Move or change Latin CorpusReader module

While doing some work with Docker, to package the web API, I came across an error that must be uncommon for devs but very common to new users.

Stems from: https://github.com/cltk/cltk/blob/master/cltk/corpus/latin/__init__.py.

On a system without any ~/cltk_data dir, anytime when CorpusImporter() is called, the following message is printed to screen: Corpus not found. Please check that the Latin Library is installed in CLTK_DATA.. Of course, not everybody wants the Latin Library corpus.

Thus, I propose making the module explicit, not hidden, import. At least until we have a strong idea about how the CorpusReader will work for other languages.

What do you think @diyclassics?

[kylepjohnson]

I think you could move its contents to a class within cltk/corpus/latin/latinlibrary.py.

Then users would import with, e.g.: from cltk.corpus.latin import latinlibrary and use with latinlibrary.LatinLibraryReader().

Just sharing an idea. A guiding principle I go by is “explicit is better than implicit”.

[todd-cook]

This has been refactored https://github.com/cltk/cltk/pull/846 you can do: from cltk.corpus.readers import get_corpus_reader latinlibrary = get_corpus_reader(language=‘latin’, corpus_name=‘latin_text_latin_library’)