snapaddy / databank Goto Github PK

That you can run queries in parallel without worrying about async/await:

db = Database()
db.execute_many(query, batch_size=5)

which will process 5 queries in parallel.

And also something like:

db.execute(query, await=False)

if you do not want the process to be blocked.

Implement a "dry run" method to print the SQL query to be executed with escaped parameters etc.

When I want to insert into a table with execute_many() I can use multiple params. But sometimes it is possible that values of a dictionary contain empty strings which can't be inserted. The solution is to change them to None.

insert_queries = QueryCollection.from_file("src/queries/insert.sql")

album_params = [{
        "artist_name": "HVOB",
        "album_name": "HVOB",
        "album_release_date": "",
        "album_type": "Album",
        "album_genre": ""
    },]
db.execute_many(insert_queries["insert_album"], album_params)

My question is: Is there maybe a solution planned to automatically convert empty strings to None or is the expected solution here to convert the empty strings to None by the user?

I would like to execute all queries of my QueryCollection like in the following example:

create_queries = QueryCollection.from_file("src/queries/create.sql")
for _, query in create_queries._queries.items():
    db.execute(query)

Does _queries has to be private? If not, one could access it directly (although it is obviously also now possible to do this, but not recommend because it is private 😉 ) and looping over it.
Alternatively, one could add the following getter function:

def get_queries(self):
    return self._queries

If a .sql file has more than one newline between each statement, one gets the following error:

databank.query.InvalidQueryHeader: '' is not a valid query header

Example

/* @name create_album */
CREATE TABLE IF NOT EXISTS album (
  ...
);



/* @name create_review */
CREATE TABLE IF NOT EXISTS review (
    ...
);

Solution ideas

• Allow more than one newline in .sql files
• Autoformat loaded .sql file before further processing
• Document this rule

Using fetch_all() with a JOIN in the query, resulting in a result containing the same column name twice, shows only one of both resulting columns in output dictionary.

Example:

Two tables contain a column called name.
We join both tables and the result table now has two columns called name.
With fetch_all() we only get one result set for a column called name (dictionary has unique keys).

Possible fixes:

• Throw a warning when the result set has a duplicate column name
• Rename columns in output dictionary when the result set has a duplicate column name
• If possible use table alias for each of the repeated column names as prefix in output dictionary key, separated by underscore