Add mimesis support

.github/workflows/ci.yml

@@ -5,7 +5,7 @@ jobs:
     strategy:
       matrix:
         os: [ubuntu-latest, windows-latest, macos-latest]
-        python-version: ["3.9", "3.10", "3.11"]
+        python-version: ["3.10", "3.11", "3.12"]
     runs-on: ${{ matrix.os }}
     defaults:
       run:

.github/workflows/pypi.yml

@@ -12,7 +12,7 @@ jobs:
       - name: Set up Python
         uses: actions/setup-python@v4
         with:
-          python-version: 3.9
+          python-version: 3.12
       - name: cache poetry install
         uses: actions/cache@v3
         with:

README.md

@@ -87,6 +87,19 @@ 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"}
 ```
 
+### Providers (beta)
+
+While [Faker](https://faker.readthedocs.io) is a sweet library, we all like options don't we? [Mimesis](https://mimesis.name/en/master/) is _also_ awesome and can be quite a bit faster than Faker. 🤫 You can use a different provider by using `-p mimesis`.
+
+> [!NOTE]  
+> Providers use their own syntax for data types, so you must change out your column names as necessary.
+
+To generate the same dataset above with Mimesis for example:
+
+```bash
+fake -p mimesis -n 10 "numeric.integer_number(0),person.username,datetime.date(2024)" -f json -c id,awesome_name,last_attention_at
+```
+
 ### 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 :)_)

faker_cli/cli.py

@@ -4,12 +4,9 @@
 from faker import Faker
 
 from faker_cli.parser import infer_column_names, parse_column_types
-from faker_cli.templates import (
-    CloudFrontLogs,
-    CloudFrontWriter,
-    S3AccessLogs,
-    S3AccessWriter,
-)
+from faker_cli.providers.faker import FakerProvider
+from faker_cli.providers.mimesis import MimesisProvider
+from faker_cli.templates import CloudFrontWriter, S3AccessWriter
 from faker_cli.writer import CSVWriter, JSONWriter
 
 KLAS_MAPPER = {
@@ -22,10 +19,6 @@
     "cloudfront": [CloudFrontWriter, "cloudfront_log"],
 }
 
-fake = Faker()
-fake.add_provider(S3AccessLogs)
-fake.add_provider(CloudFrontLogs)
-
 
 @click.command()
 @click.option("--num-rows", "-n", default=1, help="Number of rows")
@@ -40,7 +33,8 @@
 @click.option("--columns", "-c", help="Column names", default=None, required=False)
 @click.option("--template", "-t", help="Template to use", type=click.Choice(["s3access", "cloudfront"]), default=None)
 @click.argument("column_types", required=False)
-def main(num_rows, format, output, columns, template, column_types):
+@click.option("--provider", "-p", help="Fake data provider", type=click.Choice(["faker", "mimesis"]), default="faker")
+def main(num_rows, format, output, columns, template, column_types, provider):
     """
     Generate fake data, easily.
 
@@ -49,13 +43,24 @@ def main(num_rows, format, output, columns, template, column_types):
 
     You can also use --template for real-world synthetic data.
     """
+    if provider == "faker":
+        fake = FakerProvider()
+    elif provider == "mimesis":
+        fake = MimesisProvider()
+    else:
+        pass
+
     # Do some initial validation - we must have either template or column tpes
     if not template and not column_types:
         ctx = click.get_current_context()
         click.echo(ctx.get_help())
         ctx.exit()
         raise click.BadArgumentUsage("either --template or a list of Faker property names must be provided.")
 
+    # Templates are only supported with Faker at the moment
+    if template and provider != "faker":
+        raise click.BadArgumentUsage('templates are only supported with the "faker" provider.')
+
     # Parquet output requires a filename
     if format in ["parquet", "deltalake"] and output is None:
         raise click.BadArgumentUsage("parquet | deltalake formats requires --output/-o filename parameter.")
@@ -105,13 +110,14 @@ def main(num_rows, format, output, columns, template, column_types):
         raise click.ClickException(f"Format {format} not supported.")
     writer = format_klas(sys.stdout, headers, output)
     for i in range(num_rows):
-        writer.write(generate_row(fake, col_types))
+        writer.write(fake.generate_row(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/providers/base.py

@@ -0,0 +1,2 @@
+class BaseProvider:
+    pass

faker_cli/providers/faker.py

@@ -0,0 +1,21 @@
+from faker import Faker
+
+from faker_cli.templates import CloudFrontLogs, S3AccessLogs
+
+
+class FakerProvider:
+    def __init__(self) -> None:
+        self.fake = Faker()
+        self.fake.add_provider(S3AccessLogs)
+        self.fake.add_provider(CloudFrontLogs)
+
+    def generate_row(self, column_types: list[tuple[str, list]]) -> list[str]:
+        return [
+            self.fake.format(ctype, *args)
+            if not ctype.startswith("unique.")
+            else self.fake.unique.format(ctype.removeprefix("unique."), *args)
+            for ctype, args in column_types
+        ]
+
+    def format(self, log_entry) -> list[str]:
+        return self.fake.format(log_entry)

faker_cli/providers/mimesis.py

@@ -0,0 +1,15 @@
+from mimesis import Field
+
+
+class MimesisProvider:
+    def __init__(self) -> None:
+        self.field = Field()
+
+    def generate_row(self, column_types: list[tuple[str, list]]) -> list[str]:
+        return [self.field._lookup_method(ctype)(*args) for ctype, args in column_types]
+
+    def format(self, log_entry) -> list[str]:
+        raise NotImplementedError
+
+
+# field("person.username", mask="U_d", drange=(100, 1000))