I use a txt file, and I need to convert it in csv file
Saint Petersburg 0 10 0.1 - N Moscow - 9 0 - N Novgorod 0 7 1 30 Y In bash, how can I insert comma after the last letter, and after every number or "-"
For example
Saint Petersburg, 0, 10, 0.1, -, N Moscow, -, 9, 0, -, N Novgorod, 0, 7, 1, 30, Y Best
This will replace sequences of 2 or more spaces by a command and the same spaces less one:
sed 's/ \( \+\)/,\1/g' file.txt Here using the \+ extension of GNU regexps, the standard equivalent being \{1,\}, or use the -E option to switch to extended regular expressions where + is standard:
sed -E 's/ ( +)/,\1/g' file.txt It will fail if there is only one space between columns.
If you only want ", " as the field separator:
sed 's/ \{2,\}/, /g' file same as
sed -E 's/ {2,}/, /g' file Or use:
sed -E 's/ {2,}/,/g' file For no spaces around the ,s as in most CSV formats.
Using Miller (mlr) to read the input and treat any multiple of two or more spaces as a field delimiter. The output format is header-less CSV; we use Miller's cat operation to pass the data through unchanged.
$ mlr --nidx --ifs-regex ' {2,}' --ocsv --headerless-csv-output cat file Saint Petersburg,0,10,0.1,-,N Moscow,-,9,0,-,N Novgorod,0,7,1,30,Y Instead of relying of there being at least two spaces between fields, you could work on the fact that there are always 6 fields the last 5 not contain whitespace and do:
perl -lpe 's/^\s*(.*?)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s*$/$1,$2,$3,$4,$5,$6/' file Saint Petersburg,0,10,0.1,-,N Moscow,-,9,0,-,N Novgorod,0,7,1,30,Y Or to make sure that the output is valid CSV, use the Text::CSV module which will ensure fields are quoted if needed:
$ perl -MText::CSV -lne ' BEGIN{$c = Text::CSV->new({binary=>1})} $c->print(*STDOUT, \@{^CAPTURE}) if /^\s*(.*?)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s*$/' file2 "Saint Petersburg",0,10,0.1,-,N "Moscow, Russia Capital",-,9,0,-,N Novgorod,0,7,1,30,Y
bashalone? Are the fields separated by tabs?od:od -c file.csv.