Maximum call stack size exceeded

[arcturial]

Started receiving this error once I upgraded to 0.1.1

ERROR RangeError: Maximum call stack size exceeded
  at Object.stringify (native)
  at TelegramBot.request (/home/gitlab-runner/builds/8720be54/0/webservices/bot/node_modules/hubot-telegram/node_modules/telegrambot/lib/telegrambot.js:75:53)
  at TelegramBot.invoke (/home/gitlab-runner/builds/8720be54/0/webservices/bot/node_modules/hubot-telegram/node_modules/telegrambot/lib/telegrambot.js:95:32)
  at /home/gitlab-runner/builds/8720be54/0/webservices/bot/node_modules/hubot-telegram/src/telegram.coffee:143:14
  at Telegram.apiSend (/home/gitlab-runner/builds/8720be54/0/webservices/bot/node_modules/hubot-telegram/src/telegram.coffee:150:5)
  at Telegram.reply (/home/gitlab-runner/builds/8720be54/0/webservices/bot/node_modules/hubot-telegram/src/telegram.coffee:179:6)
  at /home/gitlab-runner/builds/8720be54/0/webservices/bot/node_modules/hubot/src/response.coffee:82:7
  at allDone (/home/gitlab-runner/builds/8720be54/0/webservices/bot/node_modules/hubot/src/middleware.coffee:44:37)
  at /home/gitlab-runner/builds/8720be54/0/webservices/bot/node_modules/hubot/node_modules/async/lib/async.js:274:13
  at Object.async.eachSeries (/home/gitlab-runner/builds/8720be54/0/webservices/bot/node_modules/hubot/node_modules/async/lib/async.js:142:20)
  at Object.async.reduce (/home/gitlab-runner/builds/8720be54/0/webservices/bot/node_modules/hubot/node_modules/async/lib/async.js:268:15)
  at /home/gitlab-runner/builds/8720be54/0/webservices/bot/node_modules/hubot/src/middleware.coffee:49:13
  at nextTickCallbackWith0Args (node.js:420:9)
  at process._tickCallback (node.js:349:13)

[RangeError: Maximum call stack size exceeded]

Haven't had time to debug in detail yet, rolling back to 0.1.0 solved the problem...so I am assuming it's something in the new release.

[lukefx]

@arcturial strange...my bot is running with 0.1.1 for a while and I never seen this error... By looking at the code the exception is thrown at 143 and 150, where we split messages in chunks recursively...can you try to set HUBOT_LOG_LEVEL=debug for a few days?

[arcturial]

I shall do that

[slackerzz]

i have the same problems of @arcturial i found that after removing

res.envelope.telegram = {
  parse_mode: 'HTML',
  reply_markup: {
    keyboard: [['option1', 'option2']]
  }
};

my bot works again.

Before the error was:

[Tue Jul 26 2016 01:01:35 GMT+0200 (CEST)] DEBUG Message 'bot command' matched regex //command/i/; listener.options = { id: null }
[Tue Jul 26 2016 01:01:35 GMT+0200 (CEST)] DEBUG Executing listener callback for Message 'bot command'
[Tue Jul 26 2016 01:01:36 GMT+0200 (CEST)] DEBUG Message length: 90
[Tue Jul 26 2016 01:01:36 GMT+0200 (CEST)] DEBUG Message parts: 1
[Tue Jul 26 2016 01:01:36 GMT+0200 (CEST)] ERROR RangeError: Maximum call stack size exceeded
    at Object.stringify (native)
    at TelegramBot.request (/bot/node_modules/telegrambot/lib/telegrambot.js:75:53)
    at TelegramBot.invoke (/bot/node_modules/telegrambot/lib/telegrambot.js:95:32)
    at send (/bot/node_modules/hubot-telegram/src/telegram.coffee:143:9, <js>:165:28)
    at Telegram.apiSend (/bot/node_modules/hubot-telegram/src/telegram.coffee:150:5, <js>:171:14)
    at Telegram.send (/bot/node_modules/hubot-telegram/src/telegram.coffee:160:5, <js>:189:19)
    at runAdapterSend (/bot/node_modules/hubot/src/response.coffee:82:34, <js>:87:50)
    at allDone (/bot/node_modules/hubot/src/middleware.coffee:44:37, <js>:34:16)
    at /bot/node_modules/async/lib/async.js:274:13
    at Object.async.eachSeries (/bot/node_modules/async/lib/async.js:142:20)
    at Object.async.reduce (/bot/node_modules/async/lib/async.js:268:15)
    at /bot/node_modules/hubot/src/middleware.coffee:49:7, <js>:37:22
    at _combinedTickCallback (internal/process/next_tick.js:67:7)
    at process._tickCallback (internal/process/next_tick.js:98:9)

[manoamaro]

It's happening when we put any extra params to res.envelope.telegram. Starting at line 72 (telegram.coffee), there is this method toString(object) that is called only when there is any extra.

    toString = (object) ->
      result = {}

      try
        for key in Object.keys(object)
          result[key] = toString(object[key])
      catch e
        result = "" + object;

      return result

    if extra?
      extra = toString(extra)
...

Not sure if it should work the same as JSON.stringify(obj), but actually it's returning a huge recursive object. The telegrambot already parses this extra to string using JSON.stringify(obj) (where the error occurs), so I think this code is not necessary anymore. I can submit a PR to remove this. Already tested here, and worked.

[lukefx]

Oh ok, perfect if you send a PR I will test it and merge it right away...

Am Freitag, 29. Juli 2016 schrieb Manoel de A.L. Amaro :

It's happening when we put any extra params to res.envelope.telegram. Starting at line 72 https://github.com/lukefx/hubot-telegram/blob/master/src/telegram.coffee#L72 (telegram.coffee), there is this method toString(object) that is called only when there is any extra.

toString = (object) ->
  result = {}

  try
    for key in Object.keys(object)
      result[key] = toString(object[key])
  catch e
    result = "" + object;

  return result

if extra?
  extra = toString(extra)

...

Not sure if it should work the same as JSON.stringify(obj), but actually it's returning a huge recursive object. The telegrambot already parses this extra to string using JSON.stringify(obj) (where the error occurs), so I think this code is not necessary anymore. I can submit a PR to remove this. Already tested here, and worked.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/lukefx/hubot-telegram/issues/52#issuecomment-236049039, or mute the thread https://github.com/notifications/unsubscribe-auth/AABwzsA04dQ110vSfcG1zW7F3dxD7b3rks5qaTMCgaJpZM4JQoJF .

[arcturial]

Nice catch, thanks