Minimum dependencies for building Google Cloud Platform GRPC Voice application

[thecloudist]

For the voice-controlled robot car:

Going the k.i.s.s. route and building a simple app without any special packaging.
Will worry about sharing it once it is fully operational and bug-free. (:>

What are the minimum dependencies required to incorporate voice/grpc.methods into my application?

requirements.txt:

gcloud==0.18.1
grpcio==1.0.0
PyAudio==0.2.9
grpc-google-cloud-speech-v1beta1==1.0.1

transcribe_streaming_thread.py

imports:

from gcloud.credentials import get_credentials
from google.cloud.speech.v1beta1 import cloud_speech_pb2 as cloud_speech
from google.rpc import code_pb2
from grpc.beta import implementations

Appreciate some guidance on this.

Cheers

[puneith]

@thecloudist Those are the requirements and dependencies required.

[thecloudist]

FYI - I was able to pip install the missing Pi dependencies into the virtualenv so the two libs can play together. Pi and GCO/Speech/Grpc.

(yay)

[jerjou]

Yeah - the reason we recommend installing everything into a virtualenv is to try to avoid dependency conflicts, which will be different for different platforms. It's a general problem that doesn't have a great solution because it's so specific to different environments and situations; and the cleanest way to avoid it is to just start from a clean virtualenv.

That said, there's nothing about the requirements.txt that requires being run in a virtualenv. Installing outside of a virtualenv should be fine, but may run into dependency conflict issues, as you've observed. In principle, these should be resolvable, but in practice you end up spending way more time trying to resolve them than you want (as you've discovered ;) ).

Anyway, glad you were able to figure it out. If you'd like, you can read more about virtualenvs here and requirements.txt here.