Background process taking too long.

[dega5]

Question

The issue is once the user uploads the file and clicks the submit button nothing happens for 10-12 seconds. Then suddenly they are redirected to the task list page. This isn't the best user experience as they don't even know if the button has been clicked or the background process is actually happening.

Instead I am trying to redirect the user to the 'task-list' page first and then run the background process, but it still runs validate input function before redirecting user to the other page (code below).

     ui.upload(
                    label='Upload Data Model',
                    auto_upload=True,
                    multiple=False,
                    max_files=1,
                    max_file_size=10_000_000,
                    on_rejected=lambda: ui.notify('File must a zip, tar or gzip file smaller than 10Mb !'),
                    on_upload=on_update_button_click
                ).props('accept=".gz,.zip,.tar"') 

Then I am running this code

              def on_update_button_click():
                    ui.open('/task-list')
                   validate_aci_input()


                def validate_aci_input():
                    validate_input(region.value, name.value, proj.value)
                    

[rodja]

I'm not entirely sure about what's the problem and what it is you want to archive. But maybe https://github.com/zauberzeug/nicegui/wiki/FAQs#why-is-my-long-running-function-blocking-ui-updates helps?

[rodja]

You need to make sure nothing "demanding" is executed on the main thread. What are you doing in validate_aci_input? Try using await run.cpu_bound(validate_aci_input).

[dega5]

The function calls a very demanding background process so it makes sense for it to take 10/12 seconds.
I get an error mesage - 'nicegui.run has no attribute cpi_bound'

[rodja]

Please read https://github.com/zauberzeug/nicegui/wiki/FAQs#why-is-my-long-running-function-blocking-ui-updates carefully and follow the links.

[dega5]

I have tried to follow both these links, and tried the different options but I am still struggling. I am new to programming thats why I dont really understand the deep technical stuff.

https://fastapi.tiangolo.com/async/#in-a-hurry
https://nicegui.io/documentation#running_cpu-bound_tasks

[natankeddem]

Does your processing routine consist of iterating over a dataset? If so you could yield periodically during this iteration inside an async function.
https://stackoverflow.com/questions/36647825/cooperative-yield-in-asyncio

[natankeddem]

How about using ui.spinner?

from nicegui import ui, events
import asyncio

async def on_update_button_click(e: events.UploadEventArguments):
    e.sender.visible = False
    await validate_aci_input()
    ui.open("/task-list")

async def validate_aci_input():
    await asyncio.sleep(5)

@ui.page("/")
def root():
    upload = ui.upload(
        label="Upload Data Model",
        auto_upload=True,
        multiple=False,
        max_files=1,
        max_file_size=10_000_000,
        on_rejected=lambda: ui.notify("File must a zip, tar or gzip file smaller than 10Mb !"),
        on_upload=lambda e: on_update_button_click(e),
    ).props('accept=".gz,.zip,.tar"')
    ui.spinner(size="xl").bind_visibility_from(upload, "visible", lambda x: not x)

@ui.page("/task-list")
def task_list():
    ui.label("task-list")

ui.run()

[dega5]

This doesn't work unfortunately - I don't see the spinner at all.
Its the same as before, once I click the submit button nothing happens for 10/12 seconds and then suddenly I am redirected to the task list page.

[natankeddem]

Are you running exactly what I posted above? This is how it looks on my computer for reference:
google_screen_recording_2023-11-02T17-04_03.698Z.webm

[dega5]

I am not running it exactly like that because my background process is different and won't be able to share too much because of data security.

[natankeddem]

My point is this, if you run my code unaltered you should see a spinner. If you don't there might be something wrong with your environment, since it runs on my computer correctly (see post above for video evidence). Now you have established a working baseline. After that we can talk about alterations of the code to accommodate your processing routine.

[rodja]

This might be solvable with #2410.