Convert `NETSCAPE-Bookmark-file-1` to plain text; bash script

bookmarks-to-text.sh

#!/usr/bin/env sh
# netscape-to-text.sh -- Convert Netscape (HTML) bookmarks to an indented plain-text list.
# POSIX sh compatible (no gawk extensions). 2 spaces per level. Usage:
#   ./netscape-to-text.sh bookmarks.html > bookmarks.txt
# or: cat bookmarks.html | ./netscape-to-text.sh

file=${1:-/dev/stdin}
depth=0

# Put tags onto separate lines so attributes split across lines are handled.
# Then read line by line and process.
sed -e 's/></>\
</g' "$file" | while IFS= read -r line || [ -n "$line" ]; do
  # trim leading/trailing whitespace (POSIX)
  trimmed="$(printf '%s' "$line" | sed -e 's/^[[:space:]]*//' -e 's/[[:space:]]*$//')"

  case "$trimmed" in
    *"<DL>"*|*"<dl>"*)
      depth=$((depth + 1))
      ;;

    *"</DL>"*|*"</dl>"*)
      if [ "$depth" -gt 0 ]; then
        depth=$((depth - 1))
      fi
      ;;

    *"<H3"*"</H3>"*|*"<h3"*"</h3>"*)
      # extract folder name (case-insensitive pattern)
      folder="$(printf '%s' "$trimmed" | sed -n 's/.*<[Hh]3[^>]*>\(.*\)<\/[Hh]3>.*/\1/p')"
      if [ -n "$folder" ]; then
        # indent width = depth * 2
        printf '%*s[Folder] %s\n' $((depth * 2)) '' "$folder"
      fi
      ;;

    *"<A "*"</A>"*|*"<a "*"</a>"*)
      # extract href and title. This is a best-effort: it matches href="..." and the link text.
      href="$(printf '%s' "$trimmed" | sed -n 's/.*<[Aa][^>]*[Hh][Rr][Ee][Ff]="\([^"]*\)"[^>]*>.*<\/[Aa].*/\1/p')"
      title="$(printf '%s' "$trimmed" | sed -n 's/.*<[Aa][^>]*>\([^<]*\)<\/[Aa].*/\1/p')"

      # If href or title not captured by the above (attributes order variations), try a fallback:
      if [ -z "$href" ]; then
        href="$(printf '%s' "$trimmed" | sed -n 's/.*href='\''\([^'\'']*\)'\''.*<\/[Aa].*/\1/p')"
      fi

      if [ -n "$title" ] || [ -n "$href" ]; then
        printf '%*s' $((depth * 2)) ''
        # Print "title - url" if both present, otherwise whichever exists
        if [ -n "$title" ] && [ -n "$href" ]; then
          printf '%s - %s\n' "$title" "$href"
        elif [ -n "$title" ]; then
          printf '%s\n' "$title"
        else
          printf '%s\n' "$href"
        fi
      fi
      ;;

    *)
      # ignore other lines
      ;;
  esac
done

Save gist to file, e.g. bookmarks-to-text.sh and make executable.

Run with: ./bookmarks-to-text.sh bookmarks.html > bookmarks.txt

N.b. This script was iterated with ChatGPT.