From 58bf0cf45cb1b1404ee464c5a82aea745a0f7f16 Mon Sep 17 00:00:00 2001
From: Adam Miller <adam@archive.org>
Date: Fri, 28 May 2021 23:10:11 +0000
Subject: [PATCH] Refactor Trough client URL cache put. Limit
 TroughContentDigestHistory query.

---
 .../recrawl/TroughContentDigestHistory.java   |  2 +-
 .../java/org/archive/trough/TroughClient.java | 33 ++++++++++++++++---
 2 files changed, 29 insertions(+), 6 deletions(-)

diff --git a/contrib/src/main/java/org/archive/modules/recrawl/TroughContentDigestHistory.java b/contrib/src/main/java/org/archive/modules/recrawl/TroughContentDigestHistory.java
index 424b9780..f8b3cc1b 100644
--- a/contrib/src/main/java/org/archive/modules/recrawl/TroughContentDigestHistory.java
+++ b/contrib/src/main/java/org/archive/modules/recrawl/TroughContentDigestHistory.java
@@ -197,7 +197,7 @@ public class TroughContentDigestHistory extends AbstractContentDigestHistory imp
         }
         if(!memoryDedupHit) {
             try {
-                String sql = "select * from dedup where digest_key = %s";
+                String sql = "select * from dedup where digest_key = %s limit 1";
                 List<Map<String, Object>> results = troughClient().read(getSegmentId(), sql, new String[]{persistKeyFor(curi)});
                 if (!results.isEmpty()) {
                     Map<String, Object> hist = new HashMap<String, Object>();
diff --git a/contrib/src/main/java/org/archive/trough/TroughClient.java b/contrib/src/main/java/org/archive/trough/TroughClient.java
index 15d3e58d..726f3802 100644
--- a/contrib/src/main/java/org/archive/trough/TroughClient.java
+++ b/contrib/src/main/java/org/archive/trough/TroughClient.java
@@ -318,8 +318,20 @@ public class TroughClient {
 
     protected String readUrl(String segmentId) throws TroughException {
         if (readUrlCache.get(segmentId) == null) {
-	    String url = readUrlNoCache(segmentId);
-	    readUrlCache.put(segmentId, url);
+            try {
+                readUrlCache.computeIfAbsent(segmentId, k -> {
+                    try {
+                        return readUrlNoCache(k);
+                    } catch (TroughException e) {
+                        throw new RuntimeException(e);
+                    }
+                });
+            }
+            catch(RuntimeException e) {
+                if(e.getCause() instanceof TroughException){
+                    throw (TroughException)e.getCause();
+                }
+            }
             logger.info("segment " + segmentId + " read url is " + readUrlCache.get(segmentId));
         }
         return readUrlCache.get(segmentId);
@@ -379,9 +391,20 @@ public class TroughClient {
 
     protected String writeUrl(String segmentId, String schemaId) throws IOException {
         if (writeUrlCache.get(segmentId) == null) {
-	    String url = writeUrlNoCache(segmentId, schemaId);
-	    writeUrlCache.put(segmentId, url);
-
+            try {
+                writeUrlCache.computeIfAbsent(segmentId, k -> {
+                    try {
+                        return writeUrlNoCache(k, schemaId);
+                    } catch (IOException e) {
+                        throw new RuntimeException(e);
+                    }
+                });
+            }
+            catch(RuntimeException e) {
+                if(e.getCause() instanceof IOException){
+                    throw (IOException)e.getCause();
+                }
+            }
             logger.info("segment " + segmentId + " write url is " + writeUrlCache.get(segmentId));
         }
         return writeUrlCache.get(segmentId);