从bash脚本文件读取csv文件

Question

in a folder, I have 2 files. Valuation.csv and another file myScript.sh

My CSV file has 10 lines and 5 columns. I have tried to read it multiple times but it never worked. This is what I have tried as code inside my myScript.sh:

First try:

#!/bin/bash

    while read -r line do
        field1=$(echo $line | awk -F'|' '{printf "%s", $1}' | tr -d '"')
        field2=$(echo $line | awk -F'|' '{printf "%s", $2}' | tr -d '"')

        echo $field1 $field2 done < $Valuation.csv

Result: /myScript.sh: line 10: .csv: No such file or directory

Second try:

cat Valuation.csv|while read line   do
read -d, col1 col2 < <(echo $line)
echo "I got:$col1|$col2"   done

Result: nothing

I am running the file like this:

./myScript.sh

thank you

Here the csv file opened with excel

Answer 1

here is my file data.csv

Site,Post,Subject,User,Status
stackover flow,bash,read csv,James,ok
git,linux,core,Novy,ko

in my bash script i have something like:

#!/bin/bash
INPUT=./data.csv
OLDIFS=$IFS
IFS=,
[ ! -f $INPUT ] && { echo "$INPUT file not found"; exit 99; }
while read site post subject usr status
do
    echo "Site : $site"
    echo "Post : $post"
    echo "Subject : $subject"
    echo "User : $usr"
    echo "Status : $status"
done < $INPUT
IFS=$OLDIFS

Output

Site : stackover flow Post : bash Subject : read csv User : James Status : ok Site : git Post : linux Subject : core User : Novy Status : ko

You can give a look here

Answer 2

James, avoid calling utilities within a loop in bash. That is horribly inefficient. Avoid using a fixed number of fields -- what if another is added? You can read .csv data in bash fairly easily by simply using read -a array... and setting the internal field separator to word-split on a comma (eg IFS=$',\\n' ).

For example, writing a script that takes the .csv filename to read as the first argument, you could do:

#!/bin/bash

declare -i row=0

while IFS=$',\n' read -r -a array; do
    printf "row %d - read %d fields.\n" $((row++)) ${#array[@]}
    for i in "${array[@]}"; do
        printf " %-12s" "$i"
    done
    echo ""
done < "$1"

( note: while you would in-practice want to handle the entire matter in awk , since this appears to be a bash exercise, for short .csv files, bash is more than capable)

Example Data

(I wasn't going to retype your spreadsheet, so here are the first two rows)

$ cat company.csv
Facebook,35587,55800000000,1567988,491000000000,8.80x
Uber,16000,11300000000,706250,120000000000,10.6x

Running the script and simply outputting the row count and fields-per-row followed by the field values themselves, you would get:

$ bash readcsv.sh company.csv
row 0 - read 6 fields.
 Facebook     35587        55800000000  1567988      491000000000 8.80x
row 1 - read 6 fields.
 Uber         16000        11300000000  706250       120000000000 10.6x

Look things over and let me know if you have questions.