I have created and deployed a telegram bot which can perform some simple chat and analytics functions.
I have set a couple of / commands via botfathers /setcommands functionality and they all work fine.
However in group chats I have a problem, when I add the bot he works normally however the set commands that you can execute via the menu always include an #bot_name at the end of each command which invalidates them.
If I type in the command myself and remove the #mention, the commands work normally.
How can I tackle this issue, is it something I need to set in my code or in botfather?
if ($cmd == "/command" || $cmd == "/command#BOT_USERNAME") {
// handle scenario
}
My suggestion is to modify your code and replace the '#your_bot' with an empty string '' from the command before processing it. That way you automatically handle both scenarios for each of the available commands.
Sadly, Telegram had a bug once or twice where it wouldn't register the #your_bot messages sent to it. Luckily that bug was live for a very short period of time.
Related
Using python-telegram-bot, I have a bot running with very similar settings as for the other examples. On the other side i have parallel processes that allow me to send messages to users interacting with the bot periodically. The parallel proccess communicates with the users using:
bot = Bot(token=TELEGRAM_TOKEN)
def get_button_options():
keyboard = [[ InlineKeyboardButton("reply", callback_data='reply),]]
reply_markup = InlineKeyboardMarkup(keyboard)
return reply_markup
bot.send_message(chat_id=self.telegram_id, text=text, reply_markup=get_button_options())
The thing is, even if I'm capable to send messages to the user, the above code does not allow the user to interact with the bot (By that I mean that, once the user presses the button, the bot doesn't execute any callback). This is because I need either a CallbackQueryHandler or ConversationHandler within the bot.
However for the ConversationHandler I don't seem to find any appropiate entry_point, since it is a message sent from a parallel process. On the other hand, using the following CallbackQueryHandler (taken from the example) in the main bot doesn't seem to have any effect:
def button(update: Update, _: CallbackContext) -> None:
query = update.callback_query
# CallbackQueries need to be answered, even if no notification to the user is needed
# Some clients may have trouble otherwise. See https://core.telegram.org/bots/api#callbackquery
query.answer()
query.edit_message_text(text=f"Selected option: {query.data}")
in the main function, I have added the handler using updater.dispatcher.add_handler(CallbackQueryHandler(button))
Any ideas on how to solve that?
As said by #CallMeStag, the solution is the following:
ConversationHandler(
entry_points=[CallbackQueryHandler(button)])
the correct entry point is a CallbackQueryHandler, so then it captures the selected button
With system under test, when adding a new user their initial password is emailed to them.
I could split my test into multiple sections with manual intervention but this is less than ideal.
Appreciate any suggestions on how to proceed using TestCafe as I am sure others have encountered this as well.
If you run full integration test with real email server, then you can use libraries like "mail-receive" to connect to this server and verify the email.
You can also run your backend/server logic in mock mode, and then verify the mock, that the send event happened, by calling some test-specific rest endpoint from your TestCafe test.
Alternatively, you could also use something like "smtp-receiver" to start your own email-server-mock in nodejs context, and receive event upon email arrival. However you will need to configure your app server/backend to point to this mocked email server.
I try to execute the community script "Extender/HTTP Message Logger.js". I first double click on the script to make it open in the scripting console. However, in the scripting console, the "Execute" button is disabled and I see no other way how to make it run.
What am I missing?
I think you're missing the message beneath the script which says:
Extender scripts add new functionality, including graphical elements
and new API end points.
Enabling a script installs it and disabling a script uninstalls it.
So you just need to enable the script (by right clicking it and selecting 'Enable') and then it will start working.
The actual issue was that I didn't read the script's code carefully: be default the script only logs JSON messages as defined on lines 17 and 43 and following. In order to log all the sent and received HTTP messages, I simply changed the isMessageToLog(log) function to always return true. After redeploying the script (disabling and enabling) it would log all HTTP messages.
I am a PHP dev looking to port my API over to the Parse platform.
Am I right in thinking that you only need cloud code for complex operations? For example, consider the following methods:
// Simple function to fetch a user by id
function getUser($userid) {
return (SELECT * FROM users WHERE userid=$userid LIMIT 1)
}
// another simple function, fetches all of a user's allergies (by their user id)
function getAllergies($userid) {
return (SELECT * FROM allergies WHERE userid=$userid)
}
// Creates a script (story?) about the user using their user id
// Uses their name and allergies to create the story
function getScript($userid) {
$user = getUser($userid)
$allergies = getAllergies($userid).
return "My name is {$user->getName()}. I am allergic to {$allergies}"
}
Would I need to implement getUser()/getAllergies() endpoints in Cloud Code? Or can I simply use Parse.Query("User")... thus leaving me with only the getScript() endpoint to implement in cloud code?
Cloud code is for computation heavy operations that should not be performed on the client, i.e. handling a large dataset.
It is also for performing beforeSave/afterSave and similar hooks.
In your example, providing you have set up a reasonable data model, none of the operations require cloud code.
Your approach sounds reasonable. I tend to put simply queries that will most likely not change on the client side, but it all depends on your scenario. When developing mobile apps I tend to put a lot of code in cloud code. I've found that it speeds up my development cycle. For example, if someone finds a bug and it's in cloud code, make the fix, run parse deploy, done! The change is available to all mobile environments instantly!!! If that same code is in my mobile app, it really sucks, cause now I have to fix the bug, rebuild, push it to the app store/google play, wait x number of days for it to be approved, have the users download it... you see where I'm going here.
Take for example your
SELECT * FROM allergies WHERE userid=$userid query.
Even though this is a simple query, what if you want to sort it? maybe add some additional filtering?
These are the kinds of things I think of when deciding where to put the code. Hope this helps!
As a side note, I have also found cloud code very handy when needing to add extra security to my apps.
I'm starting to develop an extension which must interact with an external application. I can run the external application as described here, but I do not see a way to get any feedback. The only information I get is the exit status, while I need to read the application output, as it would appear on a terminal (stdout). Is there a way to do this?
After running the nsiProcess, loop while checking the isRunning attribute. When it stops running, check the exitValue attribute. As I understand it, this may behave differently on different platforms, but I did use it successfully on Windows.