checker-blacklist/checker/disconnect.go

package checker

import (
	"context"
	"encoding/json"
	"fmt"
	"io"
	"net/http"
	"strings"
	"time"

	sdk "git.happydns.org/checker-sdk-go/checker"
)

const disconnectFeedURL = "https://s3.amazonaws.com/lists.disconnect.me/services.json"

func init() {
	Register(&disconnectSource{
		cache: newFeedCache(24*time.Hour, disconnectFetch(disconnectFeedURL)),
	})
}

type disconnectSource struct{ cache *feedCache }

func (*disconnectSource) ID() string   { return "disconnect" }
func (*disconnectSource) Name() string { return "Disconnect.me" }

func (*disconnectSource) Options() SourceOptions {
	return SourceOptions{}
}

func (s *disconnectSource) Query(ctx context.Context, domain, registered string, opts sdk.CheckerOptions) []SourceResult {
	if registered == "" {
		return []SourceResult{{SourceID: s.ID(), SourceName: s.Name(), Enabled: true}}
	}

	matches, size, fetched, err := s.cache.lookup(ctx, registered)
	res := SourceResult{
		SourceID: s.ID(), SourceName: s.Name(), Enabled: true,
		Reference: "https://disconnect.me/trackerprotection",
		Details:   mustJSON(map[string]any{"feed_size": size, "fetched_at": fetched}),
	}
	if err != nil {
		res.Error = err.Error()
	}

	seenCategory := map[string]bool{}
	for _, m := range matches {
		parts := strings.SplitN(m, "|", 2)
		category, company := parts[0], ""
		if len(parts) == 2 {
			company = parts[1]
		}
		if !seenCategory[category] {
			seenCategory[category] = true
			res.Reasons = append(res.Reasons, category)
		}
		extra := map[string]string{}
		if company != "" {
			extra["company"] = company
		}
		res.Evidence = append(res.Evidence, Evidence{
			Label:  "Category",
			Value:  category,
			Status: strings.ToLower(category),
			Extra:  extra,
		})
	}
	return []SourceResult{res}
}

func (*disconnectSource) Evaluate(r SourceResult) (bool, string) {
	return evidenceEval(r, SeverityWarn)
}

func (*disconnectSource) Diagnose(res SourceResult) Diagnosis {
	return Diagnosis{
		Severity: SeverityWarn,
		Title:    "Listed in Disconnect.me tracking-protection blocklist",
		Detail: fmt.Sprintf(
			"Category: %s. This domain appears in the Disconnect.me list used by Firefox Enhanced Tracking Protection, Brave, and uBlock Origin. Browsers and privacy tools will block third-party requests to this domain, which may affect analytics, ad delivery, or embedded widgets. The list is maintained by Disconnect.me; contact them if you believe the classification is incorrect.",
			joinNonEmpty(res.Reasons, ", "),
		),
		Fix:      "https://disconnect.me/contact",
		FixIsURL: true,
	}
}

// disconnectJSON mirrors the structure of services.json.
type disconnectJSON struct {
	Categories map[string][]map[string]map[string][]string `json:"categories"`
}

func disconnectFetch(feedURL string) func(context.Context) ([]string, map[string][]string, error) {
	return func(ctx context.Context) ([]string, map[string][]string, error) {
		reqCtx, cancel := context.WithTimeout(ctx, 30*time.Second)
		defer cancel()

		req, err := http.NewRequestWithContext(reqCtx, http.MethodGet, feedURL, nil)
		if err != nil {
			return nil, nil, err
		}
		req.Header.Set("User-Agent", "happydomain-checker-blacklist/1.0")

		resp, err := sharedHTTPClient.Do(req)
		if err != nil {
			return nil, nil, err
		}
		defer resp.Body.Close()

		if resp.StatusCode != http.StatusOK {
			return nil, nil, fmt.Errorf("disconnect HTTP %d", resp.StatusCode)
		}

		raw, err := io.ReadAll(io.LimitReader(resp.Body, 32<<20))
		if err != nil {
			return nil, nil, fmt.Errorf("disconnect read: %w", err)
		}

		var data disconnectJSON
		if err := json.Unmarshal(raw, &data); err != nil {
			return nil, nil, fmt.Errorf("disconnect parse: %w", err)
		}

		byHost := make(map[string][]string, 4096)
		for category, entities := range data.Categories {
			for _, entity := range entities {
				for company, sites := range entity {
					for _, domains := range sites {
						for _, d := range domains {
							d = strings.ToLower(strings.TrimSuffix(d, "."))
							if d == "" {
								continue
							}
							entry := category + "|" + company
							byHost[d] = append(byHost[d], entry)
						}
					}
				}
			}
		}

		urls := make([]string, 0, len(byHost))
		for d := range byHost {
			urls = append(urls, d)
		}
		return urls, byHost, nil
	}
}