v0.166.0
  1import os
  2from html.parser import HTMLParser
  3from typing import TYPE_CHECKING, Any, cast
  4from urllib.parse import urlparse, urlunparse
  5
  6import mistune
  7from plain.urls import reverse
  8from plain.utils.text import slugify
  9from pygments import highlight
 10from pygments.formatters import html
 11from pygments.lexers import get_lexer_by_name
 12
 13if TYPE_CHECKING:
 14    from .registry import PagesRegistry
 15
 16
 17class PagesRenderer(mistune.HTMLRenderer):
 18    def __init__(
 19        self, current_page_path: str, pages_registry: PagesRegistry, **kwargs: Any
 20    ):
 21        super().__init__(**kwargs)
 22        self.current_page_path = current_page_path
 23        self.pages_registry = pages_registry
 24
 25    def link(self, text: str, url: str, title: str | None = None) -> str:
 26        """Convert relative markdown links to proper page URLs."""
 27        # Check if it's a relative link (starts with ./ or ../, or is just a filename)
 28        is_relative = url.startswith(("./", "../")) or (
 29            not url.startswith(("http://", "https://", "/", "#")) and ":" not in url
 30        )
 31
 32        if is_relative:
 33            # Parse URL to extract components
 34            parsed_url = urlparse(url)
 35
 36            # Resolve relative to current page's directory using just the path component
 37            current_dir = os.path.dirname(self.current_page_path)
 38            resolved_path = os.path.normpath(os.path.join(current_dir, parsed_url.path))
 39            page = self.pages_registry.get_page_from_path(resolved_path)
 40
 41            # Get the primary URL name for link conversion
 42            url_name = page.get_url_name()
 43            if url_name:
 44                base_url = reverse(f"pages:{url_name}")
 45                # Reconstruct URL with preserved query params and fragment
 46                url = str(
 47                    urlunparse(
 48                        (
 49                            parsed_url.scheme,  # scheme (empty for relative)
 50                            parsed_url.netloc,  # netloc (empty for relative)
 51                            base_url,  # path (our converted URL)
 52                            parsed_url.params,  # params
 53                            parsed_url.query,  # query
 54                            parsed_url.fragment,  # fragment
 55                        )
 56                    )
 57                )
 58
 59        return super().link(text, url, title)
 60
 61    def heading(self, text: str, level: int, **attrs: Any) -> str:
 62        """Automatically add an ID to headings if one is not provided."""
 63
 64        if "id" not in attrs:
 65            inner_text = get_inner_text(text)
 66            inner_text = inner_text.replace(
 67                ".", "-"
 68            )  # Replace dots with hyphens (slugify won't)
 69            attrs["id"] = slugify(inner_text)
 70
 71        return super().heading(text, level, **attrs)
 72
 73    def block_code(self, code: str, info: str | None = None) -> str:
 74        """Highlight code blocks using Pygments."""
 75
 76        if info:
 77            lexer = get_lexer_by_name(info, stripall=True)
 78            formatter = html.HtmlFormatter(wrapcode=True)
 79            return highlight(code, lexer, formatter)
 80
 81        return "<pre><code>" + mistune.escape(code) + "</code></pre>"
 82
 83
 84def render_markdown(content: str, current_page_path: str) -> str:
 85    from .registry import pages_registry
 86
 87    renderer = PagesRenderer(
 88        current_page_path=current_page_path, pages_registry=pages_registry, escape=False
 89    )
 90    markdown = mistune.create_markdown(
 91        renderer=renderer, plugins=["strikethrough", "table"]
 92    )
 93    return cast(str, markdown(content))
 94
 95
 96class InnerTextParser(HTMLParser):
 97    def __init__(self):
 98        super().__init__()
 99        self.text_content: list[str] = []
100
101    def handle_data(self, data: str) -> None:
102        # Collect all text data
103        self.text_content.append(data.strip())
104
105
106def get_inner_text(html_content: str) -> str:
107    parser = InnerTextParser()
108    parser.feed(html_content)
109    return " ".join([text for text in parser.text_content if text])