format code and add pre commit

.pre-commit-config.yaml

@@ -0,0 +1,15 @@
+# See https://pre-commit.com for more information
+# See https://pre-commit.com/hooks.html for more hooks
+repos:
+-   repo: local
+    hooks:
+    - id: black
+      name: black
+      entry: black --check --diff --color
+      language: python
+      files: ".py"
+    - id: flake8
+      name: flake8
+      entry: flake8
+      language: python
+      files: ".py"

pdf2zh/cache.py

@@ -3,9 +3,10 @@
 import time
 import hashlib
 import shutil
-cache_dir = os.path.join(tempfile.gettempdir(), 'cache')
+
+cache_dir = os.path.join(tempfile.gettempdir(), "cache")
 os.makedirs(cache_dir, exist_ok=True)
-time_filename = 'update_time'
+time_filename = "update_time"
 max_cache = 5
 
 
@@ -16,25 +17,30 @@ def deterministic_hash(obj):
 
 
 def get_dirs():
-    dirs = [os.path.join(cache_dir, dir) for dir in os.listdir(cache_dir) if os.path.isdir(os.path.join(cache_dir, dir))]
+    dirs = [
+        os.path.join(cache_dir, dir)
+        for dir in os.listdir(cache_dir)
+        if os.path.isdir(os.path.join(cache_dir, dir))
+    ]
     return dirs
 
 
 def get_time(dir):
     try:
         timefile = os.path.join(dir, time_filename)
-        t = float(open(timefile, encoding='utf-8').read())
+        t = float(open(timefile, encoding="utf-8").read())
         return t
     except FileNotFoundError:
         # handle the error as needed, for now we'll just return a default value
-        return float('inf')  # This ensures that this directory will be the first to be removed if required
-
+        return float(
+            "inf"
+        )  # This ensures that this directory will be the first to be removed if required
 
 
 def write_time(dir):
     timefile = os.path.join(dir, time_filename)
     t = time.time()
-    print(t, file=open(timefile, "w", encoding='utf-8'), end='')
+    print(t, file=open(timefile, "w", encoding="utf-8"), end="")
 
 
 def argmin(iterable):
@@ -44,7 +50,9 @@ def argmin(iterable):
 def remove_extra():
     dirs = get_dirs()
     for dir in dirs:
-        if not os.path.isdir(dir):  # This line might be redundant now, as get_dirs() ensures only directories are returned
+        if not os.path.isdir(
+            dir
+        ):  # This line might be redundant now, as get_dirs() ensures only directories are returned
             os.remove(dir)
         try:
             get_time(dir)
@@ -73,11 +81,11 @@ def create_cache(hash_key):
 def load_paragraph(hash_key, hash_key_paragraph):
     filename = os.path.join(cache_dir, hash_key, hash_key_paragraph)
     if os.path.exists(filename):
-        return open(filename, encoding='utf-8').read()
+        return open(filename, encoding="utf-8").read()
     else:
         return None
 
 
 def write_paragraph(hash_key, hash_key_paragraph, paragraph):
     filename = os.path.join(cache_dir, hash_key, hash_key_paragraph)
-    print(paragraph, file=open(filename, "w", encoding='utf-8'), end='')
+    print(paragraph, file=open(filename, "w", encoding="utf-8"), end="")