Complete outstanding ingestion and document UX work

2026-08-22 18:54:07 +01:00
parent 932e889ee8
commit 5b5a9ecbc3
24 changed files with 696 additions and 41 deletions
@@ -0,0 +1,85 @@
+package cmd
+
+import (
+	"fmt"
+
+	"github.com/kb-search/kb/internal/api"
+	"github.com/kb-search/kb/internal/output"
+	"github.com/spf13/cobra"
+)
+
+var findCmd = &cobra.Command{
+	Use:   "find <query>",
+	Short: "Find documents by their indexed content",
+	Args:  cobra.ExactArgs(1),
+	RunE:  runFind,
+}
+
+func init() {
+	findCmd.Flags().IntP("top", "n", 10, "number of documents to return")
+	findCmd.Flags().String("tags", "", "filter by tags (comma-separated)")
+	findCmd.Flags().String("type", "", "filter by document type")
+	rootCmd.AddCommand(findCmd)
+}
+
+func runFind(cmd *cobra.Command, args []string) error {
+	top, _ := cmd.Flags().GetInt("top")
+	tags, _ := cmd.Flags().GetString("tags")
+	docType, _ := cmd.Flags().GetString("type")
+	body := map[string]interface{}{"query": args[0], "top": top}
+	if tags != "" {
+		body["tags"] = splitTags(tags)
+	}
+	if docType != "" {
+		body["doc_type"] = docType
+	}
+
+	resp, err := api.NewClient().Post("/api/v1/documents/find", body)
+	if err != nil {
+		return err
+	}
+	if err := api.CheckError(resp); err != nil {
+		return err
+	}
+	if output.IsJSON() {
+		var raw interface{}
+		if err := api.DecodeJSON(resp, &raw); err != nil {
+			return fmt.Errorf("failed to decode response: %w", err)
+		}
+		output.PrintJSON(raw)
+		return nil
+	}
+
+	var docs []struct {
+		DocumentID int     `json:"document_id"`
+		Title      string  `json:"title"`
+		Filename   string  `json:"original_filename"`
+		Type       string  `json:"doc_type"`
+		Score      float64 `json:"score"`
+		HitCount   int     `json:"hit_count"`
+		TopChunk   string  `json:"top_chunk"`
+	}
+	if err := api.DecodeJSON(resp, &docs); err != nil {
+		return fmt.Errorf("failed to decode response: %w", err)
+	}
+	if len(docs) == 0 {
+		fmt.Println("No documents found.")
+		return nil
+	}
+	for i, doc := range docs {
+		preview := doc.TopChunk
+		if len(preview) > 200 {
+			preview = preview[:200] + "..."
+		}
+		fmt.Printf("\n%d. [%.4f] %s (doc:%d, hits:%d)\n", i+1, doc.Score, doc.Title, doc.DocumentID, doc.HitCount)
+		if doc.Filename != "" {
+			fmt.Printf("   Filename: %s\n", doc.Filename)
+		}
+		if doc.Type != "" {
+			fmt.Printf("   Type: %s\n", doc.Type)
+		}
+		fmt.Printf("   %s\n", preview)
+	}
+	fmt.Println()
+	return nil
+}