ForcePilot/backend/package/yuxi/channel/extensions/jira/parse.py

from __future__ import annotations


def adf_to_plain_text(adf: dict | None) -> str:
    if not adf or not isinstance(adf, dict):
        return ""
    parts: list[str] = []
    _walk_nodes(adf.get("content", []), parts)
    return "".join(parts).strip()


def adf_to_markdown(adf: dict | None) -> str:
    if not adf or not isinstance(adf, dict):
        return ""
    parts: list[str] = []
    _walk_nodes_md(adf.get("content", []), parts)
    return "".join(parts).strip()


def _walk_nodes(nodes: list, parts: list):
    for node in nodes:
        if not isinstance(node, dict):
            continue
        node_type = node.get("type", "")
        marks = node.get("marks", [])

        if node_type == "text":
            text = node.get("text", "")
            _process_marks(parts, text, marks)
        elif node_type == "hardBreak":
            parts.append("\n")
        elif node_type == "mention":
            parts.append(f"@{node.get('attrs', {}).get('text', 'user')}")
        elif node_type == "emoji":
            parts.append(node.get("attrs", {}).get("shortName", ""))
        elif node_type == "paragraph":
            _walk_nodes(node.get("content", []), parts)
            parts.append("\n")
        elif node_type == "heading":
            level = node.get("attrs", {}).get("level", 1)
            prefix = "#" * level + " "
            parts.append(prefix)
            _walk_nodes(node.get("content", []), parts)
            parts.append("\n\n")
        elif node_type == "codeBlock":
            lang = node.get("attrs", {}).get("language", "")
            parts.append(f"```{lang}\n")
            _walk_nodes(node.get("content", []), parts)
            parts.append("\n```\n\n")
        elif node_type == "blockquote":
            content_parts: list[str] = []
            _walk_nodes(node.get("content", []), content_parts)
            for line in "".join(content_parts).split("\n"):
                if line.strip():
                    parts.append(f"> {line}\n")
            parts.append("\n")
        elif node_type == "bulletList":
            _walk_list_items(node.get("content", []), parts, "- ")
        elif node_type == "orderedList":
            _walk_list_items(node.get("content", []), parts, "1. ")
        elif node_type == "panel":
            panel_type = node.get("attrs", {}).get("panelType", "info")
            parts.append(f"> **[{panel_type}]** ")
            _walk_nodes(node.get("content", []), parts)
            parts.append("\n\n")
        elif node_type == "rule":
            parts.append("---\n\n")
        elif node_type == "table":
            _walk_table(node, parts)
            parts.append("\n")
        elif node_type == "media":
            media_attrs = node.get("attrs", {})
            display_type = media_attrs.get("type", "file")
            filename = media_attrs.get("alt", "") or f"{display_type}_attachment"
            if display_type == "image":
                url = media_attrs.get("url", "")
                parts.append(f"![{filename}]({url})\n")
            else:
                parts.append(f"[📎 {filename}]\n")
        elif node_type == "date":
            timestamp = node.get("attrs", {}).get("timestamp", "")
            parts.append(f"[{timestamp}]")
        elif node_type == "status":
            status_text = node.get("attrs", {}).get("text", "")
            color = node.get("attrs", {}).get("color", "")
            parts.append(f"[{status_text}]({color})")
        elif "content" in node:
            _walk_nodes(node["content"], parts)


def _walk_nodes_md(nodes: list, parts: list):
    _walk_nodes(nodes, parts)


def _process_marks(parts: list, text: str, marks: list):
    prefix = suffix = ""
    links: list[str] = []
    for mark in marks:
        mark_type = mark.get("type", "")
        if mark_type == "strong":
            prefix = "**" + prefix
            suffix += "**"
        elif mark_type == "em":
            prefix = "*" + prefix
            suffix += "*"
        elif mark_type == "code":
            prefix = "`" + prefix
            suffix += "`"
        elif mark_type == "strike":
            prefix = "~~" + prefix
            suffix += "~~"
        elif mark_type == "link":
            href = mark.get("attrs", {}).get("href", "")
            if href:
                links.append(href)
    if links:
        parts.append(f"[{prefix}{text}{suffix}]({links[-1]})")
    else:
        parts.append(f"{prefix}{text}{suffix}")


def _walk_list_items(items: list, parts: list, prefix: str):
    for item in items:
        if not isinstance(item, dict):
            continue
        item_content = item.get("content", [])
        for node in item_content:
            node_type = node.get("type", "")
            if node_type == "paragraph":
                parts.append(prefix)
                _walk_nodes([node], parts)
                parts.append("\n")
            elif node_type in ("bulletList", "orderedList"):
                sub_prefix = "  " + prefix
                _walk_list_items(node.get("content", []), parts, sub_prefix)
    parts.append("\n")


def _walk_table(table: dict, parts: list):
    rows = table.get("content", [])
    for row_idx, row in enumerate(rows):
        cells = row.get("content", [])
        cell_texts: list[str] = []
        for cell in cells:
            cell_parts: list[str] = []
            _walk_nodes(cell.get("content", []), cell_parts)
            cell_texts.append("".join(cell_parts).replace("\n", " ").strip())
        parts.append("| " + " | ".join(cell_texts) + " |\n")
        if row_idx == 0:
            parts.append("| " + " | ".join(["---"] * len(cells)) + " |\n")