AI/ragflow
1
0
Fork 0
mirror of https://git.mirrors.martin98.com/https://github.com/infiniflow/ragflow.git synced 2025-04-23 06:30:00 +08:00
Code Issues Packages Projects Releases Wiki Activity

Storage: Rename all the variables about get file to storage from minio. (#2497)

Browse Source 
https://github.com/infiniflow/ragflow/issues/2496

### What problem does this PR solve?

_Briefly describe what this PR aims to solve. Include background context
that will help reviewers understand the purpose of the PR._

### Type of change

- [ ] Bug Fix (non-breaking change which fixes an issue)
- [ ] New Feature (non-breaking change which adds functionality)
- [ ] Documentation Update
- [x] Refactoring
- [ ] Performance Improvement
- [ ] Other (please describe):
This commit is contained in:
Fachuan Bai 2024-09-19 19:19:27 +08:00 committed by GitHub
parent b5d1d2fec4
commit 422c229e52
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
7 changed files with 16 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
api/apps/api_app.py
View File

@ -478,7 +478,7 @@ def upload():
e, doc = DocumentService.get_by_id(doc["id"]) e, doc = DocumentService.get_by_id(doc["id"])
doc = doc.to_dict() doc = doc.to_dict()
doc["tenant_id"] = tenant_id doc["tenant_id"] = tenant_id
bucket, name = File2DocumentService.get_minio_address(doc_id=doc["id"]) bucket, name = File2DocumentService.get_storage_address(doc_id=doc["id"])
queue_tasks(doc, bucket, name) queue_tasks(doc, bucket, name)
except Exception as e: except Exception as e:
return server_error_response(e) return server_error_response(e)
@ -640,7 +640,7 @@ def document_rm():
if not tenant_id: if not tenant_id:
return get_data_error_result(retmsg="Tenant not found!") return get_data_error_result(retmsg="Tenant not found!")
b, n = File2DocumentService.get_minio_address(doc_id=doc_id) b, n = File2DocumentService.get_storage_address(doc_id=doc_id)
if not DocumentService.remove_document(doc, tenant_id): if not DocumentService.remove_document(doc, tenant_id):
return get_data_error_result( return get_data_error_result(

6
api/apps/dataset_api.py
View File

@ -420,7 +420,7 @@ def delete_document(document_id, dataset_id): # string
f" reason!", code=RetCode.AUTHENTICATION_ERROR) f" reason!", code=RetCode.AUTHENTICATION_ERROR)
# get the doc's id and location # get the doc's id and location
real_dataset_id, location = File2DocumentService.get_minio_address(doc_id=document_id) real_dataset_id, location = File2DocumentService.get_storage_address(doc_id=document_id)
if real_dataset_id != dataset_id: if real_dataset_id != dataset_id:
return construct_json_result(message=f"The document {document_id} is not in the dataset: {dataset_id}, " return construct_json_result(message=f"The document {document_id} is not in the dataset: {dataset_id}, "
@ -595,7 +595,7 @@ def download_document(dataset_id, document_id):
code=RetCode.ARGUMENT_ERROR) code=RetCode.ARGUMENT_ERROR)
# The process of downloading # The process of downloading
doc_id, doc_location = File2DocumentService.get_minio_address(doc_id=document_id) # minio address doc_id, doc_location = File2DocumentService.get_storage_address(doc_id=document_id) # minio address
file_stream = STORAGE_IMPL.get(doc_id, doc_location) file_stream = STORAGE_IMPL.get(doc_id, doc_location)
if not file_stream: if not file_stream:
return construct_json_result(message="This file is empty.", code=RetCode.DATA_ERROR) return construct_json_result(message="This file is empty.", code=RetCode.DATA_ERROR)
@ -736,7 +736,7 @@ def parsing_document_internal(id):
doc_attributes = doc_attributes.to_dict() doc_attributes = doc_attributes.to_dict()
doc_id = doc_attributes["id"] doc_id = doc_attributes["id"]
bucket, doc_name = File2DocumentService.get_minio_address(doc_id=doc_id) bucket, doc_name = File2DocumentService.get_storage_address(doc_id=doc_id)
binary = STORAGE_IMPL.get(bucket, doc_name) binary = STORAGE_IMPL.get(bucket, doc_name)
parser_name = doc_attributes["parser_id"] parser_name = doc_attributes["parser_id"]
if binary: if binary:

6
api/apps/document_app.py
View File

@ -297,7 +297,7 @@ def rm():
if not tenant_id: if not tenant_id:
return get_data_error_result(retmsg="Tenant not found!") return get_data_error_result(retmsg="Tenant not found!")
b, n = File2DocumentService.get_minio_address(doc_id=doc_id) b, n = File2DocumentService.get_storage_address(doc_id=doc_id)
if not DocumentService.remove_document(doc, tenant_id): if not DocumentService.remove_document(doc, tenant_id):
return get_data_error_result( return get_data_error_result(
@ -342,7 +342,7 @@ def run():
e, doc = DocumentService.get_by_id(id) e, doc = DocumentService.get_by_id(id)
doc = doc.to_dict() doc = doc.to_dict()
doc["tenant_id"] = tenant_id doc["tenant_id"] = tenant_id
bucket, name = File2DocumentService.get_minio_address(doc_id=doc["id"]) bucket, name = File2DocumentService.get_storage_address(doc_id=doc["id"])
queue_tasks(doc, bucket, name) queue_tasks(doc, bucket, name)
return get_json_result(data=True) return get_json_result(data=True)
@ -393,7 +393,7 @@ def get(doc_id):
if not e: if not e:
return get_data_error_result(retmsg="Document not found!") return get_data_error_result(retmsg="Document not found!")
b, n = File2DocumentService.get_minio_address(doc_id=doc_id) b, n = File2DocumentService.get_storage_address(doc_id=doc_id)
response = flask.make_response(STORAGE_IMPL.get(b, n)) response = flask.make_response(STORAGE_IMPL.get(b, n))
ext = re.search(r"\.([^.]+)$", doc.name) ext = re.search(r"\.([^.]+)$", doc.name)

2
api/apps/file_app.py
View File

@ -332,7 +332,7 @@ def get(file_id):
e, file = FileService.get_by_id(file_id) e, file = FileService.get_by_id(file_id)
if not e: if not e:
return get_data_error_result(retmsg="Document not found!") return get_data_error_result(retmsg="Document not found!")
b, n = File2DocumentService.get_minio_address(file_id=file_id) b, n = File2DocumentService.get_storage_address(file_id=file_id)
response = flask.make_response(STORAGE_IMPL.get(b, n)) response = flask.make_response(STORAGE_IMPL.get(b, n))
ext = re.search(r"\.([^.]+)$", file.name) ext = re.search(r"\.([^.]+)$", file.name)
if ext: if ext:

6
api/apps/sdk/doc.py
View File

@ -286,7 +286,7 @@ def download_document(document_id,tenant_id):
code=RetCode.ARGUMENT_ERROR) code=RetCode.ARGUMENT_ERROR)
# The process of downloading # The process of downloading
doc_id, doc_location = File2DocumentService.get_minio_address(doc_id=document_id) # minio address doc_id, doc_location = File2DocumentService.get_storage_address(doc_id=document_id) # minio address
file_stream = STORAGE_IMPL.get(doc_id, doc_location) file_stream = STORAGE_IMPL.get(doc_id, doc_location)
if not file_stream: if not file_stream:
return construct_json_result(message="This file is empty.", code=RetCode.DATA_ERROR) return construct_json_result(message="This file is empty.", code=RetCode.DATA_ERROR)
@ -373,7 +373,7 @@ def rm(tenant_id):
if not tenant_id: if not tenant_id:
return get_data_error_result(retmsg="Tenant not found!") return get_data_error_result(retmsg="Tenant not found!")
b, n = File2DocumentService.get_minio_address(doc_id=doc_id) b, n = File2DocumentService.get_storage_address(doc_id=doc_id)
if not DocumentService.remove_document(doc, tenant_id): if not DocumentService.remove_document(doc, tenant_id):
return get_data_error_result( return get_data_error_result(
@ -438,7 +438,7 @@ def run(tenant_id):
e, doc = DocumentService.get_by_id(id) e, doc = DocumentService.get_by_id(id)
doc = doc.to_dict() doc = doc.to_dict()
doc["tenant_id"] = tenant_id doc["tenant_id"] = tenant_id
bucket, name = File2DocumentService.get_minio_address(doc_id=doc["id"]) bucket, name = File2DocumentService.get_storage_address(doc_id=doc["id"])
queue_tasks(doc, bucket, name) queue_tasks(doc, bucket, name)
return get_json_result(data=True) return get_json_result(data=True)

2
api/db/services/file2document_service.py
View File

@ -69,7 +69,7 @@ class File2DocumentService(CommonService):
@classmethod @classmethod
@DB.connection_context() @DB.connection_context()
def get_minio_address(cls, doc_id=None, file_id=None): def get_storage_address(cls, doc_id=None, file_id=None):
if doc_id: if doc_id:
f2d = cls.get_by_document_id(doc_id) f2d = cls.get_by_document_id(doc_id)
else: else:

6
rag/svr/task_executor.py
View File

@ -137,7 +137,7 @@ def collect():
return tasks return tasks
def get_minio_binary(bucket, name): def get_storage_binary(bucket, name):
return STORAGE_IMPL.get(bucket, name) return STORAGE_IMPL.get(bucket, name)
@ -155,8 +155,8 @@ def build(row):
chunker = FACTORY[row["parser_id"].lower()] chunker = FACTORY[row["parser_id"].lower()]
try: try:
st = timer() st = timer()
bucket, name = File2DocumentService.get_minio_address(doc_id=row["doc_id"]) bucket, name = File2DocumentService.get_storage_address(doc_id=row["doc_id"])
binary = get_minio_binary(bucket, name) binary = get_storage_binary(bucket, name)
cron_logger.info( cron_logger.info(
"From minio({}) {}/{}".format(timer() - st, row["location"], row["name"])) "From minio({}) {}/{}".format(timer() - st, row["location"], row["name"]))
except TimeoutError as e: except TimeoutError as e: