Clients were sending me XLS files but my webs consume JSONs. So I needed to convert them to JSON easily from command-line...
- export XLS as a CSV file (I use OpenOffice.org for this)
- run
csv2json file.csv > file.json
(orjson2csv file.json > file.csv
) - there is no step 3
note: make sure your XLS table has the first row with column names, or use the -H option to provide some
This CSV file:
thumbpath,imgpath,imgsrc,width,height,thumbsrc,thumbwidth,thumbheight
thumbnails/,images/,paris_01.jpg,350,262,paris_01.jpg,75,56
thumbnails/,images/,paris_02.jpg,262,350,paris_02.jpg,75,56
gets turned into this JSON:
[
{
"thumbwidth": 75,
"imgsrc": "paris_01.jpg",
"thumbsrc": "paris_01.jpg",
"height": 262,
"imgpath": "images/",
"thumbheight": 56,
"thumbpath": "thumbnails/",
"width": 350
},
{
"thumbwidth": 75,
"imgsrc": "paris_02.jpg",
"thumbsrc": "paris_02.jpg",
"height": 350,
"imgpath": "images/",
"thumbheight": 56,
"thumbpath": "thumbnails/",
"width": 262
}
]
sudo gem install csv2json
Usage: csv2json [INPUT] [OPTIONS]
Specific options:
-s, --separator SEP Set separator character surrounded by single quotes (default is ',')
-o, --output FILE Write output to a file
-p, --pretty Pretty-format JSON output
-k, --skip-headers-from-file Ignore the headers (first line) in the file; use with --headers to set your own
-H, --headers HEADERS Supply list of headers, where no headers exist in the file, or where you're using -k to ignore them
-h, --help Show this message
-v, --version Show version
We also provide json2csv
, which converts in the opposite direction.
Usage: json2csv [INPUT] [OPTIONS]
Specific options:
-s, --separator SEP Set separator character surrounded by single quotes (default is ',')
-o, --output FILE Write output to a file
-H, --headers HEADERS Supply sorted list of headers, by which to order the columns in the CSV. These must match the key names in the JSON.
-h, --help Show this message
-v, --version Show version
Note: JSON field order is explicitly unsorted, so if you want predictable field order for CSV output, use the -H option.
common usage is csv2json file.csv > file.json
csv2json should behave like proper unix command-line utility working with pipes, redirects, etc.
cat file.csv | csv2json | gzip > file.json.gz
with files
require 'csv2json'
File.open('input.csv', 'r') do |input|
File.open('output.json', 'w') do |output|
CSV2JSON.parse(input, output)
end
end
or in-memory
require 'csv2json'
input = StringIO.new(csv_string)
output = StringIO.new()
CSV2JSON.parse(input, output)
output.pos = 0
puts output.read
https://github.com/darwin/csv2json/graphs/contributors