Home Assistant MCP

import logging import re import warnings from bs4 import BeautifulSoup from markdownify import markdownify as md logger = logging.getLogger("mcp-atlassian") class TextPreprocessor: """Handles text preprocessing for Confluence and Jira content.""" def __init__(self, base_url: str, confluence_client=None): self.base_url = base_url.rstrip("/") self.confluence_client = confluence_client def process_html_content(self, html_content: str, space_key: str = "") -> tuple[str, str]: """Process HTML content to replace user refs and page links.""" try: soup = BeautifulSoup(html_content, "html.parser") # Process user mentions user_mentions = soup.find_all("ri:user") for user in user_mentions: account_id = user.get("ri:account-id") if account_id and self.confluence_client: try: # Fetch user info using the Confluence API user_info = self.confluence_client.get_user_details_by_accountid(account_id) display_name = user_info.get("displayName", account_id) # Replace the entire ac:link structure with @mention link_tag = user.find_parent("ac:link") if link_tag: link_tag.replace_with(f"@{display_name}") except Exception as e: logger.warning(f"Could not fetch user info for {account_id}: {e}") # Fallback: just use the account ID link_tag = user.find_parent("ac:link") if link_tag: link_tag.replace_with(f"@user_{account_id}") processed_html = str(soup) processed_markdown = md(processed_html) return processed_html, processed_markdown except Exception as e: logger.error(f"Error in process_html_content: {str(e)}") raise def clean_jira_text(self, text: str) -> str: """ Clean Jira text content by: 1. Processing user mentions and links 2. Converting HTML/wiki markup to markdown """ if not text: return "" # Process user mentions mention_pattern = r"\[~accountid:(.*?)\]" text = self._process_mentions(text, mention_pattern) # Process Jira smart links text = self._process_smart_links(text) # Convert HTML to markdown if needed text = self._convert_html_to_markdown(text) return text.strip() def _process_mentions(self, text: str, pattern: str) -> str: """Process user mentions in text.""" mentions = re.findall(pattern, text) for account_id in mentions: try: # Note: This is a placeholder - actual user fetching should be injected display_name = f"User:{account_id}" text = text.replace(f"[~accountid:{account_id}]", display_name) except Exception as e: logger.error(f"Error getting user info for {account_id}: {str(e)}") return text def _process_smart_links(self, text: str) -> str: """Process Jira/Confluence smart links.""" # Pattern matches: [text|url|smart-link] link_pattern = r"\[(.*?)\|(.*?)\|smart-link\]" matches = re.finditer(link_pattern, text) for match in matches: full_match = match.group(0) link_text = match.group(1) link_url = match.group(2) # Extract issue key if it's a Jira issue link issue_key_match = re.search(r"browse/([A-Z]+-\d+)", link_url) # Check if it's a Confluence wiki link confluence_match = re.search(r"wiki/spaces/.+?/pages/\d+/(.+?)(?:\?|$)", link_url) if issue_key_match: issue_key = issue_key_match.group(1) clean_url = f"{self.base_url}/browse/{issue_key}" text = text.replace(full_match, f"[{issue_key}]({clean_url})") elif confluence_match: url_title = confluence_match.group(1) readable_title = url_title.replace("+", " ") readable_title = re.sub(r"^[A-Z]+-\d+\s+", "", readable_title) text = text.replace(full_match, f"[{readable_title}]({link_url})") else: clean_url = link_url.split("?")[0] text = text.replace(full_match, f"[{link_text}]({clean_url})") return text def _convert_html_to_markdown(self, text: str) -> str: """Convert HTML content to markdown if needed.""" if re.search(r"<[^>]+>", text): try: with warnings.catch_warnings(): warnings.filterwarnings("ignore", category=UserWarning) soup = BeautifulSoup(f"<div>{text}</div>", "html.parser") html = str(soup.div.decode_contents()) if soup.div else text text = md(html) except Exception as e: logger.warning(f"Error converting HTML to markdown: {e}") return text