kika
Explore posts from serversCDCloudflare Developers
•Created by kika on 7/20/2023 in #pages-help
Access to git tag
git describe --candidates 0 doesn't return a valid tag even if the HEAD of the corresponding branch is correctly tagged with an annotated tag. How can I get that tag?
Or, what exact command Cloudflare build system uses to clone the repository?
I'm using Gitlab if that matters.
Thanks1 replies
How to inject parentheses in the generated arithmetic expression?
https://kyse.link/?p=s&i=SD8v0Fo6NqqCrPEXMbx1
That's obviously not what I meant
8 replies
.withSchema() and raw SQL
when doing
sql <ReturnType> SELECT some_function(param);.execute(db) the function is not found because it doesn't exist in the public schema, although db was created with .withSchema('myschema').
If I change the call to SELECT myschema.some_function(param);.execute(db) it still doesn't work, because the function references tables in myschema8 replies
RRefine
•Created by exotic-emerald on 6/13/2023 in #ask-any-question
What is the idiomatic way to use `useList()` with REST API that returns a JSON array of strings?
Strings are not
BaseRecords. Should I hack dataprovider or is there a better way?4 replies
CDCloudflare Developers
•Created by kika on 6/2/2023 in #workers-help
Initialize worker
Hi, is there an idiomatic way to initialize worker outside of the
fetch()? I want to try the new connect() to connect to the database, but I don't want to do that repeatedly for every request that fetch() may handle. Is there something better than global variable and a mutex?2 replies
RRefine
•Created by wise-white on 4/21/2023 in #ask-any-question
State management
What is the canonical way to manage app state in Refine? I went through tutorial and it appears it doesn't have an app state, and components talk to backend through the data provider. I don't want to push the state to the backend, my backend is read-only for this app and I want to keep the state locally. In classical React that would be Redux or something similar.
16 replies
CDCloudflare Developers
•Created by kika on 3/13/2023 in #workers-help
API operations/invocation
It seems I'm hitting a 1000 ops per invocation limit in my scheduled worker while writing to the KV. But I still need to write more than 1000 k/v pairs and bulk writes are not supported according to the docs. What is the recommended way of dealing with this? If instead of writing to KV I'll send the data over the queue and batch into 100 messages, would every 100 messages considered a separate invocation?
6 replies