[WARNING] Could not find any valid room code about granblue-automation-pyautogui HOT 19 CLOSED

Updated the bot to utilize the Stream API as a fallback in case the bot cannot find any room codes like Lvl 120 Avatar. Let me know how it goes!

from granblue-automation-pyautogui.

When that message pops up, what Raids are you trying to find? Certain Raids are either very empty at all times of the day or only empty at certain points of the day.

from granblue-automation-pyautogui.

Because I can find the ID by some raid finder website (for example, https://gbf.life/), at the same time, I can not find any ID by the twitter-raid-finder function for the same raid. That's why I am thinking of if there is any limitation on the Twitter developer APIs.

from granblue-automation-pyautogui.

Currently, my room finder searches a total of 5 tweets from both JP and EN. It searches JP first and if it acquires 5 codes from them, then it will proceed without checking for EN tweets. If not, it will get the remaining from EN. In addition, it will skip all tweets whose room codes it already visited or that they were invalid.

I have not had an issue that I can remember with not being able to find room codes when other similar room code finders can.

from granblue-automation-pyautogui.

Thanks, I will compare the room list in the built-in room finder with the website one.

from granblue-automation-pyautogui.

Closing issue due to inactivity.

from granblue-automation-pyautogui.

Hi Steve,
I need to re-open this issue due to the issue is confirmed by the following testing:

1. I tried to use the built-in raid finder function to get the backup code on 23:55 (GMT+8), and the codes cannot be found immediately at 00:00.
2. At the same time, I can get lots of backup code from the website: https://gbf.life/?locale=ja.

Do you think the issue is related to my Twitter developer settings?

from granblue-automation-pyautogui.

What raids are you trying to find at 00:00 GMT? Does this happen at any other time?

from granblue-automation-pyautogui.

What raids are you trying to find at 00:00 GMT? Does this happen at any other time?

Currently, I only got the "end" time point of my built-in raid finder. I've not tried the start point yet.

I always use the raid function from 18:00 to 00:00. It works perfectly. I will try it after 2 hours. It's 5:50 am (GMT +8) now.

• Log file at 06:38 am (GMT +8)

0:00:19 [TWITTER] Authenticating provided consumer keys and tokens with the Twitter API...
0:00:20 [TWITTER] Successfully connected to the Twitter API.

0:00:20 [TWITTER] Now finding the 5 most recent tweets for Lvl 100 Twin Elements.

0:00:21 [TWITTER] Now cleaning up the tweets and parsing for room codes...


0:00:21 [WARNING] Could not find any valid room codes. 
Waiting 60 seconds and then trying again with 9 tries left before exiting.

from granblue-automation-pyautogui.

What raids are you trying to find at 00:00 GMT? Does this happen at any other time?

Confirmed here.

At GMT+8 7:00 (GMT 23:00), the bot didn't work either.
At GMT+8 8:45 (GMT 00:45), it works well.

Does Twitter API have a limitation on execution time?

from granblue-automation-pyautogui.

Coming back to this, I seem to not be able to find any Lvl 120 Avatar/Lv120 アバター raids at all times of the day. Even searching the query manually on Twitter seems to not bring up any results, including the Japanese query. The strangest thing is that turning off Hide sensitive content in the Search settings of Twitter makes all Avatar raids show up as normal in the Twitter search results.

Tweepy does not seem to have a way to bypass the Hide sensitive content programmatically unfortunately.

from granblue-automation-pyautogui.

can confirm, i too cant find any valid room codes. maybe it is related to the sensitive content stuffs because the gbf raid stuffs is flagged as sensitive content in twitter for no reason

from granblue-automation-pyautogui.

Poking my head in, oddly enough, searching "Lvl 120 Avata" or the like, where people put text into their backup requests, catches some results. It's as if the word "Avatar" is triggering some flag.

Also I don't know how to really use Tweepy, but would the "possibly_sensitive" parameter help at all? Or possibly using Twint as an alternative, or asking on their Github for a method to scrape content flagged as sensitive?

from granblue-automation-pyautogui.

I will take a look around for possible solutions to this.

from granblue-automation-pyautogui.

In this case gbf.life and other raid finders are probably not using the search api to constantly look up tweets, but rather get data feed from twitter's streaming apis.

Would have to probably implement the following, which can specify the filter_level parameter in the filter function for the stream.

https://docs.tweepy.org/en/latest/stream.html

from granblue-automation-pyautogui.

Thank you @lamp333 for pointing me in the right direction! Stream seems to be the way to go. Expect an update for Avatar raid farming tomorrow.

from granblue-automation-pyautogui.

Seems to be working well! However, would it be possible to make the Stream API the default mode? The bot using it as a fallback slows things down, and the Stream API is probably able to get codes from people who have accounts marked as sensitive.

from granblue-automation-pyautogui.

Thanks for the feedback. Implemented it just now. In addition, reduced number of tweets to be found from 5 to 3 to speed things up and the recovery period from finding no tweets from 30s to 15s.

from granblue-automation-pyautogui.

Thanks for your efforts!

from granblue-automation-pyautogui.