tannedcung/sbt-idp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

update index_in_request with doc_type

Browse Source
This commit is contained in:
PhanThanhTrung 2024-03-11 16:35:27 +07:00
parent 056052f1a2
commit b8e37068f6
2 changed files with 33 additions and 35 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
cope2n-api/fwd_api/celery_worker/internal_task.py
View File

@ -194,10 +194,12 @@ def process_pdf(rq_id, sub_id, p_type, user_id, files):
for j in range(len(_b_urls)): for j in range(len(_b_urls)):
_b_urls[j]["doc_type"] = file["file_type"] _b_urls[j]["doc_type"] = file["file_type"]
_b_urls[j]["page_number"] = idx _b_urls[j]["page_number"] = idx
_b_urls[j]["index_to_image_type"] = file["index_in_request"]
return idx, _b_urls[0] return idx, _b_urls[0]
elif extension in image_extensions: elif extension in image_extensions:
this_url = ProcessUtil.process_image_local_file(file["file_name"], file["file_path"], new_request, user, file["file_type"], file["index_in_request"])[0] this_url = ProcessUtil.process_image_local_file(file["file_name"], file["file_path"], new_request, user, file["file_type"], file["index_in_request"])[0]
this_url["page_number"] = idx this_url["page_number"] = idx
this_url["index_to_image_type"] = file["index_in_request"]
if file["file_type"]: if file["file_type"]:
this_url["doc_type"] = file["file_type"] this_url["doc_type"] = file["file_type"]
return idx, this_url return idx, this_url
@ -222,6 +224,7 @@ def process_pdf(rq_id, sub_id, p_type, user_id, files):
file_meta["ai_inference_profile"] = {} file_meta["ai_inference_profile"] = {}
file_meta["index_in_request"] = i file_meta["index_in_request"] = i
file_meta["preprocessing_time"] = preprocessing_time file_meta["preprocessing_time"] = preprocessing_time
file_meta["index_to_image_type"] = b_url["index_to_image_type"]
to_queue.append((fractorized_request_id, sub_id, [b_url], user_id, p_type, file_meta)) to_queue.append((fractorized_request_id, sub_id, [b_url], user_id, p_type, file_meta))
# Send to next queue # Send to next queue

63
cope2n-api/fwd_api/celery_worker/process_result_tasks.py
View File

@ -147,8 +147,11 @@ def process_invoice_sbt_result(rq_id, result, metadata):
rq_id = rq_id.split("_sub_")[0] rq_id = rq_id.split("_sub_")[0]
rq: SubscriptionRequest = SubscriptionRequest.objects.filter(request_id=rq_id).first() rq: SubscriptionRequest = SubscriptionRequest.objects.filter(request_id=rq_id).first()
image_type = metadata["doc_type"]
index_in_request = metadata.pop("index_to_image_type", 0)
result["metadata"] = metadata result["metadata"] = metadata
# status = to_status(result) _update_subscription_rq_file(request_id=rq, index_in_request=index_in_request, doc_type=image_type, result=result)
status = result.get("status", 200) status = result.get("status", 200)
redis_client.set_cache(rq_id, page_index, result) redis_client.set_cache(rq_id, page_index, result)
done = rq.pages == redis_client.get_size(rq_id) done = rq.pages == redis_client.get_size(rq_id)
@ -174,56 +177,48 @@ def process_invoice_sbt_result(rq_id, result, metadata):
rq.status = 404 # stop waiting rq.status = 404 # stop waiting
rq.predict_result = result rq.predict_result = result
rq.save() rq.save()
_update_subscription_rq_file(request_id=rq_id)
update_user(rq) update_user(rq)
except IndexError as e: except IndexError as e:
print(e) print(e)
print("NotFound request by requestId, %d", rq_id) print("NotFound request by requestId, %d", rq_id)
rq.ai_inference_time = 0 rq.ai_inference_time = 0
rq.save() rq.save()
_update_subscription_rq_file(request_id=rq_id)
except Exception as e: except Exception as e:
print(e) print(e)
print("Fail Invoice %d", rq_id) print("Fail Invoice %d", rq_id)
traceback.print_exc() traceback.print_exc()
rq.ai_inference_time = 0 rq.ai_inference_time = 0
rq.save() rq.save()
_update_subscription_rq_file(request_id=rq_id)
return "FailInvoice" return "FailInvoice"
def _update_subscription_rq_file(request_id): def _update_subscription_rq_file(request_id, index_in_request, doc_type, result):
sub_rqs = SubscriptionRequest.objects.filter(request_id=request_id).first() image = SubscriptionRequestFile.objects.filter(request=request_id, index_in_request=index_in_request, doc_type=doc_type).first()
result = sub_rqs.predict_result
if result is None:
return
files = SubscriptionRequestFile.objects.filter(request=sub_rqs) retailer_name = None
for image in files: sold_to_party = None
retailer_name = None purchase_date = []
sold_to_party = None imei_number = []
purchase_date = [] predicted_res = __get_actual_predict_result(result=result)
imei_number = [] if len(predicted_res)!=0:
predicted_res = __get_actual_predict_result(result=result) for elem in predicted_res:
if len(predicted_res)!=0: if elem["label"] == "retailername":
for elem in predicted_res: retailer_name = elem['value']
if elem["label"] == "retailername": elif elem["label"] == "sold_to_party":
retailer_name = elem['value'] sold_to_party = elem['value']
elif elem["label"] == "sold_to_party": elif elem["label"] == "purchase_date":
sold_to_party = elem['value'] purchase_date = elem['value']
elif elem["label"] == "purchase_date": else:
purchase_date=elem['value'] imei_number = elem['value']
else:
imei_number=elem['value']
_predict_result = { _predict_result = {
"retailername": retailer_name, "retailername": retailer_name,
"sold_to_party": sold_to_party, "sold_to_party": sold_to_party,
"purchase_date": purchase_date, "purchase_date": purchase_date,
"imei_number": imei_number "imei_number": imei_number
} }
image.predict_result = _predict_result image.predict_result = _predict_result
image.save() image.save()
def __get_actual_predict_result(result: dict): def __get_actual_predict_result(result: dict):
predicted_res = result.get('content', {}).get('document', []) predicted_res = result.get('content', {}).get('document', [])