Skip to content

Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, and a self-maintaining Wiki.

10.6 KB
Raw Blame
// Package handler — chunker_debug.go exposes a read-only preview endpoint
// that runs the adaptive chunker on supplied text without touching the DB
// or generating embeddings. Used by the KB editor's debug panel so users
// can experiment with chunking parameters before committing to a re-index.
package handler

import (
	"context"
	"math"
	"net/http"
	"strings"
	"time"
	"unicode/utf8"

	"github.com/Tencent/WeKnora/internal/infrastructure/chunker"
	"github.com/Tencent/WeKnora/internal/infrastructure/docparser"
	"github.com/Tencent/WeKnora/internal/logger"
	"github.com/gin-gonic/gin"
)

// previewMaxChars caps the input text size so a single preview request
// cannot tie up the splitter for long. Chosen to bound worst-case CPU
// well under the previewTimeout: at 64k runes even the heaviest tier
// chain finishes in well under a second on commodity hardware.
//
// SECURITY NOTE: the splitter is CPU-bound and does NOT accept a
// context.Context. When previewTimeout fires the handler returns to the
// caller, but the worker goroutine keeps running until the splitter
// finishes naturally. The 64k ceiling is the primary mitigation against
// goroutine pile-up under repeated authenticated requests. If the
// splitter ever gains context-awareness, the goroutine-wrapper in
// PreviewChunking should switch to it for true cancellation.
const previewMaxChars = 64 * 1024

// previewMaxChunks caps the number of chunks returned in a single preview
// response so the UI doesn't choke on pathological splits. Stats are
// computed over the full chunk set before truncation so the displayed
// avg/min/max/stddev stay representative.
const previewMaxChunks = 500

// previewTimeout caps how long the handler waits for the splitter
// goroutine before returning a 504. See note above on previewMaxChars.
const previewTimeout = 5 * time.Second

// PreviewChunkingRequest is the body shape accepted by /chunker/preview.
// Text is checked manually below so we can return a friendlier error than
// gin's default "Field validation for 'Text' failed on the 'required' tag".
type PreviewChunkingRequest struct {
	Text           string                 `json:"text"`
	ChunkingConfig PreviewChunkingPayload `json:"chunking_config"`
}

// PreviewChunkingPayload mirrors the snake_case JSON the rest of the API
// uses for ChunkingConfig fields. We don't reuse types.ChunkingConfig
// directly because it carries unrelated parser rules and table-processing
// metadata that the preview path does not use.
type PreviewChunkingPayload struct {
	ChunkSize         int      `json:"chunk_size"`
	ChunkOverlap      int      `json:"chunk_overlap"`
	Separators        []string `json:"separators"`
	EnableParentChild bool     `json:"enable_parent_child"`
	ParentChunkSize   int      `json:"parent_chunk_size"`
	ChildChunkSize    int      `json:"child_chunk_size"`
	Strategy          string   `json:"strategy"`
	TokenLimit        int      `json:"token_limit"`
	Languages         []string `json:"languages"`
}

// PreviewChunkResult describes one chunk emitted during preview.
type PreviewChunkResult struct {
	Seq              int    `json:"seq"`
	Start            int    `json:"start"`
	End              int    `json:"end"`
	SizeChars        int    `json:"size_chars"`
	SizeTokensApprox int    `json:"size_tokens_approx"`
	ContextHeader    string `json:"context_header,omitempty"`
	Content          string `json:"content"`
}

// PreviewChunkingStats summarizes chunk-size distribution. Computed over
// the FULL chunk set, even when the response truncates to previewMaxChunks
// items, so avg/min/max/stddev reflect the real distribution, not just
// the first N.
type PreviewChunkingStats struct {
	Count       int `json:"count"`
	AvgChars    int `json:"avg_chars"`
	MinChars    int `json:"min_chars"`
	MaxChars    int `json:"max_chars"`
	StddevChars int `json:"stddev_chars"`
	// TruncatedTo, when set, is the original chunk count before the
	// response was truncated to previewMaxChunks for transport.
	TruncatedTo int `json:"truncated_to,omitempty"`
}

// PreviewChunkingResponse is the body returned by /chunker/preview.
type PreviewChunkingResponse struct {
	SelectedTier chunker.StrategyTier    `json:"selected_tier"`
	TierChain    []chunker.StrategyTier  `json:"tier_chain"`
	Rejected     []chunker.TierRejection `json:"rejected"`
	Profile      *chunker.DocProfile     `json:"profile"`
	Chunks       []PreviewChunkResult    `json:"chunks"`
	Stats        PreviewChunkingStats    `json:"stats"`
}

// PreviewChunking handles POST /chunker/preview. It runs the supplied text
// through the adaptive chunker and returns the chunks plus diagnostic
// information about which tier won. Read-only: no DB writes, no embedding
// calls, no logging of the supplied text.
//
// PreviewChunking godoc
// @Summary      预览分块结果
// @Description  对提交的文本运行自适应分块器并返回分块预览,不写入数据库不生成 embedding。文本最大 64k 字符
// @Tags         分块
// @Accept       json
// @Produce      json
// @Param        request  body      handler.PreviewChunkingRequest   true  "{text, chunking_config}"
// @Success      200      {object}  handler.PreviewChunkingResponse  "分块结果"
// @Failure      400      {object}  map[string]interface{}           "请求参数错误"
// @Failure      413      {object}  map[string]interface{}           "文本超过预览限制"
// @Failure      504      {object}  map[string]interface{}           "分块超时"
// @Security     Bearer
// @Security     ApiKeyAuth
// @Router       /chunker/preview [post]
func PreviewChunking(c *gin.Context) {
	ctx, cancel := context.WithTimeout(c.Request.Context(), previewTimeout)
	defer cancel()

	var req PreviewChunkingRequest
	if err := c.ShouldBindJSON(&req); err != nil {
		c.JSON(http.StatusBadRequest, gin.H{"success": false, "error": "invalid request body: " + err.Error()})
		return
	}

	if strings.TrimSpace(req.Text) == "" {
		c.JSON(http.StatusBadRequest, gin.H{
			"success": false,
			"error":   "text is empty — paste a sample to preview chunking",
		})
		return
	}

	if utf8.RuneCountInString(req.Text) > previewMaxChars {
		c.JSON(http.StatusRequestEntityTooLarge, gin.H{
			"success": false,
			"error":   "text exceeds preview limit",
			"limit":   previewMaxChars,
		})
		return
	}
	text := chunker.NormalizeLineEndings(req.Text)
	text = docparser.NormalizeHTMLTables(text)

	cfg := chunker.NormalizeSplitterConfig(chunker.SplitterConfig{
		ChunkSize:    req.ChunkingConfig.ChunkSize,
		ChunkOverlap: req.ChunkingConfig.ChunkOverlap,
		Separators:   req.ChunkingConfig.Separators,
		Strategy:     req.ChunkingConfig.Strategy,
		TokenLimit:   req.ChunkingConfig.TokenLimit,
		Languages:    req.ChunkingConfig.Languages,
	})

	// Run the splitter on a goroutine so we can honor the request timeout.
	// The splitter is CPU-bound and doesn't accept a context — wrapping
	// here is the cheapest cancellation we can offer.
	type splitResult struct {
		chunks []chunker.Chunk
		diag   *chunker.Diagnostics
	}
	resCh := make(chan splitResult, 1)
	go func() {
		var chunks []chunker.Chunk
		var diag *chunker.Diagnostics
		if req.ChunkingConfig.EnableParentChild {
			parentCfg, childCfg := chunker.DeriveParentChildConfigs(
				cfg,
				req.ChunkingConfig.ParentChunkSize,
				req.ChunkingConfig.ChildChunkSize,
			)
			result, parentDiag := chunker.SplitParentChildWithDiagnostics(text, parentCfg, childCfg)
			chunks = make([]chunker.Chunk, len(result.Children))
			for i, child := range result.Children {
				chunks[i] = child.Chunk
			}
			diag = parentDiag
		} else {
			chunks, diag = chunker.SplitWithDiagnostics(text, cfg)
		}
		resCh <- splitResult{chunks: chunks, diag: diag}
	}()

	var sr splitResult
	select {
	case sr = <-resCh:
	case <-ctx.Done():
		c.JSON(http.StatusGatewayTimeout, gin.H{
			"success": false,
			"error":   "chunker preview timed out",
		})
		return
	}

	chunks, diag := sr.chunks, sr.diag

	// Diagnostics carries the profile when auto-strategy ran; for explicit
	// strategies the profile is nil and we materialize it here so the UI
	// can still show document stats. Avoids the previous double-pass.
	profile := diag.Profile
	if profile == nil {
		profile = chunker.ProfileDocument(text)
	}

	logger.Debugf(ctx, "chunker preview: tier=%s chunks=%d", diag.SelectedTier, len(chunks))

	lang := chunker.LangMixed
	if len(profile.DetectedLangs) > 0 {
		lang = profile.DetectedLangs[0]
	}

	// Compute rune lengths once per chunk; reused for stats and result
	// payload below. Avoids the previous triple-pass over each chunk's
	// content (stats + result + ApproxTokenCount each rune-counted).
	runeLens := make([]int, len(chunks))
	for i, ch := range chunks {
		runeLens[i] = utf8.RuneCountInString(ch.Content)
	}

	// Compute stats over the FULL chunk set first so the metrics stay
	// representative even when we trim the response to previewMaxChunks.
	totalCount := len(chunks)
	stats := computeChunkSizeStats(runeLens)
	if totalCount > previewMaxChunks {
		stats.TruncatedTo = totalCount
		chunks = chunks[:previewMaxChunks]
		runeLens = runeLens[:previewMaxChunks]
	}

	results := make([]PreviewChunkResult, 0, len(chunks))
	for i, ch := range chunks {
		results = append(results, PreviewChunkResult{
			Seq:              ch.Seq,
			Start:            ch.Start,
			End:              ch.End,
			SizeChars:        runeLens[i],
			SizeTokensApprox: chunker.ApproxTokenCountFromRuneLen(runeLens[i], lang),
			ContextHeader:    ch.ContextHeader,
			Content:          ch.Content,
		})
	}

	resp := PreviewChunkingResponse{
		SelectedTier: diag.SelectedTier,
		TierChain:    diag.TierChain,
		Rejected:     diag.Rejected,
		Profile:      profile,
		Chunks:       results,
		Stats:        stats,
	}
	c.JSON(http.StatusOK, gin.H{"success": true, "data": resp})
}

// computeChunkSizeStats summarizes count / avg / min / max / stddev from
// a pre-computed rune-length slice. Decoupling from chunker.Chunk lets
// the caller compute rune lengths once and reuse them for the response
// payload (avoids a second []rune allocation per chunk).
func computeChunkSizeStats(runeLens []int) PreviewChunkingStats {
	stats := PreviewChunkingStats{Count: len(runeLens)}
	if len(runeLens) == 0 {
		return stats
	}
	var sum, sumSq float64
	minLen, maxLen := math.MaxInt32, 0
	for _, l := range runeLens {
		sum += float64(l)
		sumSq += float64(l) * float64(l)
		if l < minLen {
			minLen = l
		}
		if l > maxLen {
			maxLen = l
		}
	}
	avg := sum / float64(len(runeLens))
	variance := sumSq/float64(len(runeLens)) - avg*avg
	if variance < 0 {
		// Float precision can push the variance slightly below zero on
		// near-uniform inputs; clamp so sqrt doesn't return NaN.
		variance = 0
	}
	stats.AvgChars = int(avg + 0.5)
	stats.MinChars = minLen
	stats.MaxChars = maxLen
	stats.StddevChars = int(math.Sqrt(variance) + 0.5)
	return stats
}