New option: csv-diff one.csv two.csv --json

Author: simonw

csv_diff/__init__.py

@@ -23,35 +23,56 @@ def compare(previous, current):
  removed_or_added = set(removed) | set(added)
  potential_changes = [id for id in current if id not in removed_or_added]
  changed = [id for id in potential_changes if current[id] != previous[id]]
- # Now generate the readable summary
- summary = []
- title = []
+ result = {"added": [], "removed": [], "changed": []}
  if added:
- fragment = "{} row{} added".format(len(added), "" if len(added) == 1 else "s")
+ result["added"] = [current[id] for id in added]
+ if removed:
+ result["removed"] = [previous[id] for id in removed]
+ if changed:
+ for id in changed:
+ d = list(diff(previous[id], current[id]))
+ result["changed"].append(
+ {
+ "key": id,
+ "changes": {
+ field: [prev_value, current_value]
+ for _, field, (prev_value, current_value) in d
+ },
+ }
+ )
+ return result
+
+
+def human_text(result):
+ title = []
+ summary = []
+ if result["added"]:
+ fragment = "{} row{} added".format(
+ len(result["added"]), "" if len(result["added"]) == 1 else "s"
+ )
  title.append(fragment)
  summary.append(fragment + "\n")
- for id in added:
- summary.append(" {}".format(json.dumps(current[id])))
+ for row in result["added"]:
+ summary.append(" {}".format(json.dumps(row)))
  summary.append("")
- if removed:
+ if result["removed"]:
  fragment = "{} row{} removed".format(
- len(removed), "" if len(removed) == 1 else "s"
+ len(result["removed"]), "" if len(result["removed"]) == 1 else "s"
  )
  title.append(fragment)
  summary.append(fragment + "\n")
- for id in removed:
- summary.append(" {}".format(json.dumps(previous[id])))
+ for row in result["removed"]:
+ summary.append(" {}".format(json.dumps(row)))
  summary.append("")
- if changed:
+ if result["changed"]:
  fragment = "{} row{} changed".format(
- len(changed), "" if len(changed) == 1 else "s"
+ len(result["changed"]), "" if len(result["changed"]) == 1 else "s"
  )
  title.append(fragment)
  summary.append(fragment + "\n")
- for id in changed:
- d = list(diff(previous[id], current[id]))
- summary.append(" Row {}".format(id))
- for _, field, (prev_value, current_value) in d:
+ for details in result["changed"]:
+ summary.append(" Row {}".format(details["key"]))
+ for field, (prev_value, current_value) in details["changes"].items():
  summary.append(
  ' {}: "{}" => "{}"'.format(field, prev_value, current_value)
  )

csv_diff/cli.py

@@ -1,5 +1,6 @@
 import click
-from . import load_csv, compare
+import json as std_json
+from . import load_csv, compare, human_text
 
 
 @click.command()
@@ -15,6 +16,13 @@
 @click.option(
  "--key", type=str, default=None, help="Column to use as a unique ID for each row"
 )
-def cli(previous, current, key):
+@click.option(
+ "--json", type=bool, default=False, help="Output changes as JSON", is_flag=True
+)
+def cli(previous, current, key, json):
  "Diff two CSV files"
- print(compare(load_csv(open(previous), key=key), load_csv(open(current), key=key)))
+ diff = compare(load_csv(open(previous), key=key), load_csv(open(current), key=key))
+ if json:
+ print(std_json.dumps(diff, indent=4))
+ else:
+ print(human_text(diff))

tests/test_cli.py

@@ -0,0 +1,31 @@
+from click.testing import CliRunner
+from csv_diff import cli
+from .test_csv_diff import ONE, TWO, THREE
+import json
+
+
+def test_human_cli(tmpdir):
+ one = tmpdir / "one.csv"
+ one.write(ONE)
+ two = tmpdir / "two.csv"
+ two.write(TWO)
+ result = CliRunner().invoke(cli.cli, [str(one), str(two), "--key", "id"])
+ assert 0 == result.exit_code
+ assert (
+ '1 row changed\n\n1 row changed\n\n Row 1\n age: "4" => "5"'
+ == result.output.strip()
+ )
+
+
+def test_human_cli_json(tmpdir):
+ one = tmpdir / "one.csv"
+ one.write(ONE)
+ two = tmpdir / "two.csv"
+ two.write(TWO)
+ result = CliRunner().invoke(cli.cli, [str(one), str(two), "--key", "id", "--json"])
+ assert 0 == result.exit_code
+ assert {
+ "added": [],
+ "removed": [],
+ "changed": [{"key": "1", "changes": {"age": ["4", "5"]}}],
+ } == json.loads(result.output.strip())

tests/test_csv_diff.py

@@ -9,9 +9,38 @@
 1,Cleo,5
 2,Pancakes,2"""
 
+THREE = """id,name,age
+1,Cleo,5"""
 
-def test_diff():
+
+def test_row_changed():
  diff = compare(
  load_csv(io.StringIO(ONE), key="id"), load_csv(io.StringIO(TWO), key="id")
  )
- assert '1 row changed\n\n1 row changed\n\n Row 1\n age: "4" => "5"' == diff
+ assert {
+ "added": [],
+ "removed": [],
+ "changed": [{"key": "1", "changes": {"age": ["4", "5"]}}],
+ } == diff
+
+
+def test_row_added():
+ diff = compare(
+ load_csv(io.StringIO(THREE), key="id"), load_csv(io.StringIO(TWO), key="id")
+ )
+ assert {
+ "changed": [],
+ "removed": [],
+ "added": [{"age": "2", "id": "2", "name": "Pancakes"}],
+ } == diff
+
+
+def test_row_removed():
+ diff = compare(
+ load_csv(io.StringIO(TWO), key="id"), load_csv(io.StringIO(THREE), key="id")
+ )
+ assert {
+ "changed": [],
+ "removed": [{"age": "2", "id": "2", "name": "Pancakes"}],
+ "added": [],
+ } == diff

tests/test_human_text.py

@@ -0,0 +1,33 @@
+from csv_diff import load_csv, compare, human_text
+from .test_csv_diff import ONE, TWO, THREE
+import io
+
+
+def test_row_changed():
+ diff = compare(
+ load_csv(io.StringIO(ONE), key="id"), load_csv(io.StringIO(TWO), key="id")
+ )
+ assert (
+ '1 row changed\n\n1 row changed\n\n Row 1\n age: "4" => "5"'
+ == human_text(diff)
+ )
+
+
+def test_row_added():
+ diff = compare(
+ load_csv(io.StringIO(THREE), key="id"), load_csv(io.StringIO(TWO), key="id")
+ )
+ assert (
+ '1 row added\n\n1 row added\n\n {"id": "2", "name": "Pancakes", "age": "2"}'
+ == human_text(diff)
+ )
+
+
+def test_row_removed():
+ diff = compare(
+ load_csv(io.StringIO(TWO), key="id"), load_csv(io.StringIO(THREE), key="id")
+ )
+ assert (
+ '1 row removed\n\n1 row removed\n\n {"id": "2", "name": "Pancakes", "age": "2"}'
+ == human_text(diff)
+ )