WeKnora deployment source with the reviewed wiki identity, summary and recovery fixes

2026-09-22 16:19:06 +08:00
commit f7fc992f57
3929 changed files with 1006607 additions and 0 deletions
@@ -0,0 +1,63 @@
+package service
+
+import (
+	"context"
+	"strings"
+	"testing"
+
+	"github.com/Tencent/WeKnora/internal/types"
+	"github.com/stretchr/testify/require"
+)
+
+type digitalHumanTestReader struct {
+	request *types.ReadRequest
+	result  *types.ReadResult
+	err     error
+}
+
+func (r *digitalHumanTestReader) Read(_ context.Context, request *types.ReadRequest) (*types.ReadResult, error) {
+	r.request = request
+	return r.result, r.err
+}
+
+func TestDigitalHumanUploadPreservesSubtitleEvidence(t *testing.T) {
+	service := &DigitalHumanService{}
+	for _, ext := range []string{"TXT", "md", "markdown", "srt", "vtt"} {
+		source, err := service.readDigitalHumanUpload(context.Background(), DigitalHumanUpload{Name: "source." + ext, Data: []byte("\ufeff1\n00:00:01,000 --> 00:00:02,000\nOriginal words\n")})
+		require.NoError(t, err)
+		require.Equal(t, "1\n00:00:01,000 --> 00:00:02,000\nOriginal words", source.Text)
+	}
+}
+
+func TestDigitalHumanUploadRejectsInvalidOrTruncatedMaterials(t *testing.T) {
+	service := &DigitalHumanService{}
+	for _, file := range []DigitalHumanUpload{
+		{Name: "source.exe", Data: []byte("text")},
+		{Name: "empty.pdf"},
+		{Name: "empty.txt", Data: []byte("  \n")},
+		{Name: "invalid.txt", Data: []byte{0xff}},
+		{Name: "large.pdf", Data: make([]byte, DigitalHumanFileLimit+1)},
+		{Name: "too-many-words.txt", Data: []byte(strings.Repeat("x", digitalHumanSourceLimit+1))},
+	} {
+		_, err := service.readDigitalHumanUpload(context.Background(), file)
+		require.Error(t, err, file.Name)
+	}
+}
+
+func TestDigitalHumanUploadDispatchesDocumentsToReader(t *testing.T) {
+	if types.DefaultParserEngine("pdf") == "anydoc" {
+		t.Skip("linked builds exercise the local parser instead of the remote transport")
+	}
+	reader := &digitalHumanTestReader{result: &types.ReadResult{MarkdownContent: "# Source\nAll source evidence"}}
+	service := &DigitalHumanService{reader: reader}
+	for _, ext := range []string{"doc", "docx", "pdf", "ppt", "pptx", "rtf", "odt"} {
+		source, err := service.readDigitalHumanUpload(context.Background(), DigitalHumanUpload{Name: "source." + ext, Data: []byte("document bytes")})
+		require.NoError(t, err, ext)
+		require.Equal(t, "# Source\nAll source evidence", source.Text)
+		require.Equal(t, ext, reader.request.FileType)
+		require.Equal(t, "false", reader.request.ParserEngineOverrides["anydoc_extract_images"])
+	}
+	reader.result = &types.ReadResult{}
+	_, err := service.readDigitalHumanUpload(context.Background(), DigitalHumanUpload{Name: "scan.pdf", Data: []byte("scan")})
+	require.Error(t, err)
+}