Merge branch 'main' of github.com:dacort/faker-cli into main

Author: dacort

README.md

@@ -66,6 +66,8 @@ fake -n 10 pyint,user_name,date_this_year -f json
 {"pyint": 5306, "user_name": "mark12", "date_this_year": "2023-04-16"}
 ```
 
+### Column Names
+
 Default column names aren't good enough for you? Fine, use your own.
 
 ```bash
@@ -85,9 +87,23 @@ fake -n 10 pyint,user_name,date_this_year -f json -c id,awesome_name,last_attent
 {"id": 1967, "awesome_name": "jmendoza", "last_attention_at": "2023-01-23"}
 ```
 
+### Provider Arguments
+
+Some [Faker providers](https://faker.readthedocs.io/en/master/providers/baseprovider.html) (like `pyint`) take arguments. You can also specify those if you like, separated by semi-colons (_because some arguments take a comma-separated string :)_)
+
+```bash
+fake -n 10 "pyint(1;100),credit_card_number(amex),pystr_format(?#-####)" -f json -c id,credit_card_number,license_plate
+```
+
+And unique values are supported as well.
+
+```bash
+fake -n 10 "unique.pyint(1;10),unique.name"
+```
+
 ### Parquet
 
-OK, it had to happen, you can even write Parquet. 
+OK, it had to happen, you can even write Parquet.
 
 Install with the `parquet` module: `pip install faker-cli[parquet]`
 
@@ -111,7 +127,6 @@ Install with the `delta` module: `pip install faker-cli[delta]`
 fake -n 10 pyint,user_name,date_this_year -f deltalake -o sample_data
 ```
 
-
 ## Templates
 
 The libary includes a couple templates that can be used to generate certain types of fake data easier.
@@ -130,4 +145,4 @@ How about CloudFront? Go ahead.
 fake -t cloudfront -n 10
 ```
 
-> **Warning**: Both of these templates are still being validated - please be cautious!
+> **Warning**: Both of these templates are still being validated - please be cautious!

faker_cli/cli.py

@@ -1,9 +1,9 @@
 import sys
-from typing import List
 
 import click
 from faker import Faker
 
+from faker_cli.parser import infer_column_names, parse_column_types
 from faker_cli.templates import (
     CloudFrontLogs,
     CloudFrontWriter,
@@ -12,18 +12,6 @@
 )
 from faker_cli.writer import CSVWriter, JSONWriter
 
-
-def infer_column_names(col_names, col_types: str) -> List[str]:
-    """
-    Infer column names from column types
-    """
-    # For now, nothing special - but eventually we need to parse things out
-    if col_names:
-        return col_names.split(",")
-
-    return col_types.split(",")
-
-
 KLAS_MAPPER = {
     "csv": CSVWriter,
     "json": JSONWriter,
@@ -108,11 +96,22 @@ def main(num_rows, format, output, columns, template, column_types):
         return
 
     # Now, if a template hasn't been provided, generate some fake data!
-    col_types = column_types.split(",")
+    # col_types = column_types.split(",")
+    # Note that if args are provided, the column headers are less than ideal
+    col_types = parse_column_types(column_types)
     headers = infer_column_names(columns, column_types)
-    writer = KLAS_MAPPER.get(format)(sys.stdout, headers, output)
+    format_klas = KLAS_MAPPER.get(format)
+    if format_klas is None:
+        raise click.ClickException(f"Format {format} not supported.")
+    writer = format_klas(sys.stdout, headers, output)
     for i in range(num_rows):
-        # TODO: Handle args
-        row = [fake.format(ctype) for ctype in col_types]
-        writer.write(row)
+        writer.write(generate_row(fake, col_types))
     writer.close()
+
+def generate_row(fake: Faker, column_types: list[tuple[str, list]]) -> list[str]:
+    return [
+        fake.format(ctype, *args)
+        if not ctype.startswith("unique.")
+        else fake.unique.format(ctype.removeprefix("unique."), *args)
+        for ctype, args in column_types
+    ]

faker_cli/parser.py

@@ -0,0 +1,39 @@
+from typing import List, Tuple
+
+def infer_column_names(col_names, col_types: str) -> List[str]:
+    """
+    Infer column names from column types
+    """
+    # For now, nothing special - but eventually we need to parse things out
+    if col_names:
+        return col_names.split(",")
+
+    return col_types.split(",")
+
+
+def parse_column_types(input_string: str) -> List[Tuple[str, List]]:
+    """
+    Parse a string of the format "pyint(1;10),datetime,profile(ssn,birthdate)" and split it by commas with optional parenthese-inclosed arguments.
+    """
+    import re
+
+    pattern = r"([\w\.]+)(\([^)]*\))?"
+    matches = re.findall(pattern, input_string)
+
+    # Extract the matched groups
+    result = [
+        (match[0], ([try_convert_to_int(val) for val in match[1].strip("()").split(";")] if match[1] else []))
+        for match in matches
+    ]
+
+    # print(result)
+    # [('pyint', ['1', '10']), ('datetime', []), ('profile', ['ssn,birthdate'])]
+
+    return result
+
+
+def try_convert_to_int(s):
+    try:
+        return int(s)
+    except ValueError:
+        return s

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "faker-cli"
-version = "0.4.0"
+version = "0.5.0"
 description = "Command-line fake data generator"
 authors = ["Damon P. Cortesi <d.lifehacker@gmail.com>"]
 readme = "README.md"

tests/test_cli.py

@@ -1,7 +1,9 @@
-from faker_cli.cli import main
-from click.testing import CliRunner
 import json
+
 import deltalake
+from click.testing import CliRunner
+
+from faker_cli.cli import main
 
 
 # Test that help is provided if the user provides no arguments
@@ -44,6 +46,7 @@ def test_numlines():
         lines = result.output.strip().splitlines()
         assert len(lines) == (6 if format == "csv" else 5)
 
+
 def test_custom_column_names():
     runner = CliRunner()
     result = runner.invoke(main, ["pyint,user_name", "-f", "json", "-c", "first,second"])
@@ -53,9 +56,10 @@ def test_custom_column_names():
     assert len(data.keys()) == 2
     assert list(data) == ["first", "second"]
 
+
 def test_deltalake_output(tmp_path):
     runner = CliRunner()
-    file = tmp_path / 'table'
+    file = tmp_path / "table"
     result = runner.invoke(main, ["pyint,user_name", "-f", "deltalake", "-o", file])
     assert result.exit_code == 0
     delta_table = deltalake.DeltaTable(file)
@@ -65,4 +69,34 @@ def test_deltalake_output(tmp_path):
 
     column_names = arrow_table.column_names
     assert column_names == ["pyint", "user_name"]
-    assert arrow_table.num_columns == 2
+    assert arrow_table.num_columns == 2
+
+
+def test_provider_args():
+    runner = CliRunner()
+    result = runner.invoke(main, ["-n", "10", "pyint(1;10)", "-f", "json", "-c", "id"])
+    assert result.exit_code == 0
+    lines = result.output.strip().splitlines()
+    for line in lines:
+        data: dict = json.loads(line)
+        assert data["id"] in range(1, 11)
+
+
+def test_unique_provider_args():
+    runner = CliRunner()
+    result = runner.invoke(main, ["-n", "10", "unique.pyint(1;10)", "-f", "json", "-c", "id"])
+    assert result.exit_code == 0
+    lines = result.output.strip().splitlines()
+    vals = []
+    for line in lines:
+        data: dict = json.loads(line)
+        assert data["id"] in range(1, 11)
+        vals.append(data["id"])
+
+    assert set(vals) == set(range(1, 11))
+
+
+def test_unique_provider_args_limit():
+    runner = CliRunner()
+    result = runner.invoke(main, ["-n", "10", "unique.pyint(1;5)", "-f", "json", "-c", "id"])
+    assert result.exit_code == 1