2 files changed, 101 insertions, 27 deletions
diff --git a/packages/shared/prompts.ts b/packages/shared/prompts.ts
index 0b79eb9a..5a6a705e 100644
--- a/packages/shared/prompts.ts
+++ b/packages/shared/prompts.ts
@@ -1,6 +1,19 @@
-import { getEncoding } from "js-tiktoken";
+import type { Tiktoken } from "js-tiktoken";
 
-const encoding = getEncoding("o200k_base");
+let encoding: Tiktoken | null = null;
+
+/**
+ * Lazy load the encoding to avoid loading the tiktoken data into memory
+ * until it's actually needed
+ */
+async function getEncodingInstance(): Promise<Tiktoken> {
+  if (!encoding) {
+    // Dynamic import to lazy load the tiktoken module
+    const { getEncoding } = await import("js-tiktoken");
+    encoding = getEncoding("o200k_base");
+  }
+  return encoding;
+}
 
 /**
  * Remove duplicate whitespaces to avoid tokenization issues
@@ -9,17 +22,22 @@ function preprocessContent(content: string) {
   return content.replace(/(\s){10,}/g, "$1");
 }
 
-function calculateNumTokens(text: string) {
-  return encoding.encode(text).length;
+async function calculateNumTokens(text: string): Promise<number> {
+  const enc = await getEncodingInstance();
+  return enc.encode(text).length;
 }
 
-function truncateContent(content: string, length: number) {
-  const tokens = encoding.encode(content);
+async function truncateContent(
+  content: string,
+  length: number,
+): Promise<string> {
+  const enc = await getEncodingInstance();
+  const tokens = enc.encode(content);
   if (tokens.length <= length) {
     return content;
   }
   const truncatedTokens = tokens.slice(0, length);
-  return encoding.decode(truncatedTokens);
+  return enc.decode(truncatedTokens);
 }
 
 export function buildImagePrompt(lang: string, customPrompts: string[]) {
@@ -35,14 +53,15 @@ ${customPrompts && customPrompts.map((p) => `- ${p}`).join("\n")}
 You must respond in valid JSON with the key "tags" and the value is list of tags. Don't wrap the response in a markdown code.`;
 }
 
-export function buildTextPrompt(
+/**
+ * Construct tagging prompt for text content
+ */
+function constructTextTaggingPrompt(
   lang: string,
   customPrompts: string[],
   content: string,
-  contextLength: number,
-) {
-  content = preprocessContent(content);
-  const constructPrompt = (c: string) => `
+): string {
+  return `
 You are an expert whose responsibility is to help with automatic tagging for a read-it-later app.
 Please analyze the TEXT_CONTENT below and suggest relevant tags that describe its key themes, topics, and main ideas. The rules are:
 - Aim for a variety of tags, including broad categories, specific keywords, and potential sub-genres.
@@ -54,30 +73,85 @@ Please analyze the TEXT_CONTENT below and suggest relevant tags that describe it
 ${customPrompts && customPrompts.map((p) => `- ${p}`).join("\n")}
 
 <TEXT_CONTENT>
-${c}
+${content}
 </TEXT_CONTENT>
 You must respond in JSON with the key "tags" and the value is an array of string tags.`;
-
-  const promptSize = calculateNumTokens(constructPrompt(""));
-  const truncatedContent = truncateContent(content, contextLength - promptSize);
-  return constructPrompt(truncatedContent);
 }
 
-export function buildSummaryPrompt(
+/**
+ * Construct summary prompt
+ */
+function constructSummaryPrompt(
   lang: string,
   customPrompts: string[],
   content: string,
-  contextLength: number,
-) {
-  content = preprocessContent(content);
-  const constructPrompt = (c: string) => `
+): string {
+  return `
 Summarize the following content responding ONLY with the summary. You MUST follow the following rules:
 - Summary must be in 3-4 sentences.
 - The summary must be in ${lang}.
 ${customPrompts && customPrompts.map((p) => `- ${p}`).join("\n")}
-    ${c}`;
+    ${content}`;
+}
+
+/**
+ * Build text tagging prompt without truncation (for previews/UI)
+ */
+export function buildTextPromptUntruncated(
+  lang: string,
+  customPrompts: string[],
+  content: string,
+): string {
+  return constructTextTaggingPrompt(
+    lang,
+    customPrompts,
+    preprocessContent(content),
+  );
+}
+
+export async function buildTextPrompt(
+  lang: string,
+  customPrompts: string[],
+  content: string,
+  contextLength: number,
+): Promise<string> {
+  content = preprocessContent(content);
+  const promptTemplate = constructTextTaggingPrompt(lang, customPrompts, "");
+  const promptSize = await calculateNumTokens(promptTemplate);
+  const truncatedContent = await truncateContent(
+    content,
+    contextLength - promptSize,
+  );
+  return constructTextTaggingPrompt(lang, customPrompts, truncatedContent);
+}
+
+export async function buildSummaryPrompt(
+  lang: string,
+  customPrompts: string[],
+  content: string,
+  contextLength: number,
+): Promise<string> {
+  content = preprocessContent(content);
+  const promptTemplate = constructSummaryPrompt(lang, customPrompts, "");
+  const promptSize = await calculateNumTokens(promptTemplate);
+  const truncatedContent = await truncateContent(
+    content,
+    contextLength - promptSize,
+  );
+  return constructSummaryPrompt(lang, customPrompts, truncatedContent);
+}
 
-  const promptSize = calculateNumTokens(constructPrompt(""));
-  const truncatedContent = truncateContent(content, contextLength - promptSize);
-  return constructPrompt(truncatedContent);
+/**
+ * Build summary prompt without truncation (for previews/UI)
+ */
+export function buildSummaryPromptUntruncated(
+  lang: string,
+  customPrompts: string[],
+  content: string,
+): string {
+  return constructSummaryPrompt(
+    lang,
+    customPrompts,
+    preprocessContent(content),
+  );
 }
diff --git a/packages/trpc/routers/bookmarks.ts b/packages/trpc/routers/bookmarks.ts
index 05ffa240..f68d5ada 100644
--- a/packages/trpc/routers/bookmarks.ts
+++ b/packages/trpc/routers/bookmarks.ts
@@ -954,7 +954,7 @@ Author: ${bookmark.author ?? ""}
         },
       });
 
-      const summaryPrompt = buildSummaryPrompt(
+      const summaryPrompt = await buildSummaryPrompt(
         serverConfig.inference.inferredTagLang,
         prompts.map((p) => p.text),
         bookmarkDetails,