Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -101,6 +101,7 @@ def extract_invoice_data(file_data, content_type, json_schema):
|
|
| 101 |
"""
|
| 102 |
system_prompt = "You are an expert in document data extraction."
|
| 103 |
base64_images = []
|
|
|
|
| 104 |
|
| 105 |
if content_type == "application/pdf":
|
| 106 |
try:
|
|
@@ -115,7 +116,9 @@ def extract_invoice_data(file_data, content_type, json_schema):
|
|
| 115 |
img.save(img_byte_arr, format="PNG", dpi=(300, 300))
|
| 116 |
base64_encoded = base64.b64encode(img_byte_arr.getvalue()).decode('utf-8')
|
| 117 |
base64_images.append(f"data:image/png;base64,{base64_encoded}")
|
| 118 |
-
|
|
|
|
|
|
|
| 119 |
|
| 120 |
except Exception as e:
|
| 121 |
logger.error(f"Error converting PDF to image: {e}")
|
|
@@ -125,7 +128,7 @@ def extract_invoice_data(file_data, content_type, json_schema):
|
|
| 125 |
# Handle direct image files
|
| 126 |
base64_encoded = base64.b64encode(file_data).decode('utf-8')
|
| 127 |
base64_images.append(f"data:{content_type};base64,{base64_encoded}")
|
| 128 |
-
base64DataResp =
|
| 129 |
|
| 130 |
# Prepare OpenAI request
|
| 131 |
openai_content = [{"type": "image_url", "image_url": {"url": img_base64}} for img_base64 in base64_images]
|
|
|
|
| 101 |
"""
|
| 102 |
system_prompt = "You are an expert in document data extraction."
|
| 103 |
base64_images = []
|
| 104 |
+
base64DataResp = ""
|
| 105 |
|
| 106 |
if content_type == "application/pdf":
|
| 107 |
try:
|
|
|
|
| 116 |
img.save(img_byte_arr, format="PNG", dpi=(300, 300))
|
| 117 |
base64_encoded = base64.b64encode(img_byte_arr.getvalue()).decode('utf-8')
|
| 118 |
base64_images.append(f"data:image/png;base64,{base64_encoded}")
|
| 119 |
+
|
| 120 |
+
# Store all images as a single JSON object
|
| 121 |
+
base64DataResp = json.dumps(base64_images)
|
| 122 |
|
| 123 |
except Exception as e:
|
| 124 |
logger.error(f"Error converting PDF to image: {e}")
|
|
|
|
| 128 |
# Handle direct image files
|
| 129 |
base64_encoded = base64.b64encode(file_data).decode('utf-8')
|
| 130 |
base64_images.append(f"data:{content_type};base64,{base64_encoded}")
|
| 131 |
+
base64DataResp = json.dumps(base64_images) # Store as a JSON object
|
| 132 |
|
| 133 |
# Prepare OpenAI request
|
| 134 |
openai_content = [{"type": "image_url", "image_url": {"url": img_base64}} for img_base64 in base64_images]
|