1. The Arcade Worker doesn't like it when the package name is different than the directory name. This PR renames the directory from `arcade_notion` to `arcade_notion_toolkit` 2. `Notion` is not a well-known provider name in `arcade-ai==1.0.5`, I've updated the dep to represent this.
201 lines
7.5 KiB
Python
201 lines
7.5 KiB
Python
import asyncio
|
|
from typing import Any, Optional
|
|
|
|
from arcade.sdk import ToolContext
|
|
|
|
from arcade_notion_toolkit.enums import BlockType
|
|
from arcade_notion_toolkit.utils import get_page_url
|
|
|
|
|
|
class BlockToMarkdownConverter:
|
|
"""
|
|
A converter class that transforms Notion blocks into Markdown.
|
|
|
|
The class registers conversion handlers for different Notion block types.
|
|
If a block type does not have a handler, then the block's plain text is returned.
|
|
"""
|
|
|
|
def __init__(self, context: ToolContext):
|
|
self.context = context
|
|
# block types whose conversion logic has been implemented
|
|
# TODO: implement conversion logic for more block types
|
|
self.handlers = {
|
|
BlockType.BULLETED_LIST_ITEM.value: self._convert_bulleted_list_item,
|
|
BlockType.EQUATION.value: self._convert_equation,
|
|
BlockType.HEADING_1.value: self._convert_heading_1,
|
|
BlockType.HEADING_2.value: self._convert_heading_2,
|
|
BlockType.HEADING_3.value: self._convert_heading_3,
|
|
BlockType.LINK_PREVIEW.value: self._convert_link_preview,
|
|
BlockType.NUMBERED_LIST_ITEM.value: self._convert_numbered_list_item,
|
|
BlockType.PARAGRAPH.value: self._convert_paragraph,
|
|
}
|
|
|
|
async def convert_block(self, block: dict[str, Any]) -> str:
|
|
"""
|
|
Convert a single Notion block to a Markdown string
|
|
|
|
Args:
|
|
block (dict[str, Any]): A Notion block.
|
|
|
|
Returns:
|
|
str: A Markdown string.
|
|
"""
|
|
block_type = block.get("type")
|
|
if block_type in self.handlers:
|
|
converter = self.handlers[block_type]
|
|
if asyncio.iscoroutinefunction(converter):
|
|
md: str = await converter(block)
|
|
return md
|
|
else:
|
|
return converter(block)
|
|
elif block_type == BlockType.CHILD_PAGE.value:
|
|
return await self._convert_child_page(block)
|
|
else:
|
|
return self._get_plaintext(block)
|
|
|
|
@staticmethod
|
|
def rich_text_to_markdown(rich_text_items: list[dict[str, Any]]) -> str:
|
|
"""
|
|
Convert a list of rich text items (from a Notion block) into Markdown.
|
|
|
|
Handles formatting such as bold, italic, strikethrough, underline (via HTML),
|
|
inline code, text coloring, hyperlinks, and equations.
|
|
"""
|
|
md = ""
|
|
for item in rich_text_items:
|
|
annotations = item.get("annotations", {})
|
|
type_val = item.get("type", "text")
|
|
link = None
|
|
|
|
# Special handling for inline equations.
|
|
if type_val == "equation":
|
|
expression = item.get("equation", {}).get("expression", "")
|
|
md += f"${expression}$"
|
|
continue
|
|
|
|
if type_val == "text":
|
|
text_obj = item.get("text", {})
|
|
text = text_obj.get("content", "")
|
|
link_obj = text_obj.get("link")
|
|
link = (
|
|
link_obj.get("url")
|
|
if (link_obj and isinstance(link_obj, dict))
|
|
else item.get("href")
|
|
)
|
|
elif type_val == "mention":
|
|
text = item.get("plain_text", "")
|
|
link = item.get("href")
|
|
else:
|
|
text = item.get("plain_text", "")
|
|
link = item.get("href")
|
|
|
|
if text.strip() == "":
|
|
continue
|
|
|
|
# Apply annotation formatting.
|
|
text = BlockToMarkdownConverter.apply_formatting(text, annotations, link)
|
|
|
|
md += text
|
|
|
|
return md
|
|
|
|
@staticmethod
|
|
def apply_formatting(text: str, annotations: dict[str, Any], link: Optional[str] = None) -> str:
|
|
"""Apply formatting to a text string based on the annotations.
|
|
Used when converting rich text to markdown
|
|
|
|
Args:
|
|
text (str): The text to format.
|
|
annotations (dict[str, Any]): The annotations to apply to the text.
|
|
link (Optional[str]): An optional link for a hyperlink.
|
|
|
|
Returns:
|
|
str: The formatted text.
|
|
"""
|
|
# If code block, wrap in backticks and skip other formatting.
|
|
if annotations.get("code"):
|
|
return f"`{text}`"
|
|
|
|
# Add underline
|
|
if annotations.get("underline"):
|
|
text = f"<u>{text}</u>"
|
|
|
|
# Apply color
|
|
color = annotations.get("color", "default")
|
|
if color != "default":
|
|
text = f'<span style="color: {color};">{text}</span>'
|
|
|
|
# Add bold, italic, and strikethrough
|
|
markers = [
|
|
marker
|
|
for key, marker in (("bold", "**"), ("italic", "*"), ("strikethrough", "~~"))
|
|
if annotations.get(key)
|
|
]
|
|
if markers:
|
|
text = "".join(markers) + text + "".join(reversed(markers))
|
|
|
|
# Add hyperlink
|
|
if link:
|
|
text = f"[{text}]({link})"
|
|
|
|
return text
|
|
|
|
def _get_plaintext(self, block: dict[str, Any]) -> str:
|
|
"""
|
|
Extract and return the plain text from a Notion block.
|
|
This acts as a fallback for unsupported block types.
|
|
"""
|
|
block_type: str = block.get("type", "")
|
|
content = block.get(block_type, {})
|
|
if isinstance(content, dict):
|
|
rich_text_items = content.get("rich_text", [])
|
|
return "".join(item.get("plain_text", "") for item in rich_text_items)
|
|
return ""
|
|
|
|
def _convert_text_block(self, block: dict[str, Any], element_key: str, prefix: str = "") -> str:
|
|
"""
|
|
Helper method to convert a Notion block's rich_text element into a Markdown string.
|
|
Optionally, a prefix (like a markdown list marker or heading hashes) is added.
|
|
"""
|
|
element = block.get(element_key, {})
|
|
rich_text_items = element.get("rich_text", [])
|
|
text = self.rich_text_to_markdown(rich_text_items)
|
|
return f"{prefix}{text} \n"
|
|
|
|
async def _convert_child_page(self, block: dict[str, Any]) -> str:
|
|
"""
|
|
Asynchronously convert a child page block. This requires fetching the page's URL.
|
|
"""
|
|
page_url = await get_page_url(self.context, block.get("id", ""))
|
|
child_page = block.get("child_page", {})
|
|
rich_text_items = child_page.get("rich_text", [])
|
|
if rich_text_items:
|
|
title = self.rich_text_to_markdown(rich_text_items)
|
|
else:
|
|
title = child_page.get("title", "")
|
|
return f"[{title}]({page_url}) \n"
|
|
|
|
def _convert_bulleted_list_item(self, block: dict[str, Any]) -> str:
|
|
return self._convert_text_block(block, "bulleted_list_item", "- ")
|
|
|
|
def _convert_equation(self, block: dict[str, Any]) -> str:
|
|
expression = block.get("equation", {}).get("expression", "")
|
|
return f"$$ {expression} $$ \n"
|
|
|
|
def _convert_heading_1(self, block: dict[str, Any]) -> str:
|
|
return self._convert_text_block(block, "heading_1", "# ")
|
|
|
|
def _convert_heading_2(self, block: dict[str, Any]) -> str:
|
|
return self._convert_text_block(block, "heading_2", "## ")
|
|
|
|
def _convert_heading_3(self, block: dict[str, Any]) -> str:
|
|
return self._convert_text_block(block, "heading_3", "### ")
|
|
|
|
def _convert_link_preview(self, block: dict[str, Any]) -> str:
|
|
return self._convert_text_block(block, "link_preview")
|
|
|
|
def _convert_numbered_list_item(self, block: dict[str, Any]) -> str:
|
|
return self._convert_text_block(block, "numbered_list_item", "1. ")
|
|
|
|
def _convert_paragraph(self, block: dict[str, Any]) -> str:
|
|
return self._convert_text_block(block, "paragraph")
|