From be9f36dbdc229e387b4a303621660188214f47ed Mon Sep 17 00:00:00 2001
From: jo <ljonas@riseup.net>
Date: Wed, 15 Jun 2022 15:56:52 +0200
Subject: [PATCH] feat(analyzer): move compute_md5 to shared library

---
 .../pipeline/analyze_metadata.py                | 17 +----------------
 .../tests/pipeline/analyze_metadata_test.py     |  6 +-----
 shared/libretime_shared/files.py                | 17 +++++++++++++++++
 shared/tests/files_test.py                      | 10 ++++++++++
 4 files changed, 29 insertions(+), 21 deletions(-)
 create mode 100644 shared/libretime_shared/files.py
 create mode 100644 shared/tests/files_test.py

diff --git a/analyzer/libretime_analyzer/pipeline/analyze_metadata.py b/analyzer/libretime_analyzer/pipeline/analyze_metadata.py
index 648b57ffa..c3df59dca 100644
--- a/analyzer/libretime_analyzer/pipeline/analyze_metadata.py
+++ b/analyzer/libretime_analyzer/pipeline/analyze_metadata.py
@@ -1,9 +1,9 @@
-import hashlib
 from datetime import timedelta
 from pathlib import Path
 from typing import Any, Dict
 
 import mutagen
+from libretime_shared.files import compute_md5
 from loguru import logger
 
 
@@ -108,18 +108,3 @@ def analyze_metadata(filepath_: str, metadata: Dict[str, Any]):
             continue
 
     return metadata
-
-
-def compute_md5(filepath: Path) -> str:
-    """
-    Compute a file md5sum.
-    """
-    with filepath.open("rb") as file:
-        buffer = hashlib.md5()  # nosec
-        while True:
-            blob = file.read(8192)
-            if not blob:
-                break
-            buffer.update(blob)
-
-        return buffer.hexdigest()
diff --git a/analyzer/tests/pipeline/analyze_metadata_test.py b/analyzer/tests/pipeline/analyze_metadata_test.py
index b1ac6c4ca..f0598d672 100644
--- a/analyzer/tests/pipeline/analyze_metadata_test.py
+++ b/analyzer/tests/pipeline/analyze_metadata_test.py
@@ -2,7 +2,7 @@ from pathlib import Path
 
 import pytest
 
-from libretime_analyzer.pipeline.analyze_metadata import analyze_metadata, compute_md5
+from libretime_analyzer.pipeline.analyze_metadata import analyze_metadata
 
 from ..fixtures import FILE_INVALID_DRM, FILE_INVALID_TXT, FILES_TAGGED
 
@@ -48,7 +48,3 @@ def test_analyze_metadata_unparsable_file():
         "hidden": False,
         "md5": "4d5e4b1c8e8febbd31fa9ce7f088beae",
     }
-
-
-def test_compute_md5():
-    assert compute_md5(FILE_INVALID_TXT) == "4d5e4b1c8e8febbd31fa9ce7f088beae"
diff --git a/shared/libretime_shared/files.py b/shared/libretime_shared/files.py
new file mode 100644
index 000000000..51f207977
--- /dev/null
+++ b/shared/libretime_shared/files.py
@@ -0,0 +1,17 @@
+import hashlib
+from pathlib import Path
+
+
+def compute_md5(filepath: Path) -> str:
+    """
+    Compute a file md5sum.
+    """
+    with filepath.open("rb") as file:
+        buffer = hashlib.md5()  # nosec
+        while True:
+            blob = file.read(8192)
+            if not blob:
+                break
+            buffer.update(blob)
+
+        return buffer.hexdigest()
diff --git a/shared/tests/files_test.py b/shared/tests/files_test.py
new file mode 100644
index 000000000..afbc462f6
--- /dev/null
+++ b/shared/tests/files_test.py
@@ -0,0 +1,10 @@
+from pathlib import Path
+
+from libretime_shared.files import compute_md5
+
+
+def test_compute_md5(tmp_path: Path) -> None:
+    tmp_file = tmp_path / "somefile.txt"
+    tmp_file.write_text("some test")
+
+    assert compute_md5(tmp_file) == "f1b75ac7689ff88e1ecc40c84b115785"