#!/usr/bin/env python3
"""
fix_page_links.py

Run this from the root of a wget-mirrored site (e.g. exhibits-library-jhu-edu/).

Problem it fixes:
    wget's --convert-links + --adjust-extension combo sometimes fails to
    rewrite <a href="..."> links to extensionless pages, even though the
    actual file was saved locally as page.html. This leaves some internal
    navigation links pointing at a URL that doesn't exist locally (e.g.
    href="/exhibits/show/some-exhibit/some-page" instead of
    href="/exhibits/show/some-exhibit/some-page.html"), causing 404s when
    browsing the mirror locally.

What it does:
    1. Walks the whole mirror and indexes every real .html file as a URL
       path (with the .html stripped), e.g. a file at
       ./exhibits/show/foo/bar.html becomes "/exhibits/show/foo/bar".
    2. Walks every .html file again, and for any href="/exhibits/show/..."
       link that exactly matches one of those indexed paths (i.e. the link
       is missing ".html" but a matching page does exist), rewrites it to
       append ".html".
    3. Leaves alone any link that doesn't match a known local page, so it
       won't touch genuinely out-of-scope or external links.

Usage:
    cd exhibits-library-jhu-edu
    python3 fix_page_links.py
"""

import os
import re

root = "."

# Step 1: Build the set of all real local pages (as URL paths, without .html)
pages = set()
for dirpath, _, files in os.walk(root):
    for fn in files:
        if fn.endswith(".html"):
            full = os.path.relpath(os.path.join(dirpath, fn), root).replace(os.sep, "/")
            pages.add("/" + full[:-5])  # strip trailing ".html"

print(f"Indexed {len(pages)} local pages")

# Step 2: Rewrite any href="/exhibits/show/..." link that's missing .html
# but matches a page we actually have.
link_re = re.compile(r'href="(/exhibits/show/[^"?#]+)"')

fixed_count = 0
for dirpath, _, files in os.walk(root):
    for fn in files:
        if fn.endswith(".html"):
            full = os.path.join(dirpath, fn)
            with open(full, "r", errors="ignore") as f:
                content = f.read()

            changed = [False]  # mutable container so repl() can update it without `nonlocal`

            def repl(m, changed=changed):
                path = m.group(1)
                if path in pages:
                    changed[0] = True
                    return f'href="{path}.html"'
                return m.group(0)

            new_content = link_re.sub(repl, content)

            if changed[0]:
                with open(full, "w") as f:
                    f.write(new_content)
                fixed_count += 1

print(f"Fixed links in {fixed_count} file(s)")