I have file like:
AA,A=14,B=356,C=845,D=4516
BB,A=65,C=255,D=841,E=5133,F=1428
CC,A=88,B=54,C=549,F=225
I never know if in the row missing A,B,C or D value. But I need to transform this file like:
AA,A=14,B=356,C=845,D=4516,-,-
BB,A=65,-,C=255,D=841,E=5133,F=1428
CC,A=88,B=54,C=549,-,-,F=225
So if any value missing print just -
mark. My plan is have the same number of columns to easy parsing. I am prefer awk solution. Thank you for any advice or help.
My first try was:
awk '{gsub(/[,]/, "\t")}; BEGIN{ FS = OFS = "\t" } { for(i=1; i<=NF; i++) if($i ~ /^ *$/) $i = "-" }; {print $0}'
But then I notice, that some values are missing.
EDIT:
From my header I know that there is value A,B,C,D,E,F...
$ cat file.txt
AA,A=14,B=356,C=845,D=4516
BB,A=65,C=255,D=841,E=5133,F=1428
CC,A=88,B=54,C=549,F=225
$ perl -F, -le '@k=(A..F);
$op[0]=$F[0]; @op[1..6]=("-")x6;
$j=0; for($i=1;$i<=$#F;){ if($F[$i] =~ m/$k[$j++]=/){$op[$j]=$F[$i]; $i++} }
print join(",",@op)
' file.txt
AA,A=14,B=356,C=845,D=4516,-,-
BB,A=65,-,C=255,D=841,E=5133,F=1428
CC,A=88,B=54,C=549,-,-,F=225
-F,
split input line on ,
and save to @F
array -l
removes newline from input line, adds newline to output @k=(A..F);
initialize @k
array with A
, B
, etc upto F
$op[0]=$F[0]; @op[1..6]=("-")x6;
initalize @op
array with first element of @F
and remaining six elements as -
@F
array, if element matches with @k
array element in corresponding index followed by =
, change @op
element print join(",",@op)
print the @op
array with ,
as separator Perl to the rescue!
You haven't specified how to obtain the header information, so in the following script, the @header array is populated directly.
%to_idx
hash maps the column names to their indices (A => 0, B => 1 etc.).
Each lines is split into fields, each field is compared to the expected one ( $next
) and dashes are printed if needed. The same happens for missing trailing fields.
#!/usr/bin/perl
use warnings;
use strict;
my @header = qw( A B C D E F );
my %to_idx = map +($header[$_] => $_), 0 .. $#header;
open my $IN, '<', shift or die $!;
while (<$IN>) {
chomp;
my @fields = split /,/;
print shift @fields;
my $next = 0;
for my $field (@fields) {
my ($name, $value) = split /=/, $field;
print ',-' x ($to_idx{$name} - $next);
print ",$name=$value";
$next = $to_idx{$name} + 1;
}
print ',-' x (1 + $#header - $next); # Missing trailing fields.
print "\n"
}
Solution in TXR
@(do (defstruct fill-missing nil strings (hash (hash :equal-based)) (:postinit (self) (each ((s self.strings)) (set [self.hash s] "-"))) (:method add (self str val) (set [self.hash str] `@str=@val`)) (:method print (self stream) (put-string `@{(mapcar self.hash self.strings) ","}` stream)))) @(repeat) @ (bind fm @(new fill-missing strings '#"A B C D E F")) @{label},@(coll)@{sym /[^,=]+/}=@{val /[^,]+/}@(do fm.(add sym val))@(end) @ (do (put-line `@label,@fm`)) @(end)
Run:
$ txr missing.txr data AA,A=14,B=356,C=845,D=4516,-,- BB,A=65,-,C=255,D=841,E=5133,F=1428 CC,A=88,B=54,C=549,-,-,F=225
BEGIN {
PROCINFO["sorted_in"]="@ind_str_asc" # order for for(i in a)
for(i=65;i<=90;i++) # create the whole alphabet to array a[]
a[sprintf("%c", i)] # you could read the header and use that as well
}
{
split($0,b,",") # split record by ","
printf "%s", b[1] # printf first element (AA, BB...)
delete b[1] # get rid of it
for(i in b)
b[substr(b[i],1,1)]=b[i] # take the first letter to use as index (A=12)
for(i in a) # go thru alphabet and printf from b[]
printf "%s%s", OFS, (i in b?b[i]:"-"); print ""
}
awk -v OFS=\, -f parsing.awk tbparsed.txt
AA,A=14,B=356,C=845,D=4516,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-
BB,A=65,-,C=255,D=841,E=5133,F=1428,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-
CC,A=88,B=54,C=549,-,-,F=225,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-,-
It prints "-" for each letter not found in the record. If the data had a header, you could split
to 2-D array b[NR]
and change the for(i in a)
to for(i in b[1]) ... printf ... b[NR][b[1][i]] ...
and if you don't need the static first column, remove the first printf
and delete
.
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.