From 5488d2538a41d79328fea0061624a127c7cfe5fd Mon Sep 17 00:00:00 2001
From: eltrac <hi@eltr.ac>
Date: Tue, 3 Nov 2026 11:29:44 +0800
Subject: migration: rehash during migrate-db #81

---
 src/weepinbell/migration.clj       | 15 ++++++++++++--
 test/weepinbell/migration_test.clj | 40 +++++++++++++++++++++++---------------
 2 files changed, 37 insertions(+), 18 deletions(-)

diff --git a/src/weepinbell/migration.clj b/src/weepinbell/migration.clj
index c59910a..701ed8f 100644
--- a/src/weepinbell/migration.clj
+++ b/src/weepinbell/migration.clj
@@ -6,6 +6,7 @@
             [weepinbell.storage.edn-storage :as edns]
             [weepinbell.storage.spec]
             [weepinbell.metadata.microformats :refer [parse-microformats]]
+            [weepinbell.utils.digest :refer [digest]]
             [clojure.java.io :as io]
             [fast-edn.core :as edn]))
 
@@ -18,6 +19,15 @@
       (assoc :jf2 (parse-microformats (:microformats wm)))
       (dissoc :microformats)))
 
+(defn rehash
+  "Given a Webmention object, recalculate its hash and return a new map."
+  [wm]
+  {:pre [(map? wm) (contains? wm :webmention)]}
+  (assoc wm :webmention
+         (assoc (:webmention wm) :id
+                (digest (str (get-in wm [:webmention :source])
+                             (get-in wm [:webmention :target]))))))
+
 (defn edn->h2
   "Given an EDN file and a SQLite DB path, migrate data from that
    edn file to SQLite DB. It appends data to DB so the destination
@@ -35,7 +45,8 @@
             (println "Appending webmention" id "to database")
             (try
               (let [wm (if (contains? wm :microformats)
-                         (mf->jf2 wm) wm)]
+                         (mf->jf2 wm) wm)
+                    wm (rehash wm)]
                 (h2/insert-webmention ds wm))
               (catch Exception e
                 (println "Something unexpected happened.")
@@ -43,7 +54,7 @@
     (println "Done")))
 
 (defn- file? [file]
-  (or (string? file) 
+  (or (string? file)
       (instance? java.io.File file)
       (instance? java.net.URL file)))
 
diff --git a/test/weepinbell/migration_test.clj b/test/weepinbell/migration_test.clj
index 07a11d6..833c066 100644
--- a/test/weepinbell/migration_test.clj
+++ b/test/weepinbell/migration_test.clj
@@ -3,14 +3,16 @@
             [weepinbell.metadata.jf2-spec]
             [fast-edn.core :as edn]
             [weepinbell.storage.h2 :as h2]
+            [weepinbell.utils.digest :refer [digest]]
             [next.jdbc :as jdbc]
+            [next.jdbc.sql :as sql]
             [clojure.spec.alpha :as s]
             [clojure.test :refer [deftest testing is]]
             [clojure.java.io :as io]))
 
 (def test-source-edn (io/resource "webmention-test.edn"))
 
-(def webmention-test-edn 
+(def webmention-test-edn
   (with-open [rdr (io/reader test-source-edn)]
     (->> rdr line-seq (map edn/read-string) doall)))
 
@@ -22,36 +24,42 @@
           converted (m/mf->jf2 data)]
       (is (contains? converted :jf2))
       (is (not (contains? converted :microformats)))
-      (is (s/valid? :weepinbell.metadata.jf2-spec/jf2 
+      (is (s/valid? :weepinbell.metadata.jf2-spec/jf2
                     (:jf2 converted))))))
 
 (deftest test-migration-from-edn-to-h2
-  (testing "Migrating EDN file storage to H2 database storage"
+  (testing "Migrating EDN file storage to H2 database storage:"
     (let [target-h2 "tmp/webmention-test-h2"]
-      ;; assert migration succeed with no error
-      (is (nil? (try
-        (m/edn->h2 test-source-edn target-h2)
-        (catch Throwable _ :fail))))
-      ;; assert data count 
-      (let [ds (h2/make-datasource target-h2)
-            result (jdbc/execute! ds ["SELECT count(*) FROM webmention"])
-            len (get (first result) (keyword "count(*)"))]
-        (is (= len (count webmention-test-edn))))
+      (testing "no error or exception should be thrown"
+        (is (nil? (try
+                    (m/edn->h2 test-source-edn target-h2)
+                    (catch Throwable _ :fail)))))
+      (testing "there should be no data loss"
+        (let [ds (h2/make-datasource target-h2)
+              result (jdbc/execute! ds ["SELECT count(*) FROM webmention"])
+              len (get (first result) (keyword "count(*)"))]
+          (is (= len (count webmention-test-edn)))))
+      (testing "should recalculate hash/id of webmentions"
+        (let [ds (h2/make-datasource target-h2)
+              result (jdbc/execute! ds ["SELECT hash, wm_source, wm_target FROM webmention"])
+              match? (fn [{:keys [hash wm_source wm_target]}]
+                       (= hash (digest (str wm_source wm_target))))]
+          (is (every? match? result))))
       (io/delete-file (str target-h2 ".mv.db"))))
   (testing "Exceptions thrown during migration does not abort program"
-    (with-redefs [weepinbell.storage.h2/insert-webmention 
+    (with-redefs [weepinbell.storage.h2/insert-webmention
                   (fn [& _] (throw (Exception. "Test exception")))]
       (let [target-h2 "tmp/webmention-test-exception-h2"]
         (is (nil? (try
                     (m/edn->h2 test-source-edn target-h2)
                     (catch Exception _ :fail)))))))
   (testing "Errors thrown during migration should fail migration"
-    (with-redefs [weepinbell.storage.h2/insert-webmention 
+    (with-redefs [weepinbell.storage.h2/insert-webmention
                   (fn [& _] (throw (Error. "Test error")))]
       (let [target-h2 "tmp/webmention-test-exception-h2"]
         (is (= :fail (try
-                    (m/edn->h2 test-source-edn target-h2)
-                    (catch Error _ :fail))))))))
+                       (m/edn->h2 test-source-edn target-h2)
+                       (catch Error _ :fail))))))))
 
 (deftest test-edn-jf2ify
   (testing "Convert webmenion-test.edn to webmention-test-jf2.edn with jf2 markup"
-- 
cgit v1.2.3