The tabtotext functions are using a universal format to store a table.
The table is a List of records. (see python typing.List)
Each record is a Dict of items. (see python typing.Dict)
Each item is a Union of basic types including str, int, float, date.
The item type is useful when handling JSON data so that the actual type names are
JSONItem = Union[str, int, float, bool, date, None, Dict[str, Any], List[Any]]
JSONDict = Dict[str, JSONItem]
JSONList = List[JSONDict]
Note that an Item can also contain more structured values but that is NOT considered for the conversion process of tabtotext. It is just useful when getting json.loads data printing them with the tabtotext utility just easily.
GFM is the github-flavoured Markdown format. This reference is needed because the original Markdown format did NOT define a table markup. This has been an extension and it was impelemented differently in different markdown tools. However the Github style has ensured a defacto standardization.
The html format here is ready to be uploaded to Confluence as a wiki page.
The tabToJSON format is extended over the python "json" library as it can handle date / datetime items. The loadJSON routine can load back the textual representation. Note that tabToJSON is a thin layer over json.loads and json.dumps routines.
The tabToCSV format is extended over the python "csv" library as it can handle date / datetime items. The loadCSV routine can load back the textual representation. Note that tabToCSV is a thin layer csv.DictReader and csv.DictWriter routines.
Unlink JSON, the TOML spec does already know about DateTime elements. Here's an implementation without tomllib. Note that tomllib is part of the Python standard library since 3.11.
The "formats" argument to each tabToFunction allows to provide string.format() specifications in the style of "{:2d}".
There is an extension that can handle "{:h}" representing a float with the fraction given in quarters. So you have "1/4" ane "1/2" and "3/4" with their latin-1 (extended-ascii) 0x00Bx unicode code points. The full strHours (.00) is encoded with "h". If you use the formatter "{:q}" then the full number is encoded with ".". Addtionally, there is a formatter with "{:M}" scaling the number by Mibi down and then using strHours with "." full encoding. Using some "{:H}" however converts to clock values like "2:15" and using "{:$}" will convert to the default currency like "2.25€".
In the newer implementions there is no "formats" argument anymore but an addition "selects" to help with print_tabtotext. And there is Excel support available in tabxlsx.py that does not require openpyxl. This file is installed as a seperate module "pip install tabxlsx" however (or via 'make insxlsx' from development).
At a later time, this will be the basis of a seperate tabtotext project.