feat: support PPTX uploads in knowledge base

astrbot/core/knowledge_base/kb_helper.py

@@ -282,6 +282,18 @@ async def upload_document(
                     parse_result = await parser.parse(file_content, file_name)
                 except KnowledgeBaseUploadError:
                     raise
+                except ValueError as exc:
+                    message = str(exc)
+                    if not message.startswith("暂时不支持的文件格式"):
+                        message = (
+                            "文档解析失败：无法读取或解析上传文件。"
+                            "请确认文件格式受支持且文件内容未损坏。"
+                        )
+                    raise KnowledgeBaseUploadError(
+                        stage="parsing",
+                        user_message=message,
+                        details={"file_name": file_name},
+                    ) from exc
                 except Exception as exc:
                     raise KnowledgeBaseUploadError(
                         stage="parsing",
@@ -334,6 +346,7 @@ async def upload_document(
                         ".md",
                         ".mdx",
                         ".mkd",
+                        ".pptx",
                         ".rst",
                         ".xls",
                         ".xlsx",

astrbot/core/knowledge_base/parsers/markitdown_parser.py

@@ -10,7 +10,7 @@
 
 
 class MarkitdownParser(BaseParser):
-    """解析 docx, xls, xlsx 格式"""
+    """解析 docx, xls, xlsx, pptx 格式"""
 
     async def parse(self, file_content: bytes, file_name: str) -> ParseResult:
         md = MarkItDown(enable_plugins=False)

astrbot/core/knowledge_base/parsers/util.py

@@ -2,7 +2,17 @@
 
 
 async def select_parser(ext: str) -> BaseParser:
-    if ext in {".md", ".txt", ".markdown", ".rst", ".adoc", ".xlsx", ".docx", ".xls"}:
+    if ext in {
+        ".md",
+        ".txt",
+        ".markdown",
+        ".rst",
+        ".adoc",
+        ".xlsx",
+        ".docx",
+        ".xls",
+        ".pptx",
+    }:
         from .markitdown_parser import MarkitdownParser
 
         return MarkitdownParser()

dashboard/src/i18n/locales/en-US/features/knowledge-base/detail.json

@@ -53,7 +53,7 @@
     "title": "Upload Document",
     "selectFile": "Select File",
     "dropzone": "Drop files here or click to select",
-    "supportedFormats": "Supported formats: .txt, .md, .markdown, .rst, .adoc, .pdf, .docx, .epub, .xls, .xlsx",
+    "supportedFormats": "Supported formats: .txt, .md, .markdown, .rst, .adoc, .pdf, .docx, .pptx, .epub, .xls, .xlsx",
     "maxSize": "Max file size: 128MB",
     "chunkSettings": "Chunk Settings",
     "batchSettings": "Batch Settings",

dashboard/src/i18n/locales/ja-JP/features/knowledge-base/detail.json

@@ -53,7 +53,7 @@
     "title": "ドキュメントをアップロード",
     "selectFile": "ファイルを選択",
     "dropzone": "ここにファイルをドラッグ＆ドロップするか、クリックして選択",
-    "supportedFormats": "対応形式：.txt, .md, .markdown, .rst, .adoc, .pdf, .docx, .epub, .xls, .xlsx",
+    "supportedFormats": "対応形式：.txt, .md, .markdown, .rst, .adoc, .pdf, .docx, .pptx, .epub, .xls, .xlsx",
     "maxSize": "最大ファイルサイズ：128 MB",
     "chunkSettings": "チャンク設定",
     "batchSettings": "バッチ処理設定",

dashboard/src/i18n/locales/ru-RU/features/knowledge-base/detail.json

@@ -53,7 +53,7 @@
         "title": "Добавление контента",
         "selectFile": "Файл",
         "dropzone": "Нажмите или перетащите файл сюда",
-        "supportedFormats": "Форматы: .txt, .md, .markdown, .rst, .adoc, .pdf, .docx, .epub, .xls, .xlsx",
+        "supportedFormats": "Форматы: .txt, .md, .markdown, .rst, .adoc, .pdf, .docx, .pptx, .epub, .xls, .xlsx",
         "maxSize": "Максимум: 128MB",
         "chunkSettings": "Фрагментация",
         "batchSettings": "Пакетная обработка",

dashboard/src/i18n/locales/zh-CN/features/knowledge-base/detail.json

@@ -53,7 +53,7 @@
     "title": "上传文档",
     "selectFile": "选择文件",
     "dropzone": "拖放文件到这里或点击选择",
-    "supportedFormats": "支持的格式: .txt, .md, .markdown, .rst, .adoc, .pdf, .docx, .epub, .xls, .xlsx",
+    "supportedFormats": "支持的格式: .txt, .md, .markdown, .rst, .adoc, .pdf, .docx, .pptx, .epub, .xls, .xlsx",
     "maxSize": "最大文件大小: 128MB",
     "chunkSettings": "分块设置",
     "batchSettings": "批处理设置",

dashboard/src/views/knowledge-base/DocumentDetail.vue

@@ -398,6 +398,7 @@ const deleteChunk = async (chunk: any) => {
 const getFileIcon = (fileType: string) => {
   const type = fileType?.toLowerCase() || ''
   if (type.includes('pdf')) return 'mdi-file-pdf-box'
+  if (type.includes('ppt')) return 'mdi-file-powerpoint-box'
   if (type.includes('epub')) return 'mdi-book-open-page-variant'
   if (type.includes('md')) return 'mdi-language-markdown'
   if (type.includes('txt')) return 'mdi-file-document-outline'

dashboard/src/views/knowledge-base/components/DocumentsTab.vue

@@ -87,7 +87,7 @@
                 <p class="mt-4 text-h6">{{ t('upload.dropzone') }}</p>
                 <p class="text-caption text-medium-emphasis mt-2">{{ t('upload.supportedFormats') }}</p>
                 <p class="text-caption text-medium-emphasis">{{ t('upload.maxSize') }}</p>
-                <input ref="fileInput" type="file" multiple hidden accept=".txt,.md,.markdown,.rst,.adoc,.pdf,.docx,.epub,.xls,.xlsx"
+                <input ref="fileInput" type="file" multiple hidden accept=".txt,.md,.markdown,.rst,.adoc,.pdf,.docx,.pptx,.epub,.xls,.xlsx"
                   @change="handleFileSelect" />
               </div>
 
@@ -730,6 +730,7 @@ const deleteDocument = async () => {
 const getFileIcon = (fileType: string) => {
   const type = fileType?.toLowerCase() || ''
   if (type.includes('pdf')) return 'mdi-file-pdf-box'
+  if (type.includes('ppt')) return 'mdi-file-powerpoint-box'
   if (type.includes('epub')) return 'mdi-book-open-page-variant'
   if (type.includes('rst') || type.includes('adoc')) return 'mdi-file-document-outline'
   if (type.includes('md') || type.includes('markdown')) return 'mdi-language-markdown'

pyproject.toml

@@ -59,7 +59,7 @@ dependencies = [
   "aiofiles>=25.1.0",
   "rank-bm25>=0.2.2",
   "jieba>=0.42.1",
-  "markitdown-no-magika[docx,xls,xlsx]>=0.1.2",
+  "markitdown-no-magika[docx,xls,xlsx,pptx]>=0.1.2",
   "xinference-client",
   "tenacity>=9.1.2",
   "shipyard-python-sdk>=0.2.4",

requirements.txt

@@ -48,7 +48,7 @@ pysocks>=1.7.1
 aiofiles>=25.1.0
 rank-bm25>=0.2.2
 jieba>=0.42.1
-markitdown-no-magika[docx,xls,xlsx]>=0.1.2
+markitdown-no-magika[docx,xls,xlsx,pptx]>=0.1.2
 xinference-client
 tenacity>=9.1.2
 shipyard-python-sdk>=0.2.4

tests/test_kb_pptx_parser.py

@@ -0,0 +1,42 @@
+from __future__ import annotations
+
+import io
+
+import pytest
+
+from astrbot.core.knowledge_base.parsers.markitdown_parser import MarkitdownParser
+from astrbot.core.knowledge_base.parsers.util import select_parser
+
+
+def _make_pptx_bytes() -> bytes:
+    from pptx import Presentation
+
+    presentation = Presentation()
+    slide = presentation.slides.add_slide(presentation.slide_layouts[0])
+    slide.shapes.title.text = "KB PPTX Slide Title"
+    buffer = io.BytesIO()
+    presentation.save(buffer)
+    return buffer.getvalue()
+
+
+@pytest.mark.asyncio
+async def test_select_parser_supports_pptx():
+    parser = await select_parser(".pptx")
+
+    assert isinstance(parser, MarkitdownParser)
+
+
+@pytest.mark.asyncio
+async def test_select_parser_rejects_legacy_ppt():
+    with pytest.raises(ValueError, match="暂时不支持的文件格式"):
+        await select_parser(".ppt")
+
+
+@pytest.mark.asyncio
+async def test_pptx_parser_extracts_slide_text():
+    pytest.importorskip("pptx")
+
+    result = await MarkitdownParser().parse(_make_pptx_bytes(), "slides.pptx")
+
+    assert result.media == []
+    assert "KB PPTX Slide Title" in result.text

tests/unit/test_kb_upload_atomicity.py

@@ -354,6 +354,7 @@ async def fake_save_media(**kwargs):
     ("file_name", "file_type"),
     [
         ("guide.docx", "docx"),
+        ("guide.pptx", "pptx"),
         ("guide.xlsx", "xlsx"),
         ("guide.xls", "xls"),
         ("guide.rst", "rst"),
@@ -419,6 +420,43 @@ async def test_upload_document_preserves_markdown_heading_paths(
     assert embedding_contents[2] == f"guide\n\n{contents[2]}"
 
 
+@pytest.mark.asyncio
+async def test_upload_document_rejects_legacy_ppt(
+    tmp_path: Path,
+    stub_provider_manager_module,
+) -> None:
+    """Legacy .ppt must fail as unsupported format, not as a corrupt file."""
+    KBHelper = _import_kb_helper()
+
+    helper = KBHelper.__new__(KBHelper)
+    helper.kb = KnowledgeBase(
+        kb_name="Test KB",
+        description="",
+        embedding_provider_id="emb",
+    )
+    helper.kb_db = MagicMock()
+    helper.kb_db.get_db = _failing_get_db()
+    helper.vec_db = AsyncMock()
+    helper.kb_medias_dir = tmp_path / "medias"
+    helper.chunker = AsyncMock()
+
+    with (
+        patch.object(helper, "_ensure_vec_db", new=AsyncMock()),
+        pytest.raises(KnowledgeBaseUploadError) as exc_info,
+    ):
+        await helper.upload_document(
+            file_name="slides.ppt",
+            file_content=b"not a pptx",
+            file_type="ppt",
+        )
+
+    assert exc_info.value.stage == "parsing"
+    assert "暂时不支持的文件格式" in exc_info.value.user_message
+    assert ".ppt" in exc_info.value.user_message
+    helper.chunker.chunk.assert_not_awaited()
+    helper.vec_db.insert_batch.assert_not_awaited()
+
+
 @pytest.mark.asyncio
 async def test_upload_document_cleans_up_on_metadata_failure(
     stub_provider_manager_module,