jacob-mathison/doczyai-pipelines
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Activity

Changed SQS message parsing

Browse Source
This commit is contained in:
Umang Mistry
2024-06-21 12:06:51 -05:00
parent 1f2f6e548e
commit c8ef3f36c7
1 changed files with 5 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
textract-pipeline/src/lambda/prompt-orchestrator/index.py
+5 -3
View File
@@ -115,8 +115,10 @@ def save_to_sf(dag_name, **kwargs):
def lambda_handler(event, context):
try:
logger.info("Processing SQS event.")
records = json.loads(event['Records'][0]['body'])
logger.info(f"PROCESSING RECORDS:: {records}")
# Process each message from the SQS event
for record in event['Records']:
for record in records['Records']:
# Extract the message body from the record
sqs_record = record['s3']
bucket_name = sqs_record['bucket']['name']
@@ -127,7 +129,7 @@ def lambda_handler(event, context):
logger.info(f"Processing document: {document_id}")
# Fetch batch_id from object tags
batch_id = get_s3_object_tags(bucket_name, object_key).get('batch_id')
batch_id = get_s3_object_tags(bucket_name, object_key).get('BatchId')
logger.info(f"Batch ID: {batch_id}")
# Query field groups from the database
@@ -387,7 +389,7 @@ def main(bucket, object_key, field_groups):
context = contents.decode("utf-8")
# Fetch batch_id from object tags
batch_id = get_s3_object_tags(bucket, object_key).get('batch_id')
batch_id = get_s3_object_tags(bucket, object_key).get('BatchId')
logger.info(f"Batch ID: {batch_id}")
df = pd.DataFrame(columns=['Contract Name', 'Field Name', 'Raw value', 'New Extracted value', 'Confidence Level', 'Snippet', 'New Page Number', 'Revised Prompt'])