Bug: the "Max. numbers of rows deletion in a single call" limitation (via System Limitations) isnt based on the amount of rows you try to delete, but the amount of rows in the table instead

Developer Talk Scripts (Python and JS)
,
1

this means deletion via batch api calls is not possible as soon as a table has more than 9999 rows.

this is the code I used to test this:

from seatable_api import Base, context
base = Base(context.api_token, context.server_url)
base.auth() 

write_limit = 1000
read_limit = 10000
delete_limit = 10000

def get_all_rows():
    query = "select _id from Table1 limit {} offset {}"

    rows = []
    index = 0

    while True:
        query_formatted = query.format(read_limit, index * read_limit)
        query_result = base.query(query_formatted)
        rows.extend(query_result)    
        index += 1
        if len(query_result) < 10000:
            break
    
    return [row["_id"] for row in rows]

def add(amount):
    rows_data = [{"filler": "filled"} for _ in range(amount)]

    for i in range(0, len(rows_data), write_limit): 
        slice = rows_data[i:i+write_limit]
        base.batch_append_rows("Table1", slice)

def remove():
    row_ids = get_all_rows()
    for i in range(0, len(row_ids), delete_limit):
        slice = row_ids[i:i+delete_limit]
        base.batch_delete_rows("Table1", row_ids)
    

 
remove()
#add(9999)
2

We will check the issue in the next week.

1 Like
3

Hi,
I think there might be some problems in the remove function. The second parameters you use in base.batch_delete_rows should be slice instead of row_ids which will take account all rows. It should be like base.batch_delete_rows("Table1", slice)

On the other hand, by defaut for the batch-delete-rows API, there is a limitation of 10000 rows for deletion operation one time. You may get an error message (status code of 413)if you exceeds the limits.

In your case , please print some message print(len(row_ids)) to check the problems.