mfer/internal/cli/check.go

// Package cli implements the mfer command-line interface.
package cli

import (
	"bytes"
	"context"
	"encoding/hex"
	"errors"
	"fmt"
	"io"
	"math"
	"path/filepath"
	"strconv"
	"strings"
	"sync"
	"time"

	"github.com/dustin/go-humanize"
	"github.com/spf13/afero"
	"github.com/urfave/cli/v3"
	"sneak.berlin/go/mfer/internal/log"
	"sneak.berlin/go/mfer/mfer"
)

// fingerprintHexLen is the length in hex characters of the fingerprint of
// an OpenPGP version 4 key, the only version mfer signs with.
const fingerprintHexLen = 40

var (
	// errNoManifestFound indicates no manifest file was found in the
	// searched directory.
	errNoManifestFound = errors.New("no manifest found")
	// errInvalidFingerprint indicates a malformed --require-signature
	// fingerprint argument. The length is spliced in from
	// fingerprintHexLen so the two cannot drift apart.
	errInvalidFingerprint = errors.New(
		"invalid fingerprint: must be exactly " +
			strconv.Itoa(fingerprintHexLen) + " hex characters")
	// errManifestNotSigned indicates a signature was required but the
	// manifest is unsigned. It is wrapped mid-sentence so that the
	// rendered message stays exactly as mfer has always printed it.
	errManifestNotSigned = errors.New("manifest is not signed")
	// errSignerMismatch indicates the embedded signing key fingerprint
	// does not match the required signer. Its text is the mid-sentence
	// fragment of the rendered message, which users grep for in CI and
	// which must therefore not change; match it with errors.Is rather
	// than by reading it.
	errSignerMismatch = errors.New("does not match required")
)

// safeUint64 converts a non-negative int64 to uint64, clamping negative
// values to zero.
func safeUint64(n int64) uint64 {
	if n < 0 {
		return 0
	}

	return uint64(n)
}

// safeRateUint64 converts a bytes-per-second rate to uint64 for display.
//
// A rate is computed as bytes/elapsed, so it is +Inf when the elapsed
// time rounds to zero and NaN when zero bytes were processed in zero
// time. Neither has a defined conversion to uint64, and on amd64 +Inf
// converts to a number that renders as "8.0 EiB/s"; both display as zero
// instead.
func safeRateUint64(rate float64) uint64 {
	if math.IsNaN(rate) || math.IsInf(rate, 0) || rate <= 0 {
		return 0
	}

	if rate >= math.MaxUint64 {
		return math.MaxUint64
	}

	return uint64(rate)
}

// findManifest returns the path of the manifest with the default name in
// dir, or an error if there is none.
func findManifest(fs afero.Fs, dir string) (string, error) {
	path := filepath.Join(dir, defaultManifestName)

	exists, err := afero.Exists(fs, path)
	if err != nil {
		return "", err
	}

	if !exists {
		return "", fmt.Errorf("%w in %s (looked for %s)",
			errNoManifestFound, dir, defaultManifestName)
	}

	return path, nil
}

// downloadManifest downloads the manifest at url into memory. It refuses a
// manifest larger than mfa.maxManifestSize.
func (mfa *CLIApp) downloadManifest(ctx context.Context, url string) ([]byte, error) {
	rc, err := mfa.openManifestReader(ctx, url)
	if err != nil {
		return nil, err
	}

	defer func() { _ = rc.Close() }()

	// Reading stops one byte past mfa.maxManifestSize, which is enough to
	// tell that the manifest is too large.
	manifestData, err := io.ReadAll(io.LimitReader(rc, mfa.maxManifestSize+1))
	if err != nil {
		return nil, fmt.Errorf("download manifest: %w", err)
	}

	if int64(len(manifestData)) > mfa.maxManifestSize {
		return nil, fmt.Errorf("download manifest: %w of %d bytes",
			errManifestTooLarge, mfa.maxManifestSize)
	}

	return manifestData, nil
}

// verifyRequiredSigner enforces the --require-signature fingerprint
// against signer, the fingerprint of the key that made the manifest's
// signature, which is empty when the manifest is not signed. Loading the
// manifest checked that signer is that key's fingerprint.
func verifyRequiredSigner(signer, requiredSigner string) error {
	// Validate fingerprint format: must be exactly 40 hex characters
	if len(requiredSigner) != fingerprintHexLen {
		return fmt.Errorf("%w, got %d", errInvalidFingerprint, len(requiredSigner))
	}

	_, err := hex.DecodeString(requiredSigner)
	if err != nil {
		return fmt.Errorf("invalid fingerprint: %w", err)
	}

	if signer == "" {
		return fmt.Errorf("%w, but signature from %s is required",
			errManifestNotSigned, requiredSigner)
	}

	// Compare fingerprints - must be exact match (case-insensitive)
	if !strings.EqualFold(signer, requiredSigner) {
		return fmt.Errorf("embedded signing key fingerprint %s %w %s",
			signer, errSignerMismatch, requiredSigner)
	}

	log.Infof("manifest signature verified (signer: %s)", signer)

	return nil
}

// reportCheckProgress renders progress updates until the channel closes.
func reportCheckProgress(progress <-chan mfer.CheckStatus, wg *sync.WaitGroup) {
	defer wg.Done()

	for status := range progress {
		if status.ETA > 0 {
			log.Progressf("Checking: %d/%d files, %s/s, ETA %s, %d failures",
				status.CheckedFiles,
				status.TotalFiles,
				humanize.IBytes(safeRateUint64(status.BytesPerSec)),
				status.ETA.Round(time.Second),
				status.Failures)
		} else {
			log.Progressf("Checking: %d/%d files, %s/s, %d failures",
				status.CheckedFiles,
				status.TotalFiles,
				humanize.IBytes(safeRateUint64(status.BytesPerSec)),
				status.Failures)
		}
	}

	log.ProgressDone()
}

// countCheckFailures consumes check results, counting failures and logging
// each at error level so that --quiet still shows them, then closes done.
func countCheckFailures(
	results <-chan mfer.Result, failures *int64, done chan<- struct{},
) {
	for result := range results {
		if result.Status != mfer.StatusOK {
			*failures++

			log.Errorf("%s: %s (%s)", result.Status, result.Path, result.Message)
		} else {
			log.Verbosef("%s: %s", result.Status, result.Path)
		}
	}

	close(done)
}

// findExtraFiles reports files present on disk but absent from the
// manifest, and anything the search cannot read: each is a failure, logged
// at error level, under --no-extra-files, otherwise a warning. index.mf at
// the top of the base directory is never reported as absent from the
// manifest, even when check is given another manifest or a URL, since gen
// and fetch write the manifest there.
func findExtraFiles(
	ctx context.Context, cmd *cli.Command, chk *mfer.Checker, failures *int64,
) error {
	extraResults := make(chan mfer.Result, 1)
	extraDone := make(chan struct{})

	go func() {
		for result := range extraResults {
			if result.Status == mfer.StatusExtra && result.Path == defaultManifestName {
				continue
			}

			if cmd.Bool("no-extra-files") {
				*failures++

				log.Errorf("%s: %s (%s)", result.Status, result.Path, result.Message)
			} else {
				log.Warnf("%s: %s (%s)", result.Status, result.Path, result.Message)
			}
		}

		close(extraDone)
	}()

	err := chk.FindExtraFiles(ctx, extraResults)
	if err != nil {
		return fmt.Errorf("find extra files: %w", err)
	}

	<-extraDone

	return nil
}

// runCheck runs the manifest check with progress and result reporting
// and returns the number of failures.
func runCheck(
	ctx context.Context, cmd *cli.Command, chk *mfer.Checker, showProgress bool,
) (int64, error) {
	// Set up results channel
	results := make(chan mfer.Result, 1)

	// Set up progress channel
	var (
		progress   chan mfer.CheckStatus
		progressWg sync.WaitGroup
	)

	if showProgress {
		progress = make(chan mfer.CheckStatus, 1)

		progressWg.Add(1)

		go reportCheckProgress(progress, &progressWg)
	}

	// Process results in a goroutine
	var failures int64

	done := make(chan struct{})

	go countCheckFailures(results, &failures, done)

	// Run check
	err := chk.Check(ctx, results, progress)

	progressWg.Wait()

	if err != nil {
		return 0, fmt.Errorf("check files: %w", err)
	}

	// Wait for results processing to complete
	<-done

	err = findExtraFiles(ctx, cmd, chk, &failures)
	if err != nil {
		return 0, err
	}

	return failures, nil
}

func (mfa *CLIApp) checkManifestOperation(
	ctx context.Context, cmd *cli.Command,
) error {
	log.Debug("checkManifestOperation()")

	manifestPath, err := mfa.resolveManifestArg(cmd)
	if err != nil {
		return err
	}

	basePath := resolveBasePath(cmd, manifestPath)

	opts := &mfer.CheckerOptions{BasePath: basePath, Fs: mfa.Fs}

	// A manifest given by URL is downloaded into memory and read from
	// there.
	if isHTTPURL(manifestPath) {
		manifestData, downloadErr := mfa.downloadManifest(ctx, manifestPath)
		if downloadErr != nil {
			return downloadErr
		}

		opts.ManifestReader = bytes.NewReader(manifestData)
	} else {
		opts.ManifestPath = manifestPath
	}

	showProgress := cmd.Bool("progress")

	log.Infof("checking manifest %s with base %s", manifestPath, basePath)

	chk, err := mfer.NewChecker(opts)
	if err != nil {
		return fmt.Errorf("load manifest: %w", err)
	}

	// Check signature requirement. An empty fingerprint is checked too,
	// and refused as one of the wrong length.
	if cmd.IsSet(flagRequireSignature) {
		err = verifyRequiredSigner(string(chk.Signer()), cmd.String(flagRequireSignature))
		if err != nil {
			return err
		}
	}

	log.Infof("manifest contains %d files, %s", chk.FileCount(),
		humanize.IBytes(safeUint64(int64(chk.TotalBytes()))))

	failures, err := runCheck(ctx, cmd, chk, showProgress)
	if err != nil {
		return err
	}

	elapsed := time.Since(mfa.startupTime).Seconds()

	rate := float64(chk.TotalBytes()) / elapsed
	if failures == 0 {
		log.Infof("checked %d files (%s) in %.1fs (%s/s): all OK",
			chk.FileCount(), humanize.IBytes(safeUint64(int64(chk.TotalBytes()))),
			elapsed, humanize.IBytes(safeRateUint64(rate)))
	} else {
		log.Infof("checked %d files (%s) in %.1fs (%s/s): %d failed",
			chk.FileCount(), humanize.IBytes(safeUint64(int64(chk.TotalBytes()))),
			elapsed, humanize.IBytes(safeRateUint64(rate)), failures)
	}

	if failures > 0 {
		mfa.exitCode = 1
	}

	return nil
}