feat(imitation): accept pasted source content as an alternative to URL

- Make source_url optional and add source_content to the request; the two are normalized into a source kind ("url" | "content"), URL wins when both are present, and missing both returns source_required - Skip web fetching in the generation worker when pasted content is provided; truncate it to the same 24,000-rune cap - Label untitled pasted-content jobs with a locale-aware display title and mark the source kind in wizard state and prompt metadata - Cover source normalization, truncation, and display title with tests
2026-07-21 08:41:55 +08:00
parent 403ff91789
commit d62f7b3c91
3 changed files with 151 additions and 22 deletions
@@ -26,6 +26,8 @@ import (
 const (
 	articleImitationTaskType       = "imitation"
 	articleImitationSourceType     = "imitation"
+	articleImitationSourceURL      = "url"
+	articleImitationSourceContent  = "content"
 	articleImitationFallbackTitle  = "仿写文章生成中"
 	maxImitationSourceContentRunes = 24000
 	defaultImitationWebSearchLimit = int32(5)
@@ -44,7 +46,8 @@ type ArticleImitationService struct {
 }

 type GenerateImitationRequest struct {
-	SourceURL         string   `json:"source_url" binding:"required"`
+	SourceURL         string   `json:"source_url"`
+	SourceContent     string   `json:"source_content"`
 	SourceTitle       string   `json:"source_title"`
 	Locale            string   `json:"locale"`
 	BrandName         string   `json:"brand_name"`
@@ -130,7 +133,7 @@ func (s *ArticleImitationService) Generate(ctx context.Context, req GenerateImit
 		return nil, response.ErrServiceUnavailable(50311, "llm_unavailable", err.Error())
 	}

-	sourceURL, err := normalizeImitationSourceURL(req.SourceURL)
+	sourceURL, sourceContent, sourceKind, err := normalizeImitationSource(req)
 	if err != nil {
 		return nil, err
 	}
@@ -148,15 +151,17 @@ func (s *ArticleImitationService) Generate(ctx context.Context, req GenerateImit
 		return nil, response.ErrForbidden(40301, "quota_insufficient", "generation quota insufficient, please upgrade your plan")
 	}

-	inputParams := buildImitationInputParams(sourceURL, req)
+	inputParams := buildImitationInputParams(sourceURL, sourceContent, sourceKind, req)
 	inputParams["brand_id"] = brandID
 	inputParams = attachBrandPromptContext(inputParams, brandPromptContext)
 	inputJSON, _ := json.Marshal(inputParams)
 	initialTitle := buildImitationInitialTitle(req.SourceTitle)
+	sourceDisplayTitle := imitationSourceDisplayTitle(sourceKind, req.SourceTitle, req.Locale)
 	wizardStateJSON, _ := json.Marshal(map[string]interface{}{
 		"title":        initialTitle,
 		"source_url":   sourceURL,
-		"source_title": strings.TrimSpace(req.SourceTitle),
+		"source_kind":  sourceKind,
+		"source_title": sourceDisplayTitle,
 		"current_step": 0,
 	})

@@ -332,24 +337,28 @@ func (s *ArticleImitationService) executeGeneration(ctx context.Context, job art
 	job.InputParams = attachBrandPromptContext(job.InputParams, brandPromptContext)

 	sourceURL := strings.TrimSpace(extractString(job.InputParams, "source_url"))
-	if sourceURL == "" {
-		s.failGeneration(ctx, job, "task_load", fmt.Errorf("source_url is required"))
-		return
-	}
-	if s.knowledge == nil {
-		s.failGeneration(ctx, job, "source_fetch", fmt.Errorf("knowledge service is not configured"))
-		return
-	}
+	sourceContent := ""
+	if sourceURL != "" {
+		if s.knowledge == nil {
+			s.failGeneration(ctx, job, "source_fetch", fmt.Errorf("knowledge service is not configured"))
+			return
+		}

-	parsed, err := s.knowledge.parseWebsiteKnowledge(ctx, sourceURL)
-	if err != nil {
-		s.failGeneration(ctx, job, "source_fetch", err)
-		return
-	}
+		parsed, parseErr := s.knowledge.parseWebsiteKnowledge(ctx, sourceURL)
+		if parseErr != nil {
+			s.failGeneration(ctx, job, "source_fetch", parseErr)
+			return
+		}

-	sourceContent := truncateRunes(strings.TrimSpace(parsed.Markdown), maxImitationSourceContentRunes)
-	if sourceContent == "" {
-		sourceContent = truncateRunes(strings.TrimSpace(parsed.Text), maxImitationSourceContentRunes)
+		sourceContent = truncateRunes(strings.TrimSpace(parsed.Markdown), maxImitationSourceContentRunes)
+		if sourceContent == "" {
+			sourceContent = truncateRunes(strings.TrimSpace(parsed.Text), maxImitationSourceContentRunes)
+		}
+	} else {
+		sourceContent = truncateRunes(
+			strings.TrimSpace(extractString(job.InputParams, "source_content")),
+			maxImitationSourceContentRunes,
+		)
 	}
 	if sourceContent == "" {
 		s.failGeneration(ctx, job, "source_fetch", fmt.Errorf("source article content is empty"))
@@ -625,9 +634,27 @@ func normalizeImitationSourceURL(raw string) (string, error) {
 	return parsed.String(), nil
 }

-func buildImitationInputParams(sourceURL string, req GenerateImitationRequest) map[string]interface{} {
+func normalizeImitationSource(req GenerateImitationRequest) (string, string, string, error) {
+	if strings.TrimSpace(req.SourceURL) != "" {
+		sourceURL, err := normalizeImitationSourceURL(req.SourceURL)
+		if err != nil {
+			return "", "", "", err
+		}
+		return sourceURL, "", articleImitationSourceURL, nil
+	}
+
+	sourceContent := truncateRunes(strings.TrimSpace(req.SourceContent), maxImitationSourceContentRunes)
+	if sourceContent == "" {
+		return "", "", "", response.ErrBadRequest(40001, "source_required", "source_url or source_content is required")
+	}
+	return "", sourceContent, articleImitationSourceContent, nil
+}
+
+func buildImitationInputParams(sourceURL, sourceContent, sourceKind string, req GenerateImitationRequest) map[string]interface{} {
 	params := map[string]interface{}{
 		"source_url":          sourceURL,
+		"source_content":      sourceContent,
+		"source_kind":         sourceKind,
 		"source_title":        strings.TrimSpace(req.SourceTitle),
 		"locale":              normalizeImitationLocale(req.Locale),
 		"brand_name":          strings.TrimSpace(req.BrandName),
@@ -654,10 +681,22 @@ func buildImitationInitialTitle(sourceTitle string) string {
 	return "仿写：" + title
 }

+func imitationSourceDisplayTitle(sourceKind, sourceTitle, locale string) string {
+	title := strings.TrimSpace(sourceTitle)
+	if title != "" || sourceKind != articleImitationSourceContent {
+		return title
+	}
+	if normalizeImitationLocale(locale) == "en-US" {
+		return "Pasted content"
+	}
+	return "粘贴正文"
+}
+
 func buildImitationGenerationPrompt(params map[string]interface{}, sourceContent string, knowledgePrompt string) string {
 	locale := normalizeImitationLocale(extractString(params, "locale"))
 	sourceTitle := strings.TrimSpace(extractString(params, "source_title"))
 	sourceURL := strings.TrimSpace(extractString(params, "source_url"))
+	sourceKind := strings.TrimSpace(extractString(params, "source_kind"))
 	keywords := extractStringList(params["keywords"], 16)
 	knowledgePrompt = strings.TrimSpace(knowledgePrompt)

@@ -667,6 +706,8 @@ func buildImitationGenerationPrompt(params map[string]interface{}, sourceContent
 	}
 	if sourceURL != "" {
 		appendRawPromptLine(&sourceMeta, "URL", sourceURL)
+	} else if sourceKind == articleImitationSourceContent {
+		appendRawPromptLine(&sourceMeta, "来源", "用户粘贴正文")
 	}

 	var settings strings.Builder
@@ -0,0 +1,87 @@
+package app
+
+import (
+	"strings"
+	"testing"
+)
+
+func TestNormalizeImitationSourcePrefersURL(t *testing.T) {
+	t.Parallel()
+
+	sourceURL, sourceContent, sourceKind, err := normalizeImitationSource(GenerateImitationRequest{
+		SourceURL:     " https://example.com/article ",
+		SourceContent: "pasted content should be ignored",
+	})
+	if err != nil {
+		t.Fatalf("normalizeImitationSource() error = %v", err)
+	}
+	if sourceURL != "https://example.com/article" {
+		t.Fatalf("sourceURL = %q, want normalized URL", sourceURL)
+	}
+	if sourceContent != "" {
+		t.Fatalf("sourceContent = %q, want empty when URL is present", sourceContent)
+	}
+	if sourceKind != articleImitationSourceURL {
+		t.Fatalf("sourceKind = %q, want %q", sourceKind, articleImitationSourceURL)
+	}
+}
+
+func TestNormalizeImitationSourceAcceptsPastedContent(t *testing.T) {
+	t.Parallel()
+
+	sourceURL, sourceContent, sourceKind, err := normalizeImitationSource(GenerateImitationRequest{
+		SourceContent: "  article body  ",
+	})
+	if err != nil {
+		t.Fatalf("normalizeImitationSource() error = %v", err)
+	}
+	if sourceURL != "" {
+		t.Fatalf("sourceURL = %q, want empty", sourceURL)
+	}
+	if sourceContent != "article body" {
+		t.Fatalf("sourceContent = %q, want trimmed content", sourceContent)
+	}
+	if sourceKind != articleImitationSourceContent {
+		t.Fatalf("sourceKind = %q, want %q", sourceKind, articleImitationSourceContent)
+	}
+}
+
+func TestNormalizeImitationSourceRejectsMissingSource(t *testing.T) {
+	t.Parallel()
+
+	_, _, _, err := normalizeImitationSource(GenerateImitationRequest{})
+	if err == nil {
+		t.Fatal("normalizeImitationSource() error = nil, want source_required")
+	}
+	if err.Error() != "source_required" {
+		t.Fatalf("normalizeImitationSource() error = %q, want source_required", err.Error())
+	}
+}
+
+func TestNormalizeImitationSourceTruncatesPastedContent(t *testing.T) {
+	t.Parallel()
+
+	_, sourceContent, _, err := normalizeImitationSource(GenerateImitationRequest{
+		SourceContent: strings.Repeat("文", maxImitationSourceContentRunes+100),
+	})
+	if err != nil {
+		t.Fatalf("normalizeImitationSource() error = %v", err)
+	}
+	if got := len([]rune(sourceContent)); got != maxImitationSourceContentRunes {
+		t.Fatalf("source content rune count = %d, want %d", got, maxImitationSourceContentRunes)
+	}
+}
+
+func TestImitationSourceDisplayTitleLabelsUntitledPastedContent(t *testing.T) {
+	t.Parallel()
+
+	if got := imitationSourceDisplayTitle(articleImitationSourceContent, "", "zh-CN"); got != "粘贴正文" {
+		t.Fatalf("Chinese display title = %q, want 粘贴正文", got)
+	}
+	if got := imitationSourceDisplayTitle(articleImitationSourceContent, "", "en-US"); got != "Pasted content" {
+		t.Fatalf("English display title = %q, want Pasted content", got)
+	}
+	if got := imitationSourceDisplayTitle(articleImitationSourceContent, "Original title", "en-US"); got != "Original title" {
+		t.Fatalf("explicit display title = %q, want Original title", got)
+	}
+}