Update app.py
Browse files
app.py
CHANGED
|
@@ -79,9 +79,9 @@ async def ocr_endpoint(files: List[UploadFile] = File(...), max_pages: int | Non
|
|
| 79 |
if len(files) > 15:
|
| 80 |
raise HTTPException(status_code=400, detail="Maximum 15 files allowed.")
|
| 81 |
|
| 82 |
-
|
| 83 |
|
| 84 |
-
for file in files:
|
| 85 |
filename = file.filename.lower()
|
| 86 |
ext = filename.split(".")[-1]
|
| 87 |
|
|
@@ -91,17 +91,40 @@ async def ocr_endpoint(files: List[UploadFile] = File(...), max_pages: int | Non
|
|
| 91 |
with open(temp_path, "wb") as f:
|
| 92 |
f.write(await file.read())
|
| 93 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 94 |
# PDF
|
|
|
|
| 95 |
if filename.endswith(".pdf"):
|
| 96 |
img_paths = pdf_to_images(temp_path, max_pages=max_pages)
|
| 97 |
-
for img_path in img_paths:
|
| 98 |
-
final_output.extend(extract_text(img_path))
|
| 99 |
|
| 100 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 101 |
elif filename.endswith((".jpg", ".jpeg", ".png")):
|
| 102 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 103 |
|
| 104 |
else:
|
| 105 |
raise HTTPException(status_code=400, detail=f"Unsupported type: {filename}")
|
| 106 |
|
| 107 |
-
|
|
|
|
|
|
|
|
|
| 79 |
if len(files) > 15:
|
| 80 |
raise HTTPException(status_code=400, detail="Maximum 15 files allowed.")
|
| 81 |
|
| 82 |
+
structured_output = {"files": []}
|
| 83 |
|
| 84 |
+
for index, file in enumerate(files, start=1):
|
| 85 |
filename = file.filename.lower()
|
| 86 |
ext = filename.split(".")[-1]
|
| 87 |
|
|
|
|
| 91 |
with open(temp_path, "wb") as f:
|
| 92 |
f.write(await file.read())
|
| 93 |
|
| 94 |
+
file_record = {
|
| 95 |
+
"file_id": f"file_{index}",
|
| 96 |
+
"filename": filename,
|
| 97 |
+
"pages": []
|
| 98 |
+
}
|
| 99 |
+
|
| 100 |
+
# -------------------------------
|
| 101 |
# PDF
|
| 102 |
+
# -------------------------------
|
| 103 |
if filename.endswith(".pdf"):
|
| 104 |
img_paths = pdf_to_images(temp_path, max_pages=max_pages)
|
|
|
|
|
|
|
| 105 |
|
| 106 |
+
for page_idx, img_path in enumerate(img_paths):
|
| 107 |
+
ocr_results = extract_text(img_path)
|
| 108 |
+
|
| 109 |
+
file_record["pages"].append({
|
| 110 |
+
"page_index": page_idx,
|
| 111 |
+
"results": ocr_results
|
| 112 |
+
})
|
| 113 |
+
|
| 114 |
+
# -------------------------------
|
| 115 |
+
# IMAGE
|
| 116 |
+
# -------------------------------
|
| 117 |
elif filename.endswith((".jpg", ".jpeg", ".png")):
|
| 118 |
+
ocr_results = extract_text(temp_path)
|
| 119 |
+
|
| 120 |
+
file_record["pages"].append({
|
| 121 |
+
"page_index": 0,
|
| 122 |
+
"results": ocr_results
|
| 123 |
+
})
|
| 124 |
|
| 125 |
else:
|
| 126 |
raise HTTPException(status_code=400, detail=f"Unsupported type: {filename}")
|
| 127 |
|
| 128 |
+
structured_output["files"].append(file_record)
|
| 129 |
+
|
| 130 |
+
return JSONResponse(structured_output)
|