generate_json.py

from json import dumps
from sys import argv
from re import compile
from pathlib import Path

if __name__ != "__main__":
    exit(0)

if len(argv) != 3:
    print("Usage: generate_json.py <TDLib_version> <Last_TDLib_commit_hash>")
    exit(1)

data = {
    "name": "Auto-generated JSON TDLib API for Pytdbot ~ https://github.com/pytdbot/client",
    "version": argv[1],
    "commit_hash": argv[2],
    "classes": {},
    "types": {},
    "updates": {},
    "functions": {},
}
params = {}
start = False
is_functions = False
description = ""
class_regex = compile(r"//@class\s(?P<name>.*)\s@description\s(?P<description>.*)")
description_regex = compile(r"//@description\s(?P<description>.*)$")
parameter_regex = compile(r"@(.*?)\s+([^@]+)")
end_param_regex = compile(r"(?P<name>\w+):(?P<type>[<\w>]+)")
end_regex = compile(r"^(?P<name>.*?)\s(?P<params>.*)=\s(?P<type>\w+);$")
tl = Path("td_api.tl").read_text().replace("\n//-", " ")


def is_optional(d):
    return (
        "; may be null" in d
        or "; pass null" in d
        or "; may be empty" in d
        or "If non-empty," in d
    )


for line in tl.split("\n"):
    if "--functions--" in line:
        is_functions = True
        continue

    if line.startswith("//"):
        start = True

    if line != "" and start:
        if _class := class_regex.match(line):
            data["classes"][_class.group("name").strip()] = {
                "description": _class.group("description").strip(),
                "types": [],
                "functions": [],
            }
        elif _param := parameter_regex.findall(line):
            for name, _description in _param:
                if name.strip() == "description":
                    description = _description.strip()
                else:
                    params[name.replace("param_", "").strip()] = _description.strip()
        elif _end := end_regex.match(line):
            _data = {
                "description": description,
                "args": {},
                "type": _end.group("type").strip(),
            }

            for keyv in end_param_regex.finditer(_end.group("params")):
                k, v = keyv.group("name").strip(), keyv.group("type").strip()
                _data["args"][k] = {
                    "description": params[k],
                    "is_optional": is_optional(params[k]),
                    "type": v,
                }

            json_name = _end.group("name").strip()
            if is_functions:
                data["functions"][json_name] = _data
                if _data["type"] in data["classes"]:
                    data["classes"][_data["type"]]["functions"].append(json_name)
            elif json_name.startswith("update"):
                data["updates"][json_name] = _data
            else:
                data["types"][json_name] = _data
                if _data["type"] in data["classes"]:
                    data["classes"][_data["type"]]["types"].append(json_name)

            params = {}
            description = ""

with open("td_api.json", "w") as f:
    f.write(dumps(data, indent=4))
    print(
        "Classes: {}\nTypes: {}\nFunctions: {}\nUpdates: {}".format(
            len(data["classes"]),
            len(data["types"]),
            len(data["functions"]),
            len(data["updates"]),
        )
    )