Add a json parser/reader

Goals:

[x] import dumped json from a conversation -> Conversation
[x] import jumped json from a corpus -> Corpus

In order to make the review easier we recommend the following workflow.

Import cha and write both a conversation and corpus to json. (Download a .cha fil e.g. here)
```
import sktalk
```

conversation_cha = sktalk.ChaFile('path/to/conversation_file.cha').parse() conversation_cha.write_json(path = "path/to/conversation_file.json")

corpus_obj = sktalk.Corpus(name = "Example Corpus from Griffith Corpus of Spoken Australian English", url = "https://ca.talkbank.org/data-orig/GCSAusE/") corpus_obj.append(conversation_cha)

corpus_obj.write_json(path = "path/to/corpus_file.json")


2. Here you import the json files as Corpus and Conversation objects. This is the functionality that has been added.

corpus_obj_from_json = Corpus.from_json("path/to/corpus_file.json") conversation_obj_from_json = Conversation.from_json("path/to/conversation_file.json")


3. You can verify them by checking for instance the metadata.

corpus_obj_from_json.metadata conversation_obj_from_json.metadata



Closes #34
Should also address #41

elpaco-escience / scikit-talk

Add a json parser/reader #39