Skip to content

AI turns documents or topics into real, native PowerPoint decks—with native shapes, transitions and animations, data-backed charts and tables on demand, audio narration from speaker notes, and support for your own .pptx templates. · by Hugo He

fix(tools): 支持混合格式导出。

shandianchengzi048f2a885daa
diff --git a/tools/svg_to_pptx.py b/tools/svg_to_pptx.py
index ac96c561..2c22428d 100644
--- a/tools/svg_to_pptx.py
+++ b/tools/svg_to_pptx.py
@@ -251,8 +251,8 @@ def find_notes_files(project_path: Path, svg_files: List[Path] = None) -> dict:
"""
查找项目中的备注文件
- 支持两种匹配模式:
- 1. 按文件名匹配(推荐):notes/01_封面.md 对应 01_封面.svg
+ 支持两种匹配模式(支持混合匹配):
+ 1. 按文件名匹配(优先):notes/01_封面.md 对应 01_封面.svg
2. 按序号匹配(向后兼容):notes/slide01.md 对应第1个 SVG
Args:
@@ -268,10 +268,13 @@ def find_notes_files(project_path: Path, svg_files: List[Path] = None) -> dict:
if not notes_dir.exists():
return notes
- # 收集所有 notes 文件信息
- notes_by_name = {} # 按文件名
- notes_by_num = {} # 按序号(向后兼容)
+ svg_stems_mapping = {}
+
+ if svg_files:
+ for i, svg_path in enumerate(svg_files, 1):
+ svg_stems_mapping[svg_path.stem] = i
+ # 收集所有 notes 文件信息
for notes_file in notes_dir.glob('*.md'):
try:
with open(notes_file, 'r', encoding='utf-8') as f:
@@ -280,29 +283,18 @@ def find_notes_files(project_path: Path, svg_files: List[Path] = None) -> dict:
continue
stem = notes_file.stem
- notes_by_name[stem] = content
-
+
# 尝试提取序号(向后兼容 slide01.md 格式)
match = re.search(r'slide[_]?(\d+)', stem)
if match:
- notes_by_num[int(match.group(1))] = content
+ notes[int(match.group(1))] = content
+
+ # 按文件名提取(覆盖向后兼容的格式)
+ if stem in svg_stems_mapping:
+ notes[svg_stems_mapping[stem]] = content
except Exception:
pass
- # 如果提供了 SVG 文件列表,优先按文件名匹配
- if svg_files:
- for i, svg_path in enumerate(svg_files, 1):
- svg_stem = svg_path.stem
- if svg_stem in notes_by_name:
- # 按文件名匹配
- notes[svg_stem] = notes_by_name[svg_stem]
- elif i in notes_by_num:
- # 回退到按序号匹配
- notes[svg_stem] = notes_by_num[i]
- else:
- # 没有 SVG 文件列表时,返回序号索引(向后兼容)
- notes = notes_by_num
-
return notes