From 90f2c68c6e718f60f7b84f4ea8c840801245ddc8 Mon Sep 17 00:00:00 2001
From: eltrac <hi@eltr.ac>
Date: Wed, 6 May 2026 11:56:09 +0800
Subject: refactor(processor): separate verification from processor logic

---
 src/weepinbell/processor.clj | 28 ++++++----------------------
 src/weepinbell/utils.clj     |  6 ------
 src/weepinbell/verifier.clj  | 39 +++++++++++++++++++++++++++++++++++++++
 3 files changed, 45 insertions(+), 28 deletions(-)
 create mode 100644 src/weepinbell/verifier.clj

(limited to 'src')

diff --git a/src/weepinbell/processor.clj b/src/weepinbell/processor.clj
index fe2d600..4466745 100644
--- a/src/weepinbell/processor.clj
+++ b/src/weepinbell/processor.clj
@@ -1,29 +1,13 @@
 (ns weepinbell.processor
+  "Process in-coming webmentions.
+   If no such webmention exists and request is verified, create.
+   If exists and verification fails, delete it.
+   If exists and verification passes, update it."
   (:require [clojure.core.async :refer [go-loop <!]]
-            [clojure.string :refer [includes?]]
-            [clj-http.client :as client]
-            [weepinbell.utils :refer [validate-content-types]]
             [weepinbell.wrapper :as wrapper]
+            [weepinbell.verifier :refer [verify-mention verify-content-types]]
             [taoensso.timbre :as timbre]))
 
-(defn verify-mention
-  "Get source and follow redirects to check if target link is present."
-  [source target]
-  (try
-    (timbre/info (str "Verifying mention from: " source))
-    ;; get first 1MB of source URL and verify
-    (let [{:keys [status body]} (client/get source {:throw false
-                                                    :headers {"Range" "bytes=0-1048575"}})]
-      (if (and (>= status 200) (< status 300))
-        ;; TODO needs to be more robust
-        (if (includes? (str body) target)
-          true
-          (do (timbre/error "Target link not found in source body.") false))
-        (do (timbre/error "Source returned status:" status) false)))
-    (catch Exception e
-      (timbre/error "Error verifying mention:" (.getMessage e))
-      false)))
-
 (defn process-new-webmention
   "Store new webmention if verifed. Do nothing otherwise."
   [source target callbacks storage]
@@ -62,7 +46,7 @@
   (let [{:keys [s t]} wm]
    ;; first check Content-Types of both urls
    ;; only proceed when it's HTML or plain text
-    (if (validate-content-types s t)
+    (if (verify-content-types s t)
      ;; second check if such webmention exists
      ;; if so, update it or delete it
      ;; if not, write to data
diff --git a/src/weepinbell/utils.clj b/src/weepinbell/utils.clj
index 9afcc49..4611225 100644
--- a/src/weepinbell/utils.clj
+++ b/src/weepinbell/utils.clj
@@ -43,12 +43,6 @@
     (string/includes? content-type "text/plain") :text
     :else false))
 
-(defn validate-content-types [source target]
-  (let [validate #(-> % get-content-type plain-or-html? boolean)
-        source-valid (validate source)
-        target-valid (validate target)]
-    (and source-valid target-valid)))
-
 (defn javamap->map [map]
   (->> (seq map) (into {}) walk/keywordize-keys))
 
diff --git a/src/weepinbell/verifier.clj b/src/weepinbell/verifier.clj
new file mode 100644
index 0000000..0928fb5
--- /dev/null
+++ b/src/weepinbell/verifier.clj
@@ -0,0 +1,39 @@
+(ns weepinbell.verifier
+  "Verifies in-coming webmentions after receiver checks URL.
+   This makes sure: 
+   1) source contains target URL.
+   2) source respond with 2** status code.
+   3) source's Content-Type is either HTML or plain text."
+  (:require [weepinbell.utils :refer [get-content-type plain-or-html?]]
+            [clojure.string :refer [includes?]]
+            [clj-http.client :as client]
+            [taoensso.timbre :as timbre]))
+
+(defn verify-content-types [source target]
+  (let [validate #(-> % get-content-type plain-or-html? boolean)]
+    (and (validate source) (validate target))))
+
+(defn verify-mention
+  "Get source and follow redirects to check if target link is present."
+  [source target]
+  (try
+    (timbre/info (str "Verifying mention from: " source))
+    ;; get first 1MB of source URL and verify
+    (let [{:keys [status body]} (client/get source {:throw false
+                                                    :headers {"Range" "bytes=0-1048575"}})]
+      ;; Verify status code
+      (if (and (>= status 200) (< status 300))
+        ;; Verify inclusion of target URL
+        ;; TODO needs to be more robust
+        (if (includes? (str body) target)
+          true
+          (do 
+            (timbre/error "Target link not found in source body.") 
+            false))
+        (do 
+          (timbre/error "Source returned status:" status) 
+          false)))
+    (catch Exception e
+      (timbre/error "Error verifying mention:" (.getMessage e))
+      false)))
+
-- 
cgit v1.2.3