fix: use usage_key instead of block_id as location identifier for scorm data

[Danyal-Faheem]

Fixes #70.

Changes

• Added a new base_path for the scorm folder. The previous one was using the block_id as an identifier for the directory name but when exporting a course and importing it, the block_id stayed the same causing both the courses to share the scorm directory. Now, the base_path is a sha1 hash of the usage_key which is unique across different courses as it also takes into account the course_key as well.
• The previous base_path is still kept for backwards compatibility and that path is returned in case storage exists on that path, meaning that the scorm module was created before this change.

Caveats

• If a scorm unit has already been initialized, i.e, scorm file has been uploaded, then duplicating that subsection or exporting and importing that course will still break the scorm module as earlier due to the path not being updated to the new version.
• Due to above point, this change will only effect the scorm modules initialized after this PR has been merged. Any course already created will still be facing the issue mentioned in #70 until the course author reupload scorm files and save again, i.e studio_submit() is called again.

[Danyal-Faheem]

@Danyal-Faheem I think we should avoid uploading of content file it does not make sense to extract old package when course is imported. I think in case of course import scorm package should become unavailable in new course whereas in previous course it should work fine.

@ziafazal okay, I will remove the code for saving the content file and just fix the path then.