diffinite 0.4.0__tar.gz → 0.6.0__tar.gz

{diffinite-0.4.0/src/diffinite.egg-info → diffinite-0.6.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: diffinite
-Version: 0.4.0
+Version: 0.6.0
 Summary: Forensic source-code comparison tool — Winnowing fingerprints and professional PDF reports for IP litigation & code audit
 Author: nash-dir
 License: Apache-2.0

{diffinite-0.4.0 → diffinite-0.6.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "diffinite"
-version = "0.4.0"
+version = "0.6.0"
 description = "Forensic source-code comparison tool — Winnowing fingerprints and professional PDF reports for IP litigation & code audit"
 readme = "README.md"
 license = {text = "Apache-2.0"}

diffinite-0.6.0/src/diffinite/__main__.py

@@ -0,0 +1,11 @@
+"""Allow ``python -m diffinite``."""
+
+import multiprocessing
+
+from diffinite.cli import main
+
+# Required for PyInstaller frozen executables.
+# Without this, ProcessPoolExecutor child processes crash on Windows.
+# Ref: https://docs.python.org/3/library/multiprocessing.html#multiprocessing.freeze_support
+multiprocessing.freeze_support()
+main()

{diffinite-0.4.0 → diffinite-0.6.0}/src/diffinite/cli.py

@@ -194,7 +194,7 @@ def main(argv: list[str] | None = None) -> None:
         dest="window",
         help=(
             f"Winnowing window size (default: {DEFAULT_W}). "
-            "Density guarantee: ≥(W+K−1) shared tokens always detected."
+            "Density guarantee: >=(W+K-1) shared tokens always detected."
         ),
     )
     deep_group.add_argument(

{diffinite-0.4.0 → diffinite-0.6.0}/src/diffinite/deep_compare.py

@@ -31,6 +31,7 @@ Winnowing 핑거프린트의 **역 인덱스(Inverted Index)** 를 활용하여,
 from __future__ import annotations
 
 import logging
+import multiprocessing
 from collections import defaultdict
 from concurrent.futures import ProcessPoolExecutor
 from pathlib import Path
@@ -176,7 +177,8 @@ def run_deep_compare(
     fp_b: dict[str, set[int]] = {}
 
     all_items = items_a + items_b
-    with ProcessPoolExecutor(max_workers=workers) as pool:
+    ctx = multiprocessing.get_context("spawn")
+    with ProcessPoolExecutor(max_workers=workers, mp_context=ctx) as pool:
         results = list(pool.map(_extract_one, all_items))
 
     for i, (rel, hset, cnt) in enumerate(results):

{diffinite-0.4.0 → diffinite-0.6.0}/src/diffinite/fingerprint.py

@@ -4,9 +4,9 @@ Stanford MOSS 스타일 문서 핑거프린팅 파이프라인을 구현한다.
 전체 흐름: Tokenize → K-gram → Rolling Hash → Winnow → Fingerprint Set
 
 핵심 보장:
-    **밀도 보장 (Density Guarantee)** — 두 문서가 ≥ (W + K − 1) 토큰의
+    **밀도 보장 (Density Guarantee)** -- 두 문서가 >= (W + K - 1) 토큰의
     공통 부분 문자열을 공유하면, 반드시 1개 이상의 공통 핑거프린트가 생성된다.
-    현재 설정(K=5, W=4)에서 ≥ 8 토큰 공유 시 탐지 보장.
+    현재 설정(K=5, W=4)에서 >= 8 토큰 공유 시 탐지 보장.
 
 참조:
     Schleimer, Wilkerson, Aiken. "Winnowing: Local Algorithms for Document

{diffinite-0.4.0 → diffinite-0.6.0}/src/diffinite/models.py

@@ -163,7 +163,7 @@ class AnalysisMetadata:
     """K-gram 크기. 커질수록 정밀도 ↑ / 재현율 ↓."""
 
     w: int
-    """Winnowing 윈도우 크기. 밀도 보장: ≥(W+K-1) 토큰 공유 시 반드시 탐지."""
+    """Winnowing 윈도우 크기. 밀도 보장: >=(W+K-1) 토큰 공유 시 반드시 탐지."""
 
     threshold: float
     """최소 Jaccard 유사도 임계값. 이 미만의 매칭은 결과에서 제외."""

{diffinite-0.4.0 → diffinite-0.6.0/src/diffinite.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: diffinite
-Version: 0.4.0
+Version: 0.6.0
 Summary: Forensic source-code comparison tool — Winnowing fingerprints and professional PDF reports for IP litigation & code audit
 Author: nash-dir
 License: Apache-2.0

diffinite-0.4.0/src/diffinite/__main__.py

@@ -1,5 +0,0 @@
-"""Allow ``python -m diffinite``."""
-
-from diffinite.cli import main
-
-main()