fix: support rst and adoc knowledge uploads (#8255)

Author: he-yufeng

astrbot/core/knowledge_base/parsers/util.py

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

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

@@ -49,7 +49,7 @@
     "title": "Upload Document",
     "selectFile": "Select File",
     "dropzone": "Drop files here or click to select",
-    "supportedFormats": "Supported formats: .txt, .md, .pdf, .docx, .epub, .xls, .xlsx",
+    "supportedFormats": "Supported formats: .txt, .md, .markdown, .rst, .adoc, .pdf, .docx, .epub, .xls, .xlsx",
     "maxSize": "Max file size: 128MB",
     "chunkSettings": "Chunk Settings",
     "batchSettings": "Batch Settings",

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

@@ -49,7 +49,7 @@
         "title": "Добавление контента",
         "selectFile": "Файл",
         "dropzone": "Нажмите или перетащите файл сюда",
-        "supportedFormats": "Форматы: .txt, .md, .pdf, .docx, .epub, .xls, .xlsx",
+        "supportedFormats": "Форматы: .txt, .md, .markdown, .rst, .adoc, .pdf, .docx, .epub, .xls, .xlsx",
         "maxSize": "Максимум: 128MB",
         "chunkSettings": "Фрагментация",
         "batchSettings": "Пакетная обработка",

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

@@ -49,7 +49,7 @@
     "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, .pdf, .docx, .epub, .xls, .xlsx",
+    "supportedFormats": "支持的格式: .txt, .md, .markdown, .rst, .adoc, .pdf, .docx, .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/components/DocumentsTab.vue

@@ -85,7 +85,7 @@
                 <p class="text-caption text-medium-emphasis mt-2">{{ t('upload.supportedFormats') }}</p>
                 <p class="text-caption text-medium-emphasis">{{ t('upload.maxSize') }}</p>
                 <p class="text-caption text-medium-emphasis">最多可上传 10 个文件</p>
-                <input ref="fileInput" type="file" multiple hidden accept=".txt,.md,.pdf,.docx,.epub,.xls,.xlsx"
+                <input ref="fileInput" type="file" multiple hidden accept=".txt,.md,.markdown,.rst,.adoc,.pdf,.docx,.epub,.xls,.xlsx"
                   @change="handleFileSelect" />
               </div>
 
@@ -709,6 +709,7 @@ const getFileIcon = (fileType: string) => {
   const type = fileType?.toLowerCase() || ''
   if (type.includes('pdf')) return 'mdi-file-pdf-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'
   if (type.includes('txt')) return 'mdi-file-document-outline'
   if (type.includes('url')) return 'mdi-link-variant'
@@ -719,6 +720,7 @@ const getFileColor = (fileType: string) => {
   const type = fileType?.toLowerCase() || ''
   if (type.includes('pdf')) return 'error'
   if (type.includes('epub')) return 'warning'
+  if (type.includes('rst') || type.includes('adoc')) return 'success'
   if (type.includes('md')) return 'info'
   if (type.includes('txt')) return 'success'
   if (type.includes('url')) return 'primary'

tests/test_epub_parser.py

@@ -6,6 +6,7 @@
 import pytest
 
 from astrbot.core.knowledge_base.parsers.epub_parser import EpubParser
+from astrbot.core.knowledge_base.parsers.markitdown_parser import MarkitdownParser
 from astrbot.core.knowledge_base.parsers.util import select_parser
 
 
@@ -174,6 +175,14 @@ async def test_select_parser_supports_epub():
     assert isinstance(parser, EpubParser)
 
 
+@pytest.mark.asyncio
+@pytest.mark.parametrize("ext", [".rst", ".adoc"])
+async def test_select_parser_supports_text_markup_formats(ext):
+    parser = await select_parser(ext)
+
+    assert isinstance(parser, MarkitdownParser)
+
+
 @pytest.mark.asyncio
 async def test_epub_parser_reads_spine_order_as_text():
     result = await EpubParser().parse(_make_epub_bytes(), "book.epub")