Merge pull request #118 from seanpedrick-case/dev

Corrected base href references for FastAPI implementation

Author: seanpedrick-case

README.md

@@ -11,7 +11,7 @@ short_description: OCR / redact PDF documents and tabular data
 ---
 # Document redaction
 
-version: 1.6.6
+version: 1.6.7
 
 Redact personally identifiable information (PII) from documents (pdf, png, jpg), Word files (docx), or tabular data (xlsx/csv/parquet). Please see the [User Guide](#user-guide) for a full walkthrough of all the features in the app.
 

app.py

@@ -300,7 +300,10 @@ async def lifespan(app: FastAPI):
 
 
 # 3. Initialize the App with the lifespan parameter
-app = FastAPI(lifespan=lifespan)
+# Clean the ROOT_PATH for FastAPI
+# Ensure it starts with / and has no trailing /
+CLEAN_ROOT = f"/{FASTAPI_ROOT_PATH.strip('/')}" if FASTAPI_ROOT_PATH.strip("/") else ""
+app = FastAPI(lifespan=lifespan, root_path=CLEAN_ROOT)
 
 # Added to pass lint check, no effect
 spaces.annotations
@@ -521,16 +524,26 @@ async def lifespan(app: FastAPI):
 }
 """
 
-# Create the gradio interface
-blocks = gr.Blocks(
-    theme=gr.themes.Default(primary_hue="blue"),
-    head=head_html,
-    css=css,
-    analytics_enabled=False,
-    title="Document Redaction App",
-    delete_cache=(43200, 43200),  # Temporary file cache deleted every 12 hours
-    fill_width=True,
-)
+# Create the gradio interface. If running in FastAPI mode, we don't need the head_html and base_href.
+if RUN_FASTAPI:
+    blocks = gr.Blocks(
+        theme=gr.themes.Default(primary_hue="blue"),
+        css=css,
+        analytics_enabled=False,
+        title="Document Redaction App",
+        delete_cache=(43200, 43200),  # Temporary file cache deleted every 12 hours
+        fill_width=True,
+    )
+else:
+    blocks = gr.Blocks(
+        theme=gr.themes.Default(primary_hue="blue"),
+        head=head_html,
+        css=css,
+        analytics_enabled=False,
+        title="Document Redaction App",
+        delete_cache=(43200, 43200),  # Temporary file cache deleted every 12 hours
+        fill_width=True,
+    )
 
 with blocks:
 
@@ -6959,7 +6972,7 @@ def health_check():
                 show_error=True,
                 auth=authenticate_user if COGNITO_AUTH else None,
                 max_file_size=MAX_FILE_SIZE,
-                path=FASTAPI_ROOT_PATH,
+                path="",
                 favicon_path=Path(FAVICON_PATH),
                 mcp_server=RUN_MCP_SERVER,
             )

entrypoint.sh

@@ -25,7 +25,8 @@ else
         exec uvicorn app:app \
             --host $GRADIO_SERVER_NAME \
             --port $GRADIO_SERVER_PORT \
-            --proxy-headers
+            --proxy-headers \
+            --forwarded-allow-ips "*"
     else
         echo "Starting in Gradio mode..."
         exec python app.py

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "doc_redaction"
-version = "1.6.6"
+version = "1.6.7"
 description = "Redact PDF/image-based documents, Word, or CSV/XLSX files using a Gradio-based GUI interface"
 readme = "README.md"
 authors = [

tools/file_conversion.py

@@ -323,7 +323,6 @@ def convert_pdf_to_images(
     heights = [result[3] for result in results]
 
     print("PDF has been converted to images.")
-    print("images:", images)
     return images, widths, heights, results