Organize testcontent dir + fix for description length issue #235

[ivanistheone]

The channel validation/truncate_fileds normally happens later on in the upload process, so I added an early call to truncate_fields to avoid #235

I tested manually on command line. If description is > 400 chars, prints:

   Creating channel Sample Ricecooker Channel
		Channel sample-ricecooker-channel: description This channel was created from the files in the content/ dir and the metadata provided in Pythonalkjsalksjlakjs salksj lakjs lakjs lkajs lkaj alkjsalksjlakjs salksj lakjs lakjs lkajs lkaj alkjsalksjlakjs salksj lakjs lakjs lkajs lkaj alkjsalksjlakjs salksj lakjs lakjs lkajs lkaj alkjsalksjlakjs salksj lakjs lakjs lkajs lkaj alkjsalksjlakjs salksj lakjs lakjs lkajs lkaj babababaabab ab ababasj babababaabab ab ababasjbabababaabab ab ababasjbabababaabab ab ababasjbabababaabab ab ababasjbabababaabab ab ababasjbabababaabab ab ababasj babababaabab ab ababasj caococo caococo caococo caococo caococo caococo caococo caococo caococo caococo caococo caococo caococo caococo caococo caococo caococo caococo  dodoodododoaodao doao dodoodododoaodao doao dodoodododoaodao doao dodoodododoaodao doao dodoodododoaodao doao dodoodododoaodao doao dodoodododoaodao doao  is too long - max 400 characters (truncating)

and only first 400 chars got uploaded to Studio, see here.

[ivanistheone]

@kollivier This morning I was trying to fix a flaky test file, and ended up implementing a good part of the cleanup you suggested in #243

This is what testcontent looks like before running any tests:

> tree -a tests/testcontent/

tests/testcontent/
├── downloaded
│   └── .gitkeep
├── exercises
│   ├── eb3f3bf7c317408ee90995b5bcf4f3a59606aedd-data.json
│   ├── eb3f3bf7c317408ee90995b5bcf4f3a59606aedd.svg
│   ├── no-wifi.png
│   ├── perseus_question_x43bbec76d5f14f88_bg.json
│   ├── perseus_question_x43bbec76d5f14f88_en.json
│   └── test_image_base64.data
├── generated
│   └── .gitkeep
└── samples
    ├── sample_doc_with_toc.pdf
    ├── testdocument.epub
    ├── testsubtitles_ar.srt
    ├── thumbnail.jpg
    └── thumbnail.png

And after running tests:

> tree -a tests/testcontent/

tests/testcontent/
├── downloaded
│   ├── .gitkeep
│   ├── 41568-pdf.pdf
│   ├── Beyond-Good-and-Evil-Galbraithcolor.pdf
│   ├── basic.srt
│   ├── basic.vtt
│   ├── encapsulated.sami
│   ├── encapsulated.vtt
│   ├── high_res_video.mp4
│   ├── low_res_ogv_video.ogv
│   ├── low_res_video.mp4
│   ├── testaudio.mp3
│   ├── testdocument.pdf
│   ├── testhtml.zip
│   ├── testsubtitles_ar
│   └── testsubtitles_ar.ttml
├── exercises
│   ├── eb3f3bf7c317408ee90995b5bcf4f3a59606aedd-data.json
│   ├── eb3f3bf7c317408ee90995b5bcf4f3a59606aedd.svg
│   ├── no-wifi.png
│   ├── perseus_question_x43bbec76d5f14f88_bg.json
│   ├── perseus_question_x43bbec76d5f14f88_en.json
│   └── test_image_base64.data
├── generated
│   ├── .gitkeep
│   ├── bad_video.mp4
│   ├── invalid_audio.mp3
│   ├── invalid_document.epub
│   ├── invalid_document.pdf
│   ├── invalid_video.mp4
│   ├── invalidimage.png
│   ├── slide0.jpg
│   ├── slide1.jpg
│   ├── slide2.jpg
│   ├── slide3.jpg
│   ├── slide4.jpg
│   ├── slide5.jpg
│   ├── slide6.jpg
│   ├── slide7.jpg
│   ├── slide8.jpg
│   ├── slide9.jpg
│   ├── testinvalidhtml.zip
│   ├── testsubtitles.vtt
│   └── unconvetible.sub
└── samples
    ├── sample_doc_with_toc.pdf
    ├── testdocument.epub
    ├── testsubtitles_ar.srt
    ├── thumbnail.jpg
    └── thumbnail.png

Running make clean-test puts us back into the initial state.

[ivanistheone]

This fixture was not used by any tests, so removing.

[kollivier]

LGTM, thanks!