This repository hosts all the data that powers SanskritSahitya.org.
The structure of the repo is straightforward:
code/holds any scripts that might be useful for reading / writing this data- For a typical Kaavya named
kaavya, there would be filekavya/kaavya.jsonthat has the entire text of the Kaavya.
For large kaavyas such as the Mahabharatam, the data has been sharded, with the number appended to the filename. eg. mahabharatam/mahabharatam1.json etc.
The typical structure of the JSON is:
{
"title": "Name of the Kaavya",
"books": [{"number": "1", "name": "Name of Book 1"},
{"number": "2", "name": "Name of Book 2"}, ...]
"chapters": [{"number": "1", "name": "Name of Chapter 1"},
{"number": "2", "name": "Name of Chapter 2"}, ...]
"data": [
{"c": "1", "n": "1", "i": 0, "v": "Shlok 1.1"},
{"c": "1", "n": "2", "i": 1, "v": "Shlok 1.2"},
...
]
}
The format currently supports kaavyas that are either:
- Not sub-divided into chapters / sections. A shloka in such a kaavya is referenced with just the number of the shloka. Example
/meghadutam/1. - Divided into sections / chapters that comprise of shlokas. A shloka in such a kaavya is referenced with the chapter number combined with the shloka number. Example
/raghuvansham/1.1. - Divided into books which are sub-divided into sections, which then comprise of shlokas. A chapter in such a kaavya is referenced with the book number and chapter number combined (e.g.
1.1) and a shloka is represented with all three combined. Example/mahabharatam/1.1.1.
The data key has a list of objects (referred to as Shloka objects) representing the entire text.
The structure of a Shloka object is:
{
// [String] Chapter number, matches the chapter numbers declared in the `chapters` key.
// For Kaavyas with books, this would include the book number.
"c": "1",
// [String] Shloka number
"n": "1",
// [Integer] Index. This represents the overall index of this shloka / line
// in the text, and runs monotonically from start to end.
"i": 0,
// [String] Verse [Shloka]. The actual text of the shloka.
"v": "वागर्थाविव संपृक्तौ..."
// [String] Text. Refers to non-shloka text such as 'धृतराष्ट्र उवाच ।' that is
// meant to be interspersed with shlokas.
// A Shloka object should have only one of |v| and |t| keys.
// A text entry doesn't have the "n" or "i" keys present.
"t": "धृतराष्ट्र उवाच ।"
// [String] Speaker. This is used in plays to annotate the speaker of this particular text.
// Using this makes it easier to show the speaker in the UI and allows searching for dialogues
// based on speakers.
"sp": "शकुन्तला"
}
The focus is to make this easy to access through computational means, but it's written out in a specifically pretty-printed JSON format to be relatively easily human readable and to lead to clean diffs when modifying any line. It is still a pure JSON object that can be read by any parser, but if you are planning to submit changes to the repository, use the smart_json_dump helper function in code/smart_json_dump.py to pretty-print your output to the repo style.