Wrong requirements in samples/*/requirements.txt

[zub2]

The requirements.txt in samples still specify version google-cloud-speech==1.3.2. But the code itself (at least in python-speech/samples/microphone) needs version 2.0.0.

As a result, when one blindly runs pip install -r samples/microphone/requirements.txt and then proceeds to run e.g. samples/microphone/transcribe_streaming_mic.py, the result is:

$ python samples/microphone/transcribe_streaming_mic.py 
Traceback (most recent call last):
  File "samples/microphone/transcribe_streaming_mic.py", line 197, in <module>
    main()
  File "samples/microphone/transcribe_streaming_mic.py", line 172, in main
    config = speech.RecognitionConfig(
AttributeError: module 'google.cloud.speech' has no attribute 'RecognitionConfig'

... which provides a suboptimal "getting started" experience.

Steps to reproduce

1. virtualenv testenv
2. . testenv/bin/activate
3. pip install -r samples/microphone/requirements.txt
4. python samples/microphone/transcribe_streaming_mic.py

The fix is to install google-cloud-speech==2.0.0 instead of the old 1.3.2.

I think this issue is fixed in #78 but as it's not yet merged and others might run into this (and spend time figuring out what's wrong), I'm creating this issue.

[b-loved-dreamer]

Hi @zub2, thank you for making note of that. As you noted, we are going thorough the process of updating our samples to reflect the changes in 2.0.0.