aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rw-r--r--wqflask/wqflask/correlation/correlation_file_generator.py25
1 files changed, 9 insertions, 16 deletions
diff --git a/wqflask/wqflask/correlation/correlation_file_generator.py b/wqflask/wqflask/correlation/correlation_file_generator.py
index 4d8a2f23..f207f2c6 100644
--- a/wqflask/wqflask/correlation/correlation_file_generator.py
+++ b/wqflask/wqflask/correlation/correlation_file_generator.py
@@ -146,8 +146,6 @@ def generate_csv_file(conn,db_name,txt_dir,file_name):
raise e
-
-
def array_to_bytes(x:np.ndarray) -> bytes:
np_bytes = BytesIO()
@@ -162,37 +160,32 @@ def bytes_to_array(b: bytes) -> np.ndarray:
return np.load(np_bytes, allow_pickle=True)
-def lmdb_file_generator(file_name,data:np.ndarray):
-
- # create unique filename
-
- # pass
-
-
-
def lmdb_error_handler(func):
@functools.wraps(func)
def inner_function(*args, **kwargs):
try:
return func(*args, **kwargs)
except lmdb.Error as error:
- print(f"{func.__name__} >>>> error . {str(error)}")
+ print(f"{func.__name__} >>>> error . {error}")
return None
return inner_function
-
-def create_dataset(file_name, db_name, dataset: np.ndarray):
+@lmdb_error_handler
+def create_dataset(file_path, db_name,cols ,dataset: np.ndarray):
# map size int(ie12)
with lmdb.open(file_name, map_size=dataset.nbytes*10) as env:
with env.begin(write=True) as txn:
+ txn.put(f"db_name:col".encode(), array_to_bytes(cols))
txn.put(db_name.encode(), array_to_bytes(dataset))
return (file_name, db_name)
-
@lmdb_error_handler
-def read_dataset(file_name, db_name):
+def read_dataset(file_path, db_name):
with lmdb.open(file_name, readonly=True, create=False) as env:
with env.begin() as txn:
results = txn.get(db_name.encode())
- return bytes_to_array(results) if results else None \ No newline at end of file
+ cols = txn.get(f"{db_name}:col".encode())
+ if (cols and results):
+ return (bytes_to_array(cols),bytes_to_array(results))
+