import posixpath
import re
from urllib.parse import urlparse

import bleach
from django.conf import settings
from markdown import markdown

ALLOWED_TAGS = [
    "p",
    "br",
    "strong",
    "em",
    "ul",
    "ol",
    "li",
    "table",
    "thead",
    "tbody",
    "tr",
    "th",
    "td",
    "h1",
    "h2",
    "h3",
    "h4",
    "a",
    "code",
    "pre",
    "blockquote",
    "hr",
    "img",
    "span",
    "div",
]

ALLOWED_ATTRIBUTES = {
    "a": ["href", "title"],
    "img": ["src", "alt", "title"],
    "th": ["align"],
    "td": ["align"],
}

ALLOWED_PROTOCOLS = ["http", "https", "mailto"]

EXTENSIONS = ["tables", "fenced_code", "sane_lists"]

WIKI_IMAGE_PREFIXES = ("/buyer-guide/", "/ops-guide/")

_IMAGE_SRC_RE = re.compile(r'(<img\b[^>]*?\bsrc=")([^"]*)(")', re.IGNORECASE)
_ANCHOR_HREF_RE = re.compile(r'(<a\b[^>]*?\bhref=")([^"]*)(")', re.IGNORECASE)


def _rewrite_anchor(match, base_folder, help_root):
    prefix, href, suffix = match.groups()
    parsed = urlparse(href)
    if parsed.scheme or href.startswith(("#", "/")):
        return match.group(0)
    resolved = posixpath.normpath(posixpath.join(base_folder, href))
    if resolved.startswith("user-guides/"):
        suffix_path = resolved[len("user-guides/"):]
        return f'{prefix}{help_root}{suffix_path}/{suffix}'
    return match.group(0)


def _rewrite_image(match):
    prefix, src, suffix = match.groups()
    if src.startswith(WIKI_IMAGE_PREFIXES):
        return f'{prefix}{settings.WIKI_PUBLIC_BASE}{src}{suffix}'
    return match.group(0)


def render_markdown(content, page_path, help_root):
    html = markdown(content or "", extensions=EXTENSIONS)
    base_folder = posixpath.dirname(page_path)
    html = _ANCHOR_HREF_RE.sub(lambda m: _rewrite_anchor(m, base_folder, help_root), html)
    html = _IMAGE_SRC_RE.sub(_rewrite_image, html)
    return bleach.clean(
        html,
        tags=ALLOWED_TAGS,
        attributes=ALLOWED_ATTRIBUTES,
        protocols=ALLOWED_PROTOCOLS,
    )