webscan: Prüfung auf Playwright und Browser mit Installationshinweisen

- fehlt Playwright oder der Browser, werden die passenden Befehle für
  Ubuntu/Debian bzw. Arch ausgegeben (erkannt über /etc/os-release),
  Abbruch mit Exit-Code 2
- doc/README.md: Einrichtung getrennt nach Ubuntu und Arch

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

doc/README.md

@@ -154,13 +154,27 @@ Konfiguration (in `cfg/webscan.cfg`: 800×500).
 Einmalig einrichten (Python-Umgebung im Projektverzeichnis):
 
 ```sh
-python3 -m venv .venv
+python3 -m venv .venv          # Ubuntu: vorher sudo apt install python3-venv
 .venv/bin/pip install -r requirements.txt
+```
+
+Browser unter Ubuntu/Debian:
+
+```sh
+sudo .venv/bin/playwright install-deps chromium
 .venv/bin/playwright install chromium
 ```
 
-Statt des Playwright-Chromiums kann auch ein installiertes Chromium benutzt
-werden: `-b /usr/bin/chromium` oder `export PICCREATE_BROWSER=/usr/bin/chromium`.
+Browser unter Arch Linux (Playwright unterstützt Arch nicht offiziell, daher
+besser das Chromium der Distribution):
+
+```sh
+sudo pacman -S chromium
+export PICCREATE_BROWSER=/usr/bin/chromium   # oder Option -b /usr/bin/chromium
+```
+
+Fehlen Playwright oder der Browser, gibt `webscan.py` die passenden Befehle
+für die erkannte Distribution aus und bricht mit Exit-Code 2 ab.
 
 Eingabedatei, z. B. `in/kunden.csv` (Kopfzeile optional, `http://` wird bei
 Bedarf ergänzt):

libs/webscan.py

@@ -18,7 +18,70 @@ import optparse
 import os
 import sys
 
-from playwright.async_api import async_playwright, Error as PlaywrightError
+PROJECT_DIR = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+VENV = os.path.join(PROJECT_DIR, '.venv')
+
+INSTALL_HINTS = {
+    'ubuntu': '''Ubuntu / Debian:
+    sudo apt install python3-venv
+    python3 -m venv {venv}
+    {venv}/bin/pip install -r {project}/requirements.txt
+    sudo {venv}/bin/playwright install-deps chromium
+    {venv}/bin/playwright install chromium''',
+    'arch': '''Arch Linux:
+    python -m venv {venv}
+    {venv}/bin/pip install -r {project}/requirements.txt
+  then either the browser from Arch (recommended, Playwright does not
+  officially support Arch):
+    sudo pacman -S chromium
+    export PICCREATE_BROWSER=/usr/bin/chromium
+  or the Playwright browser:
+    {venv}/bin/playwright install chromium''',
+}
+
+BROWSER_HINTS = {
+    'ubuntu': '''Ubuntu / Debian:
+    sudo {venv}/bin/playwright install-deps chromium
+    {venv}/bin/playwright install chromium''',
+    'arch': '''Arch Linux:
+    sudo pacman -S chromium
+    export PICCREATE_BROWSER=/usr/bin/chromium
+  or the Playwright browser:
+    {venv}/bin/playwright install chromium''',
+}
+
+
+def linux_distro():
+    ''' liefert 'ubuntu' oder 'arch' (aus /etc/os-release), sonst None '''
+    try:
+        with open('/etc/os-release', encoding='utf-8') as fh:
+            info = dict(line.rstrip('\n').split('=', 1) for line in fh if '=' in line)
+    except OSError:
+        return None
+    ids = (info.get('ID', '') + ' ' + info.get('ID_LIKE', '')).replace('"', '').split()
+    if 'arch' in ids:
+        return 'arch'
+    if 'ubuntu' in ids or 'debian' in ids:
+        return 'ubuntu'
+    return None
+
+
+def print_hint(hints):
+    ''' gibt die Installationshinweise für diese Distribution aus
+    (bei unbekannter Distribution alle) '''
+    distro = linux_distro()
+    keys = [distro] if distro in hints else sorted(hints)
+    for key in keys:
+        print(hints[key].format(venv=VENV, project=PROJECT_DIR), file=sys.stderr)
+
+
+try:
+    from playwright.async_api import async_playwright, Error as PlaywrightError
+except ImportError:
+    print('Playwright is not installed for', sys.executable, file=sys.stderr)
+    print('Install it into the project environment:', file=sys.stderr)
+    print_hint(INSTALL_HINTS)
+    sys.exit(2)
 
 from utils import CompleteEnding
 
@@ -92,8 +155,17 @@ async def run(jobs, pic_config, options):
     async with async_playwright() as p:
         launch_args = {}
         if options.browser:
+            if not os.path.isfile(options.browser):
+                print('browser', options.browser, 'does not exist.', file=sys.stderr)
+                print_hint(BROWSER_HINTS)
+                return None
             launch_args['executable_path'] = options.browser
+        try:
             browser = await p.chromium.launch(**launch_args)
+        except PlaywrightError as e:
+            print('could not start the browser:', str(e).splitlines()[0], file=sys.stderr)
+            print_hint(BROWSER_HINTS)
+            return None
 
         async def worker(customer_id, url):
             out_path = os.path.join(out_dir, customer_id + '.' + pic_config['format'])
@@ -133,6 +205,8 @@ def main():
         print('no urls found in', options.infile)
         return 1
     failed = asyncio.run(run(jobs, pic_config, options))
+    if failed is None:
+        return 2
     if failed:
         print(len(failed), 'of', len(jobs), 'pages failed:', ', '.join(failed), file=sys.stderr)
         return 1