From bb93e8d7c0a37a38d80b75d55117a75aa53e1de1 Mon Sep 17 00:00:00 2001
From: St33v <github@f3rr3t.com>
Date: Sun, 1 Feb 2026 14:53:04 +1100
Subject: robots / sitemap demo version

---
 forge/script/gen-robots-sitemap.sh | 48 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 48 insertions(+)
 create mode 100755 forge/script/gen-robots-sitemap.sh

(limited to 'forge/script/gen-robots-sitemap.sh')

diff --git a/forge/script/gen-robots-sitemap.sh b/forge/script/gen-robots-sitemap.sh
new file mode 100755
index 0000000..30dbdea
--- /dev/null
+++ b/forge/script/gen-robots-sitemap.sh
@@ -0,0 +1,48 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+OUT_DIR="${1:-.faircamp_build}"
+SITE_URL="${2:-https://st33v.com}"
+
+cd "$OUT_DIR"
+
+# --- robots.txt ---
+cat > robots.txt <<EOF
+User-agent: *
+Allow: /
+
+Sitemap: ${SITE_URL%/}/sitemap.xml
+EOF
+
+# --- sitemap.xml ---
+# Include HTML pages + common content types; exclude obvious junk.
+# If you have multiple languages/hosts, we can expand later.
+tmp="$(mktemp)"
+find . -type f \( -name '*.html' -o -name '*.pdf' -o -name '*.mp3' -o -name '*.flac' \) \
+  ! -path './.git/*' ! -path './assets/*' ! -path './static/*' \
+  -print0 \
+| sort -z \
+| while IFS= read -r -d '' f; do
+    # Turn ./path/index.html into /path/index.html
+    path="${f#./}"
+    # Basic lastmod (UTC) from file mtime
+    lastmod="$(date -u -r "$f" +%Y-%m-%dT%H:%M:%SZ)"
+    printf '%s\t%s\n' "$path" "$lastmod"
+  done > "$tmp"
+
+{
+  printf '%s\n' '<?xml version="1.0" encoding="UTF-8"?>'
+  printf '%s\n' '<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">'
+  while IFS=$'\t' read -r path lastmod; do
+    # Escape ampersands minimally
+    url="${SITE_URL%/}/$(printf '%s' "$path" | sed 's/&/\&amp;/g')"
+    printf '  <url><loc>%s</loc><lastmod>%s</lastmod></url>\n' "$url" "$lastmod"
+  done < "$tmp"
+  printf '%s\n' '</urlset>'
+} > sitemap.xml
+
+rm -f "$tmp"
+
+echo "Wrote: $OUT_DIR/robots.txt"
+echo "Wrote: $OUT_DIR/sitemap.xml"
+
-- 
cgit v1.3