Prevent reimports for TF

[zhiltsov-max]

Motivation and context

• Updated TF import check so that TF couldn't be re-imported with default import expression if an import error happens

How has this been tested?

Manual test

Checklist

• I submit my changes into the develop branch
• I have added description of my changes into CHANGELOG file
• I have updated the documentation accordingly
• I have added tests to cover my changes
• I have linked related issues (read github docs)
• I have increased versions of npm packages if it is necessary (cvat-canvas,
  cvat-core, cvat-data and cvat-ui)

License

• I submit my code changes under the same MIT License that covers the project.
  Feel free to contact the maintainers if that's a concern.
• I have updated the license header for each file (see an example below)

# Copyright (C) 2020 Intel Corporation
#
# SPDX-License-Identifier: MIT

[nmanovic]

Here is an overview of what got changed by this pull request:

Complexity increasing per file
==============================
- datumaro/datumaro/util/tf_util.py  2
         

See the complete overview on Codacy

[coveralls]

Pull Request Test Coverage Report for Build 5760

• 6 of 10 (60.0%) changed or added relevant lines in 1 file are covered.
• 7 unchanged lines in 1 file lost coverage.
• Overall coverage decreased (-0.04%) to 65.756%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
datumaro/datumaro/util/tf_util.py	6	10	60.0%

Files with Coverage Reduction	New Missed Lines	%
cvat/apps/engine/media_extractors.py	7	75.59%

Totals
Change from base Build 5743:	-0.04%
Covered Lines:	10920
Relevant Lines:	16214

---

💛 - Coveralls

[nmanovic]

Is it performance optimisation? Could you please add a comment?

[zhiltsov-max]

Not entirely. It is a way to be sure about the result of get() call. The problem with default None is that it is a valid value for a module, and, moreover, we need it later. Using some specific object() we can be sure about existence of a key and its value.

[nmanovic]

LGTM