From e9258af7421bb873a6a4ea0a9e0043f7dc25c75c Mon Sep 17 00:00:00 2001
From: "yanlong.wang" <yanlong.wang@naiver.org>
Date: Wed, 9 Oct 2024 17:47:23 +0800
Subject: [PATCH] fix: pdf mode and google web cache

---
 backend/functions/src/services/snapshot-formatter.ts | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/backend/functions/src/services/snapshot-formatter.ts b/backend/functions/src/services/snapshot-formatter.ts
index 292848b..0e94568 100644
--- a/backend/functions/src/services/snapshot-formatter.ts
+++ b/backend/functions/src/services/snapshot-formatter.ts
@@ -136,7 +136,8 @@ export class SnapshotFormatter extends AsyncService {
         }
 
         let pdfMode = false;
-        if (snapshot.pdfs?.length && !snapshot.title) {
+        // in case of Google Web Cache content
+        if (snapshot.pdfs?.length && (!snapshot.title || snapshot.title.startsWith('cache:'))) {
             const pdf = await this.pdfExtractor.cachedExtract(snapshot.pdfs[0],
                 this.threadLocal.get('cacheTolerance')
             );
@@ -330,7 +331,7 @@ export class SnapshotFormatter extends AsyncService {
             const n = code - 200;
             if (n < 0 || n >= 200) {
                 const text = snapshot.statusText || STATUS_CODES[code];
-                formatted.warning = `Target URL returned error ${code}${text? `: ${text}` : ''}`;
+                formatted.warning = `Target URL returned error ${code}${text ? `: ${text}` : ''}`;
             }
         }