From c9fd16899322ba393abfbeb1846be23cf816e53f Mon Sep 17 00:00:00 2001
From: Carl Gieringer <78054+carlgieringer@users.noreply.github.com>
Date: Mon, 5 Aug 2024 00:13:54 -0700
Subject: [PATCH] Normalize UrlLocator.domAnchors so that they merge correctly.

Signed-off-by: Carl Gieringer <78054+carlgieringer@users.noreply.github.com>
---
 premiser-ui/src/normalizationSchemas.ts | 23 +++++++++++++++++++++++
 premiser-ui/src/reducers/entities.ts    |  8 +++++---
 premiser-ui/src/util.tsx                | 21 +++++++++++++++++++++
 3 files changed, 49 insertions(+), 3 deletions(-)

diff --git a/premiser-ui/src/normalizationSchemas.ts b/premiser-ui/src/normalizationSchemas.ts
index 23d1056f..2b3857f0 100644
--- a/premiser-ui/src/normalizationSchemas.ts
+++ b/premiser-ui/src/normalizationSchemas.ts
@@ -6,6 +6,7 @@ import {
   AppearanceView,
   ContextTrailItem,
   Domain,
+  DomAnchor,
   JustificationView,
   JustificationVote,
   MediaExcerptCitationIdentifier,
@@ -38,6 +39,7 @@ import {
 } from "howdju-common";
 
 import { applyCustomizations, momentConversion } from "./normalizationUtil";
+import { hashString } from "./util";
 
 export const userSchema = new schema.Entity<UserOut>("users");
 export const usersSchema = new schema.Array(userSchema);
@@ -176,10 +178,31 @@ export const sourceExcerptParaphraseSchema =
 export const urlSchema = new schema.Entity<UrlOut>("urls");
 export const urlsSchema = new schema.Array(urlSchema);
 
+export const domAnchorSchema = new schema.Entity<DomAnchor>(
+  "domAnchors",
+  {},
+  {
+    idAttribute: domAnchorKey,
+  }
+);
+
+function domAnchorKey(anchor: DomAnchor) {
+  return [
+    anchor.urlLocatorId,
+    anchor.startOffset,
+    anchor.endOffset,
+    hashString(anchor.prefixText),
+    hashString(anchor.exactText),
+    hashString(anchor.suffixText),
+  ].join("-");
+}
+export const domAnchorsSchema = new schema.Array(domAnchorSchema);
+
 export const urlLocatorSchema = new schema.Entity<UrlLocatorOut>(
   "urlLocators",
   {
     url: urlSchema,
+    anchors: domAnchorsSchema,
   },
   {
     processStrategy: (value) =>
diff --git a/premiser-ui/src/reducers/entities.ts b/premiser-ui/src/reducers/entities.ts
index 340c1a4d..76157376 100644
--- a/premiser-ui/src/reducers/entities.ts
+++ b/premiser-ui/src/reducers/entities.ts
@@ -37,6 +37,7 @@ import {
   mediaExcerptSpeakerKey,
   propositionTagVoteSchema,
   domainSchema,
+  domAnchorSchema,
 } from "@/normalizationSchemas";
 import { MergeDeep } from "type-fest";
 
@@ -91,6 +92,7 @@ export const initialState = {
   appearances: {} as SchemaEntityState<typeof appearanceSchema>,
   contextTrailItems: {} as SchemaEntityState<typeof contextTrailItemSchema>,
   domains: {} as SchemaEntityState<typeof domainSchema>,
+  domAnchors: {} as SchemaEntityState<typeof domAnchorSchema>,
   justifications: {} as SchemaEntityState<
     typeof justificationSchema,
     {
@@ -174,7 +176,7 @@ const slice = createSlice({
           return;
         }
         const payload = action.payload as unknown as ApiErrorPayload;
-        // If a proposition is not found (e.g., another user deleted it), then remove it.
+        // If a statement is not found (e.g., another user deleted it), then remove it.
         if (payload.httpStatusCode === httpStatusCodes.NOT_FOUND) {
           const { rootTargetId } = action.meta.requestMeta;
           delete state.statements[rootTargetId];
@@ -548,8 +550,8 @@ export function deepMerge<T1, T2>(x: Partial<T1>, y: Partial<T2>): T1 & T2 {
 }
 
 export const deepMergeOptions: DeepMergeOptions = {
-  // Overwrite arrays. This prevents us from merging arrays of objects, but since we store our
-  // entities normalized, most objects we care about merging will be top-level entities.
+  // Union arrays so that we deduplicate related entity IDs. (The default behavior is to
+  // concatenate arrays elements.)
   arrayMerge: (targetArray, sourceArray, _options) =>
     union(targetArray, sourceArray),
   // Don't copy the properties of Moment objects (or else we lose their methods.)
diff --git a/premiser-ui/src/util.tsx b/premiser-ui/src/util.tsx
index 8dea49c1..78d8c555 100644
--- a/premiser-ui/src/util.tsx
+++ b/premiser-ui/src/util.tsx
@@ -156,3 +156,24 @@ export function toCompatibleTagVotes(
       } as TagVote)
   );
 }
+
+/**
+ * Hashes a string into a number.
+ *
+ * https://github.com/bryc/code/blob/da36a3e07acfbd07f930a9212a2df9e854ff56e4/jshash/experimental/cyrb53.js
+ */
+export function hashString(str: string, seed = 0) {
+  let h1 = 0xdeadbeef ^ seed,
+    h2 = 0x41c6ce57 ^ seed;
+  for (let i = 0, ch; i < str.length; i++) {
+    ch = str.charCodeAt(i);
+    h1 = Math.imul(h1 ^ ch, 2654435761);
+    h2 = Math.imul(h2 ^ ch, 1597334677);
+  }
+  h1 = Math.imul(h1 ^ (h1 >>> 16), 2246822507);
+  h1 ^= Math.imul(h2 ^ (h2 >>> 13), 3266489909);
+  h2 = Math.imul(h2 ^ (h2 >>> 16), 2246822507);
+  h2 ^= Math.imul(h1 ^ (h1 >>> 13), 3266489909);
+
+  return 4294967296 * (2097151 & h2) + (h1 >>> 0);
+}