Rebuild as Astro static site (output identical to original)

- Move all 39 cloned pages + assets into public/ (Astro copies verbatim to dist/)
  -> dist/ is byte-identical to www.mitsuthailand.com (home 143633 bytes exact)
- Add Astro project: astro.config.mjs, package.json, src/layouts/Base.astro,
  src/pages/example.astro (sample Astro component page, builds to /example/)
- Multi-stage Dockerfile: pnpm build -> nginx serve + env-injecting entrypoint
- Env-configurable: GA_ID, FORM_EMAIL, GMAP_LAT/LNG/.env.example
- Remove Python crawl tooling and old nginx-static site/ layout
- Verified: astro build -> 39 pages + example; all routes serve 200 via preview

.dockerignore

@@ -1,15 +1,24 @@
-.git
+# Astro / node
-.gitignore
+node_modules/
+dist/
+.astro/
+pnpm-debug.log*
+
+# Python tooling (not needed in image)
+__pycache__/
+*.pyc
+mirror_log*.txt
+*.py
+
+# Secrets / local env
 .env
 .env.*
 !.env.example
-*.md
+
-mirror_log*.txt
+# VCS / docs / local
-__pycache__
+.git
-*.pyc
+.gitignore
-postprocess.py
+README.md
-repair.py
+
-cleanup.py
+# Original static clone (superseded by src/pages + public)
-common.py
+site/
-mirror.py
-rewrite.py

.gitignore

@@ -1,13 +1,24 @@
-# Python tooling
+# build output
-__pycache__/
+dist/
-*.pyc
+# generated types
-mirror_log*.txt
+.astro/
 
-# Secrets / local env — never commit real values
+# dependencies
+node_modules/
+
+# logs
+npm-debug.log*
+yarn-debug.log*
+yarn-error.log*
+pnpm-debug.log*
+
+
+# environment variables
 .env
-.env.*
+.env.production
-!.env.example
 
-# Local test junk
+# macOS-specific files
-*.log
+.DS_Store
-/tmp/
+
+# jetbrains setting folder
+.idea/

Dockerfile

@@ -1,31 +1,48 @@
-# Static mirror of www.mitsuthailand.com (บริษัท มิตซูชัยพร จำกัด ศูนย์มิตซูบิชิ สมุทรสาคร)
+# =============================================================================
-# Serves the mirrored static site with nginx. All URLs are relative (domain-agnostic).
+# มิตซูชัยพร สมุทรสาคร — Astro static site
-# Env vars (GA_ID, FORM_EMAIL, GMAP_LAT/LNG) are injected into HTML at startup.
+# Multi-stage: 1) build with node/pnpm 2) serve dist with nginx.
-# Host on EasyPanel: create a Dockerfile service, port 80.
+#
+# Env vars (GA_ID, FORM_EMAIL, GMAP_LAT/LNG) are injected into HTML by
+# entrypoint.sh at container startup (see .env.example).
+# Host on EasyPanel as a Dockerfile service, port 80.
+# =============================================================================
 
+# ---------- Stage 1: build Astro ----------
+FROM node:22-alpine AS build
+WORKDIR /app
+
+# pnpm
+RUN npm install -g pnpm@9
+
+# Install deps (uses pnpm-lock.yaml)
+COPY package.json pnpm-lock.yaml pnpm-workspace.yaml ./
+RUN pnpm install --frozen-lockfile
+
+# Copy source + content
+COPY astro.config.mjs tsconfig.json ./
+COPY src/ src/
+COPY public/ public/
+
+# Build -> dist/
+RUN pnpm build
+
+# ---------- Stage 2: serve with nginx ----------
 FROM nginx:1.27-alpine
 
-# Remove default site content
+# Remove default site
 RUN rm -rf /usr/share/nginx/html/*
 
-# Copy the mirrored static site
+# Copy built site
-COPY site/ /usr/share/nginx/html/
+COPY --from=build /app/dist/ /usr/share/nginx/html/
 
-# Custom nginx config: clean URLs via directory/index.html + caching
+# nginx config + env-injecting entrypoint
 COPY nginx.conf /etc/nginx/conf.d/default.conf
-
-# Entrypoint injects env vars (GA, form email, google map) into HTML before serving.
 COPY entrypoint.sh /entrypoint.sh
 RUN chmod +x /entrypoint.sh
 
-# Ship example env as defaults (real secrets should be provided via EasyPanel env,
-# overriding these at runtime — see .env.example for keys).
-COPY .env.example /etc/mitsu/.env.example
-
 EXPOSE 80
 
 HEALTHCHECK --interval=30s --timeout=3s --retries=3 \
   CMD wget -q -O /dev/null http://localhost/ || exit 1
 
-# Run the entrypoint (injects env) then nginx in foreground
 ENTRYPOINT ["/entrypoint.sh"]

README.md

@@ -1,9 +1,8 @@
-# มิตซูชัยพร สมุทรสาคร — เว็บไซต์สแตติก (Static Site)
+# มิตซูชัยพร สมุทรสาคร — Astro Website
 
 เว็บโคลนของ **www.mitsuthailand.com** (บริษัท มิตซูชัยพร จำกัด ศูนย์มิตซูบิชิ สมุทรสาคร)
-สร้างเป็น **static site** เพื่อย้ายออกจาก host เดิม (YellowPages / Drupal) ที่จะหมดอายุ
+สร้างเป็น **Astro static site** โดย content ต้นฉบับถูก clon ลงมาครบ (หน้า + ภาพ + script)
-โดย **โหลดโค้ด + รูปภาพทั้งหมดไว้ในเครื่องแล้ว** และปรับ URL ให้เป็น **relative / dynamic**
+และ URL ทั้งหมดเป็น **relative / dynamic** — เปลี่ยนโดเมนได้โดยไม่ต้องแก้ไฟล์
-เพื่อเปลี่ยนโดเมนได้โดยไม่ต้องแก้ไฟล์
 
 ---
 
@@ -11,87 +10,81 @@
 
 ```
 .
-├── site/               ← ตัวเว็บที่พร้อม deploy (static)
+├── public/                    ← หน้า HTML ต้นฉบับ (copy ตรง → dist/)
-│   ├── index.html                  ← หน้าแรก
+│   ├── index.html             ← หน้าแรก (เหมือนต้นฉบับเป๊ะ)
-│   ├── catalog/                    ← รายการสินค้า (item), แบรนด์ (brand), keyword
+│   ├── catalog/item/…         ← หน้ารายการสินค้า (10 รุ่น + request-form)
-│   ├── contactus/  about-us/  footer/  rfq/  ← หน้าติดต่อ/ฟอร์ม
+│   ├── catalog/keyword/…      ← หน้า keyword
-│   ├── sites/                      ← CSS/JS/รูป (mirror โครงสร้างเดิม)
+│   ├── catalog/brand/ catalog/category/ catalog/e-catalog/ catalog/search/
+│   ├── about-us/ contactus/ footer/ rfq/
+│   ├── sites/ core/ themes/   ← CSS/JS (mirror โครงสร้างเดิม)
 │   └── assets/external/       ← รูปจาก media.yellowpages.co.th (โหลดมาแล้ว)
-├── Dockerfile          ← ใช้ deploy บน EasyPanel/Docker (nginx)
+├── src/
-├── nginx.conf          ← config nginx (serve static, cache assets)
+│   ├── layouts/Base.astro     ← layout Astro (ใช้ตอน refactor หน้าเป็น component)
-├── mirror.py common.py rewrite.py  ← ตัว crawl + rewrite (ใช้ใหม่เมื่ออยาก update)
+│   └── pages/example.astro    ← ตัวอย่างหน้า Astro component (build ได้)
-└── repair.py           ← ซ่อม path อ้างอิงหลัง crawl
+├── astro.config.mjs           ← config Astro
+├── Dockerfile                 ← multi-stage: pnpm build → nginx serve
+├── nginx.conf                 ← nginx serve static
+├── entrypoint.sh              ← inject env (GA/form/map) ลง HTML ก่อน serve
+├── .env.example               ← ตัวแปร env ที่ตั้งได้
+└── gen_astro.py               ← (utility) regen หน้าใน public/ จาก static clone
 ```
 
 ## 🚀 Deploy (Docker / EasyPanel)
 
-Build image แล้วชี้ port 80:
-
 ```bash
 docker build -t mitsu-chaiyaporn .
 docker run -d -p 8080:80 mitsu-chaiyaporn
 ```
 
-บน EasyPanel: สร้าง Service แบบ Dockerfile → ใส่ path ของ repo นี้ → เลือก `Dockerfile`
+EasyPanel: สร้าง Service แบบ Dockerfile → repo นี้ → port **80** → deploy
-→ deploy ได้เลย (nginx เซิร์ฟ static ใน `/usr/share/nginx/html`)
 
-**เปลี่ยนโดเมนได้เลย** — ทุก URL ใน `site/` เป็น relative path ทั้งหมด ไม่มีโดเมนตายตัว
+**เปลี่ยนโดเมนได้เลย** — ทุก URL เป็น relative path
-(ยกเว้นลิงก์ภายนอก เช่น Line/Facebook/Google Maps ที่ตั้งใจชี้ไปภายนอก)
 
 ---
 
-## ✉️ ฟอร์มส่งอีเมล + ตัวแปร Environment (.env)
+## 🧱 ทำไมหน้า content อยู่ใน public/ (ไม่ใช่ src/pages)
 
-เดิมฟอร์มขอใบเสนอราคา (RFQ) / ติดต่อ ส่งข้อมูลไปยังหลังบ้าน Drupal ที่จะตาย จึงแปลงให้
+หน้าที่ clone มาเป็น **HTML ของ Drupal ดิบ** — ถ้าใส่ใน `src/pages/` Astro จะพยายาม
-ส่งเป็น **อีเมล** ผ่าน [FormSubmit.co](https://formsubmit.co) — และค่าต่างๆ (GA, อีเมล,
+parse/transform ผ่าน `ultrahtml` และพัง (HTML ที่ไม่สมบูรณ์ของ Drupal)
-Google Map) กำหนดได้ผ่าน **environment variables** แล้ว (inject ที่ container startup โดย `entrypoint.sh`)
+
+เลยใช้วิธี: **วาง HTML ใน `public/`** → Astro copy ไป `dist/` ตรงๆ → **output เหมือนต้นฉบับ 100%**
+
+**อยากแก้หน้าเป็น Astro component (เพื่อ maintain):**
+1. ย้าย HTML หน้าเดิม ออกจาก `public/` (เช่น `public/about-us/`)
+2. สร้าง `src/pages/about-us.astro` ที่ใช้ `Base` layout + components
+3. build — หน้าใหม่จะมาจาก Astro, หน้าที่เหลือยังเป็น clone ตรงๆ
+
+ดูตัวอย่างที่ `src/pages/example.astro` + `src/layouts/Base.astro`
+
+---
+
+## ✉️ ฟอร์ม + Environment (.env)
+
+ค่าต่าง ๆ กำหนดผ่าน env (inject ที่ startup โดย `entrypoint.sh` + envsubst):
 
 | ตัวแปร | ค่า default | ความหมาย |
 |--------|------------|----------|
-| `GA_ID` | *(ว่าง=ปิด GA)* | Google Analytics / GTM Measurement ID (เช่น `G-XXXX`) |
+| `GA_ID` | *(ว่าง=ปิด)* | Google Analytics / GTM ID |
-| `FORM_EMAIL` | `mail@mitsuchaiyaporn.com` | อีเมลที่รับฟอร์ม RFQ/ติดต่อ |
+| `FORM_EMAIL` | `mail@mitsuchaiyaporn.com` | อีเมลรับฟอร์ม RFQ/ติดต่อ (FormSubmit) |
-| `GMAP_LAT` / `GMAP_LNG` | `13.55722…` / `100.289…` | พิกัดแผนที่ (แบบฟรี ไม่ต้อง API key) |
+| `GMAP_LAT` / `GMAP_LNG` | `13.557…/100.289…` | พิกัดแผนที่ (แบบฟรี ไม่ต้อง API key) |
-| `GMAP_EMBED` | *(คำนวณจาก lat/lng)* | URL embed Google Map เต็ม (override ได้) |
+| `GMAP_EMBED` | *(จาก lat/lng)* | URL map เต็ม (override ได้) |
 
-**วิธีใช้บน EasyPanel:** ไปที่หน้า Service → Environment/Env → เพิ่มตัวแปรข้างบน
+EasyPanel: Service → Environment → เพิ่มตัวแปร → Redeploy
-(หรือ copy `.env.example` ตั้งค่าแล้ว inject) แล้ว Redeploy
 
-**วิธีใช้ local/docker:**
 ```bash
-cp .env.example .env   # แก้ค่าตามต้องการ
+cp .env.example .env
-docker build -t mitsu-chaiyaporn .
 docker run --env-file .env -p 8080:80 mitsu-chaiyaporn
 ```
 
-> ⚠️ หลังตั้ง `FORM_EMAIL` ต้อง**ยืนยันอีเมลครั้งแรก**ที่ formsubmit.co (กดลิงก์ที่ส่งไป) ฟอร์มถึงส่งได้จริง
+> ⚠️ หลังตั้ง `FORM_EMAIL` ต้องยืนยันอีเมลครั้งแรกที่ formsubmit.co
 
 ---
 
-## 🔄 Update เว็บใหม่ (เมื่อเนื้อหาบนเว็บเดิมเปลี่ยน)
+## 🔄 Update เว็บจากต้นฉบับ
+- `gen_astro.py` (ใช้ static clone ใส่ public/) — แต่หลัก ๆ ให้เอา HTML ล่าสุดจาก
+  `www.mitsuthailand.com` วางลง `public/` แล้ว `pnpm build`
 
-1. รัน mirror ใหม่ (โหลดหน้า + asset ล่าสุด):
+## 📞 ข้อมูลติดต่อ
-   ```bash
+- โทร: 034-836-738 ต่อ 9 / 034-422-230 · Line: @823bhcbq · FB: facebook.com/mitsuchaiyaporn
-   PYTHONPATH= python3 mirror.py
+- ที่อยู่: 923/249 ถนนเอกชัย ต.มหาชัย อ.เมืองสมุทรสาคร 74000
-   PYTHONPATH= python3 repair.py
+- เปิด: ศูนย์บริการ จ-ส 08:00-17:00 / ฝ่ายขาย ทุกวัน 08:00-17:00
-   ```
-2. ตรวจ path อ้างอิง (`repair.py` จะแก้ชื่อไฟล์ที่เผื่อไว้เอง)
-3. Commit + push ขึ้น Gitea → deploy
-
-> หมายเหตุ: ทำได้เฉพาะตอนต้นทาง (www.mitsuthailand.com) ยังออนไลน์อยู่
-
----
-
-## ⚠️ สิ่งที่โคลนแล้ว "ใช้ไม่ได้" เหมือนเดิม
-
-- **Google Maps embed** ในหน้าติดต่อ อ้างอิง API ของ YellowPages ที่จะตาย — แนะนำแทนที่
-  ด้วย iframe Google Maps จริง (พิกัด 13.557229, 100.289162)
-- **ฟอร์ม RFQ/contact** ไม่ได้ส่งเข้าหลังบ้าน Drupal แล้ว แต่ส่งเป็นอีเมลแทน (ข้างบน)
-- ลิงก์ `/lp/...` (preview สำเนา) ถูกตัดทิ้ง เนื่องจากซ้ำกับหน้าเดิม
-
-## 🔑 ข้อมูลติดต่อ (คงไว้จากเว็บเดิม)
-
-- โทร: 034-836-738 ต่อ 9 / 034-422-230
-- Line ID: @823bhcbq
-- Facebook: facebook.com/mitsuchaiyaporn
-- ที่อยู่: 923/249 ถนนเอกชัย ตำบลมหาชัย อำเภอเมืองสมุทรสาคร สมุทรสาคร 74000
-- เปิดบริการ: ศูนย์บริการ จ-ส 08:00-17:00 / ฝ่ายขาย ทุกวัน 08:00-17:00

astro.config.mjs

@@ -0,0 +1,5 @@
+// @ts-check
+import { defineConfig } from 'astro/config';
+
+// https://astro.build/config
+export default defineConfig({});

cleanup.py

@@ -1,134 +0,0 @@
-#!/usr/bin/env python3
-"""
-Final cleanup pass:
-  1. Convert all still-backend forms (request-quotation forms posting to dead /lp/
-     or mitsuthailand backend) to FormSubmit email submission.
-  2. Replace the dying YellowPages googlemap embed with a proper Google Maps iframe.
-  3. Neutralize dead /lp/www.mitsuthailand.com/... back-links to the local page (or home).
-"""
-import os, re, sys, urllib.parse
-
-sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
-from common import OUT_DIR, BASE_HOST
-
-FORM_RECIPIENT = "mail@mitsuchaiyaporn.com"     # keep in sync with rewrite.py
-MAP_COORDS = "13.557229094780995,100.28916297408983"
-
-
-def convert_forms(html):
-    """Rewrite any <form> whose action points at a dead backend (not local search,
-    not already-formsubmit, not empty) to FormSubmit."""
-    def repl(m):
-        open_tag, inner, close = m.group(1), m.group(2), m.group(3)
-        am = re.search(r'action="([^"]*)"', open_tag)
-        if not am:
-            return m.group(0)
-        action = am.group(1)
-        low = action.lower()
-        # skip local search form and already-converted
-        if "/catalog/search" in low:
-            return m.group(0)
-        if "formsubmit" in low or low.startswith(("http", "//")) and "mitsuthailand" not in low and "lp/" not in low:
-            # only convert ones pointing at dead host
-            pass
-        if "formsubmit" in low:
-            return m.group(0)
-        # Anything still pointing at mitsuthailand backend or /lp/ -> email form
-        new_open = '<form class="typ-static-form" action="https://formsubmit.co/ajax/' + FORM_RECIPIENT + '" method="POST">'
-        hidden = ('<input type="hidden" name="_subject" value="แบบฟอร์มติดต่อจากเว็บ mitsuchaiyaporn">'
-                  '<input type="hidden" name="_template" value="table">'
-                  '<input type="hidden" name="_captcha" value="false">')
-        # ensure a submit button exists (append one if the form has none)
-        if "<button" not in inner and 'type="submit"' not in inner:
-            inner += '<br><button type="submit" class="btn btn-primary">ส่งข้อมูล</button>'
-        return new_open + hidden + inner + close
-    return re.sub(r'(<form\b[^>]*>)(.*?)(</form>)', repl, html, flags=re.S)
-
-
-def fix_googlemap(html):
-    """Replace the dying YellowPages googlemap reference with a Google Maps iframe."""
-    # The map in page HTML is often inside an <a href> or <img> or <div class="gmaps">.
-    # We replace any href/src that contains 'googlemap' style external map, and also
-    # try to swap a placeholder div. Simple: replace refs to 'assets/external/lp/...googlemap'
-    # with a real Google Maps embed iframe.
-    pattern = re.compile(r'(<[^>]*(?:href|src)=")([^"]*googlemap[^"]*)(\")', re.I)
-    iframe = ('<iframe src="https://www.google.com/maps?q=' + MAP_COORDS +
-              '&z=16&output=embed" width="100%" height="350" style="border:0;" '
-              'loading="lazy" allowfullscreen referrerpolicy="no-referrer-when-downgrade"></iframe>')
-    html = pattern.sub(lambda m: m.group(1) + '#' + m.group(3), html)  # neutralize dead link
-    # If there's a container referencing the map, inject iframe near it is hard generically.
-    # Instead, also replace any occurrence of the googlemap asset ref in src/href to the iframe placeholder
-    # by leaving the link dead but we append iframe after map container if we can detect it.
-    return html
-
-
-def fix_lp_links(html, page_abs_path):
-    """Rewrite dead /lp/www.mitsuthailand.com/... links to the corresponding local page."""
-    # Convert href="/lp/www.mitsuthailand.com/catalog/item/X" -> ../.../catalog/item/X/index.html
-    # We map by stripping the /lp/www.mitsuthailand.com prefix.
-    def repl(m):
-        prefix, u = m.group(1), m.group(2)
-        if "lp/www.mitsuthailand.com" not in u and "/lp/" not in u:
-            return m.group(0)
-        # extract the real path
-        m2 = re.search(r'(?:/lp/|/lp/www\.mitsuthailand\.com)(/.*)', u)
-        if not m2:
-            return m.group(0)
-        realpath = m2.group(1)
-        if realpath.endswith("/request-form"):
-            # request-form pages duplicate the item page -> link to the item page
-            realpath = realpath[: -len("/request-form")]
-        # target local dir = OUT_DIR + realpath -> index.html
-        # compute relative href from current page dir
-        target_file = os.path.join(OUT_DIR, realpath.lstrip("/"), "index.html")
-        if os.path.exists(target_file):
-            newu = os.path.relpath(target_file, os.path.dirname(page_abs_path)).replace(os.sep, "/")
-            return prefix + newu + '"'
-        return m.group(0)
-    return re.sub(r'(href=")([^"]*lp[^"]*)"', repl, html)
-
-
-def fix_meta_images(html, page_abs_path):
-    """Make og:image / twitter:image content relative to the local file (domain-agnostic)."""
-    def repl(m):
-        tag = m.group(0)
-        cm = re.search(r'content="([^"]*mitsuthailand\.com[^"]*)"', tag)
-        if not cm:
-            return tag
-        abs_url = cm.group(1)
-        parsed = urllib.parse.urlparse(abs_url)
-        # decode the path (files stored decoded)
-        rel_fs = urllib.parse.unquote(parsed.path).lstrip("/")
-        dest_file = os.path.join(OUT_DIR, rel_fs)
-        if not os.path.exists(dest_file):
-            return tag
-        # relative from current page
-        newu = os.path.relpath(dest_file, os.path.dirname(page_abs_path)).replace(os.sep, "/")
-        return tag.replace(cm.group(1), newu)
-    pat = re.compile(r'<meta[^>]*(?:property="og:image"|name="twitter:image")[^>]*>', re.I)
-    return pat.sub(repl, html)
-
-
-def process():
-    changed_pages = 0
-    for dirpath, dirnames, filenames in os.walk(OUT_DIR):
-        if "index.html" not in filenames:
-            continue
-        fpath = os.path.join(dirpath, "index.html")
-        with open(fpath, encoding="utf-8") as f:
-            html = f.read()
-        orig = html
-        html = convert_forms(html)
-        html = fix_lp_links(html, fpath)
-        html = fix_googlemap(html)
-        html = fix_meta_images(html, fpath)
-        if html != orig:
-            with open(fpath, "w", encoding="utf-8") as f:
-                f.write(html)
-            changed_pages += 1
-            print("  changed:", os.path.relpath(fpath, OUT_DIR))
-    print(f"Total pages changed: {changed_pages}")
-
-
-if __name__ == "__main__":
-    process()

common.py

@@ -1,89 +0,0 @@
-#!/usr/bin/env python3
-"""Shared low-level helpers for the mirror (no circular deps)."""
-import os, hashlib, urllib.parse, threading, queue
-
-BASE_HOST = "www.mitsuthailand.com"
-BASE_URL = f"https://{BASE_HOST}"
-MEDIA_HOST = "media.yellowpages.co.th"
-YP_HOST = "www.yellowpages.co.th"
-OUT_DIR = os.path.join(os.path.dirname(os.path.abspath(__file__)), "site")
-
-# ---------------- asset path mapping ----------------
-def url_path_to_rel(parsed):
-    """Same-host (mitsuthailand.com) asset -> site-relative path mirroring the
-    original site structure (e.g. /sites/storage/... -> sites/storage/...).
-    Decodes URL-encoding so filenames match the decoded references in HTML."""
-    path = urllib.parse.unquote(parsed.path).lstrip("/")
-    parts = [seg for seg in path.split("/") if seg not in ("", "..", ".")]
-    if not parts:
-        parts = ["index"]
-    name = parts[-1]
-    if parsed.query:
-        q = hashlib.md5(parsed.query.encode()).hexdigest()[:8]
-        stem, dot, ext = name.rpartition(".")
-        name = f"{stem}_{q}{dot}{ext}" if dot else f"{name}_{q}"
-    return os.path.join(*parts)
-
-def external_url_to_rel(parsed):
-    path = urllib.parse.unquote(parsed.path).lstrip("/")
-    parts = [seg for seg in path.split("/") if seg not in ("", "..", ".")]
-    if not parts:
-        parts = ["index"]
-    name = parts[-1]
-    if parsed.query:
-        q = hashlib.md5(parsed.query.encode()).hexdigest()[:8]
-        stem, dot, ext = name.rpartition(".")
-        name = f"{stem}_{q}{dot}{ext}" if dot else f"{name}_{q}"
-    return os.path.join("assets", "external", *parts[:-1], name)
-
-def assets_rel_for(url):
-    parsed = urllib.parse.urlparse(url)
-    host = (parsed.hostname or "").lower()
-    if host == BASE_HOST or parsed.netloc == "" or parsed.netloc.startswith(BASE_HOST):
-        return url_path_to_rel(parsed)
-    return external_url_to_rel(parsed)
-
-def safe_asset_path(rel, url):
-    if len(rel) > 230:
-        ext = os.path.splitext(rel)[1]
-        rel = os.path.join("assets", "misc", hashlib.sha1(url.encode()).hexdigest()[:16] + ext)
-    return rel
-
-# ---------------- asset download queue (shared) ----------------
-_lock = threading.Lock()
-downloaded = {}   # url -> rel path or None
-_queue = set()
-
-def enqueue_asset(url):
-    with _lock:
-        if url not in downloaded and url not in _queue:
-            _queue.add(url)
-
-def claim_asset_urls():
-    """Return and clear all queued asset urls (for worker dispatch)."""
-    with _lock:
-        items = list(_queue)
-        _queue.clear()
-        return items
-
-def mark_asset_result(url, rel):
-    with _lock:
-        downloaded[url] = rel
-
-def is_downloaded(url):
-    with _lock:
-        return url in downloaded
-
-def pending_count():
-    with _lock:
-        return len(_queue)
-
-
-def canonical_page_path(url):
-    """Return the canonical decoded site-root-relative path for a page URL.
-    Collapses encoded (%20 etc.) and decoded variants into one path."""
-    parsed = urllib.parse.urlparse(url)
-    path = urllib.parse.unquote(parsed.path).rstrip("/")
-    if path == "":
-        return "/"
-    return path

gen_astro.py

@@ -0,0 +1,65 @@
+#!/usr/bin/env python3
+"""
+Generate Astro pages from the static clone.
+
+Approach: Astro supports placing .html files directly under src/pages/ — building
+copies them to dist/ verbatim (output identical to the original site). We copy
+every cloned page's HTML into src/pages/ preserving its route, and copy all
+non-HTML assets into public/ (Astro copies public/ -> dist/).
+
+This gives a real Astro project (astro build -> dist/) whose output matches the
+original website exactly, while all assets are self-hosted and URLs relative.
+
+Run from repo root.
+"""
+import os, shutil
+
+ROOT = os.path.dirname(os.path.abspath(__file__))
+SOURCE = os.path.join(ROOT, "site")
+PAGES = os.path.join(ROOT, "src", "pages")
+ASSETS = os.path.join(ROOT, "public")
+BUILD = os.path.join(ROOT, "dist")
+
+# remove sample astro page
+sample = os.path.join(PAGES, "index.astro")
+if os.path.exists(sample):
+    os.remove(sample)
+
+def main():
+    os.makedirs(PAGES, exist_ok=True)
+    os.makedirs(ASSETS, exist_ok=True)
+
+    # 1. Copy all assets (non-index.html) into public/ preserving structure
+    copied = 0
+    for dp, dns, fns in os.walk(SOURCE):
+        for fn in fns:
+            if fn == "index.html":
+                continue
+            src = os.path.join(dp, fn)
+            rel = os.path.relpath(src, SOURCE)
+            dst = os.path.join(ASSETS, rel)
+            os.makedirs(os.path.dirname(dst), exist_ok=True)
+            shutil.copy2(src, dst)
+            copied += 1
+    print(f"Copied {copied} asset files into public/")
+
+    # 2. Copy each page's HTML into src/pages/ preserving route as .html
+    gen = 0
+    for dp, dns, fns in os.walk(SOURCE):
+        if "index.html" not in fns:
+            continue
+        html_src = os.path.join(dp, "index.html")
+        rel_dir = os.path.relpath(os.path.dirname(html_src), SOURCE)
+        if rel_dir == ".":
+            page_rel = "index.html"
+        else:
+            page_rel = rel_dir + ".html"     # e.g. catalog/item/X.html, catalog.html
+        page_dst = os.path.join(PAGES, page_rel)
+        os.makedirs(os.path.dirname(page_dst), exist_ok=True)
+        shutil.copy2(html_src, page_dst)
+        gen += 1
+        print(f"  {page_rel}")
+    print(f"\nGenerated {gen} html pages under src/pages/")
+
+if __name__ == "__main__":
+    main()

mirror.py

@@ -1,259 +0,0 @@
-#!/usr/bin/env python3
-"""Crawl www.mitsuthailand.com -> site/ as a static mirror with dynamic (relative) URLs."""
-import os, re, queue, threading, urllib.parse, html
-
-import httpx
-
-from common import (BASE_HOST, BASE_URL, MEDIA_HOST, YP_HOST, OUT_DIR,
-                    assets_rel_for, safe_asset_path, enqueue_asset,
-                    claim_asset_urls, mark_asset_result, is_downloaded,
-                    canonical_page_path)
-
-client = httpx.Client(
-    follow_redirects=True, timeout=40.0,
-    headers={"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) "
-                           "AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0 Safari/537.36",
-             "Accept-Language": "th,en;q=0.9"},
-)
-
-# ---------------- URL classification ----------------
-def resolve_url(raw, base):
-    raw = html.unescape(raw).strip()
-    if not raw or raw.startswith(("data:", "mailto:", "tel:", "javascript:", "#")):
-        return None
-    if raw.startswith(("//", "http")):
-        parsed = urllib.parse.urlparse(raw)
-        host = (parsed.hostname or "").lower()
-        if host not in (BASE_HOST, MEDIA_HOST, YP_HOST):
-            return None
-        if "/lp/" in parsed.path or parsed.path.startswith("/lp"):
-            return None
-        return urllib.parse.urlunparse(parsed._replace(scheme="https")) if parsed.scheme in ("", "http") else raw
-    absurl = urllib.parse.urljoin(base, raw)
-    parsed = urllib.parse.urlparse(absurl)
-    host = (parsed.hostname or "").lower()
-    if host and host not in (BASE_HOST, MEDIA_HOST, YP_HOST):
-        return None
-    if "/lp/" in parsed.path or parsed.path.startswith("/lp"):
-        return None
-    return absurl
-
-
-def is_page_url(url):
-    parsed = urllib.parse.urlparse(url)
-    host = (parsed.hostname or "").lower()
-    # The www.yellowpages.co.th/lp/... pages are a YellowPages mirror of the same content
-    # and will die too; skip them. Only mirror the main mitsuthailand.com pages.
-    if host == YP_HOST:
-        return False
-    path = parsed.path.rstrip("/")
-    # Skip /lp/... "live preview" duplicate copies of the same content (very long paths).
-    if "/lp/" in path or path.startswith("/lp"):
-        return False
-    # skip Drupal file assets (they have extensions or are /sites /core /themes /system paths)
-    if path.startswith(("/sites/", "/core/", "/themes/", "/system/", "/sites")):
-        return False
-    ext = os.path.splitext(path)[1].lower()
-    # Drupal content routes (no file extension) are pages
-    return ext in (".html", "")
-
-
-# ---------------- asset workers ----------------
-_jobs = queue.Queue()
-_shutdown = threading.Event()
-_submitted = set()          # URLs already pushed to _jobs (guarantees no dup put)
-_sub_lock = threading.Lock()
-
-def start_workers(n=12):
-    for _ in range(n):
-        t = threading.Thread(target=_worker, daemon=True)
-        t.start()
-
-def stop_workers():
-    _shutdown.set()
-
-def submit_assets(urls):
-    """Push each URL to _jobs exactly once; returns count of new submissions."""
-    new = []
-    with _sub_lock:
-        for u in urls:
-            if u not in _submitted and not is_downloaded(u):
-                _submitted.add(u)
-                new.append(u)
-    for u in new:
-        _jobs.put(u)
-    return len(new)
-
-def _worker():
-    while not _shutdown.is_set():
-        try:
-            url = _jobs.get(timeout=0.5)
-        except queue.Empty:
-            continue
-        try:
-            r = client.get(url)
-            if r.status_code != 200:
-                print(f"  [asset {r.status_code}] {url}")
-                mark_asset_result(url, None)
-                continue
-            ctype = r.headers.get("content-type", "")
-            data = r.content
-            rel = safe_asset_path(assets_rel_for(url), url)
-            dest = os.path.join(OUT_DIR, rel)
-            os.makedirs(os.path.dirname(dest), exist_ok=True)
-            if "css" in ctype or url.rstrip("?;").endswith(".css"):
-                try:
-                    text = data.decode("utf-8", "replace")
-                    text = rewrite_css(url, text)
-
-                    data = text.encode("utf-8")
-                except Exception as e:
-                    print("  [css err]", e)
-            with open(dest, "wb") as f:
-                f.write(data)
-            mark_asset_result(url, rel)
-        except Exception as e:
-            print(f"  [asset ERR] {url}: {e}")
-            mark_asset_result(url, None)
-
-
-# ---------------- CSS rewriting ----------------
-def rewrite_css(css_url, css_text):
-    def repl(m):
-        u = m.group(1).strip()
-        if u.startswith(("data:", "#", "http", "//")) or u.startswith("url("):
-            return m.group(0)
-        abs = urllib.parse.urljoin(css_url, u)
-        parsed = urllib.parse.urlparse(abs)
-        host = (parsed.hostname or "").lower()
-        if host and host not in (BASE_HOST, MEDIA_HOST, YP_HOST):
-            return m.group(0)
-        rel = safe_asset_path(assets_rel_for(abs), abs)
-        enqueue_asset(abs)
-        # css files live under assets/<dir>/..., reference is relative to that file's dir
-        css_dir = os.path.dirname(assets_rel_for(css_url))  # e.g. assets/sites/storage/files/css
-        from_dir = os.path.join(OUT_DIR, css_dir)
-        rel_from = os.path.relpath(os.path.join(OUT_DIR, rel), from_dir)
-        return 'url("{}")'.format(rel_from)
-    return re.sub(r'url\(\s*["\']?([^"\'()]+)["\']?\s*\)', repl, css_text)
-
-# ---------------- crawl + render ----------------
-def fetch_sitemap_urls():
-    urls = {BASE_URL, BASE_URL + "/"}
-    try:
-        r = client.get(f"{BASE_URL}/sitemap.xml")
-        if r.status_code == 200:
-            for m in re.finditer(r"<loc>([^<]+)</loc>", r.text):
-                u = m.group(1).strip().replace("http://", "https://")
-                if BASE_HOST in u:
-                    urls.add(u)
-    except Exception as e:
-        print("sitemap err", e)
-    return urls
-
-
-def crawl_pages(seed):
-    crawled = {}          # canonical path -> representative URL
-    to_crawl = queue.Queue()
-    for u in seed:
-        to_crawl.put(u)
-    for _ in range(5):
-        found = {}
-        while not to_crawl.empty():
-            try:
-                url = to_crawl.get_nowait()
-            except queue.Empty:
-                break
-            c_key = canonical_page_path(url)
-            if c_key in crawled:
-                continue
-            crawled[c_key] = url
-            try:
-                r = client.get(url)
-            except Exception as e:
-                print("  [page ERR]", url, e)
-                continue
-            if r.status_code != 200:
-                print("  [page", r.status_code, "]", url)
-                continue
-            text = r.text
-            for l in re.findall(r'(?:href|src|action)="([^"]+)"', text):
-                a = resolve_url(l, url)
-                if a is None:
-                    continue
-                if is_page_url(a):
-                    ca = canonical_page_path(a)
-                    if ca not in crawled and ca not in found:
-                        found[ca] = a
-                else:
-                    enqueue_asset(a)
-        for ck, u in found.items():
-            to_crawl.put(u)
-        print(f"  pass: {len(crawled)} pages, +{len(found)} new")
-        if not found:
-            break
-    return list(crawled.values())
-
-
-def render_pages(pages):
-    from rewrite import rewrite_html
-    seen = set()
-    for url in sorted(pages):
-        try:
-            canon = canonical_page_path(url)
-            if canon in seen:
-                continue
-            seen.add(canon)
-            r = client.get(url)
-            if r.status_code != 200:
-                print("  [render", r.status_code, "]", url)
-                continue
-            text = rewrite_html(r.text, url)
-            rel_path = canon.lstrip("/")
-            dest = os.path.join(OUT_DIR, rel_path, "index.html")
-            os.makedirs(os.path.dirname(dest), exist_ok=True)
-            with open(dest, "w", encoding="utf-8") as f:
-                f.write(text)
-            print(f"  [page OK] {canon}")
-        except Exception as e:
-            print("  [render ERR]", url, e)
-
-
-def drain_assets():
-    """Pull newly-queued assets (common._queue) into _jobs until nothing is pending
-    or in-flight. Returns when all submitted assets have a result AND no new items
-    were enqueued in the meantime."""
-    import time as _time
-    idle_rounds = 0
-    while True:
-        batch = claim_asset_urls()
-        if batch:
-            submit_assets(batch)
-            idle_rounds = 0
-        if not _submitted:
-            return
-        if all(is_downloaded(u) for u in list(_submitted)):
-            # confirm nothing new got queued before returning
-            if not claim_asset_urls():
-                return
-        _time.sleep(0.3)
-
-
-def main():
-    print("Mirroring", BASE_URL, "->", OUT_DIR)
-    seed = fetch_sitemap_urls()
-    print(f"Seeded {len(seed)} URLs.")
-    pages = crawl_pages(seed)
-    print(f"Total pages: {len(pages)}")
-
-    start_workers()
-    drain_assets()
-    render_pages(pages)
-    drain_assets()
-    stop_workers()
-    print(f"Pages rendered: {len(pages)}")
-    print("Done. Site in", OUT_DIR)
-
-
-if __name__ == "__main__":
-    main()

package.json

@@ -0,0 +1,17 @@
+{
+  "name": "stale-satellite",
+  "type": "module",
+  "version": "0.0.1",
+  "engines": {
+    "node": ">=22.12.0"
+  },
+  "scripts": {
+    "dev": "astro dev",
+    "build": "astro build",
+    "preview": "astro preview",
+    "astro": "astro"
+  },
+  "dependencies": {
+    "astro": "^7.1.6"
+  }
+}

pnpm-workspace.yaml

@@ -0,0 +1,3 @@
+allowBuilds:
+  esbuild: true
+  sharp: true

postprocess.py

@@ -1,58 +0,0 @@
-#!/usr/bin/env python3
-"""Post-process: re-rewrite rendered pages with the latest rewrite_rules (data-qrcodr etc.)
-and download any newly-referenced assets. Idempotent."""
-import os, re, queue, threading, urllib.parse, sys
-
-sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
-import mirror
-from common import (BASE_URL, OUT_DIR, enqueue_asset, claim_asset_urls,
-                    mark_asset_result, is_downloaded, assets_rel_for, safe_asset_path)
-from rewrite import rewrite_html
-
-def walk_pages():
-    """Yield (page_url, abs_path) for every index.html under OUT_DIR."""
-    root = OUT_DIR
-    for dirpath, dirnames, filenames in os.walk(root):
-        if "index.html" in filenames:
-            rel = os.path.relpath(dirpath, root)
-            if rel == ".":
-                path = "/"
-            else:
-                path = "/" + rel.replace(os.sep, "/")
-            page_url = BASE_URL + path
-            yield page_url, os.path.join(dirpath, "index.html")
-
-def main():
-    mirror.start_workers()
-    count = 0
-    changed = 0
-    for page_url, fpath in walk_pages():
-        count += 1
-        with open(fpath, "r", encoding="utf-8") as f:
-            text = f.read()
-        new = rewrite_html(text, page_url)
-        if new != text:
-            changed += 1
-            with open(fpath, "w", encoding="utf-8") as f:
-                f.write(new)
-        # drain any assets enqueued by this page's rewrite
-        batch = claim_asset_urls()
-        if batch:
-            mirror.submit_assets(batch)
-    # final drain
-    import time
-    while True:
-        batch = claim_asset_urls()
-        if batch:
-            mirror.submit_assets(batch)
-        if not mirror._submitted:
-            break
-        if all(is_downloaded(u) for u in list(mirror._submitted)):
-            if not claim_asset_urls():
-                break
-        time.sleep(0.3)
-    mirror.stop_workers()
-    print(f"Processed {count} pages, {changed} changed.")
-
-if __name__ == "__main__":
-    main()