m->d: Fix HTML entities showing in reply preview

m2d/converters/event-to-message.js

@@ -5,6 +5,7 @@ const DiscordTypes = require("discord-api-types/v10")
 const chunk = require("chunk-text")
 const TurndownService = require("turndown")
 const assert = require("assert").strict
+const entities = require("entities")
 
 const passthrough = require("../../passthrough")
 const {sync, db, discord, select, from} = passthrough
@@ -349,11 +350,13 @@ async function eventToMessage(event, guild, di) {
 			} else {
 				const repliedToContent = repliedToEvent.content.formatted_body || repliedToEvent.content.body
 				const contentPreviewChunks = chunk(
+					entities.decodeHTML5Strict( // Remove entities like & "
 						repliedToContent.replace(/.*<\/mx-reply>/, "") // Remove everything before replies, so just use the actual message body
 						.replace(/<blockquote>.*?<\/blockquote>/, "") // If the message starts with a blockquote, don't count it and use the message body afterwards
 						.replace(/(?:\n|<br>)+/g, " ") // Should all be on one line
 						.replace(/<span [^>]*data-mx-spoiler\b[^>]*>.*?<\/span>/g, "[spoiler]") // Good enough method of removing spoiler content. (I don't want to break out the HTML parser unless I have to.)
-					.replace(/<[^>]+>/g, ""), 50) // Completely strip all other formatting.
+						.replace(/<[^>]+>/g, "") // Completely strip all HTML tags and formatting.
+					), 50)
 				contentPreview = ":\n> "
 				contentPreview += contentPreviewChunks.length > 1 ? contentPreviewChunks[0] + "..." : contentPreviewChunks[0]
 			}

m2d/converters/event-to-message.test.js

@@ -813,6 +813,55 @@ test("event2message: should include a reply preview when message ends with a blo
 	)
 })
 
+test("event2message: entities are not escaped in main message or reply preview", async t => {
+	// Intended result: Testing? in italics, followed by the sequence "':.`[]&things
+	t.deepEqual(
+		await eventToMessage({
+			type: "m.room.message",
+			sender: "@cadence:cadence.moe",
+			content: {
+				msgtype: "m.text",
+				body: "> <@cadence:cadence.moe> _Testing?_ \"':.`[]&things\n\n_Testing?_ \"':.`[]&things",
+				format: "org.matrix.custom.html",
+				formatted_body: "<mx-reply><blockquote><a href=\"https://matrix.to/#/!fGgIymcYWOqjbSRUdV:cadence.moe/$yIWjZPi6Xk56fBxJwqV4ANs_hYLjnWI2cNKbZ2zwk60?via=cadence.moe&via=feather.onl&via=mythic.onl\">In reply to</a> <a href=\"https://matrix.to/#/@cadence:cadence.moe\">@cadence:cadence.moe</a><br><em>Testing?</em> \"':.`[]&amp;things</blockquote></mx-reply><em>Testing?</em> &quot;':.`[]&amp;things",
+				"m.relates_to": {
+					"m.in_reply_to": {
+						event_id: "$yIWjZPi6Xk56fBxJwqV4ANs_hYLjnWI2cNKbZ2zwk60"
+					}
+				}
+			},
+			event_id: "$2I7odT9okTdpwDcqOjkJb_A3utdO4V8Cp3LK6-Rvwcs",
+			room_id: "!fGgIymcYWOqjbSRUdV:cadence.moe"
+		}, data.guild.general, {
+			api: {
+				getEvent: mockGetEvent(t, "!fGgIymcYWOqjbSRUdV:cadence.moe", "$yIWjZPi6Xk56fBxJwqV4ANs_hYLjnWI2cNKbZ2zwk60", {
+					type: "m.room.message",
+					sender: "@cadence:cadence.moe",
+					content: {
+					  "msgtype": "m.text",
+					  "body": "_Testing?_ \"':.`[]&things",
+					  "format": "org.matrix.custom.html",
+					  "formatted_body": "<em>Testing?</em> &quot;':.`[]&amp;things"
+					},
+					event_id: "$yIWjZPi6Xk56fBxJwqV4ANs_hYLjnWI2cNKbZ2zwk60",
+					room_id: "!fGgIymcYWOqjbSRUdV:cadence.moe"
+				})
+			}
+		}),
+		{
+			messagesToDelete: [],
+			messagesToEdit: [],
+			messagesToSend: [{
+				username: "cadence [they]",
+				content: "> <:L1:1144820033948762203><:L2:1144820084079087647>Ⓜ️**cadence [they]**:"
+					+ "\n> Testing? \"':.`[]&things"
+					+ "\n_Testing?_ \"':.\\`\\[\\]&things",
+				avatar_url: "https://matrix.cadence.moe/_matrix/media/r0/download/cadence.moe/azCAhThKTojXSZJRoWwZmhvU"
+			}]
+		}
+	)
+})
+
 test("event2message: editing a rich reply to a sim user", async t => {
 	const eventsFetched = []
 	t.deepEqual(

package-lock.json

@@ -14,6 +14,7 @@
         "chunk-text": "^2.0.1",
         "cloudstorm": "^0.8.0",
         "discord-markdown": "git+https://git.sr.ht/~cadence/nodejs-discord-markdown#abc56d544072a1dc5624adfea455b0e902adf7b3",
+        "entities": "^4.5.0",
         "giframe": "github:cloudrac3r/giframe#v0.4.1",
         "heatsync": "^2.4.1",
         "js-yaml": "^4.1.0",
@@ -1096,6 +1097,17 @@
         "once": "^1.4.0"
       }
     },
+    "node_modules/entities": {
+      "version": "4.5.0",
+      "resolved": "https://registry.npmjs.org/entities/-/entities-4.5.0.tgz",
+      "integrity": "sha512-V0hjH4dGPh9Ao5p0MoRY6BVqtwCjhz6vI5LT8AJ55H+4g9/4vbHx1I54fS0XuclLhDHArPQCiMjDxjaL8fPxhw==",
+      "engines": {
+        "node": ">=0.12"
+      },
+      "funding": {
+        "url": "https://github.com/fb55/entities?sponsor=1"
+      }
+    },
     "node_modules/es-get-iterator": {
       "version": "1.1.3",
       "resolved": "https://registry.npmjs.org/es-get-iterator/-/es-get-iterator-1.1.3.tgz",

package.json

@@ -20,6 +20,7 @@
     "chunk-text": "^2.0.1",
     "cloudstorm": "^0.8.0",
     "discord-markdown": "git+https://git.sr.ht/~cadence/nodejs-discord-markdown#abc56d544072a1dc5624adfea455b0e902adf7b3",
+    "entities": "^4.5.0",
     "giframe": "github:cloudrac3r/giframe#v0.4.1",
     "heatsync": "^2.4.1",
     "js-yaml": "^4.1.0",

readme.md

@@ -164,6 +164,7 @@ Follow these steps:
 * (1) discord-markdown: This is my fork! I make sure it does what I want.
 * (0) giframe: This is my fork! It should do what I want.
 * (1) heatsync: Module hot-reloader that I trust.
+* (0) entities: Looks fine. No dependencies.
 * (1) js-yaml: It seems to do what I want, and it's already pulled in by matrix-appservice.
 * (70) matrix-appservice: I wish it didn't pull in express :(
 * (0) minimist: It's already pulled in by better-sqlite3->prebuild-install