Python module: Revised JSON structure and filters

[kavitharaju]

The python module, especially the to_dict() or --format json API, has been re-written completely with the following effects

• The JSON output has been changed

USFM	New JSON(with all contents)	Old JSON(with filter=all)
\id GEN \h Genesis \c 1 \p \v 1 test verse	{"book": { "bookCode": "GEN", "headers": [ {"h": "Genesis"} ], "chapters": [ {"chapterNumber": "1", "contents": [ {"p": [{"verseNumber": "1"}, {"verseText": "test verse"}] }] }] }}	[{"book": [ {"id": ["\\id ", {"bookcode": ["GEN"]}, {"description": "\n"}]}] }, {"hBlock": [{"h": [ {"hTag": ["\\h"," "]}, {"text": "Genesis\n"}]}] }, {"chapter": [ {"c": ["\\c ",{"chapterNumber": "1"}]}, {"paragraph": [{"p": [ "\\p", {"v": ["\\v ",{"verseNumber": "1 "}]}, {"verseText": [{"text": "test verse\n"}]} ]}] }] } ]

• Filter options changed
  [--filter {scripture-bcv,notes,scripture-paragraph,all}]
  to
  [--filter {BOOK_HEADERS,PARAGRAPHS,TITLES,SCRIPTURE_TEXT,NOTES,ATTRIBUTES,MILESTONES,STUDY_BIBLE}]

Also, the way filter is used has been changed. Instead of selecting one option, now user can add more than one filter choice and include different combinations of the contents he needs to extract from the USFM. Example, scripture-and-notes, notes-and-milestones, scripture-and-titles, scripture-paragraphs-and-titles etc.

• The to_list() or --format table API, used for getting TSV outputs, has been re-written to work according to the JSON changes. This API is internally dependant on the to_dict() api and just re-formats the dict output to a convenient list form. But this api cannot handle the paragraph filter option, as the nested structure of paragraphs is hard to implement in a table output format.

• Added linting check on gitactions for the python module

• As part of the above to_dict refactoring, one grammar rule had to be changed(toc) to follow the pattern of other marker rules. This results in a slight change in the syntax tree.