Automatically generate discord object parse code from their docs

hurrhnn_autogen.py

import re
import requests

# Do check for: ex) "integer or string"

files = ["Application.md", "Audit_Log.md", "Channel.md", "Emoji.md", "Guild.md", "Guild_Scheduled_Event.md", "Guild_Template.md", "Invite.md", "Stage_Instance.md", "Sticker.md", "User.md", "Voice.md", "Webhook.md"]

known_types = {
    # Type: [funcname, casting]
    "boolean": ["boolean", None],
    "integer": ["number", "int"],
    "string": ["string", "char *"],
    "array": ["!array", "char *"],
    "default": ["!object", "char *"]
}

aliases = {
    "snowflake": "string",
    "iso8601 timestamp": "string",
    "integer or string": "string",
    "int": "integer",
    "string (can be null only in reaction emoji objects)": "string"
}

unregistered_type = []

def get_func(funcname, casting):
    if funcname.startswith("!"):
        funcstr = f"json_serialize_to_string(json_{funcname[1:]}_get_wrapping_value(json_object_get_{funcname[1:]}({{}}, {{}})))"
    else:
        funcstr = f"json_object_get_{funcname}({{}}, {{}})"

    if casting is not None:
        funcstr = f"({casting}) {funcstr}"

    return funcstr


def get_typedata(typename):
    typename = typename.lower().strip().rstrip()

    if typename in aliases:
        typename = aliases.get(typename)

    type_ = known_types.get(typename, None)

    if type_ is None:
        unregistered_type.append(typename)
        if typename.startswith("array") or typename.startswith("list"):
            type_ = known_types.get("array")
        else:
            type_ = known_types.get("default")

    return type_


def format_full(s_name, s_data, baseindent=4):
    indent = baseindent + 16 + len(s_name)

    structstr = f"""\
{s_name}_t *parse_{s_name}(JSON_Value *{s_name}_value) {{
    JSON_Object *{s_name} = json_object({s_name}_value);

    *result = ({s_name}_t) @;

    return result;
}}
"""

    substr = f",\n{' '*indent}".join([get_func(*get_typedata(t_type)).format(s_name, f"\"{t_name}\"") for t_name, t_type in s_data.items()])

    print(structstr.replace("@", f"{{{substr}\n{' '*baseindent}}}"))
    print("\n")



def purify_item(text):
    rep_txt = text.replace("?", "").replace("| ", "|").replace(" |", "|")
    pure_txt = re.sub("[\\\\]+\\*", "", rep_txt)
    return re.sub("\\s{2,}", "", pure_txt).split("|", 3)[1:3]


def get_structures(text):
    i = 0
    check = re.compile("#+ (.+?) Structure")
    structures = {}

    lines = text.split("\n")

    while i < len(lines):
        line = lines[i]

        struct_match = check.match(line)
        if struct_match:
            struct_name = re.sub("\\s+", "_", struct_match.groups()[0].lower())
            #print("###", struct_name, "###")
            struct_dict = {}

            if ">" in lines[i+2]:
                i += 4
            i += 4

            while True:
                line = lines[i]
                if not line:
                    break

                items = purify_item(line)
                #print(items)
                key, val = items
                struct_dict[key] = val

                i += 1

            structures[struct_name] = struct_dict

        i += 1

    return structures

check = []

for file in files:
    structures = get_structures(requests.get(f"https://raw.githubusercontent.com/discord/discord-api-docs/master/docs/resources/{file}").text)
    
    for s_name, s_data in structures.items():
        if s_name in check:
            print(f"### WARNING: {s_name} already exists")
        check.append(s_name)
        format_full(s_name, s_data)

# print("|\n".join(check))
# print("#####")
# print("\n".join(unregistered_type))

hurrhnn_autogen_declare.py

import re
import requests

# Do check for: ex) "integer or string"

files = ["Application.md", "Audit_Log.md", "Channel.md", "Emoji.md", "Guild.md", "Guild_Scheduled_Event.md", "Guild_Template.md", "Invite.md", "Stage_Instance.md", "Sticker.md", "User.md", "Voice.md", "Webhook.md"]

known_types = {
    # Type: [funcname, casting]
    "boolean": ["boolean", None],
    "integer": ["number", "int"],
    "string": ["string", "char *"],
    "array": ["!array", "char *"],
    "default": ["!object", "char *"]
}

aliases = {
    "snowflake": "string",
    "iso8601 timestamp": "string",
    "integer or string": "string",
    "int": "integer",
    "string (can be null only in reaction emoji objects)": "string"
}

unregistered_type = []

def get_func(funcname, casting):
    if funcname.startswith("!"):
        funcstr = f"json_serialize_to_string(json_{funcname[1:]}_get_wrapping_value(json_object_get_{funcname[1:]}({{}}, {{}})))"
    else:
        funcstr = f"json_object_get_{funcname}({{}}, {{}})"

    if casting is not None:
        funcstr = f"({casting}) {funcstr}"

    return funcstr


def get_typedata(typename):
    typename = typename.lower().strip().rstrip()

    if typename in aliases:
        typename = aliases.get(typename)

    type_ = known_types.get(typename, None)

    if type_ is None:
        unregistered_type.append(typename)
        if typename.startswith("array") or typename.startswith("list"):
            type_ = known_types.get("array")
        else:
            type_ = known_types.get("default")

    return type_


def format_full(s_name, s_data, baseindent=4):
    indent = baseindent + 16 + len(s_name)

    structstr = f"""\
{s_name}_t *parse_{s_name}(JSON_Value *{s_name}_value) {{
    JSON_Object *{s_name} = json_object({s_name}_value);

    *result = ({s_name}_t) @;

    return result;
}}
"""

    substr = f",\n{' '*indent}".join([get_func(*get_typedata(t_type)).format(s_name, f"\"{t_name}\"") for t_name, t_type in s_data.items()])

    print(structstr.replace("@", f"{{{substr}\n{' '*baseindent}}}"))
    print("\n")


def format_declare_full(s_name, s_data, baseindent=4):
    structstr = f"""\
typedef struct yadl_{s_name} {{
    @
}} {s_name}_t;
"""
    
    substr_list = []

    for t_name, t_type in s_data.items():
        t_type = re.sub("\\[(.*?)\\]\\(.*?\\)", lambda x:  x.groups()[0], t_type)
        t_name = t_name.strip().rstrip()
        funcname, casting = get_typedata(t_type)
        if casting == None:
            casting = "bool"
        subsubstr = f"{casting}{' ' if casting[-1] != '*' else ''}{t_name};\n{' '*baseindent}"
        substr_list.append(subsubstr)

        if funcname.startswith("!"):
            substr_list.append(f"/* {t_type} */\n{' '*baseindent}")

    substr = "".join(substr_list)

    print(structstr.replace("@", substr))


def purify_item(text):
    rep_txt = text.replace("?", "").replace("| ", "|").replace(" |", "|")
    pure_txt = re.sub("([\\\\]+)?\\*", "", rep_txt)
    return re.sub("\\s{2,}", "", pure_txt).split("|", 3)[1:3]


def get_structures(text):
    i = 0
    check = re.compile("#+ (.+?) Structure")
    structures = {}

    lines = text.split("\n")

    while i < len(lines):
        line = lines[i]

        struct_match = check.match(line)
        if struct_match:
            struct_name = re.sub("\\s+", "_", struct_match.groups()[0].lower())
            #print("###", struct_name, "###")
            struct_dict = {}

            if ">" in lines[i+2]:
                i += 4
            i += 4

            while True:
                line = lines[i]
                if not line:
                    break

                items = purify_item(line)
                #print(items)
                key, val = items
                struct_dict[key] = val

                i += 1

            structures[struct_name] = struct_dict

        i += 1

    return structures

check = []

for file in files:
    structures = get_structures(requests.get(f"https://raw.githubusercontent.com/discord/discord-api-docs/master/docs/resources/{file}").text)
    
    for s_name, s_data in structures.items():
        if s_name in check:
            print(f"### WARNING: {s_name} already exists")
        check.append(s_name)
        format_declare_full(s_name, s_data)

# print("|\n".join(check))
# print("#####")
# print("\n".join(unregistered_type))