knox/internal/ingest/git.go

package ingest

import (
	"fmt"
	"os"
	"os/exec"
	"path/filepath"
	"regexp"
	"sort"
	"strings"
)

// GitIngester tracks project status for local git repositories.
//
// For each repo under one or more scan roots (default ~/src) it records the
// current branch, whether the working tree is clean, ahead/behind counts vs
// the upstream, the HEAD hash, and the last commit time + subject. The entry's
// CreatedAt is the last commit time, so staleness/dedup works naturally and
// an entry updates when new commits land.
//
// When Recursive is set, repos are discovered at any depth under a root
// (following symlinked directories), not just direct children — this picks up
// nested repos like ~/assistant/kitmaker or ~/Arduino/*/... Repos already have
// a stable identity via their real (symlink-resolved) path, so overlapping
// roots across symlinks do not duplicate entries.
type GitIngester struct {
	Version   string
	Root      string   // single scan root; overridden by Roots (empty => $HOME/src)
	Roots     []string // multi-scan roots; overrides Root when non-empty
	Recursive bool     // recurse into subdirs (and follow dir symlinks) to find nested .git
}

func NewGitIngester() *GitIngester {
	return &GitIngester{Version: "git/v1"}
}

func (g *GitIngester) SourceID() string { return "git" }

// scanRoots returns the effective list of scan roots.
func (g *GitIngester) scanRoots() []string {
	if len(g.Roots) > 0 {
		return g.Roots
	}
	if g.Root != "" {
		return []string{g.Root}
	}
	home, err := os.UserHomeDir()
	if err != nil {
		return nil
	}
	return []string{filepath.Join(home, "src")}
}

func (g *GitIngester) IngestAll() ([]*IngestResult, error) {
	// Dedup by real path so overlapping roots / symlinked roots don't double-add.
	seen := make(map[string]struct{})
	var results []*IngestResult
	for _, root := range g.scanRoots() {
		for _, repoPath := range collectGitRepos(root, g.Recursive) {
			real, err := filepath.EvalSymlinks(repoPath)
			if err != nil {
				real = repoPath
			}
			if _, dup := seen[real]; dup {
				continue
			}
			seen[real] = struct{}{}
			if r := g.repoToResult(real); r != nil {
				results = append(results, r)
			}
		}
	}
	return results, nil
}

func isGitRepo(path string) bool {
	info, err := os.Stat(filepath.Join(path, ".git"))
	return err == nil && info.IsDir()
}

// collectGitRepos returns paths of git repos under root. In non-recursive mode
// only direct children are considered (root itself is not counted), matching
// the original behavior. In recursive mode the root is walked to any depth,
// descending into symlinked directories, up to and including the root if it is
// itself a git repo. Descending stops once a git repo is found, so nested
// repos (e.g. a vendored copy inside another repo) are treated as leaves.
func collectGitRepos(root string, recursive bool) []string {
	visited := make(map[string]struct{})
	var out []string

	var walk func(dir string)
	walk = func(dir string) {
		real, err := filepath.EvalSymlinks(dir)
		if err != nil {
			return
		}
		if _, ok := visited[real]; ok {
			return
		}
		visited[real] = struct{}{}

		if isGitRepo(real) {
			out = append(out, real)
			return
		}

		entries, err := os.ReadDir(real)
		if err != nil {
			return
		}
		for _, e := range entries {
			if strings.HasPrefix(e.Name(), ".") {
				continue
			}
			child := filepath.Join(real, e.Name())
			if recursive {
				walk(child)
				continue
			}
			if e.IsDir() && isGitRepo(child) {
				out = append(out, child)
			}
		}
	}

	walk(root)
	return out
}

type GitStatus struct {
	Branch     string
	Head       string
	Dirty      int
	Ahead      int
	Behind     int
	LastCommit string
	Subject    string
}

func (g *GitIngester) repoToResult(repoPath string) *IngestResult {
	status := gitStatus(repoPath)
	name := filepath.Base(repoPath)
	fp := Fingerprint([]byte("git:repo:" + gitIdentity(repoPath, status.Branch)))

	clean := "clean"
	if status.Dirty > 0 {
		clean = fmt.Sprintf("%d change(s)", status.Dirty)
	}

	summary := fmt.Sprintf("%s @ %s — %s", name, status.Branch, clean)
	if status.Ahead > 0 || status.Behind > 0 {
		summary += fmt.Sprintf(", ahead %d / behind %d", status.Ahead, status.Behind)
	}
	if status.Subject != "" {
		summary += " — “" + truncate(status.Subject, 80) + "”"
	}

	createdAt := status.LastCommit

	return &IngestResult{
		Fingerprint:     fp,
		SourceID:        g.SourceID(),
		SourcePath:      repoPath,
		Project:         name,
		ContentType:     "git-status",
		Title:           name,
		Summary:         summary,
		CreatedAt:       createdAt,
		Confidence:      0.9,
		IngesterVersion: g.Version,
		Provenance: map[string]any{
			"type":        "git_status",
			"path":        repoPath,
			"branch":      status.Branch,
			"head":        status.Head,
			"dirty":       status.Dirty,
			"ahead":       status.Ahead,
			"behind":      status.Behind,
			"last_commit": status.LastCommit,
			"subject":     status.Subject,
		},
	}
}

func gitStatus(repoPath string) GitStatus {
	out, err := gitOut(repoPath, "status", "--porcelain")
	if err != nil {
		return GitStatus{}
	}
	dirty := 0
	if out != "" {
		dirty = len(strings.Split(strings.TrimRight(out, "\n"), "\n"))
	}

	branch, _ := gitOut(repoPath, "rev-parse", "--abbrev-ref", "HEAD")
	head, _ := gitOut(repoPath, "rev-parse", "--short", "HEAD")
	lastCommit, _ := gitOut(repoPath, "log", "-1", "--format=%cI")
	subject, _ := gitOut(repoPath, "log", "-1", "--format=%s")

	ahead, behind := 0, 0
	// ahead/behind only meaningful if there's an upstream
	if upOut, _ := gitOut(repoPath, "rev-parse", "--abbrev-ref", "HEAD@{upstream}"); upOut != "" {
		if rev, err := gitOut(repoPath, "rev-list", "--left-right", "--count", "HEAD...@{upstream}"); err == nil {
			fields := strings.Fields(rev)
			if len(fields) == 2 {
				var a, b int
				if _, e1 := fmt.Sscanf(fields[0], "%d", &a); e1 == nil {
					ahead = a
				}
				if _, e2 := fmt.Sscanf(fields[1], "%d", &b); e2 == nil {
					behind = b
				}
			}
		}
	}

	return GitStatus{
		Branch:     strings.TrimSpace(branch),
		Head:       strings.TrimSpace(head),
		Dirty:      dirty,
		Ahead:      ahead,
		Behind:     behind,
		LastCommit: strings.TrimSpace(lastCommit),
		Subject:    strings.TrimSpace(subject),
	}
}

// sanitizeGitField strips ANSI codes and control bytes from git output
// (colour/unicode modes can inject escape sequences).
var ansiRe = regexp.MustCompile(`\x1b\[[0-9;]*m`)

func sanitizeGitField(s string) string {
	s = ansiRe.ReplaceAllString(s, "")
	s = strings.ReplaceAll(s, "\x1b", "")
	return strings.TrimSpace(s)
}

// gitOut runs git in repoPath and returns trimmed stdout (or "" on error).
func gitOut(repoPath string, args ...string) (string, error) {
	cmd := exec.Command("git", append([]string{"-C", repoPath}, args...)...)
	cmd.Env = append(os.Environ(), "GIT_CONFIG_NOSYSTEM=1")
	out, err := cmd.Output()
	if err != nil {
		return "", err
	}
	return sanitizeGitField(string(out)), nil
}

// gitIdentity returns a machine-independent identity for a repo: the remote
// origin URL when configured (works across clones/machines), else the repo
// directory name. The branch is appended so entries are per-branch like the
// underlying gitStatus.
func gitIdentity(repoPath, branch string) string {
	if out, err := gitOut(repoPath, "remote", "get-url", "origin"); err == nil && out != "" {
		return out + ":" + branch
	}
	return filepath.Base(repoPath) + ":" + branch
}

func (g *GitIngester) Ingest(_ string) (*IngestResult, error) {
	return nil, fmt.Errorf("use IngestAll() for git")
}

var _ Ingester = (*GitIngester)(nil)

// SortResults orders results deterministically by project name (helper for CLI output).
func SortResults(res []*IngestResult) {
	sort.Slice(res, func(i, j int) bool { return res[i].Project < res[j].Project })
}