There seems like there should be a clean way specify engine options for a bind? Based on the answers here, I am guessing not?
I see 2 options but don't really care for either:
Use apply_driver_hacks as suggested in the comments of the above link.
Use SQLAlchemy event hooks to modify the connection, passing the bind engine like db.engines['bind_name'] to the event decorator. Note: you do need to be within the application context for this to work.
I ended up using 2 because I was already comfortable with the method having used it for dynamic tokens but would really like a cleaner solution.
Related
I am trying decide which tool works best for my organization. For that reason I am testing performance of powerbi, looker and tableau against bigQuery etc. Since this is a benchmarking exercise and I am planning to test it for multiple iterations, I want to disable the result set caching property of big Query. In the official documentation, they are letting us disable by passing query config use_query_cache=False
Since I am connecting from front end tools, I am not really sure how to pass this parameter. Can someone help in achieving this? or alternative options if available?
I've not tried this, but I'd think it could be passed like other options.
In this documentation, they have code that looks like this:
Source = GoogleBigQuery.Database(
[BillingProject="Include-Billing-Project-Id-Here", UseStorageApi=false])
I'd expect your parameter to look similar. I.e.
Source = GoogleBigQuery.Database(
[BillingProject="Include-Billing-Project-Id-Here", UseQueryCache=false])
I am locking for a read-write lock in Django, using PostgreSQL.
I know select_for_update
On top, I need SELECT FOR SHARE
I founds this super old Django-ticket
I couldn't find any third party library implementing this for Django
Now I am wondering:
Is there a good reason this was not implemented in the Django-ORM?
Does anyone know any third party library for this?
Or any easy work around?
Can't you wrap the logic in pl/pgsql function that uses select for share and you then call the function from django?
Is there a good reason this was not implemented in the Django-ORM?
The ticket you've posted probably provides the reason: no one is motivated enough to write a patch.
Does anyone know any third party library for this?
Not me, sorry.
And if by any chance you are thinking about ditching Django for some other ORM then you must ask yourself: "There is a feature I need that's missing in Django... what features will I miss in this other ORM?"
Or any easy work around?
Probably not. But here are some options:
Every ORM I know has an escape hatch to raw SQL. You probably know that, but reluctant to use it... But, as you also lack the motivation to make a pull request, that probably means that you do not have hundreds of requests that require SELECT FOR SHARE functionality. So you should consider it. Performing raw SQL queries
Stored procedures as steve mentioned
https://docs.djangoproject.com/en/3.0/topics/db/sql/#calling-stored-procedures
The last comment on the ticked you've posted is from a man (David Schwärzle) who claims that he has a solution (not for PostgreSQL specifically but a solution nevertheless)... maybe you should try and contact him.
Haven't tried it, but probably the way you can add the desired functionality is by Writing your own Query Expressions
You can easily implement with a raw query.
from django.db import connection
query = f"""SELECT * FROM "appname_modelname" WHERE id={obj_id} FOR SHARE"""
with connection.cursor() as cursor:
cursor.execute(query, None)
obj_id is python variable.
appname_modelname is name of table created in your db by django. Django by default combines lowercased app name and model name with underscore.
Since there is no Top_Hits aggregation in NEST, what is the best way to get around this? I tried looking for an AggregationRaw like there is a QueryRaw, but couldn't find it. What I've done for now is to revert to using the ElasticsearchClient and writting a raw request. I'd prefer to write fluent as much as possible, so something like an AggregationRaw would be best, short of implementing it completely in NEST.
Thanks!
Unfortunately, using the low-level ElasticsearchClient and writing the request manually is the only way, until it's added to NEST. You can access the low-level client via the Raw property on the high-level ElasticClient though, which is a bit cleaner than creating a separate instance:
var client = new ElasticClient();
client.Raw.Search("{ \"match_all\" : { } }");
The good news is that the top hits aggregation is planned to be included in the 1.1 release, which is coming very soon. Here's a link the GitHub issue so you can keep an eye on it: #820.
Is there a way to list all keys in a velocity context from the template that is currently being expanded? I've done this in the past by including a reference to the context in the context. I guess this method is okay, but I'm wondering if there's a better way to do it.
That's the traditional way, and there is nothing wrong with it. If you are using VelocityTools, you can use the ContextTool to accomplish that and a bit more.
I'm currently using NHibernate, for the first time, with Fluent NHibernate. I've gotten everything setup nicely, however now I've come to actual doing some data retrieval, it seems to have fallen short.
I was expecting NHibernate, to allow me to do something like:
session.CreateCriteria<TblDocket>()
.Add(Restrictions.Eq(x=> x.DocketNumber, "10101"));
However, this doesn't appear to be the case and I seem to have to write:
session.CreateCriteria<TblDocket>()
.Add(Restrictions.Eq("DocketNumber", "10101"));
That'll be less-than-wonderful when I rename any properties! I've always though hard coded strings in code is bad, especially when the strings relate to property names.
Is there any way I can strongly type these restrictions? I had a look at this blog post, but it seems quite messy, is there a nicer solution?
I decided to use NHibernate.Linq instead. I found a brilliant tutorial here.
You can't using out-of-the-box NHibernate.
There is a project called NHibernate Lambda Extensions which allows you to do this with some limitations.
since NHibernate 3.0 there is also QueryOver available which are a nice typesafe wrapper around the criteria API.
session.QueryOver<TblDocket>()
.Where(x => x.DocketNumber, "10101");
For anyone who comes along this post and doesn't like linq or isn't too familiar with lambda you can still safely use ICrierta's such as
session.CreateCriteria<TblDocket>().Add(Restrictions.Eq("DocketNumber", "10101"));
what you need is helper classes so you can remove the magic strings such as "DocketNumber" so that if you do change your property names or column names these are taken care of for you or will atleast produce a build error so you know before you publish your code.
Anyone wanting to see an example can have a look at NhGen ( http://sourceforge.net/projects/nhgen/ ) and the query examples at https://sourceforge.net/projects/nhgen/forums/forum/1169117/topic/3789112 which show how helpers classes can be used like.
// Find using a simple entity query
IList<IMessage> messageList3 = messageDao.Find( new [] { Restrictions.Le(MessageHelper.Columns.Date, dateLastChecked) } );
Note that this project also created entity wrapper classes which group all your common CRUD methods into one class (xxxDao as show above) so you don't have to keep copying the same code over and over again.