keyhunter/pkg/recon/sources/wayback.go

package sources

import (
	"bufio"
	"context"
	"fmt"
	"net/http"
	"net/url"
	"strings"
	"time"

	"golang.org/x/time/rate"

	"github.com/salvacybersec/keyhunter/pkg/providers"
	"github.com/salvacybersec/keyhunter/pkg/recon"
)

// WaybackMachineSource implements recon.ReconSource against the Wayback Machine
// CDX Server API. It queries web.archive.org/cdx/search/cdx for historical
// snapshots of pages matching provider keywords (e.g. domains known to host
// API key documentation or configuration files).
//
// RECON-ARCH-01: Each matching CDX record yields a Finding pointing at the
// archived snapshot URL. The source is credentialless and always enabled.
type WaybackMachineSource struct {
	// BaseURL defaults to https://web.archive.org. Tests override with httptest URL.
	BaseURL string
	// Registry drives the keyword query list via BuildQueries.
	Registry *providers.Registry
	// Limiters is the shared recon.LimiterRegistry.
	Limiters *recon.LimiterRegistry
	// Client is the shared retry HTTP wrapper. If nil, a default is used.
	Client *Client
}

// Compile-time assertion that WaybackMachineSource satisfies recon.ReconSource.
var _ recon.ReconSource = (*WaybackMachineSource)(nil)

func (s *WaybackMachineSource) Name() string          { return "wayback" }
func (s *WaybackMachineSource) RateLimit() rate.Limit { return rate.Every(5 * time.Second) }
func (s *WaybackMachineSource) Burst() int            { return 1 }
func (s *WaybackMachineSource) RespectsRobots() bool  { return true }

// Enabled always returns true: CDX API is unauthenticated.
func (s *WaybackMachineSource) Enabled(_ recon.Config) bool { return true }

// Sweep iterates provider keywords, queries the CDX API for each, and emits
// a Finding for every archived snapshot URL returned. The CDX API returns
// plain-text lines with space-separated fields; we extract the original URL
// and timestamp to construct the full Wayback snapshot link.
func (s *WaybackMachineSource) Sweep(ctx context.Context, _ string, out chan<- recon.Finding) error {
	base := s.BaseURL
	if base == "" {
		base = "https://web.archive.org"
	}
	client := s.Client
	if client == nil {
		client = NewClient()
	}

	queries := BuildQueries(s.Registry, "wayback")
	if len(queries) == 0 {
		return nil
	}

	for _, q := range queries {
		if err := ctx.Err(); err != nil {
			return err
		}

		if s.Limiters != nil {
			if err := s.Limiters.Wait(ctx, s.Name(), s.RateLimit(), s.Burst(), false); err != nil {
				return err
			}
		}

		// CDX API: output=text, fl=timestamp,original limits response to two fields per line.
		// limit=50 keeps the response bounded per keyword.
		endpoint := fmt.Sprintf("%s/cdx/search/cdx?url=*&output=text&fl=timestamp,original&limit=50&matchType=prefix&filter=statuscode:200&query=%s",
			base, url.QueryEscape(q))
		req, err := http.NewRequestWithContext(ctx, http.MethodGet, endpoint, nil)
		if err != nil {
			return fmt.Errorf("wayback: build req: %w", err)
		}
		req.Header.Set("Accept", "text/plain")

		resp, err := client.Do(ctx, req)
		if err != nil {
			// Non-fatal: skip this keyword on transient errors.
			continue
		}

		scanner := bufio.NewScanner(resp.Body)
		for scanner.Scan() {
			line := strings.TrimSpace(scanner.Text())
			if line == "" {
				continue
			}
			// CDX text output: "timestamp original-url"
			parts := strings.SplitN(line, " ", 2)
			if len(parts) < 2 {
				continue
			}
			ts := parts[0]
			origURL := parts[1]

			snapshotURL := fmt.Sprintf("%s/web/%s/%s", base, ts, origURL)

			f := recon.Finding{
				ProviderName: "",
				Source:       snapshotURL,
				SourceType:   "recon:wayback",
				Confidence:   "low",
				DetectedAt:   time.Now(),
			}
			select {
			case out <- f:
			case <-ctx.Done():
				_ = resp.Body.Close()
				return ctx.Err()
			}
		}
		_ = resp.Body.Close()
	}
	return nil
}