884 downloads this week        Star



PyPI Changelog Tests License

Tool for viewing the difference between two CSV, TSV or JSON files. See Generating a commit log for San Francisco’s official list of trees (and the sf-tree-history repo commit log) for background information on this project.


pip install csv-diff


Consider two CSV files:





csv-diff can show a human-readable summary of differences between the files:

$ csv-diff one.csv two.csv --key=id
1 row changed, 1 row added, 1 row removed

1 row changed

  Row 1
    age: "4" => "5"

1 row added

  id: 3
  name: Bailey
  age: 1

1 row removed

  id: 2
  name: Pancakes
  age: 2

The --key=id option means that the id column should be treated as the unique key, to identify which records have changed.

The tool will automatically detect if your files are comma- or tab-separated. You can over-ride this automatic detection and force the tool to use a specific format using --format=tsv or --format=csv.

You can also feed it JSON files, provided they are a JSON array of objects where each object has the same keys. Use --format=json if your input files are JSON.

Use --show-unchanged to include full details of the unchanged values for rows with at least one change in the diff output:

% csv-diff one.csv two.csv --key=id --show-unchanged
1 row changed

  id: 1
    age: "4" => "5"

      name: "Cleo"

You can use the --json option to get a machine-readable difference:

$ csv-diff one.csv two.csv --key=id --json
    "added": [
            "id": "3",
            "name": "Bailey",
            "age": "1"
    "removed": [
            "id": "2",
            "name": "Pancakes",
            "age": "2"
    "changed": [
            "key": "1",
            "changes": {
                "age": [
    "columns_added": [],
    "columns_removed": []

As a Python library

You can also import the Python library into your own code like so:

from csv_diff import load_csv, compare
diff = compare(
    load_csv(open("one.csv"), key="id"),
    load_csv(open("two.csv"), key="id")

diff will now contain the same data structure as the output in the --json example above.

If the columns in the CSV have changed, those added or removed columns will be ignored when calculating changes made to specific rows.

As a Docker container

Build the image

$ docker build -t csvdiff .

Run the container

$ docker run --rm -v $(pwd):/files csvdiff

Suppose current directory contains two csv files : one.csv two.csv

$ docker run --rm -v $(pwd):/files csvdiff one.csv two.csv