file-converter-fgsm 0.1.3

Converting a single file

Suppose you have a file sample_file.csv in the current directory with the contents

header1,header2,header3,header4
1,a,1.5,
2,b,3.2,info

To convert it into a json format, run

file_converter --conversion=csv2json --input_path=./sample_file.csv

and then you should have a sample_file.json in the same directory with the contents:

[
    {
        "header1": 1,
        "header2": "a",
        "header3": 1.5,
        "header4": null
    },
    {
        "header1": 2,
        "header2": "b",
        "header3": 3.2,
        "header4": "info"
    }
]

Notice how the missing info became null and each type was properly parsed.

The previous command is short for

file_converter --conversion=csv2json --input_path=./sample_file.csv --output_path=. --separator=, --prefix=""

Experiment running again but with a different --output_path.

The command supports the other way for the conversion as well. To test that, let us convert the recently converted json file but prefixing the resulting file name with a new_:

file_converter --conversion=json2csv --input_path=./sample_file.json --prefix="new_"

Notice the csv2json became json2csv in the --conversion attribute. Also, we should have a new_sample_file.csv with the same contents of the first.

Converting all files in a folder

If instead of a single file path you pass a folder in the --input_path parameter, the command will attempt to convert all files inside of it. Notice they all should have the same separator.

Experiment passing the current directory like

file_converter --conversion=csv2json --input_path=. --prefix="from_folder_"

If you followed the previous example, you should have two new json files starting with from_folder_.