it's a new use case. The line with id 8 causes an Illegal quoting.
I tried setting :quote_char to nil but it needs a one-char string.
Is there some way to achieve this internally, or should I just preprocess the quote before myself (I don't need it at all) ?
thanks for your support (and work on FasterCSV),
Thibaut
--
http://www.learnivore.com
|
|
|
