A Bash script tool for filtering and splitting CSV file. This tool is to help you split a large CSV into several CSV files based on filtered criteria(s).
I faced a situation recently in my work where i need to split up a csv file into several csv files, based on several criterias. And I needed each file to be name after last criteria specified.
./dataSplitter <source-csv> <filters-csv> [-d]
This will split the <source-csv>
, making sure that every output matched all criteria(s) in <filters-csv>
.
The only option is -d
, which allows you to get the details of running process instead of plain progress bar.
"title 1","title 2","title 3","title 4","title 5"
"data r1c1","data r1c2","data r1c3","data r1c4","data r1c5"
"data r2c1","data r2c2","data r2c3","data r2c4","data r2c5"
"data r3c1","data r3c2","data r3c3","data r3c4","data r3c5"
Line 1 must be the title or column name.
Line 2 and below is the raw data.
You need to make sure the last line is an empty line.
$2,$4,$5
"filter 1", "filter 2", "filter 3"
"filter 1", "filter 2", "filter 3"
"filter 1", "filter 2", "filter 3"
Line 1 must be the column number of filters to be applied based on source-csv, starting with $
Line 2 and below is the filter values.
You need to make sure the last line is an empty line.
Currently you can use up to 5 filters.
The <source-csv>
will be splitting into several CSVs in output
folder. You will be prompt the output
folder name during the execution of the script.
Naming conversion for the files is {second_last_filter_value}_{last_filter_value}.csv. And the folder structure also will follow the filter values.
output/filter_1/filter_2/filter2_filter_3.csv