@lark-apaas/coding-steering 0.1.32-dev.5abff3b → 0.1.32-dev.942e73f

package/steering/design-html/skills/pptx-style-extract/scripts/test_logo_scope.py

@@ -0,0 +1,148 @@
+#!/usr/bin/env python3
+"""Regression tests for template logo placement in generated deck HTML."""
+import os
+import sys
+import tempfile
+import unittest
+
+sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
+
+from verify_logo_scope import validate_logo_scope
+
+
+class LogoScopeTest(unittest.TestCase):
+    def write_pack(self, root):
+        with open(os.path.join(root, 'design.md'), 'w', encoding='utf-8') as stream:
+            stream.write(
+                '---\n'
+                'assets:\n'
+                '  logo-1:\n'
+                '    path: assets/logos/1.png\n'
+                '    kind: logo\n'
+                'layouts: layouts.md\n'
+                '---\n'
+            )
+        with open(os.path.join(root, 'layouts.md'), 'w', encoding='utf-8') as stream:
+            stream.write(
+                '---\n'
+                'canvas: 1920x1080\n'
+                'layouts:\n'
+                '  cover:\n'
+                '    role: cover\n'
+                '    slots:\n'
+                '      - {role: logo, box: [64, 64, 224, 48], type: pic, asset: logo-1}\n'
+                '  content:\n'
+                '    role: content\n'
+                '    slots:\n'
+                '      - {role: title, box: [100, 100, 800, 100], type: title}\n'
+                '  closing:\n'
+                '    role: closing\n'
+                '    slots:\n'
+                '      - {role: logo, box: [64, 64, 224, 48], type: pic, asset: logo-1}\n'
+                '---\n'
+            )
+
+    def write_html(self, root, content):
+        path = os.path.join(root, 'index.html')
+        with open(path, 'w', encoding='utf-8') as stream:
+            stream.write(content)
+        return path
+
+    def test_logo_is_limited_to_every_archetype_with_its_slot(self):
+        with tempfile.TemporaryDirectory() as root:
+            self.write_pack(root)
+            html = self.write_html(
+                root,
+                '<deck-stage>'
+                '<section data-pptx-layout="cover">'
+                '<img src="assets/pptx-volcengine/logos/1.png"></section>'
+                '<section data-pptx-layout="content"><h1>正文</h1></section>'
+                '<section data-pptx-layout="closing">'
+                '<img src="assets/pptx-volcengine/logos/1.png"></section>'
+                '</deck-stage>',
+            )
+
+            self.assertEqual(
+                [],
+                validate_logo_scope(root, html, 'assets/pptx-volcengine'),
+            )
+
+    def test_logo_on_an_unowned_archetype_is_rejected(self):
+        with tempfile.TemporaryDirectory() as root:
+            self.write_pack(root)
+            html = self.write_html(
+                root,
+                '<deck-stage>'
+                '<section data-pptx-layout="content">'
+                '<img src="assets/pptx-volcengine/logos/1.png"></section>'
+                '</deck-stage>',
+            )
+
+            problems = validate_logo_scope(root, html, 'assets/pptx-volcengine')
+
+            self.assertEqual(1, len(problems))
+            self.assertIn('content', problems[0])
+            self.assertIn('logo-1', problems[0])
+
+    def test_nested_content_section_does_not_become_a_slide(self):
+        with tempfile.TemporaryDirectory() as root:
+            self.write_pack(root)
+            html = self.write_html(
+                root,
+                '<deck-stage><section data-pptx-layout="cover">'
+                '<section class="content"><img src="assets/pptx-volcengine/logos/1.png">'
+                '</section></section></deck-stage>',
+            )
+
+            self.assertEqual(
+                [],
+                validate_logo_scope(root, html, 'assets/pptx-volcengine'),
+            )
+
+    def test_each_slide_declares_its_template_archetype(self):
+        with tempfile.TemporaryDirectory() as root:
+            self.write_pack(root)
+            html = self.write_html(
+                root,
+                '<deck-stage><section><h1>未声明页型</h1></section></deck-stage>',
+            )
+
+            self.assertEqual(
+                ['第 1 页缺少 data-pptx-layout，无法核验模板 logo 归属'],
+                validate_logo_scope(root, html, 'assets/pptx-volcengine'),
+            )
+
+    def test_logo_in_global_css_is_rejected(self):
+        with tempfile.TemporaryDirectory() as root:
+            self.write_pack(root)
+            html = self.write_html(
+                root,
+                '<style>.slide-logo { background-image: url('
+                '"assets/pptx-volcengine/logos/1.png") }</style>'
+                '<deck-stage><section data-pptx-layout="content">正文</section></deck-stage>',
+            )
+
+            self.assertEqual(
+                ['模板 logo logo-1 出现在 slide section 外，无法核验页型归属'],
+                validate_logo_scope(root, html, 'assets/pptx-volcengine'),
+            )
+
+    def test_logo_in_a_slide_style_block_is_rejected(self):
+        with tempfile.TemporaryDirectory() as root:
+            self.write_pack(root)
+            html = self.write_html(
+                root,
+                '<deck-stage><section data-pptx-layout="content"><style>'
+                '.slide-logo { background-image: url('
+                '"assets/pptx-volcengine/logos/1.png") }</style>'
+                '正文</section></deck-stage>',
+            )
+
+            self.assertEqual(
+                ['模板 logo logo-1 出现在 slide section 外，无法核验页型归属'],
+                validate_logo_scope(root, html, 'assets/pptx-volcengine'),
+            )
+
+
+if __name__ == '__main__':
+    unittest.main()

package/steering/design-html/skills/pptx-style-extract/scripts/verify_logo_scope.py

@@ -0,0 +1,188 @@
+#!/usr/bin/env python3
+"""Verify that generated slides use template logos only on owning archetypes."""
+import argparse
+import posixpath
+import re
+import sys
+from html.parser import HTMLParser
+from urllib.parse import urlsplit
+
+from check_v2 import Pack
+
+
+URL_RE = re.compile(r'url\(\s*[\'"]?([^\'")\s]+)', re.I)
+VOID_TAGS = {
+    'area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input',
+    'link', 'meta', 'param', 'source', 'track', 'wbr',
+}
+
+
+def normalized_path(value):
+    path = urlsplit(value).path.replace('\\', '/')
+    return posixpath.normpath(path).lstrip('./')
+
+
+def asset_urls(pack, asset_prefix):
+    prefix = normalized_path(asset_prefix).rstrip('/')
+    urls = {}
+    for asset_id, (entry, _) in pack.assets.items():
+        if not isinstance(entry, dict) or entry.get('kind') != 'logo':
+            continue
+        path = entry.get('path')
+        if not isinstance(path, str) or not path:
+            continue
+        relative = normalized_path(path)
+        if relative.startswith('assets/'):
+            relative = relative[len('assets/'):]
+        urls[posixpath.join(prefix, relative)] = asset_id
+    return urls
+
+
+def contains_asset(value, asset_id):
+    if isinstance(value, dict):
+        return (value.get('asset') == asset_id
+                or any(contains_asset(child, asset_id) for child in value.values()))
+    if isinstance(value, list):
+        return any(contains_asset(child, asset_id) for child in value)
+    return False
+
+
+def logo_owners(pack):
+    owners = {}
+    for asset_url, asset_id in asset_urls(pack, '').items():
+        del asset_url
+        owners[asset_id] = {
+            name
+            for name, (layout, _) in pack.layouts.items()
+            if contains_asset(layout, asset_id)
+        }
+    return owners
+
+
+def urls_from_attrs(attrs):
+    urls = []
+    for key, value in attrs:
+        if not value:
+            continue
+        if key.lower() in ('src', 'href'):
+            urls.append(value)
+        elif key.lower() == 'srcset':
+            urls.extend(item.strip().split(' ', 1)[0] for item in value.split(','))
+        elif key.lower() == 'style':
+            urls.extend(URL_RE.findall(value))
+    return urls
+
+
+class SlideAssetParser(HTMLParser):
+    def __init__(self):
+        super().__init__()
+        self.slides = []
+        self._tags = []
+        self._sections = []
+        self._active_slides = []
+        self._style_depth = 0
+        self.outside_urls = []
+
+    def handle_starttag(self, tag, attrs):
+        tag = tag.lower()
+        urls = urls_from_attrs(attrs)
+        parent = self._tags[-1] if self._tags else None
+        if tag == 'section':
+            slide = (
+                {'layout': dict(attrs).get('data-pptx-layout'), 'urls': urls}
+                if parent == 'deck-stage' else None
+            )
+            self._sections.append(slide)
+            if slide is not None:
+                self._active_slides.append(slide)
+        elif self._active_slides:
+            self._active_slides[-1]['urls'].extend(urls)
+        else:
+            self.outside_urls.extend(urls)
+        if tag == 'style':
+            self._style_depth += 1
+        if tag not in VOID_TAGS:
+            self._tags.append(tag)
+
+    def handle_startendtag(self, tag, attrs):
+        if self._active_slides:
+            self._active_slides[-1]['urls'].extend(urls_from_attrs(attrs))
+        else:
+            self.outside_urls.extend(urls_from_attrs(attrs))
+
+    def handle_endtag(self, tag):
+        tag = tag.lower()
+        if tag == 'style' and self._style_depth:
+            self._style_depth -= 1
+        if tag == 'section' and self._sections:
+            slide = self._sections.pop()
+            if slide is not None:
+                self.slides.append(slide)
+                self._active_slides.pop()
+        if tag not in VOID_TAGS and self._tags:
+            self._tags.pop()
+
+    def handle_data(self, data):
+        if not self._style_depth:
+            return
+        self.outside_urls.extend(URL_RE.findall(data))
+
+
+def validate_logo_scope(pack_dir, html_path, asset_prefix):
+    """Return every template-logo placement that does not match the slide archetype."""
+    pack = Pack(pack_dir)
+    known_urls = asset_urls(pack, asset_prefix)
+    owners = logo_owners(pack)
+    if not known_urls:
+        return []
+
+    with open(html_path, encoding='utf-8') as stream:
+        text = stream.read()
+    parser = SlideAssetParser()
+    parser.feed(text)
+    parser.close()
+
+    problems = []
+    for url in parser.outside_urls:
+        asset_id = known_urls.get(normalized_path(url))
+        if asset_id:
+            problems.append(
+                '模板 logo %s 出现在 slide section 外，无法核验页型归属' % asset_id)
+    for number, slide in enumerate(parser.slides, 1):
+        layout = slide['layout']
+        if not layout:
+            problems.append('第 %d 页缺少 data-pptx-layout，无法核验模板 logo 归属' % number)
+            continue
+        if layout not in pack.layouts:
+            problems.append('第 %d 页声明了不存在的模板页型: %s' % (number, layout))
+            continue
+        for url in slide['urls']:
+            asset_id = known_urls.get(normalized_path(url))
+            if asset_id and layout not in owners.get(asset_id, set()):
+                allowed = '、'.join(sorted(owners.get(asset_id) or ())) or '（无）'
+                problems.append(
+                    '第 %d 页页型 %s 不得使用 %s；只允许: %s'
+                    % (number, layout, asset_id, allowed))
+    return problems
+
+
+def main(argv=None):
+    parser = argparse.ArgumentParser(
+        description='Verify that generated deck HTML places PPTX logos only on owning layouts.')
+    parser.add_argument('pack_dir')
+    parser.add_argument('html_path')
+    parser.add_argument('--asset-prefix', required=True)
+    args = parser.parse_args(argv)
+
+    problems = validate_logo_scope(args.pack_dir, args.html_path, args.asset_prefix)
+    if not problems:
+        print('PPTX_LOGO_SCOPE: PASS')
+        return 0
+    print('PPTX_LOGO_SCOPE: FAIL count=%d' % len(problems))
+    for problem in problems:
+        print('[logoScope] %s' % problem)
+    return 1
+
+
+if __name__ == '__main__':
+    sys.exit(main())