Initial commit

internal/ingestion/image.go

@@ -0,0 +1,63 @@
+package ingestion
+
+import (
+	"log/slog"
+	"net/http"
+	"strconv"
+	"strings"
+	"time"
+)
+
+var imageClient = &http.Client{
+	Timeout: 10 * time.Second,
+	CheckRedirect: func(req *http.Request, via []*http.Request) error {
+		if len(via) >= 3 {
+			return http.ErrUseLastResponse
+		}
+		return nil
+	},
+}
+
+// ValidateImageURL checks that an image URL returns a valid image response.
+// Returns false (with no error) on any failure — story posts without image.
+func ValidateImageURL(url string) bool {
+	if url == "" {
+		return false
+	}
+
+	req, err := http.NewRequest("HEAD", url, nil)
+	if err != nil {
+		slog.Debug("image validation: bad url", "url", url, "err", err)
+		return false
+	}
+	req.Header.Set("User-Agent", userAgent)
+
+	resp, err := imageClient.Do(req)
+	if err != nil {
+		slog.Warn("image validation: request failed", "url", url, "err", err)
+		return false
+	}
+	defer resp.Body.Close()
+
+	if resp.StatusCode != http.StatusOK {
+		slog.Warn("image validation: bad status", "url", url, "status", resp.StatusCode)
+		return false
+	}
+
+	contentType := resp.Header.Get("Content-Type")
+	if !strings.HasPrefix(contentType, "image/") {
+		slog.Warn("image validation: not an image", "url", url, "content_type", contentType)
+		return false
+	}
+
+	// Filter tracking pixels: Content-Length must be > 5KB if present
+	if cl := resp.Header.Get("Content-Length"); cl != "" {
+		size, err := strconv.ParseInt(cl, 10, 64)
+		if err == nil && size <= 5120 {
+			slog.Warn("image validation: too small (likely tracking pixel)", "url", url, "size", size)
+			return false
+		}
+	}
+
+	return true
+}