yandex ocr + error handling fixes

bun.lock

@@ -8,6 +8,7 @@
         "@sapphire/discord.js-utilities": "^7.3.2",
         "cheerio": "^1.0.0",
         "discord.js": "^14.17.3",
+        "file-type": "^20.1.0",
         "ky": "^1.7.4",
         "lru-cache": "^11.0.2",
         "nanoid": "^5.0.9",
@@ -88,6 +89,10 @@
 
     "@sapphire/utilities": ["@sapphire/utilities@3.18.2", "", {}, "sha512-QGLdC9+pT74Zd7aaObqn0EUfq40c4dyTL65pFnkM6WO1QYN7Yg/s4CdH+CXmx0Zcu6wcfCWILSftXPMosJHP5A=="],
 
+    "@tokenizer/inflate": ["@tokenizer/inflate@0.2.6", "", { "dependencies": { "debug": "^4.3.7", "fflate": "^0.8.2", "token-types": "^6.0.0" } }, "sha512-SdR/i05U7Xhnsq36iyIq/ZiGGw4PKzw4ww3bOq80Pjj4wyXpqyTcgrgdDdGlcatnlvzNJx8CQw3hp6QZvkUwhA=="],
+
+    "@tokenizer/token": ["@tokenizer/token@0.3.0", "", {}, "sha512-OvjF+z51L3ov0OyAU0duzsYuvO01PH7x4t6DJx+guahgTnBHkhJdG7soQeTSFLWN3efnHyibZ4Z8l2EuWwJN3A=="],
+
     "@types/bun": ["@types/bun@1.2.2", "", { "dependencies": { "bun-types": "1.2.2" } }, "sha512-tr74gdku+AEDN5ergNiBnplr7hpDp3V1h7fqI2GcR/rsUaM39jpSeKH0TFibRvU0KwniRx5POgaYnaXbk0hU+w=="],
 
     "@types/estree": ["@types/estree@1.0.6", "", {}, "sha512-AYnb1nQyY49te+VRAVgmzfcgjYS91mY5P0TKUDCLEM+gNnA+3T6rWITXRLYCpahpqSQbN5cE+gHpnPyXjHWxcw=="],
@@ -218,8 +223,12 @@
 
     "fecha": ["fecha@4.2.3", "", {}, "sha512-OP2IUU6HeYKJi3i0z4A19kHMQoLVs4Hc+DPqqxI2h/DPZHTm/vjsfC6P0b4jCMy14XizLBqvndQ+UilD7707Jw=="],
 
+    "fflate": ["fflate@0.8.2", "", {}, "sha512-cPJU47OaAoCbg0pBvzsgpTPhmhqI5eJjh/JIu8tPj5q+T7iLvW/JAYUqmE7KOB4R1ZyEhzBaIQpQpardBF5z8A=="],
+
     "file-entry-cache": ["file-entry-cache@8.0.0", "", { "dependencies": { "flat-cache": "^4.0.0" } }, "sha512-XXTUwCvisa5oacNGRP9SfNtYBNAMi+RPwBFmblZEF7N7swHYQS6/Zfk7SRwx4D5j3CH211YNRco1DEMNVfZCnQ=="],
 
+    "file-type": ["file-type@20.1.0", "", { "dependencies": { "@tokenizer/inflate": "^0.2.6", "strtok3": "^10.2.0", "token-types": "^6.0.0", "uint8array-extras": "^1.4.0" } }, "sha512-XoxU+lETfCf+bYK3SXkxFusAvmtYQl1u/ZC4zw1DBLEsHUvh339uwYucgQnnSMz1mRCWYJrCzsbJJ95hsQbZ8A=="],
+
     "fill-range": ["fill-range@7.1.1", "", { "dependencies": { "to-regex-range": "^5.0.1" } }, "sha512-YsGpe3WHLK8ZYi4tWDg2Jy3ebRz2rXowDxnld4bkQB00cc/1Zw9AWnC0i9ztDJitivtQvaI9KaLyKrc+hBW0yg=="],
 
     "find-up": ["find-up@5.0.0", "", { "dependencies": { "locate-path": "^6.0.0", "path-exists": "^4.0.0" } }, "sha512-78/PXT1wlLLDgTzDs7sjq9hzz0vXD+zn+7wypEe4fXQxCmdmqfGsEPQxmiCSQI3ajFV91bVSsvNtrJRiW6nGng=="],
@@ -242,6 +251,8 @@
 
     "iconv-lite": ["iconv-lite@0.6.3", "", { "dependencies": { "safer-buffer": ">= 2.1.2 < 3.0.0" } }, "sha512-4fCk79wshMdzMp2rH06qWrJE4iolqLhCUH+OiuIgU++RB0+94NlDL81atO7GX55uUKueo0txHNtvEyI6D7WdMw=="],
 
+    "ieee754": ["ieee754@1.2.1", "", {}, "sha512-dcyqhDvX1C46lXZcVqCpK+FtMRQVdIMN6/Df5js2zouUsqG7I6sFxitIC+7KYK29KdXOLHdu9zL4sFnoVQnqaA=="],
+
     "ignore": ["ignore@5.3.2", "", {}, "sha512-hsBTNUqQTDwkWtcdYI2i06Y/nUBEsNEDJKjWdigLvegy8kDuJAS8uRlpkkcQpyEXL0Z/pjDy5HBmMjRCJ2gq+g=="],
 
     "import-fresh": ["import-fresh@3.3.1", "", { "dependencies": { "parent-module": "^1.0.0", "resolve-from": "^4.0.0" } }, "sha512-TR3KfrTZTYLPB6jUjfx6MF9WcWrHL9su5TObK4ZkYgBdWKPOFoSoQIdEuTuR82pmtxH2spWG9h6etwfr1pLBqQ=="],
@@ -326,6 +337,8 @@
 
     "path-key": ["path-key@3.1.1", "", {}, "sha512-ojmeN0qd+y0jszEtoY48r0Peq5dwMEkIlCOu6Q5f41lfkswXuKtYrhgoTpLnyIcHm24Uhqx+5Tqm2InSwLhE6Q=="],
 
+    "peek-readable": ["peek-readable@6.1.1", "", {}, "sha512-7QmvgRKhxM0E2PGV4ocfROItVode+ELI27n4q+lpufZ+tRKBu/pBP8WOmw9HXn2ui/AUizqtvaVQhcJrOkRqYg=="],
+
     "picomatch": ["picomatch@2.3.1", "", {}, "sha512-JU3teHTNjmE2VCGFzuY8EXzCDVwEqB2a8fsIvwaStHhAWJEeVd1o1QD80CU6+ZdEXXSLbSsuLwJjkCBWqRQUVA=="],
 
     "prelude-ls": ["prelude-ls@1.2.1", "", {}, "sha512-vkcDPrRZo1QZLbn5RLGPpg/WmIQ65qoWWhcGKf/b5eplkkarX0m9z8ppCat4mlOqUsWpyNuYgO3VRyrYHSzX5g=="],
@@ -362,12 +375,16 @@
 
     "strip-json-comments": ["strip-json-comments@3.1.1", "", {}, "sha512-6fPc+R4ihwqP6N/aIv2f1gMH8lOVtWQHoqC4yK6oSDVVocumAsfCqjkXnqiYMhmMwS/mEHLp7Vehlt3ql6lEig=="],
 
+    "strtok3": ["strtok3@10.2.1", "", { "dependencies": { "@tokenizer/token": "^0.3.0", "peek-readable": "^6.1.1" } }, "sha512-Q2dTnW3UXokAvXmXvrvMoUj/me3LyJI76HNHeuGMh2o0As/vzd7eHV3ncLOyvu928vQIDbE7Vf9ldEnC7cwy1w=="],
+
     "supports-color": ["supports-color@7.2.0", "", { "dependencies": { "has-flag": "^4.0.0" } }, "sha512-qpCAvRl9stuOHveKsn7HncJRvv501qIacKzQlO/+Lwxc9+0q2wLyv4Dfvt80/DPn2pqOBsJdDiogXGR9+OvwRw=="],
 
     "text-hex": ["text-hex@1.0.0", "", {}, "sha512-uuVGNWzgJ4yhRaNSiubPY7OjISw4sw4E5Uv0wbjp+OzcbmVU/rsT8ujgcXJhn9ypzsgr5vlzpPqP+MBBKcGvbg=="],
 
     "to-regex-range": ["to-regex-range@5.0.1", "", { "dependencies": { "is-number": "^7.0.0" } }, "sha512-65P7iz6X5yEr1cwcgvQxbbIw7Uk3gOy5dIdtZ4rDveLqhrdJP+Li/Hx6tyK0NEb+2GCyneCMJiGqrADCSNk8sQ=="],
 
+    "token-types": ["token-types@6.0.0", "", { "dependencies": { "@tokenizer/token": "^0.3.0", "ieee754": "^1.2.1" } }, "sha512-lbDrTLVsHhOMljPscd0yitpozq7Ga2M5Cvez5AjGg8GASBjtt6iERCAJ93yommPmz62fb45oFIXHEZ3u9bfJEA=="],
+
     "triple-beam": ["triple-beam@1.4.1", "", {}, "sha512-aZbgViZrg1QNcG+LULa7nhZpJTZSLm/mXnHXnbAbjmN5aSa0y7V+wvv6+4WaBtpISJzThKy+PIPxc1Nq1EJ9mg=="],
 
     "ts-api-utils": ["ts-api-utils@2.0.1", "", { "peerDependencies": { "typescript": ">=4.8.4" } }, "sha512-dnlgjFSVetynI8nzgJ+qF62efpglpWRk8isUEWZGWlJYySCTD6aKvbUDu+zbPeDakk3bg5H4XpitHukgfL1m9w=="],
@@ -382,6 +399,8 @@
 
     "typescript-eslint": ["typescript-eslint@8.23.0", "", { "dependencies": { "@typescript-eslint/eslint-plugin": "8.23.0", "@typescript-eslint/parser": "8.23.0", "@typescript-eslint/utils": "8.23.0" }, "peerDependencies": { "eslint": "^8.57.0 || ^9.0.0", "typescript": ">=4.8.4 <5.8.0" } }, "sha512-/LBRo3HrXr5LxmrdYSOCvoAMm7p2jNizNfbIpCgvG4HMsnoprRUOce/+8VJ9BDYWW68rqIENE/haVLWPeFZBVQ=="],
 
+    "uint8array-extras": ["uint8array-extras@1.4.0", "", {}, "sha512-ZPtzy0hu4cZjv3z5NW9gfKnNLjoz4y6uv4HlelAjDK7sY/xOkKZv9xK/WQpcsBB3jEybChz9DPC2U/+cusjJVQ=="],
+
     "undici": ["undici@6.19.8", "", {}, "sha512-U8uCCl2x9TK3WANvmBavymRzxbfFYG+tAu+fgx3zxQy3qdagQqBLwJVrdyO1TBfUXvfKveMKJZhpvUYoOjM+4g=="],
 
     "undici-types": ["undici-types@6.20.0", "", {}, "sha512-Ny6QZ2Nju20vw1SRHe3d9jVu6gJ+4e3+MMpqu7pqE5HT6WsTSlce++GQmK5UXS8mzV8DSYHrQH+Xrf2jVcuKNg=="],

package.json

@@ -14,6 +14,7 @@
     "@sapphire/discord.js-utilities": "^7.3.2",
     "cheerio": "^1.0.0",
     "discord.js": "^14.17.3",
+    "file-type": "^20.1.0",
     "ky": "^1.7.4",
     "lru-cache": "^11.0.2",
     "nanoid": "^5.0.9",

src/api.ts

@@ -1,6 +0,0 @@
-import { env } from "./env";
-import { API } from "@discordjs/core";
-import { REST } from "discord.js";
-
-const rest = new REST().setToken(env.DISCORD_TOKEN);
-export const api = new API(rest);

src/client.ts

@@ -1,7 +1,6 @@
-import { Client, Collection, GatewayIntentBits } from "discord.js";
+import { Client, Collection, GatewayIntentBits, REST } from "discord.js";
 import { env } from "./env";
 import { ActivityType, API } from "@discordjs/core";
-import { api } from "./api";
 import type { Command } from "./types/command";
 import path from "node:path";
 import fs from "node:fs/promises";
@@ -31,7 +30,8 @@ export class ArtemisClient extends Client {
       },
     });
 
-    this.api = api;
+    const rest = new REST().setToken(env.DISCORD_TOKEN);
+    this.api = new API(rest);
 
     this.on("error", (err) => {
       log.error("Unhandled Client Error", err);

src/commands/language/wiktionary.ts

@@ -4,6 +4,7 @@ import { getDefinitions, getSuggestions } from "../../utils/wiktionary";
 import { stripHtml } from "../../utils/functions";
 import { PaginatedMessage } from "@sapphire/discord.js-utilities";
 import { LRUCache } from "lru-cache";
+import { abort } from "../../utils/error";
 
 const titleCache = new LRUCache<string, string>({ max: 100 });
 
@@ -51,10 +52,7 @@ export default defineCommand({
 
     const definitions = await getDefinitions(term);
     if (!definitions?.length) {
-      await interaction.reply({
-        content: "No definitions found",
-      });
-      return;
+      abort("No definitions found");
     }
 
     const title = titleCache.get(term) ?? term;

src/commands/owner/sync.ts

@@ -1,6 +1,7 @@
 import { MessageFlags, SlashCommandBuilder } from "discord.js";
 import { client } from "../../client";
 import { defineCommand } from "..";
+import { abort } from "../../utils/error";
 
 export default defineCommand({
   data: new SlashCommandBuilder()
@@ -13,10 +14,7 @@ export default defineCommand({
     const counts = await client.syncCommands();
 
     if (!counts) {
-      await interaction.followUp({
-        content: "No commands to sync",
-      });
-      return;
+      abort("No commands to sync");
     }
 
     const { guildCount, globalCount } = counts;

src/commands/utility/ocr.ts

@@ -0,0 +1,61 @@
+import { codeBlock, inlineCode, SlashCommandBuilder } from "discord.js";
+import { defineCommand } from "..";
+import { downloadFile } from "../../utils/http";
+import { abort } from "../../utils/error";
+import { yandexOcr } from "../../utils/api";
+
+export default defineCommand({
+  data: new SlashCommandBuilder()
+    .setName("ocr")
+    .setDescription("OCR an image using Yandex")
+    .addAttachmentOption((option) =>
+      option
+        .setName("image")
+        .setDescription("The image to OCR")
+        .setRequired(true)
+    ),
+
+  async execute(interaction) {
+    const attachment = interaction.options.getAttachment("image", true);
+    if (!attachment.contentType?.startsWith("image/")) {
+      abort("The file must be an image!");
+    }
+
+    const { data, type } = await downloadFile(attachment.url);
+    if (!type?.mime.startsWith("image/")) {
+      abort("The file must be an image!");
+    }
+
+    await interaction.deferReply();
+
+    const { text, detected_lang } = await yandexOcr(data, type.mime);
+
+    const languageName =
+      new Intl.DisplayNames(["en"], { type: "language" }).of(
+        detected_lang ?? "unknown"
+      ) ?? "unknown";
+
+    const content = `Detected language: ${inlineCode(
+      languageName
+    )}\n${codeBlock(text)}`;
+
+    if (content.length > 2000) {
+      await interaction.editReply({
+        content: `Detected language: ${inlineCode(languageName)}`,
+        files: [
+          {
+            name: "ocr.txt",
+            attachment: text,
+          },
+          attachment,
+        ],
+      });
+      return;
+    }
+
+    await interaction.editReply({
+      content,
+      files: [attachment],
+    });
+  },
+});

src/commands/utility/ping.ts

@@ -1,6 +1,7 @@
 import { inlineCode, SlashCommandBuilder } from "discord.js";
 import { client } from "../../client";
 import { defineCommand } from "..";
+import { abort } from "../../utils/error";
 
 export default defineCommand({
   data: new SlashCommandBuilder()
@@ -9,10 +10,9 @@ export default defineCommand({
 
   async execute(interaction) {
     if (client.ws.ping < 1) {
-      await interaction.reply(
+      abort(
         ":ping_pong: Pong!\nThe bot is still starting up, accurate latency will be available shortly."
       );
-      return;
     }
 
     const msg = (

src/env.ts

@@ -10,6 +10,8 @@ const envSchema = z.object({
     .default("development"),
   DEV_GUILD_ID: z.string(),
   DEV_CHANNEL_ID: z.string(),
+  API_URL: z.string(),
+  API_TOKEN: z.string(),
 });
 
 export const env = envSchema.parse(process.env);

src/events/interactionCreate.ts

@@ -8,7 +8,7 @@ import {
 import { client } from "../client";
 import { log } from "../utils/logger";
 import { defineEvent } from ".";
-import { isCommandError, notifyError } from "../utils/error";
+import { isExplicitCommandError, notifyError } from "../utils/error";
 import { nanoid } from "../utils/functions";
 
 const running = new Map<string, number>();
@@ -54,11 +54,11 @@ async function handleChatInputCommand(
   try {
     await command.execute(interaction);
   } catch (err) {
-    let content = isCommandError(err)
+    let content = isExplicitCommandError(err)
       ? err.message
       : "An unknown error occurred!";
 
-    if (!isCommandError(err)) {
+    if (!isExplicitCommandError(err)) {
       const trace = nanoid();
       content += `\ntrace: ${inlineCode(trace)}`;
       log.error("Unhandled Command Error", { trace, err });

src/utils/api.ts

@@ -0,0 +1,26 @@
+import ky from "ky";
+import { env } from "../env";
+import { version } from "discord.js";
+
+type OCRResult = {
+  text: string;
+  detected_lang?: string;
+};
+
+const client = ky.create({
+  prefixUrl: env.API_URL,
+  headers: {
+    "User-Agent": `artemis (discord.js v${version})`,
+    Authorization: `Bearer ${env.API_TOKEN}`,
+  },
+});
+
+export async function yandexOcr(image: Buffer, mime: string) {
+  const res = await client.post("ocr/yandex", {
+    json: {
+      file: image.toString("base64"),
+      mime,
+    },
+  });
+  return res.json<OCRResult>();
+}

src/utils/error.ts

@@ -2,10 +2,16 @@ import { codeBlock, type TextChannel } from "discord.js";
 import { client } from "../client";
 import { env } from "../env";
 
-export class CommandError extends Error {}
+export class ExplicitCommandError extends Error {}
 
-export function isCommandError(error: any): error is CommandError {
-  return error instanceof CommandError;
+export function abort(message: string): never {
+  throw new ExplicitCommandError(message);
+}
+
+export function isExplicitCommandError(
+  error: any
+): error is ExplicitCommandError {
+  return error instanceof ExplicitCommandError;
 }
 
 export async function notifyError(trace: string, error: any) {

src/utils/http.ts

@@ -0,0 +1,8 @@
+import { fileTypeFromBuffer } from "file-type";
+
+export async function downloadFile(url: string) {
+  const res = await fetch(url);
+  const data = Buffer.from(await res.arrayBuffer());
+  const type = await fileTypeFromBuffer(data);
+  return { data, type };
+}