gh-4215 Fix NPE in context lookup with no data loaded

Author: at055612

stroom-pipeline/src/main/java/stroom/pipeline/refdata/ReferenceData.java

@@ -422,7 +422,7 @@ private void loadContextData(
             final long byteCount = contextStream.size();
             // Only use context data if we actually have some.
             if (byteCount > MINIMUM_BYTE_COUNT) {
-                // load the context data into the RefDataStore so it is available for lookups
+                // load the context data into the RefDataStore, so it is available for lookups
                 contextDataLoader.load(
                         contextStream,
                         meta,

stroom-pipeline/src/main/java/stroom/pipeline/refdata/store/onheapstore/RefDataOnHeapStore.java

@@ -11,6 +11,7 @@
 import stroom.pipeline.refdata.store.RefStoreEntry;
 import stroom.pipeline.refdata.store.RefStreamDefinition;
 import stroom.pipeline.refdata.store.offheapstore.TypedByteBuffer;
+import stroom.util.NullSafe;
 import stroom.util.logging.LambdaLogger;
 import stroom.util.logging.LambdaLoggerFactory;
 import stroom.util.logging.LogUtil;
@@ -107,17 +108,23 @@ public Optional<RefDataValue> getValue(final MapDefinition mapDefinition, final
                 // just do the range lookup
                 final long keyLong = Long.parseLong(key);
 
-                // look up our long key in the range store to see if it is part of a range
-                NavigableMap<Range<Long>, RefDataValue> rangeSubMap = rangeValueNestedMap.get(mapDefinition);
-                result = getValueByRange(rangeSubMap, keyLong);
+                // Look up our long key in the range store to see if it is part of a range.
+                // It is possible for an event stream to have no associated context stream, or
+                // the context stream may not contain the map and/or key requested in the lookup
+                final NavigableMap<Range<Long>, RefDataValue> rangeSubMap = rangeValueNestedMap.get(mapDefinition);
+                if (NullSafe.hasEntries(rangeSubMap)) {
+                    result = getValueByRange(rangeSubMap, keyLong);
+                } else {
+                    result = Optional.empty();
+                }
 
             } catch (NumberFormatException e) {
                 // key could not be converted to a long, either this mapdef has no ranges or
                 // an invalid key was used. See if we have any ranges at all for this mapdef
                 // to determine whether to error or not.
                 boolean doesStoreContainRanges = rangeValueNestedMap.containsKey(mapDefinition);
                 if (doesStoreContainRanges) {
-                    // we have ranges for this map def so we would expect to be able to convert the key
+                    // we have ranges for this map def, so we would expect to be able to convert the key
                     throw new RuntimeException(LogUtil.message(
                             "Key {} cannot be used with the range store as it cannot be converted to a long",
                             key),
@@ -169,9 +176,7 @@ private Optional<RefDataValue> getValueByRange(final NavigableMap<Range<Long>, R
 
         // create a sub-set of the map starting at the key of interest or the next smallest
         // range from value. Note the map is reverse sorted to increase confusion.
-
         final SortedMap<Range<Long>, RefDataValue> tailMap = rangeSubMap.tailMap(startKey);
-
         for (Map.Entry<Range<Long>, RefDataValue> entry : tailMap.entrySet()) {
             // see if our key is in the found range
             if (entry.getKey().contains(key)) {
@@ -209,7 +214,7 @@ public List<RefStoreEntry> list(final int limit,
         throw new UnsupportedOperationException("Not yet implemented");
     }
 
-//    @Override
+    //    @Override
 //    public <T> T consumeEntryStream(final Function<Stream<RefStoreEntry>, T> streamFunction) {
 //        throw new UnsupportedOperationException("Not yet implemented");
 //    }

stroom-pipeline/src/test/java/stroom/pipeline/refdata/store/onheapstore/TestRefDataOnHeapStore.java

@@ -228,6 +228,32 @@ void testOverwrite_doNotOverwrite_rangeValueStore() {
         doKeyRangeValueOverwriteTest(false, value1, value2, expectedFinalValue);
     }
 
+    @Test
+    void testGetWithNoLoad_stringKey() {
+        final RefStreamDefinition refStreamDefinition = buildUniqueRefStreamDefinition();
+        MapDefinition mapDefinition = new MapDefinition(refStreamDefinition, "map1");
+        String key = "myKey";
+
+        assertThat(refDataStore.getKeyValueEntryCount())
+                .isEqualTo(0);
+
+        assertThat(refDataStore.getValue(mapDefinition, key))
+                .isEmpty();
+    }
+
+    @Test
+    void testGetWithNoLoad_longKey() {
+        final RefStreamDefinition refStreamDefinition = buildUniqueRefStreamDefinition();
+        MapDefinition mapDefinition = new MapDefinition(refStreamDefinition, "map1");
+        String key = "123";
+
+        assertThat(refDataStore.getKeyValueEntryCount())
+                .isEqualTo(0);
+
+        assertThat(refDataStore.getValue(mapDefinition, key))
+                .isEmpty();
+    }
+
     private void doKeyValueOverwriteTest(final boolean overwriteExisting,
                                          final StringValue value1,
                                          final StringValue value2,

unreleased_changes/20240416_142821_562__4215.md

@@ -0,0 +1,24 @@
+* Issue **#4215** : Fix NPE during context lookup with no context data loaded.
+
+
+```sh
+# ********************************************************************************
+# Issue title: ERROR during lookup()
+# Issue link:  https://github.com/gchq/stroom/issues/4215
+# ********************************************************************************
+
+# ONLY the top line will be included as a change entry in the CHANGELOG.
+# The entry should be in GitHub flavour markdown and should be written on a SINGLE
+# line with no hard breaks. You can have multiple change files for a single GitHub issue.
+# The  entry should be written in the imperative mood, i.e. 'Fix nasty bug' rather than
+# 'Fixed nasty bug'.
+#
+# Examples of acceptable entries are:
+#
+#
+# * Issue **123** : Fix bug with an associated GitHub issue in this repository
+#
+# * Issue **namespace/other-repo#456** : Fix bug with an associated GitHub issue in another repository
+#
+# * Fix bug with no associated GitHub issue.
+```