Set default user agent for HTTP requests

In order to behave like a good bot, set a default user agent to identify Cupdate requests.
AlexGustafsson · Dec 31, 2024 · 8ba6433 · 8ba6433
1 parent 74728ff
commit 8ba6433
Show file tree

Hide file tree

Showing 3 changed files with 15 additions and 0 deletions.
diff --git a/cmd/cupdate/main.go b/cmd/cupdate/main.go
@@ -47,6 +47,10 @@ type Config struct {
 		Address  string `env:"ADDRESS"`
 	} `envPrefix:"WEB_"`
 
+	HTTP struct {
+		UserAgent string `env:"USER_AGENT" envDefault:"Cupdate/1.0"`
+	} `envPrefix:"HTTP_"`
+
 	Cache struct {
 		Path   string        `env:"PATH" envDefault:"cachev1.boltdb"`
 		MaxAge time.Duration `env:"MAX_AGE" envDefault:"24h"`
@@ -214,6 +218,7 @@ func main() {
 
 	wg.Go(func() error {
 		httpClient := httputil.NewClient(cache, config.Cache.MaxAge)
+		httpClient.UserAgent = config.HTTP.UserAgent
 		prometheus.DefaultRegisterer.MustRegister(httpClient)
 
 		worker := worker.New(httpClient, writeStore)

diff --git a/docs/config.md b/docs/config.md
@@ -12,6 +12,7 @@ done using environment variables.
 | `CUPDATE_API_PORT`                        | The port to expose the API on.                                                                                                                          | `8080`                      |
 | `CUPDATE_WEB_DISABLED`                    | Whether or not to disable the web UI.                                                                                                                   | `false`                     |
 | `CUPDATE_WEB_ADDRESS`                     | The URL at which the UI is available (such as `https://example.com`). Used for RSS feeds, should generally not be set                                   | Automatically resolved      |
+| `CUPDATE_HTTP_USER_AGENT`                 | The User Agent string to use for HTTP requests.                                                                                                         | `Cupdate/1.0`               |
 | `CUPDATE_CACHE_PATH`                      | A path to the boltdb file in which to store cache.                                                                                                      | `cachev1.boltdb`            |
 | `CUPDATE_CACHE_MAX_AGE`                   | The maximum age of cache entries.                                                                                                                       | `24h`                       |
 | `CUPDATE_DB_PATH`                         | A path to the sqlite file in which to store data.                                                                                                       | `dbv1.sqlite`               |

diff --git a/internal/httputil/client.go b/internal/httputil/client.go
@@ -20,6 +20,9 @@ var _ prometheus.Collector = (*Client)(nil)
 
 type Client struct {
 	http.Client
+
+	UserAgent string
+
 	cache       cache.Cache
 	cacheMaxAge time.Duration
 
@@ -58,6 +61,12 @@ func NewClient(cache cache.Cache, maxAge time.Duration) *Client {
 
 // See [http.Client.Do].
 func (c *Client) Do(req *http.Request) (*http.Response, error) {
+	if _, ok := req.Header["User-Agent"]; !ok {
+		if c.UserAgent != "" {
+			req.Header.Set("User-Agent", c.UserAgent)
+		}
+	}
+
 	res, err := c.Client.Do(req)
 	if err == nil {
 		c.requestsCounter.WithLabelValues(req.URL.Host, req.Method, strconv.FormatInt(int64(res.StatusCode), 10)).Inc()