awk delete line from csv if char length of second column is less than 12 - awk

Question

I have a csv which looks like so:
42242,"France."
2343242,"Laboratoire de Virologie, AP-HP, Hôpital Necker-Enfants malades, Paris, France."
234234,"brazil"
23432423,"colombia"
234234,"Université de Paris, Institut Cochin, INSERM U1016, CNRS UMR8104, Paris, France."
234234,"Paris."
I would like to delete rows where the char length of the second column is <12.
I think awk can do this:
awk -F , '$2=length>12' file >filout
but this seems wrong.. :(
I want to delete the line to get:
2343242,"Laboratoire de Virologie, AP-HP, Hôpital Necker-Enfants malades, Paris, France."
234234,"Université de Paris, Institut Cochin, INSERM U1016, CNRS UMR8104, Paris, France."

Answer 1

$ awk -F, 'length($2)>=12' input_file
2343242,"Laboratoire de Virologie, AP-HP, Hôpital Necker-Enfants malades, Paris, France."
234234,"Université de Paris, Institut Cochin, INSERM U1016, CNRS UMR8104, Paris, France."

Answer 2

As your second field is contained within double quotes, you must use the double quote, rather than the comma, as the separator to determine the length of the second field:
awk -F\" 'length($2)>=12' file
If you just print the length of the second field, you will see what I mean. First using the comma as separator:
awk -F, '{print length($2)}' file
9
25
8
10
20
8
Second, using the double quote as the separator:
awk -F\" '{print length($2)}' file
7
79
6
8
80
6

Answer 3

adding gsub(/[\200-\277]/, "&") to properly measure # of UTF-8 characters in byte mode, assuming well-formed UTF-8 input. skip this part if you're using gawk in unicode mode
echo '42242,"France."
2343242,"Laboratoire de Virologie, AP-HP, Hôpital Necker-Enfants malades, Paris, France."
234234,"brazil"
23432423,"colombia"
234234,"Université de Paris, Institut Cochin, INSERM U1016, CNRS UMR8104, Paris, France."
234234,"Paris."' |
— to do it the CSV way without a proper parser (or an awk that's even unicode aware) - it measures full row length, minus string index position of 1st comma, then minus 2 more for the quotation marks :
mawk '(_+=++_)^_^_<(-_-- + length($--_) \
- index($_, FS) \
- gsub(/[\200-\277]/, "&"))' FS=','
to do it the double-quotes ("...") way :
gawk '((_+=++_)^_^_-_^_)<length($(--_+_--))' FS='"'
1 2343242,"Laboratoire de Virologie, AP-HP, Hôpital Necker-Enfants malades, Paris, France."
2 234234,"Université de Paris, Institut Cochin, INSERM U1016, CNRS UMR8104, Paris, France."

Answer 4

Using ruby with CSV.parse on slightly modified data to show correct output on commas within quotes.
% ruby -r 'csv' -ne 'if CSV.parse($_).map {|i|
if i[1].length >= 12 then true end}[0] then puts $_ end' file
2343242,"Labor, atoire de Virologie, AP-HP, Hôpital Necker-Enfants malades, Paris, France."
23432423,Institut Cochin,"colombia"
234234,"Université de Paris, Institut Cochin, INSERM U1016, CNRS UMR8104, Paris, France."
Data
% cat file
42242,"France."
2343242,"Labor, atoire de Virologie, AP-HP, Hôpital Necker-Enfants malades, Paris, France."
234234,"brazil"
23432423,Institut Cochin,"colombia"
234234,"Université de Paris, Institut Cochin, INSERM U1016, CNRS UMR8104, Paris, France."
234234,"Paris."

Answer 5

42242,"France."
2343242,"Laboratoire de Virologie, AP-HP, Hôpital Necker-Enfants malades, Paris, France."
234234,"brazil"
23432423,"colombia"
234234,"Université de Paris, Institut Cochin, INSERM U1016, CNRS UMR8104, Paris, France."
234234,"Paris."
does have , inside quoted values, therefore just setting field separator to , would not work, observe that e.g. 2nd columnd of 2nd row will be "Laboratoire de Virologie, to counter that you might use FPAT provided in More CSV chapter of GNU AWK manual as follows
awk 'BEGIN{FPAT="([^,]*)|(\"([^\"]|\"\")+\")"}length($2)>12' file.csv
though keep in mind that " are included in quotes fields, so you might need to adjust value inside comparison.