ForcePilot/backend/package/yuxi/channel/extensions/zalouser/format.py

from __future__ import annotations

import re
from typing import Any

ZALO_TEXT_LIMIT = 2000

_FORMAT_TAG_PATTERN = re.compile(r"\{(\w+)\}(.*?)\{/(\w+)\}", re.DOTALL)


def parse_zalouser_text_styles(markdown: str) -> list[dict[str, Any]]:
    styles: list[dict[str, Any]] = []

    code_block_pattern = re.compile(r"```[\s\S]*?```")
    code_blocks = code_block_pattern.findall(markdown)
    placeholder = "\u0000"
    text = markdown
    for cb in code_blocks:
        text = text.replace(cb, placeholder, 1)

    def _capture_style(start: int, end: int, style_type: str) -> None:
        styles.append(
            {
                "start": start,
                "end": end,
                "type": style_type,
            }
        )

    def _apply_regex(pat: re.Pattern, style_type: str, text_cursor: str) -> str:
        offset = 0
        result = list(text_cursor)
        for m in pat.finditer(text_cursor):
            inner = m.group(2) if pat.groups >= 2 else m.group(1)
            raw_start = m.start() - offset
            raw_end = raw_start + len(inner)
            _capture_style(raw_start, raw_end, style_type)
            result[m.start() - offset : m.end() - offset] = [inner]
            offset += m.end() - m.start() - len(inner)
        return "".join(result)

    text = _apply_regex(re.compile(r"\*\*(.+?)\*\*"), "b", text)
    text = _apply_regex(re.compile(r"__(.+?)__"), "u", text)
    text = _apply_regex(re.compile(r"\*(.+?)\*"), "i", text)
    text = _apply_regex(re.compile(r"~~(.+?)~~"), "s", text)

    heading_pat = re.compile(r"^#{1,3}\s+(.+)$", re.MULTILINE)
    for m in heading_pat.finditer(text):
        _capture_style(m.start(1), m.end(1), "b")
        _capture_style(m.start(1), m.end(1), "f_18")

    custom_tag_pat = re.compile(r"\{(\w+)\}(.*?)\{/(\w+)\}", re.DOTALL)
    tag_style_map = {
        "red": "c_db342e",
        "orange": "c_f27806",
        "yellow": "c_f7b503",
        "green": "c_15a85f",
        "blue": "c_4391d9",
        "big": "f_18",
    }
    for m in custom_tag_pat.finditer(text):
        tag = m.group(1)
        if tag in tag_style_map:
            _capture_style(m.start(2), m.end(2), tag_style_map[tag])

    pid = 0
    for pid, ph in enumerate(code_blocks):
        text = text.replace(placeholder, ph, 1)

    return styles


def chunk_text_for_outbound(text: str, limit: int = ZALO_TEXT_LIMIT) -> list[str]:
    if len(text) <= limit:
        return [text]

    chunks: list[str] = []
    paragraphs = text.split("\n\n")

    current = ""
    for para in paragraphs:
        if len(current) + len(para) + 2 <= limit:
            current = f"{current}\n\n{para}" if current else para
        else:
            if current:
                chunks.append(current)
            if len(para) <= limit:
                current = para
            else:
                current = ""
                para_remain = para
                while len(para_remain) > limit:
                    split_at = para_remain.rfind("\n", 0, limit)
                    if split_at == -1:
                        split_at = para_remain.rfind(" ", 0, limit)
                    if split_at == -1:
                        split_at = limit
                    chunks.append(para_remain[:split_at].strip())
                    para_remain = para_remain[split_at:].strip()
                if para_remain:
                    current = para_remain

    if current:
        chunks.append(current)

    return chunks if chunks else [text[:limit]]


def resolve_reaction_icon(raw: str) -> str | None:
    alias = raw.strip().lower()
    from yuxi.channel.extensions.zalouser.types import REACTION_ALIAS_MAP

    if alias in REACTION_ALIAS_MAP:
        return REACTION_ALIAS_MAP[alias].value
    for icon in REACTION_ALIAS_MAP.values():
        if icon.value.lower() == alias:
            return icon.value
    return None


def strip_own_mentions(content: str, bot_name: str = "") -> str:
    if not bot_name:
        return content
    pattern = re.compile(rf"@{re.escape(bot_name)}\s*", re.IGNORECASE)
    return pattern.sub("", content).strip()


def extract_mention_ids(content: str) -> list[str]:
    pattern = re.compile(r"@(\w+)")
    return pattern.findall(content)


def format_for_zalouser_markdown(markdown: str) -> str:
    return markdown