Call flaskgroup main, not app.cli.main

[revmischa]

Because app.cli.main is replaced by click.core.main (I don't know why)

[codecov]

Codecov Report

Merging #99 into master will decrease coverage by 0.4%.
The diff coverage is 33.33%.

@@            Coverage Diff             @@
##           master      #99      +/-   ##
==========================================
- Coverage     100%   99.59%   -0.41%     
==========================================
  Files           5        5              
  Lines         492      491       -1     
  Branches       60       60              
==========================================
- Hits          492      489       -3     
- Misses          0        2       +2

Impacted Files	Coverage Δ
wsgi_handler.py	96.87% <33.33%> (-3.13%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 3d30831...cab200f. Read the comment docs.

[codecov]

Codecov Report

Merging #99 into master will decrease coverage by 0.4%.
The diff coverage is 33.33%.

@@            Coverage Diff             @@
##           master      #99      +/-   ##
==========================================
- Coverage     100%   99.59%   -0.41%     
==========================================
  Files           5        5              
  Lines         492      491       -1     
  Branches       60       60              
==========================================
- Hits          492      489       -3     
- Misses          0        2       +2

Impacted Files	Coverage Δ
wsgi_handler.py	96.87% <33.33%> (-3.13%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 3d30831...cab200f. Read the comment docs.

[revmischa]

I don't understand the test error here but it doesn't look related to my changes? https://travis-ci.com/jetbridge/serverless-wsgi/jobs/211630464#L717

[logandk]

Thanks for the PR! The failing test was fixed in e53c701

[revmischa]

It's slightly improved but still broken. Need help.

[logandk]

I did a few additional changes on master. When testing your PR, it worked fine locally - but when deployed to Lambda, I'd see an error like:

Traceback (most recent call last):
  File "/var/task/wsgi_handler.py", line 95, in handler
    fg.main(shlex.split(meta.get("data", "")), standalone_mode=False)
  File "/var/task/flask/cli.py", line 586, in main
    return super(FlaskGroup, self).main(*args, **kwargs)
  File "/var/task/click/core.py", line 717, in main
    rv = self.invoke(ctx)
  File "/var/task/click/core.py", line 1132, in invoke
    cmd_name, cmd, args = self.resolve_command(ctx, args)
  File "/var/task/click/core.py", line 1188, in resolve_command
    ctx.fail('No such command "%s".' % original_cmd_name)
  File "/var/task/click/core.py", line 496, in fail
    raise UsageError(message, self)
click.exceptions.UsageError: No such command "testcommand".

I added the explicit create_app argument back in, which sorted it out.

Do you see anything else that's not working properly on master?

[revmischa]

i'll try master now. i was able to get to the point where custom and built-in commands worked on local, but custom didn't work on lambda, i think. something like that.

[logandk]

Alright - I tried both custom and built-in commands, locally and remotely. Let me know if it works out for you too.

[revmischa]

pulled master, linked it, deployed.

sls wsgi flask local -c routes - works

sls wsgi flask -c routes -

  Error --------------------------------------------------
 
  Invoked function failed
 
     For debugging logs, run again after setting the "SLS_DEBUG=*" environment variable.
 
  Stack Trace --------------------------------------------
 
Error: Invoked function failed
    at AwsInvoke.log (/Users/cyber/.config/yarn/global/node_modules/serverless/lib/plugins/aws/invoke/index.js:97:31)
From previous event:
    at Object.invoke:invoke [as hook] (/Users/cyber/.config/yarn/global/node_modules/serverless/lib/plugins/aws/invoke/index.js:22:10)
    at /Users/cyber/.config/yarn/global/node_modules/serverless/lib/classes/PluginManager.js:448:55
From previous event:
    at PluginManager.invoke (/Users/cyber/.config/yarn/global/node_modules/serverless/lib/classes/PluginManager.js:448:22)
    at PluginManager.run (/Users/cyber/.config/yarn/global/node_modules/serverless/lib/classes/PluginManager.js:479:17)
    at /Users/cyber/dev/jb/serverless-wsgi/index.js:472:12
    at ServerlessWSGI.invokeHandler (/Users/cyber/dev/jb/serverless-wsgi/index.js:463:12)
    at ServerlessWSGI.flask (/Users/cyber/dev/jb/serverless-wsgi/index.js:531:17)
    at ServerlessWSGI.<anonymous> (/Users/cyber/dev/jb/serverless-wsgi/index.js:749:28)
    at processImmediate (internal/timers.js:439:21)
    at process.topLevelDomainCallback (domain.js:126:23)

[revmischa]

actually this error may be caused by a timeout. stand by

[logandk]

Ok, what does sls logs -f yourfunction say?

[revmischa]

 sls invoke -f app  -c routes 
{
    "errorMessage": "'headers'",
    "errorType": "KeyError",
    "stackTrace": [
        "  File \"/var/task/wsgi_handler.py\", line 108, in handler\n    return serverless_wsgi.handle_request(wsgi_app, event, context)\n",
        "  File \"/var/task/serverless_wsgi.py\", line 94, in handle_request\n    headers = Headers(event[u\"headers\"])\n"
    ]
}

[logandk]

You should be doing sls wsgi flask -c routes and sls wsgi flask local -c routes instead of sls invoke?

[revmischa]

you are completely right sorry lol

[logandk]

No worries :)

[revmischa]

⇒ sls wsgi flask local -c config
{'API_SPEC_OPTIONS': {'components': {'securitySchemes': {'bearerAuth': {'bearerFormat': 'JWT',
                                                                        'scheme': 'bearer',
                                                                        'type': 'http'}}},
                      'security': [{'bearerAuth': []}]},
 'APPLICATION_ROOT': '/',
 'CAN_SEED_DB': True,
 'DATABASE_URL': '[object Object]',
 'DEBUG': False,
 'ENV': 'production',
....

 ⇒ sls wsgi flask  -c config     
Traceback (most recent call last):
  File "/var/task/wsgi_handler.py", line 96, in handler
    shlex.split(meta.get("data", "")), standalone_mode=False
  File "/var/task/flask/cli.py", line 586, in main
    return super(FlaskGroup, self).main(*args, **kwargs)
  File "/var/task/click/core.py", line 717, in main
    rv = self.invoke(ctx)
  File "/var/task/click/core.py", line 1132, in invoke
    cmd_name, cmd, args = self.resolve_command(ctx, args)
  File "/var/task/click/core.py", line 1188, in resolve_command
    ctx.fail('No such command "%s".' % original_cmd_name)
  File "/var/task/click/core.py", line 496, in fail
    raise UsageError(message, self)
click.exceptions.UsageError: No such command "config".

source is in https://github.com/jetbridge/sls-flask/

[logandk]

Ok, I suppose the create_app argument to FlaskGroup is somehow not working with your app which uses the factory pattern. I'll dig a bit further into it later this week

[revmischa]

I was experimenting with:

                from flask.cli import cli
                cmd = shlex.split(meta.get("data", ""))
                cli.main(cmd)

[logandk]

I found the culprit: https://github.com/jetbridge/sls-flask/blob/master/TEMPLATE/commands.py#L6

Remove if app.debug: and everything works 👍