Merge pull request #7 from FlacSy/develop

2.3.1 — PyPI sdist fix, ML example

Author: FlacSy

.github/workflows/release.yml

@@ -3,6 +3,7 @@ name: Release
 on:
   release:
     types: [published]
+  workflow_dispatch:
 
 permissions:
   contents: read

Cargo.lock

@@ -1,6 +1,6 @@
 # Release Checklist
 
-## Versions (2.2.0)
+## Versions (2.3.1)
 
 | Package | File | Publication |
 |---------|------|-------------|

RELEASE_CHECKLIST.md

@@ -0,0 +1,47 @@
+#!/usr/bin/env python3
+"""ML-based toxicity detection example.
+
+Requires: pip install 'badwords-py[ml]'
+
+The ML model (XLM-RoBERTa) detects toxicity in multiple languages.
+First run downloads the model from GitHub Releases (~135MB).
+"""
+
+import sys
+
+try:
+    from badwords.ml import ToxicityPredictor
+except ImportError:
+    print("Error: badwords-py[ml] required. Install with:", file=sys.stderr)
+    print("  pip install 'badwords-py[ml]'", file=sys.stderr)
+    sys.exit(1)
+
+
+def main() -> None:
+    # Создаём предиктор (модель скачается при первом вызове predict)
+    predictor = ToxicityPredictor()
+
+    texts = [
+        "Hello, how are you today?",
+        "Have a nice day!",
+        "You are stupid and worthless",
+        "Поздравляю, теперь ты не тупой",
+        "Иди нахуй",
+    ]
+
+    print("=" * 60)
+    print("ML Toxicity Detection (0.0 = clean, 1.0 = toxic)")
+    print("Threshold: 0.5")
+    print("=" * 60)
+
+    for text in texts:
+        prob = predictor.predict(text)
+        label = "TOXIC" if prob >= 0.5 else "clean"
+        bar = "█" * int(prob * 20) + "░" * (20 - int(prob * 20))
+        print(f"  {prob:.2f} [{label:5}] {bar}  {text!r}")
+
+    print("=" * 60)
+
+
+if __name__ == "__main__":
+    main()

examples/python/ml_moderation.py

@@ -4,7 +4,7 @@ build-backend = "maturin"
 
 [project]
 name = "badwords-py"
-version = "2.2.0"
+version = "2.3.1"
 requires-python = ">=3.10"
 authors = [
   {name = "iamlostshe", email = "vanamelcikov7275@gmail.com"},

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "maturin"
 
 [project]
 name = "badwords-py"
-version = "2.2.0"
+version = "2.3.1"
 requires-python = ">=3.10"
 authors = [
   {name = "iamlostshe", email = "vanamelcikov7275@gmail.com"},

python/pyproject.toml

@@ -1,6 +1,6 @@
 [package]
 name = "badwords-py"
-version = "2.2.0"
+version = "2.3.1"
 edition = "2021"
 description = "Python bindings for badwords profanity filter"
 license = "MIT"