Properly check if download() is needed in get_corpus_path()

[bact]

• Remove corpus download code that scattered around and put them in one single get_corpus_path().
• Make sure the download starts when necessary as intended (check both listing in the corpus catalog and the actual existence of the file)
• In the future, will there be any change (like to resolve Considerations for language model inclusion in default package or download them later #298 and pythainlp.corpus.get_corpus_path() should not try to download the corpus automatically #385), hopefully one can change it only once at get_corpus_path() and need to

Also:

• Refactor royin romanization function
• Add more test cases for royin romanization (found few bugs along the way - see 'royin' engine gives wrong romanization in a lot of cases #415)
• Refactor ThaiNameTagger(), remove unnessary self. references

This will partly close #411

[pep8speaks]

Hello @bact! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

There are currently no PEP 8 issues detected in this Pull Request. Cheers! 🍻

Comment last updated at 2020-05-27 15:24:34 UTC

[coveralls]

Coverage remained the same at 92.319% when pulling f2a343d on more-test-cases into 4c22970 on dev.