sundynix-agentix/sundynix-gateway/internal/handler/kb.go

package handler

import (
	"context"
	"crypto/rand"
	"encoding/base64"
	"encoding/hex"
	"encoding/json"
	"errors"
	"io"
	"net/http"
	"path/filepath"
	"strings"
	"time"

	"github.com/gin-gonic/gin"

	"github.com/sundynix/sundynix-shared/contract"
)

// KbIngest: POST /api/v1/kb/ingest —— 文本入库（异步，返回 job_id；进度经 SSE 看）。
func (h *Handler) KbIngest(c *gin.Context) {
	var body struct {
		KB   string `json:"kb"`
		Text string `json:"text"`
	}
	if err := c.ShouldBindJSON(&body); err != nil || body.Text == "" {
		c.JSON(http.StatusBadRequest, gin.H{"error": "text required"})
		return
	}
	job := newJobID()
	go h.runIngest(job, body.KB, "", nil, body.Text)
	c.JSON(http.StatusAccepted, gin.H{"job_id": job})
}

// KbIngestFile: POST /api/v1/kb/ingest_file（multipart）—— 文件入库（异步，返回 job_id）。
// 流水线（解析→切块→向量化→写入）的进度经 sundynix.streams.<job_id> 回流，UI 用 SSE 看。
func (h *Handler) KbIngestFile(c *gin.Context) {
	kb := c.PostForm("kb")
	fh, err := c.FormFile("file")
	if err != nil {
		c.JSON(http.StatusBadRequest, gin.H{"error": "file required"})
		return
	}
	f, err := fh.Open()
	if err != nil {
		c.JSON(http.StatusBadGateway, gin.H{"error": err.Error()})
		return
	}
	defer f.Close()
	data, err := io.ReadAll(f)
	if err != nil {
		c.JSON(http.StatusBadGateway, gin.H{"error": err.Error()})
		return
	}
	job := newJobID()
	go h.runIngest(job, kb, fh.Filename, data, "")
	c.JSON(http.StatusAccepted, gin.H{"job_id": job, "file": fh.Filename})
}

// runIngest 后台跑入库流水线，逐阶段把进度发到 sundynix.streams.<job>。
// filename 非空表示文件入库（先经 mcp-py 解析）；否则用 rawText。
func (h *Handler) runIngest(job, kb, filename string, data []byte, rawText string) {
	ctx := context.Background()
	emit := func(ev contract.IngestEvent) { _ = h.bus.PublishIngest(job, &ev) }
	time.Sleep(400 * time.Millisecond) // 给 SSE 客户端订阅时间（core NATS 无缓冲）

	text := rawText
	if filename != "" {
		emit(contract.IngestEvent{Stage: "解析", Msg: filename})
		parsed, err := h.parseFile(ctx, filename, data)
		if err != nil {
			emit(contract.IngestEvent{Stage: "失败", Error: "解析失败: " + err.Error()})
			_ = h.bus.CompleteStream(job)
			return
		}
		emit(contract.IngestEvent{Stage: "解析完成", Msg: "解析出 " + itoa(len([]rune(parsed))) + " 字"})
		text = parsed
	}

	// 调 mcp-go kb_ingest（带 job_id）：它会发 切块/向量化/写入/完成 事件 + CompleteStream。
	res, err := h.bus.CallTool(ctx, contract.ToolSubjectGo("kb_ingest"),
		&contract.ToolCall{Tool: "kb_ingest", Args: map[string]any{"kb": kb, "text": text, "job_id": job}})
	if err != nil || res == nil || !res.OK {
		msg := "kb_ingest 失败"
		if err != nil {
			msg = err.Error()
		} else if res != nil {
			msg = res.Error
		}
		emit(contract.IngestEvent{Stage: "失败", Error: msg})
		_ = h.bus.CompleteStream(job)
	}
}

// KbIngestStream: GET /api/v1/kb/ingest/:id/stream —— SSE 实时推送入库进度事件。
func (h *Handler) KbIngestStream(c *gin.Context) {
	job := c.Param("id")
	c.Writer.Header().Set("Content-Type", "text/event-stream")
	c.Writer.Header().Set("Cache-Control", "no-cache")
	c.Writer.Header().Set("Connection", "keep-alive")

	events := make(chan []byte, 64)
	done := make(chan struct{})
	unsub, err := h.bus.SubscribeTokens(job,
		func(ev []byte) {
			select {
			case events <- ev:
			default:
			}
		},
		func() { close(done) },
	)
	if err != nil {
		c.JSON(http.StatusBadGateway, gin.H{"error": err.Error()})
		return
	}
	defer func() { _ = unsub() }()

	c.Stream(func(w io.Writer) bool {
		select {
		case ev := <-events:
			c.SSEvent("progress", string(ev))
			return true
		case <-done:
			c.SSEvent("done", job)
			return false
		case <-c.Request.Context().Done():
			return false
		}
	})
}

func newJobID() string {
	var b [8]byte
	_, _ = rand.Read(b[:])
	return "ingest_" + hex.EncodeToString(b[:])
}

// itoa 简易整数转字符串（避免引入 strconv）。
func itoa(n int) string {
	if n == 0 {
		return "0"
	}
	neg := n < 0
	if neg {
		n = -n
	}
	var b []byte
	for n > 0 {
		b = append([]byte{byte('0' + n%10)}, b...)
		n /= 10
	}
	if neg {
		b = append([]byte{'-'}, b...)
	}
	return string(b)
}

// parseFile 把文件字节转为纯文本：文本类直读，其余经 mcp-py parse_document（算法层）。
func (h *Handler) parseFile(ctx context.Context, filename string, data []byte) (string, error) {
	switch strings.ToLower(filepath.Ext(filename)) {
	case ".txt", ".md", ".markdown", ".text":
		return string(data), nil
	}
	res, err := h.bus.CallTool(ctx, contract.ToolSubjectPy("parse_document"),
		&contract.ToolCall{Tool: "parse_document", Args: map[string]any{
			"filename": filename, "content_b64": base64.StdEncoding.EncodeToString(data),
		}})
	if err != nil {
		return "", err
	}
	if res == nil || !res.OK {
		if res != nil && res.Error != "" {
			return "", errors.New(res.Error)
		}
		return "", errors.New("parse_document 无响应（mcp-py 未运行？）")
	}
	return res.Content, nil
}

// KbSearch: POST /api/v1/kb/search —— 检索台：查某知识库，返回带分数的命中（→ mcp-go kb_search）。
func (h *Handler) KbSearch(c *gin.Context) {
	var body struct {
		KB   string `json:"kb"`
		Q    string `json:"q"`
		TopK int    `json:"topK"`
	}
	if err := c.ShouldBindJSON(&body); err != nil || body.Q == "" {
		c.JSON(http.StatusBadRequest, gin.H{"error": "q required"})
		return
	}
	args := map[string]any{"kb": body.KB, "q": body.Q}
	if body.TopK > 0 {
		args["topK"] = body.TopK
	}
	res, err := h.bus.CallTool(c.Request.Context(), contract.ToolSubjectGo("kb_search"),
		&contract.ToolCall{Tool: "kb_search", Args: args})
	if err != nil {
		c.JSON(http.StatusBadGateway, gin.H{"error": err.Error()})
		return
	}
	if !res.OK {
		c.JSON(http.StatusUnprocessableEntity, gin.H{"error": res.Error})
		return
	}
	var hits []map[string]any
	_ = json.Unmarshal([]byte(res.Content), &hits)
	c.JSON(http.StatusOK, gin.H{"hits": hits})
}

// KbGraph: GET /api/v1/kb/graph?kb= —— 某知识库的图谱三元组（→ mcp-go kb_graph，Neo4j）。
func (h *Handler) KbGraph(c *gin.Context) {
	res, err := h.bus.CallTool(c.Request.Context(), contract.ToolSubjectGo("kb_graph"),
		&contract.ToolCall{Tool: "kb_graph", Args: map[string]any{"kb": c.Query("kb"), "limit": 100}})
	if err != nil {
		c.JSON(http.StatusBadGateway, gin.H{"error": err.Error()})
		return
	}
	var triples []map[string]any
	_ = json.Unmarshal([]byte(res.Content), &triples)
	c.JSON(http.StatusOK, gin.H{"triples": triples})
}