Spaces:
Paused
Paused
Update app.py
Browse files
app.py
CHANGED
|
@@ -81,29 +81,39 @@ def ocr_pdf_with_gemini(pdf_path):
|
|
| 81 |
return f"エラーが発生しました: {e}" # ユーザー向けのエラーメッセージ
|
| 82 |
|
| 83 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 84 |
def process_pdf(pdf_file):
|
| 85 |
"""PDFファイルを処理するメイン関数"""
|
|
|
|
|
|
|
| 86 |
# 一時ディレクトリを作成
|
| 87 |
with tempfile.TemporaryDirectory() as temp_dir:
|
| 88 |
-
#
|
| 89 |
-
temp_pdf_path =
|
| 90 |
-
with open(temp_pdf_path, "wb") as f:
|
| 91 |
-
f.write(pdf_file.read()) # bytesではなく、.read()で読み込んだ内容を書き込む
|
| 92 |
|
| 93 |
-
|
|
|
|
|
|
|
|
|
|
| 94 |
split_pdf_paths = split_pdf(temp_pdf_path, temp_dir)
|
|
|
|
| 95 |
|
| 96 |
# 並列処理でOCR変換
|
| 97 |
markdown_results = []
|
| 98 |
with ThreadPoolExecutor() as executor:
|
| 99 |
markdown_results = list(executor.map(ocr_pdf_with_gemini, split_pdf_paths))
|
|
|
|
| 100 |
|
| 101 |
# 結果を結合
|
| 102 |
combined_markdown = "\n\n".join(markdown_results)
|
| 103 |
|
| 104 |
return combined_markdown
|
| 105 |
|
| 106 |
-
|
| 107 |
# Gradioインターフェースの作成
|
| 108 |
def create_interface():
|
| 109 |
with gr.Blocks() as demo:
|
|
|
|
| 81 |
return f"エラーが発生しました: {e}" # ユーザー向けのエラーメッセージ
|
| 82 |
|
| 83 |
|
| 84 |
+
import logging
|
| 85 |
+
|
| 86 |
+
# ロギング設定
|
| 87 |
+
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
|
| 88 |
+
|
| 89 |
+
|
| 90 |
def process_pdf(pdf_file):
|
| 91 |
"""PDFファイルを処理するメイン関数"""
|
| 92 |
+
logging.info(f"Received file: {pdf_file.name if hasattr(pdf_file, 'name') else pdf_file}") # ファイル名/オブジェクトのログ
|
| 93 |
+
|
| 94 |
# 一時ディレクトリを作成
|
| 95 |
with tempfile.TemporaryDirectory() as temp_dir:
|
| 96 |
+
# Gradioから渡されるのはNamedStringオブジェクトなので、.nameでファイルパスを取得
|
| 97 |
+
temp_pdf_path = pdf_file.name
|
|
|
|
|
|
|
| 98 |
|
| 99 |
+
|
| 100 |
+
logging.info(f"Temporary PDF path: {temp_pdf_path}")
|
| 101 |
+
|
| 102 |
+
# PDFを分割 (split_pdf に渡すパスは変更なし)
|
| 103 |
split_pdf_paths = split_pdf(temp_pdf_path, temp_dir)
|
| 104 |
+
logging.info(f"Split PDF paths: {split_pdf_paths}")
|
| 105 |
|
| 106 |
# 並列処理でOCR変換
|
| 107 |
markdown_results = []
|
| 108 |
with ThreadPoolExecutor() as executor:
|
| 109 |
markdown_results = list(executor.map(ocr_pdf_with_gemini, split_pdf_paths))
|
| 110 |
+
logging.info(f"Markdown results length: {len(markdown_results)}")
|
| 111 |
|
| 112 |
# 結果を結合
|
| 113 |
combined_markdown = "\n\n".join(markdown_results)
|
| 114 |
|
| 115 |
return combined_markdown
|
| 116 |
|
|
|
|
| 117 |
# Gradioインターフェースの作成
|
| 118 |
def create_interface():
|
| 119 |
with gr.Blocks() as demo:
|