Agent-System/internal/docrevision/diff.go

package docrevision

import (
	"fmt"
	"strings"
)

// ComputeDiff produces a line-oriented structured diff between two markdown bodies.
func ComputeDiff(fromMarkdown, toMarkdown string, fromBlocks, toBlocks []BlockNode) DiffResult {
	fromLines := splitLines(fromMarkdown)
	toLines := splitLines(toMarkdown)
	hunks := myersLineDiff(fromLines, toLines, fromBlocks, toBlocks)
	// Detect moves on raw delete/insert pairs before collapsing modifies.
	hunks = annotateMoves(hunks, fromBlocks, toBlocks)
	hunks = collapseAdjacentModify(hunks)
	sum := DiffSummary{}
	for _, h := range hunks {
		switch h.Kind {
		case "insert":
			sum.Inserts++
		case "delete":
			sum.Deletes++
		case "modify":
			sum.Modifies++
		case "move":
			sum.Moves++
		}
	}
	return DiffResult{Summary: sum, Hunks: hunks}
}

func splitLines(s string) []string {
	s = strings.TrimSuffix(s, "\n")
	if s == "" {
		return nil
	}
	return strings.Split(s, "\n")
}

func myersLineDiff(a, b []string, fromBlocks, toBlocks []BlockNode) []DiffHunk {
	n, m := len(a), len(b)
	// Simple LCS DP for correctness on moderate docs.
	dp := make([][]int, n+1)
	for i := range dp {
		dp[i] = make([]int, m+1)
	}
	for i := n - 1; i >= 0; i-- {
		for j := m - 1; j >= 0; j-- {
			if a[i] == b[j] {
				dp[i][j] = dp[i+1][j+1] + 1
			} else if dp[i+1][j] >= dp[i][j+1] {
				dp[i][j] = dp[i+1][j]
			} else {
				dp[i][j] = dp[i][j+1]
			}
		}
	}
	var hunks []DiffHunk
	i, j := 0, 0
	idx := 0
	for i < n && j < m {
		if a[i] == b[j] {
			i++
			j++
			continue
		}
		if dp[i+1][j] >= dp[i][j+1] {
			h := DiffHunk{
				ID:         fmt.Sprintf("h%d", idx),
				Kind:       "delete",
				BeforeText: a[i],
				LineStart:  i + 1,
				LineEnd:    i + 1,
				Section:    sectionForLine(i, fromBlocks),
			}
			if blk := blockForLine(i, a, fromBlocks); blk != nil {
				h.BlockIDBefore = blk.ID
			}
			hunks = append(hunks, h)
			idx++
			i++
		} else {
			h := DiffHunk{
				ID:        fmt.Sprintf("h%d", idx),
				Kind:      "insert",
				AfterText: b[j],
				LineStart: j + 1,
				LineEnd:   j + 1,
				Section:   sectionForLine(j, toBlocks),
			}
			if blk := blockForLine(j, b, toBlocks); blk != nil {
				h.BlockIDAfter = blk.ID
			}
			hunks = append(hunks, h)
			idx++
			j++
		}
	}
	for i < n {
		h := DiffHunk{ID: fmt.Sprintf("h%d", idx), Kind: "delete", BeforeText: a[i], LineStart: i + 1, LineEnd: i + 1, Section: sectionForLine(i, fromBlocks)}
		if blk := blockForLine(i, a, fromBlocks); blk != nil {
			h.BlockIDBefore = blk.ID
		}
		hunks = append(hunks, h)
		idx++
		i++
	}
	for j < m {
		h := DiffHunk{ID: fmt.Sprintf("h%d", idx), Kind: "insert", AfterText: b[j], LineStart: j + 1, LineEnd: j + 1, Section: sectionForLine(j, toBlocks)}
		if blk := blockForLine(j, b, toBlocks); blk != nil {
			h.BlockIDAfter = blk.ID
		}
		hunks = append(hunks, h)
		idx++
		j++
	}
	return hunks
}

func collapseAdjacentModify(hunks []DiffHunk) []DiffHunk {
	if len(hunks) == 0 {
		return hunks
	}
	out := make([]DiffHunk, 0, len(hunks))
	for i := 0; i < len(hunks); i++ {
		if i+1 < len(hunks) && hunks[i].Kind == "delete" && hunks[i+1].Kind == "insert" {
			out = append(out, DiffHunk{
				ID:            hunks[i].ID,
				Kind:          "modify",
				BeforeText:    hunks[i].BeforeText,
				AfterText:     hunks[i+1].AfterText,
				BlockIDBefore: hunks[i].BlockIDBefore,
				BlockIDAfter:  hunks[i+1].BlockIDAfter,
				Section:       firstNonEmpty(hunks[i].Section, hunks[i+1].Section),
				LineStart:     hunks[i].LineStart,
				LineEnd:       hunks[i+1].LineEnd,
			})
			i++
			continue
		}
		out = append(out, hunks[i])
	}
	return out
}

func annotateMoves(hunks []DiffHunk, fromBlocks, toBlocks []BlockNode) []DiffHunk {
	fromByFP := map[string]string{}
	for _, b := range fromBlocks {
		fromByFP[b.Fingerprint] = b.ID
	}
	toByFP := map[string]string{}
	for _, b := range toBlocks {
		toByFP[b.Fingerprint] = b.ID
	}
	deletedFP := map[string]int{}
	insertedFP := map[string]int{}
	for i, h := range hunks {
		if h.Kind == "delete" {
			deletedFP[ContentFingerprint(h.BeforeText)] = i
		}
		if h.Kind == "insert" {
			insertedFP[ContentFingerprint(h.AfterText)] = i
		}
	}
	used := map[int]bool{}
	out := make([]DiffHunk, 0, len(hunks))
	for i, h := range hunks {
		if used[i] {
			continue
		}
		if h.Kind == "delete" {
			fp := ContentFingerprint(h.BeforeText)
			if j, ok := insertedFP[fp]; ok && !used[j] {
				used[i] = true
				used[j] = true
				mh := DiffHunk{
					ID:            h.ID,
					Kind:          "move",
					BeforeText:    h.BeforeText,
					AfterText:     hunks[j].AfterText,
					BlockIDBefore: h.BlockIDBefore,
					BlockIDAfter:  hunks[j].BlockIDAfter,
					Section:       fmt.Sprintf("%s → %s", h.Section, hunks[j].Section),
					LineStart:     h.LineStart,
					LineEnd:       hunks[j].LineEnd,
				}
				out = append(out, mh)
				continue
			}
		}
		if h.Kind == "insert" {
			fp := ContentFingerprint(h.AfterText)
			if _, ok := deletedFP[fp]; ok {
				// paired as move above
				if used[i] {
					continue
				}
			}
		}
		out = append(out, h)
	}
	return out
}

func sectionForLine(lineIdx int, blocks []BlockNode) string {
	// Approximate: find block whose order matches line; fallback heading path of nearest prior.
	if lineIdx < 0 || len(blocks) == 0 {
		return ""
	}
	best := ""
	for _, b := range blocks {
		if b.Order <= lineIdx {
			if b.HeadingPath != "" {
				best = b.HeadingPath
			}
		}
	}
	return best
}

func blockForLine(lineIdx int, lines []string, blocks []BlockNode) *BlockNode {
	if lineIdx < 0 || lineIdx >= len(lines) {
		return nil
	}
	fp := ContentFingerprint(lines[lineIdx])
	for i := range blocks {
		if blocks[i].Fingerprint == fp {
			return &blocks[i]
		}
	}
	if lineIdx < len(blocks) {
		return &blocks[lineIdx]
	}
	return nil
}

func firstNonEmpty(a, b string) string {
	if a != "" {
		return a
	}
	return b
}