@lark-apaas/coding-steering 0.1.32-dev.4f80f68 → 0.1.32-dev.942e73f

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

@@ -299,6 +299,22 @@ NV_TAGS = {'sp': 'p:nvSpPr', 'pic': 'p:nvPicPr', 'grpSp': 'p:nvGrpSpPr',
            'cxnSp': 'p:nvCxnSpPr', 'graphicFrame': 'p:nvGraphicFramePr'}
 
 
+def blip_opacity(blip):
+    """Combine the alpha modulation transforms attached to one local picture."""
+    opacity = 1.0
+    found = False
+    for effect in blip:
+        if local(effect.tag) not in ('alphaMod', 'alphaModFix'):
+            continue
+        try:
+            amount = int(effect.get('amt'))
+        except (TypeError, ValueError):
+            continue
+        opacity *= max(0.0, min(amount / 100000.0, 1.0))
+        found = True
+    return round(opacity, 6) if found else None
+
+
 def walk_tree(el, ctx, out, path=(), xf=(1.0, 1.0, 0.0, 0.0), depth=0):
     sx, sy, dx, dy = xf
     U, W, H = ctx.units, ctx.units.w, ctx.units.h
@@ -393,9 +409,9 @@ def walk_tree(el, ctx, out, path=(), xf=(1.0, 1.0, 0.0, 0.0), depth=0):
             blip = sp.find('p:blipFill/a:blip', NS)
             if blip is not None:
                 rec['media'] = ctx.media_of(blip.get(R_EMBED)) or ctx.media_of(blip.get(R_LINK))
-                alpha = blip.find('a:alphaModFix', NS)
-                if alpha is not None and alpha.get('amt') is not None:
-                    rec['opacity'] = round(int(alpha.get('amt')) / 100000.0, 6)
+                opacity = blip_opacity(blip)
+                if opacity is not None:
+                    rec['opacity'] = opacity
                 svg = blip.find('a:extLst//asvg:svgBlip', NS)
                 if svg is not None:
                     rec['media_svg'] = ctx.media_of(svg.get(R_EMBED))

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

@@ -88,13 +88,15 @@ def css_gradient(grad):
 def background_css(bg):
     """页底 <p:bg> 描述符 → CSS background 值。解析不出返回 None，交上层回退到下一层底色。
 
-    bgPr solid / bgRef 都带 `color`；gradient 与形状渐变共用 css_gradient；
-    背景图不重建（媒体未必导出），按无底色处理。
+    bgPr solid / bgRef 都带 `color`；线性 gradient 与形状渐变共用 css_gradient；
+    path 渐变和背景图不重建，按无底色处理，不能伪装成线性渐变。
     """
     if not bg:
         return None
     kind = bg.get('type')
     if kind == 'gradient':
+        if bg.get('path'):
+            return None
         return css_gradient(bg)
     if kind in ('image', 'none', 'pattern'):
         return None

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

@@ -0,0 +1,556 @@
+#!/usr/bin/env python3
+"""Integration regression tests for model-decided image roles."""
+import json
+import os
+import sys
+import tempfile
+import unittest
+
+sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
+
+from check_v2 import Pack, RULES  # noqa: E402
+from package import (Fail, apply_asset_decisions, bind_asset_vision_group_scopes,
+                     load_layout_blocks, main as package_main,  # noqa: E402
+                     validate_asset_vision_groups)
+
+
+class AssetJudgmentPackageTest(unittest.TestCase):
+    def test_asset_vision_index_requires_selected_candidates(self):
+        with tempfile.TemporaryDirectory() as lout:
+            with open(os.path.join(lout, 'asset-vision-groups.json'),
+                      'w', encoding='utf-8') as stream:
+                json.dump({'version': 2}, stream)
+
+            with self.assertRaisesRegex(Fail, '缺少 selected'):
+                validate_asset_vision_groups(lout, {'asset_vision_groups': []})
+
+    def test_asset_vision_groups_must_keep_every_extracted_candidate(self):
+        with tempfile.TemporaryDirectory() as lout:
+            with open(os.path.join(lout, 'asset-vision-groups.json'),
+                      'w', encoding='utf-8') as stream:
+                json.dump({
+                    'version': 2,
+                    'selected': [{
+                        'candidates': [{
+                            'id': 'asset-cover',
+                            'source_media': 'cover.png',
+                            'placements': [{
+                                'box': [100, 200, 300, 400],
+                                'archetype': 'cover',
+                            }],
+                        }, {
+                            'id': 'asset-ornament',
+                            'source_media': 'ornament.png',
+                            'placements': [{
+                                'box': [10, 20, 30, 40],
+                                'archetype': 'content',
+                            }],
+                        }],
+                    }],
+                }, stream)
+            manifest = {
+                'asset_vision_groups': [{
+                    'id': 'asset-cover',
+                    'source_media': 'cover.png',
+                    'layout': 'cover',
+                    'box': '[100, 200, 300, 400]',
+                    'visual_kind': 'illustration',
+                }, {
+                    'id': 'asset-ornament',
+                    'source_media': 'ornament.png',
+                    'layout': 'content',
+                    'box': '[10, 20, 30, 40]',
+                    'visual_kind': 'decorative',
+                }],
+            }
+
+            validate_asset_vision_groups(lout, manifest)
+            manifest['asset_vision_groups'].pop()
+
+            with self.assertRaisesRegex(Fail, '逐项保留抽取候选.*asset-ornament'):
+                validate_asset_vision_groups(lout, manifest)
+
+    def test_asset_vision_groups_reject_rewritten_extracted_candidate_scope(self):
+        with tempfile.TemporaryDirectory() as lout:
+            with open(os.path.join(lout, 'asset-vision-groups.json'),
+                      'w', encoding='utf-8') as stream:
+                json.dump({
+                    'version': 2,
+                    'selected': [{
+                        'candidates': [{
+                            'id': 'asset-ornament',
+                            'source_media': 'ornament.png',
+                            'placements': [{
+                                'box': [10, 20, 30, 40],
+                                'archetype': None,
+                            }],
+                        }],
+                    }],
+                }, stream)
+            manifest = {
+                'asset_vision_groups': [{
+                    'id': 'asset-ornament',
+                    'source_media': 'other.png',
+                    'box': '[10, 20, 30, 40]',
+                    'visual_kind': 'decorative',
+                }],
+            }
+
+            with self.assertRaisesRegex(Fail, '改写了抽取候选的图片或位置'):
+                validate_asset_vision_groups(lout, manifest)
+
+    def test_v2_candidate_ids_resolve_same_box_on_different_layouts(self):
+        with tempfile.TemporaryDirectory() as lout:
+            with open(os.path.join(lout, 'asset-vision-groups.json'),
+                      'w', encoding='utf-8') as stream:
+                json.dump({
+                    'version': 2,
+                    'selected': [{
+                        'candidates': [{
+                            'id': 'asset-cover-logo',
+                            'source_media': 'shared.png',
+                            'placements': [{
+                                'box': [100, 200, 80, 80],
+                                'archetype': 'cover',
+                            }],
+                        }, {
+                            'id': 'asset-content-ornament',
+                            'source_media': 'shared.png',
+                            'placements': [{
+                                'box': [100, 200, 80, 80],
+                                'archetype': 'content',
+                            }],
+                        }],
+                    }],
+                }, stream)
+            manifest = {
+                'assets': [],
+                'asset_vision_groups': [{
+                    'id': 'asset-cover-logo',
+                    'source_media': 'shared.png',
+                    'box': '[100, 200, 80, 80]',
+                    'visual_kind': 'logo',
+                }, {
+                    'id': 'asset-content-ornament',
+                    'source_media': 'shared.png',
+                    'box': '[100, 200, 80, 80]',
+                    'visual_kind': 'decorative',
+                }],
+            }
+
+            bind_asset_vision_group_scopes(lout, manifest)
+            decisions, _ = apply_asset_decisions(
+                manifest,
+                bound_sources={'shared.png'},
+                bound_slots={
+                    'shared.png': {
+                        ('cover', (100, 200, 80, 80)),
+                        ('content', (100, 200, 80, 80)),
+                    },
+                },
+            )
+
+        self.assertEqual(
+            decisions.for_scope('shared.png', 'cover', (100, 200, 80, 80)),
+            'logo',
+        )
+        self.assertEqual(
+            decisions.for_scope('shared.png', 'content', (100, 200, 80, 80)),
+            'texture',
+        )
+
+    def test_legacy_asset_vision_index_accepts_a_group_default(self):
+        with tempfile.TemporaryDirectory() as lout:
+            with open(os.path.join(lout, 'asset-vision-groups.json'),
+                      'w', encoding='utf-8') as stream:
+                json.dump({
+                    'version': 1,
+                    'selected': [{
+                        'candidates': [{
+                            'id': 'asset-cover',
+                            'source_media': 'cover.png',
+                            'placements': [{'box': [0, 0, 100, 100]}],
+                        }, {
+                            'id': 'asset-ornament',
+                            'source_media': 'ornament.png',
+                            'placements': [{'box': [10, 20, 30, 40]}],
+                        }],
+                    }],
+                }, stream)
+            manifest = {
+                'asset_vision_groups': [{
+                    'id': 'vision-1',
+                    'source_media': '[cover.png, ornament.png]',
+                    'visual_kind': 'decorative',
+                }],
+            }
+
+            validate_asset_vision_groups(lout, manifest)
+
+    def test_layout_controls_override_legacy_judgment_without_reading_layout_todos(self):
+        with tempfile.TemporaryDirectory() as root:
+            lout = os.path.join(root, 'l-out')
+            os.makedirs(lout)
+            with open(os.path.join(lout, 'layouts.yaml'), 'w', encoding='utf-8') as stream:
+                stream.write(
+                    'names:\n'
+                    '  content: TODO中文名\n'
+                    'roles:\n'
+                    '  content: TODO角色\n'
+                    'layouts:\n'
+                    '  content:\n'
+                    '    slots:\n'
+                    '      - {role: body, box: [10, 20, 30, 40], type: body}\n'
+                    '    confidence: high\n'
+                )
+            with open(os.path.join(lout, 'layout-controls.yaml'), 'w', encoding='utf-8') as stream:
+                stream.write(
+                    'names:\n'
+                    '  content: "内容页"\n'
+                    'roles:\n'
+                    '  content: content\n'
+                )
+
+            blocks = load_layout_blocks(lout)
+
+            self.assertEqual([key for key, _, _ in blocks],
+                             ['names', 'roles', 'layouts'])
+            self.assertIn('内容页', blocks[0][2][0])
+            self.assertIn('content', blocks[1][2][0])
+            self.assertNotIn('TODO', '\n'.join(
+                line for _, _, lines in blocks for line in lines
+            ))
+
+    def test_layout_controls_reject_non_judgment_block(self):
+        with tempfile.TemporaryDirectory() as root:
+            lout = os.path.join(root, 'l-out')
+            os.makedirs(lout)
+            with open(os.path.join(lout, 'layouts.yaml'), 'w', encoding='utf-8') as stream:
+                stream.write('layouts:\n  content:\n    role: content\n')
+            with open(os.path.join(lout, 'layout-controls.yaml'), 'w', encoding='utf-8') as stream:
+                stream.write('layouts:\n  cover:\n    role: cover\n')
+
+            with self.assertRaisesRegex(Fail, '只能包含'):
+                load_layout_blocks(lout)
+
+    def test_layout_controls_must_cover_all_legacy_judgment_blocks(self):
+        with tempfile.TemporaryDirectory() as root:
+            lout = os.path.join(root, 'l-out')
+            os.makedirs(lout)
+            with open(os.path.join(lout, 'layouts.yaml'), 'w', encoding='utf-8') as stream:
+                stream.write(
+                    'names:\n'
+                    '  content: "内容页"\n'
+                    'roles:\n'
+                    '  content: content\n'
+                    'layouts:\n'
+                    '  content:\n'
+                    '    role: content\n'
+                )
+            with open(os.path.join(lout, 'layout-controls.yaml'), 'w',
+                      encoding='utf-8') as stream:
+                stream.write('roles:\n  content: content\n')
+
+            with self.assertRaisesRegex(Fail, '缺：names'):
+                load_layout_blocks(lout)
+
+    def test_layout_controls_must_keep_each_legacy_judgment_entry(self):
+        with tempfile.TemporaryDirectory() as root:
+            lout = os.path.join(root, 'l-out')
+            os.makedirs(lout)
+            with open(os.path.join(lout, 'layouts.yaml'), 'w', encoding='utf-8') as stream:
+                stream.write(
+                    'names:\n'
+                    '  cover: "封面"\n'
+                    '  content: "内容页"\n'
+                    'roles:\n'
+                    '  cover: cover\n'
+                    '  content: content\n'
+                    'layouts:\n'
+                    '  cover:\n'
+                    '    role: cover\n'
+                    '  content:\n'
+                    '    role: content\n'
+                )
+            with open(os.path.join(lout, 'layout-controls.yaml'), 'w',
+                      encoding='utf-8') as stream:
+                stream.write(
+                    'names:\n'
+                    '  cover: "封面"\n'
+                    'roles:\n'
+                    '  cover: cover\n'
+                    '  content: content\n'
+                )
+
+            with self.assertRaisesRegex(Fail, 'names 段缺少旧判断条目：content'):
+                load_layout_blocks(lout)
+
+    def test_package_strips_judgment_fields_and_preserves_content_slot(self):
+        try:
+            from PIL import Image
+        except ImportError:
+            self.skipTest('Pillow unavailable')
+
+        with tempfile.TemporaryDirectory() as root:
+            stage1 = os.path.join(root, 'stage1')
+            lout = os.path.join(stage1, 'l-out')
+            pack = os.path.join(root, 'pack')
+            media = os.path.join(stage1, 'media-out')
+            ref = os.path.join(stage1, 'ref')
+            os.makedirs(lout)
+            os.makedirs(media)
+            os.makedirs(ref)
+
+            Image.new('RGBA', (40, 20), (20, 40, 60, 255)).save(
+                os.path.join(media, 'partner-logo.png'))
+            Image.new('RGBA', (24, 24), (60, 40, 20, 255)).save(
+                os.path.join(media, 'ornament.png'))
+
+            box_rows = [
+                ('ppt/media/image1.png', 'partner-logo.png', [100, 200, 400, 200]),
+                ('ppt/media/image2.png', 'ornament.png', [40, 40, 120, 120]),
+            ]
+            extract = {
+                'source': {'filename': 'fixture.pptx'},
+                'canvas': {
+                    'px': [1920, 1080],
+                    'source': {'cx': 12192000, 'cy': 6858000, 'unit': 'EMU'},
+                },
+                'media': [
+                    {
+                        'media': source,
+                        'out': 'media-out/' + output,
+                        'source_px': [40, 20] if output == 'partner-logo.png' else [24, 24],
+                    }
+                    for source, output, _ in box_rows
+                ],
+                'images': [
+                    {
+                        'media': source,
+                        'boxes': [{
+                            'count': 1,
+                            'box': {'x': box[0], 'y': box[1], 'w': box[2], 'h': box[3]},
+                            'parts': ['ppt/slides/slide1.xml'],
+                        }],
+                    }
+                    for source, _, box in box_rows
+                ],
+                'color_freq': [],
+                'text_scale': [],
+            }
+            with open(os.path.join(stage1, 'extract.json'), 'w', encoding='utf-8') as stream:
+                json.dump(extract, stream)
+            with open(os.path.join(ref, 'shapes.json'), 'w', encoding='utf-8') as stream:
+                json.dump({
+                    'shapes': [
+                        {
+                            'part': 'ppt/slides/slide1.xml',
+                            'kind': 'pic',
+                            'name': output,
+                            'box': {'x': box[0], 'y': box[1], 'w': box[2], 'h': box[3]},
+                        }
+                        for _, output, box in box_rows
+                    ],
+                }, stream)
+
+            with open(os.path.join(lout, 'manifest.yaml'), 'w', encoding='utf-8') as stream:
+                stream.write(
+                    'version: alpha\n'
+                    'name: asset-judgment-fixture\n'
+                    'description: >\n'
+                    '  A restrained fixture with explicit image-role judgments and traced layout slots.\n'
+                    'assets:\n'
+                    '  - id: logo-primary\n'
+                    '    source_media: partner-logo.png\n'
+                    '    kind: logo\n'
+                    '    on-bg: light\n'
+                    'asset_vision_groups:\n'
+                    '  - id: vision-1\n'
+                    '    source_media: [partner-logo.png, ornament.png]\n'
+                    '    visual_kind: content-image\n'
+                    'asset_decisions:\n'
+                    '  - source_media: ornament.png\n'
+                    '    visual_kind: decorative\n'
+                )
+            with open(os.path.join(lout, 'frontmatter.yaml'), 'w', encoding='utf-8') as stream:
+                stream.write('')
+            with open(os.path.join(lout, 'layouts.yaml'), 'w', encoding='utf-8') as stream:
+                stream.write(
+                    'roles:\n'
+                    '  content: TODO角色\n'
+                    'layouts:\n'
+                    '  content:\n'
+                    '    name: "内容页"\n'
+                    '    role: content\n'
+                    '    slots:\n'
+                    '      - {role: logo, box: [100, 200, 400, 200], type: pic, '
+                    'asset: logo-primary, source_media: partner-logo.png}\n'
+                    '      - {role: asset-candidate, box: [40, 40, 120, 120], type: pic, '
+                    'source_media: ornament.png}\n'
+                    '    confidence: high\n'
+                )
+            with open(os.path.join(lout, 'layout-controls.yaml'), 'w',
+                      encoding='utf-8') as stream:
+                stream.write('roles:\n  content: content\n')
+            with open(os.path.join(lout, 'body.md'), 'w', encoding='utf-8') as stream:
+                stream.write(
+                    '## Usage\n\n'
+                    'Read `layouts.md` before composing a slide.\n\n'
+                    '## Assets\n\n'
+                    '{{ASSET_TABLE}}\n\n'
+                    '## Hard Rules\n\n'
+                    '{{LOGO_RULES}}\n'
+                )
+            check_v1 = os.path.join(root, 'check_v1.py')
+            with open(check_v1, 'w', encoding='utf-8') as stream:
+                stream.write('raise SystemExit(0)\n')
+
+            rc = package_main([stage1, lout, pack, '--check-v1', check_v1])
+
+            self.assertEqual(rc, 0)
+            with open(os.path.join(pack, 'design.md'), encoding='utf-8') as stream:
+                design = stream.read()
+            with open(os.path.join(pack, 'layouts.md'), encoding='utf-8') as stream:
+                layouts = stream.read()
+            self.assertNotIn('asset_decisions', design)
+            self.assertNotIn('asset_vision_groups', design)
+            self.assertNotIn('visual_kind', design)
+            self.assertNotIn('source_media', design)
+            self.assertNotIn('source_media', layouts)
+            self.assertNotIn('TODO', layouts)
+            self.assertNotIn('{{LOGO_RULES}}', design)
+            self.assertNotIn('logo-primary', design)
+            self.assertNotIn('logo-primary', layouts)
+            self.assertIn(
+                '- {role: pic, box: [100, 200, 400, 200], type: pic}',
+                layouts,
+            )
+            self.assertIn('texture-1:', design)
+            self.assertIn('kind: texture', design)
+            self.assertIn(
+                '- {role: texture, box: [40, 40, 120, 120], type: pic, asset: texture-1}',
+                layouts,
+            )
+            gate = Pack(pack)
+            self.assertFalse([
+                result
+                for result in (rule(gate) for rule in RULES)
+                if result.fails
+            ])
+
+    def test_package_places_a_rendered_picture_instance_as_a_logo_asset(self):
+        try:
+            from PIL import Image
+        except ImportError:
+            self.skipTest('Pillow unavailable')
+
+        with tempfile.TemporaryDirectory() as root:
+            stage1 = os.path.join(root, 'stage1')
+            lout = os.path.join(stage1, 'l-out')
+            pack = os.path.join(root, 'pack')
+            media = os.path.join(stage1, 'media-out')
+            ref = os.path.join(stage1, 'ref')
+            os.makedirs(lout)
+            os.makedirs(media)
+            os.makedirs(ref)
+
+            source = 'generated/instance/picture-directory-logo.webp'
+            output = 'picture-directory-logo.webp'
+            box = [1813, 18, 59, 228]
+            Image.new('RGBA', (59, 228), (20, 60, 100, 255)).save(
+                os.path.join(media, output), 'WEBP')
+            extract = {
+                'source': {'filename': 'fixture.pptx'},
+                'canvas': {
+                    'px': [1920, 1080],
+                    'source': {'cx': 12192000, 'cy': 6858000, 'unit': 'EMU'},
+                },
+                'media': [{
+                    'media': source,
+                    'out': 'media-out/' + output,
+                    'generated': True,
+                    'rendered_from': 'ppt/media/original-logo.png',
+                }],
+                'images': [{
+                    'media': source,
+                    'boxes': [{
+                        'count': 1,
+                        'box': dict(zip(('x', 'y', 'w', 'h'), box)),
+                        'parts': ['ppt/slides/slide2.xml'],
+                    }],
+                }],
+                'color_freq': [],
+                'text_scale': [],
+            }
+            with open(os.path.join(stage1, 'extract.json'), 'w', encoding='utf-8') as stream:
+                json.dump(extract, stream)
+            with open(os.path.join(ref, 'shapes.json'), 'w', encoding='utf-8') as stream:
+                json.dump({'shapes': [{
+                    'part': 'ppt/slides/slide2.xml',
+                    'kind': 'pic',
+                    'box': dict(zip(('x', 'y', 'w', 'h'), box)),
+                }]}, stream)
+            with open(os.path.join(lout, 'asset-vision-groups.json'),
+                      'w', encoding='utf-8') as stream:
+                json.dump({
+                    'version': 2,
+                    'selected': [{
+                        'candidates': [{
+                            'id': 'asset-directory-logo',
+                            'source_media': output,
+                            'placements': [{
+                                'slide': 2,
+                                'archetype': 'content',
+                                'box': box,
+                            }],
+                        }],
+                    }],
+                    'omitted': [],
+                }, stream)
+            with open(os.path.join(lout, 'manifest.yaml'), 'w', encoding='utf-8') as stream:
+                stream.write(
+                    'version: alpha\n'
+                    'name: rendered-instance-fixture\n'
+                    'description: Rendered local PPT image instances remain reusable assets.\n'
+                    'asset_vision_groups:\n'
+                    '  - id: asset-directory-logo\n'
+                    '    source_media: picture-directory-logo.webp\n'
+                    '    box: [1813, 18, 59, 228]\n'
+                    '    visual_kind: logo\n'
+                )
+            for name, content in (
+                ('frontmatter.yaml', ''),
+                ('body.md', 'Read `layouts.md` before composing a slide.\n\n{{ASSET_TABLE}}\n'),
+                ('layouts.yaml',
+                 'layouts:\n'
+                 '  content:\n'
+                 '    name: "目录"\n'
+                 '    role: content\n'
+                 '    slots:\n'
+                 '      - {role: asset-candidate, box: [1813, 18, 59, 228], type: pic, '
+                 'source_media: picture-directory-logo.webp}\n'
+                 '    confidence: high\n'),
+            ):
+                with open(os.path.join(lout, name), 'w', encoding='utf-8') as stream:
+                    stream.write(content)
+            check_v1 = os.path.join(root, 'check_v1.py')
+            with open(check_v1, 'w', encoding='utf-8') as stream:
+                stream.write('raise SystemExit(0)\n')
+
+            self.assertEqual(package_main([stage1, lout, pack, '--check-v1', check_v1]), 0)
+
+            with open(os.path.join(pack, 'design.md'), encoding='utf-8') as stream:
+                design = stream.read()
+            with open(os.path.join(pack, 'layouts.md'), encoding='utf-8') as stream:
+                layouts = stream.read()
+            self.assertTrue(os.path.isfile(os.path.join(pack, 'assets', 'logos', '1.webp')))
+            self.assertIn('logo-1:', design)
+            self.assertIn('assets/logos/1.webp', design)
+            self.assertIn(
+                '- {role: logo, box: [1813, 18, 59, 228], type: pic, asset: logo-1}',
+                layouts,
+            )
+
+
+if __name__ == '__main__':
+    unittest.main()