Refactor URL import to reuse cached markdown

server/src/controllers/items-controller.ts

@@ -6,6 +6,11 @@ import {
   listItems as listItemsRepo,
   updateItem as updateItemRepo
 } from '../db/items-repository.js';
+import {
+  createUrl,
+  findUrlByValue,
+  updateUrlBodyText
+} from '../db/urls-repository.js';
 import { findImageById } from '../db/images-repository.js';
 import { HttpError } from '../errors/http-error.js';
 import { resolveUrlId } from './url-helpers.js';
@@ -17,6 +22,7 @@ import {
   parseItemUpdatePayload
 } from '../validation/items.js';
 import { readUrlMarkdown } from '../services/url-reader-service.js';
+import { normalizeUrl } from '../utils/url-normalizer.js';
 
 export const listItems = async (req: Request, res: Response) => {
   const projectId = parseUuid(req.params.projectId, 'projectId');
@@ -162,7 +168,14 @@ export const importItemFromUrl = async (req: Request, res: Response) => {
     // If the URL constructor fails here, readUrlMarkdown will surface the error.
   }
 
+  const normalizedDbUrl = normalizeUrl(url);
+  let urlRecord = await findUrlByValue(normalizedDbUrl);
   let markdown: string;
+
+  const existingBody = urlRecord?.bodyText;
+  if (existingBody && existingBody.trim().length > 0) {
+    markdown = existingBody;
+  } else {
     try {
       markdown = await readUrlMarkdown(normalizedUrl);
     } catch (error) {
@@ -175,6 +188,16 @@ export const importItemFromUrl = async (req: Request, res: Response) => {
       );
     }
 
+    if (urlRecord) {
+      const updated = await updateUrlBodyText(urlRecord.id, markdown);
+      if (updated) {
+        urlRecord = updated;
+      }
+    } else {
+      urlRecord = await createUrl({ url: normalizedDbUrl, bodyText: markdown });
+    }
+  }
+
   try {
     const parsed = new URL(normalizedUrl);
     hostname = parsed.hostname.replace(/^www\./, '');

server/src/db/urls-repository.ts

@@ -90,3 +90,25 @@ export const createUrl = async (params: CreateUrlParams): Promise<UrlRecord> =>
 
   return mapUrlRow(result.rows[0]);
 };
+
+export const updateUrlBodyText = async (
+  id: string,
+  bodyText: string
+): Promise<UrlRecord | null> => {
+  const result = await query<UrlRow>(
+    `
+      UPDATE urls
+      SET body_text = $2,
+          updated_at = NOW()
+      WHERE id = $1
+      RETURNING id, url, body_text, attributes, has_price, created_at, updated_at
+    `,
+    [id, bodyText]
+  );
+
+  if (result.rows.length === 0) {
+    return null;
+  }
+
+  return mapUrlRow(result.rows[0]);
+};

server/src/services/url-reader-service.ts

@@ -40,7 +40,7 @@ export const readUrlMarkdown = async (rawUrl: string): Promise<string> => {
         Authorization: `Bearer ${env.urlReader.apiKey}`
       }
     });
-  } catch (error) {
+  } catch {
     throw new HttpError(
       502,
       'We could not reach the URL reader service. Please try again later.'