get_page

Instructions

Implementation Reference

• src/of_mcp/server.py:77-101 (handler)

  The MCP tool handler for 'get_page'. Accepts page_ref (Confluence id, exact title, or substring) and optional section heading. Looks up the page via db.find_page_by_slug_or_title, then optionally slices to a heading section via _slice_section.

  @mcp.tool()
  def get_page(page_ref: str, section: str | None = None) -> str:
      """Return the full markdown of a page.
  
      Args:
          page_ref: Confluence page id, exact title, or substring of the title.
          section: Optional heading text. If provided, only the matching section
                   (and its sub-sections) is returned.
      """
      with closing(_conn()) as conn:
          row = db.find_page_by_slug_or_title(conn, page_ref)
      if not row:
          return f"No page found for {page_ref!r}."
      header = (
          f"# {row['title']}\n\n"
          f"_Breadcrumb:_ {row['breadcrumb']}\n"
          f"_URL:_ {row['url']}\n"
          f"_Page id:_ {row['id']}\n\n---\n\n"
      )
      body = row["body_md"]
      if section:
          body = _slice_section(body, section)
          if not body:
              return header + f"_(section {section!r} not found in page)_"
      return header + body

• src/of_mcp/server.py:77-78 (registration)

  The @mcp.tool() decorator registers get_page as an MCP tool on the FastMCP server instance (line 34: mcp = FastMCP('of-mcp')).

  @mcp.tool()
  def get_page(page_ref: str, section: str | None = None) -> str:

• src/of_mcp/server.py:150-174 (helper)

  Helper function _slice_section used by get_page to extract a single heading section (and its sub-sections) from the markdown body, based on case-insensitive heading text matching.

  def _slice_section(body_md: str, section: str) -> str:
      """Return the heading whose text contains `section` (case-insensitive),
      plus everything until the next heading of equal or higher level.
      """
      import re
  
      needle = section.strip().lower()
      lines = body_md.splitlines()
      start = -1
      start_level = 0
      for i, line in enumerate(lines):
          m = re.match(r"^(#{1,6})\s+(.*?)\s*$", line)
          if m and needle in m.group(2).lower():
              start = i
              start_level = len(m.group(1))
              break
      if start < 0:
          return ""
      end = len(lines)
      for j in range(start + 1, len(lines)):
          m = re.match(r"^(#{1,6})\s+", lines[j])
          if m and len(m.group(1)) <= start_level:
              end = j
              break
      return "\n".join(lines[start:end]).strip()

• src/of_mcp/db.py:180-191 (helper)

  Database helper find_page_by_slug_or_title used by get_page handler. Attempts lookup by exact id, exact title, then substring (LIKE) match on title.

  def find_page_by_slug_or_title(conn: sqlite3.Connection, needle: str) -> sqlite3.Row | None:
      # 1) exact id, 2) exact title, 3) LIKE on title
      row = conn.execute("SELECT * FROM pages WHERE id = ? LIMIT 1", (needle,)).fetchone()
      if row:
          return row
      row = conn.execute("SELECT * FROM pages WHERE title = ? LIMIT 1", (needle,)).fetchone()
      if row:
          return row
      return conn.execute(
          "SELECT * FROM pages WHERE title LIKE ? ORDER BY length(title) ASC LIMIT 1",
          (f"%{needle}%",),
      ).fetchone()

• src/of_mcp/db.py:24-35 (schema)

  The SQLite schema for the 'pages' table defines the shape of page data (id, title, url, parent_id, breadcrumb, space_key, version, updated_at, body_md) returned by the get_page tool.

  SCHEMA = """
  CREATE TABLE IF NOT EXISTS pages (
      id            TEXT PRIMARY KEY,           -- Confluence content id
      title         TEXT NOT NULL,
      url           TEXT NOT NULL,
      parent_id     TEXT,                       -- Confluence parent page id (or NULL)
      breadcrumb    TEXT NOT NULL,              -- "Home > Section > Page"
      space_key     TEXT NOT NULL,
      version       INTEGER NOT NULL DEFAULT 0,
      updated_at    TEXT,                       -- ISO8601
      body_md       TEXT NOT NULL,              -- full page as markdown
      fetched_at    TEXT NOT NULL DEFAULT (datetime('now'))