Fixed issue 207. Thanks for the review, Yingyi.
git-svn-id: https://asterixdb.googlecode.com/svn/branches/asterix_stabilization@1016 eaa15691-b419-025a-1212-ee371bd00084
diff --git a/asterix-algebra/src/main/java/edu/uci/ics/asterix/optimizer/base/FuzzyUtils.java b/asterix-algebra/src/main/java/edu/uci/ics/asterix/optimizer/base/FuzzyUtils.java
index 89d78f2..baf16c2 100644
--- a/asterix-algebra/src/main/java/edu/uci/ics/asterix/optimizer/base/FuzzyUtils.java
+++ b/asterix-algebra/src/main/java/edu/uci/ics/asterix/optimizer/base/FuzzyUtils.java
@@ -35,6 +35,7 @@
return AsterixBuiltinFunctions.COUNTHASHED_WORD_TOKENS;
case UNORDEREDLIST:
case ORDEREDLIST:
+ case ANY:
return null;
default:
throw new NotImplementedException("No tokenizer for type " + inputTag);
diff --git a/asterix-algebra/src/main/java/edu/uci/ics/asterix/optimizer/rules/FuzzyEqRule.java b/asterix-algebra/src/main/java/edu/uci/ics/asterix/optimizer/rules/FuzzyEqRule.java
index 77c719d..345f68a 100644
--- a/asterix-algebra/src/main/java/edu/uci/ics/asterix/optimizer/rules/FuzzyEqRule.java
+++ b/asterix-algebra/src/main/java/edu/uci/ics/asterix/optimizer/rules/FuzzyEqRule.java
@@ -99,9 +99,16 @@
inputExprTypes.add(t.getTypeTag());
} else if (inputExp.getExpressionTag() == LogicalExpressionTag.FUNCTION_CALL) {
// Hack to make sure that we will add the func call as is, without wrapping a tokenizer around.
- inputTypeTag = ATypeTag.UNORDEREDLIST;
+ IAType type = (IAType) context.getExpressionTypeComputer().getType(inputExp, metadataProvider, env);
+ inputTypeTag = type.getTypeTag();
+ // Only auto-tokenize strings.
+ if (inputTypeTag == ATypeTag.STRING) {
+ // Strings will be auto-tokenized.
+ inputTypeTag = ATypeTag.UNORDEREDLIST;
+ } else {
+ useExprAsIs = true;
+ }
inputExprTypes.add(inputTypeTag);
- useExprAsIs = true;
} else if (inputExp.getExpressionTag() == LogicalExpressionTag.CONSTANT) {
ConstantExpression inputConst = (ConstantExpression) inputExp;
AsterixConstantValue constVal = (AsterixConstantValue) inputConst.getValue();
diff --git a/asterix-algebra/src/main/java/edu/uci/ics/asterix/optimizer/rules/FuzzyJoinRule.java b/asterix-algebra/src/main/java/edu/uci/ics/asterix/optimizer/rules/FuzzyJoinRule.java
index 035f1df..8936425 100644
--- a/asterix-algebra/src/main/java/edu/uci/ics/asterix/optimizer/rules/FuzzyJoinRule.java
+++ b/asterix-algebra/src/main/java/edu/uci/ics/asterix/optimizer/rules/FuzzyJoinRule.java
@@ -191,6 +191,10 @@
List<LogicalVariable> leftInputPKs = context.findPrimaryKey(leftInputVar);
List<LogicalVariable> rightInputPKs = context.findPrimaryKey(rightInputVar);
+ // Bail if primary keys could not be inferred.
+ if (leftInputPKs == null || rightInputPKs == null) {
+ return false;
+ }
// primary key has only one variable
if (leftInputPKs.size() != 1 || rightInputPKs.size() != 1) {
return false;
diff --git a/asterix-app/src/test/resources/runtimets/queries/fuzzyjoin/dblp-aqlplus_2.aql b/asterix-app/src/test/resources/runtimets/queries/fuzzyjoin/dblp-aqlplus_2.aql
new file mode 100644
index 0000000..7bf7d5f
--- /dev/null
+++ b/asterix-app/src/test/resources/runtimets/queries/fuzzyjoin/dblp-aqlplus_2.aql
@@ -0,0 +1,32 @@
+/*
+ * Description : Tests that a proper error messags is returned for this scenario.
+ * Since we cannot statically know the type of the field 'title', the FuzzyEqRule
+ * cannot auto-inject a tokenizer, and hence we expect an error saying that we cannot
+ * scan over a string as if it were a collection.
+ * Guards against regression to issue 207.
+ * Success : Yes
+ */
+
+drop dataverse fuzzyjoin if exists;
+create dataverse fuzzyjoin;
+use dataverse fuzzyjoin;
+
+create type DBLPType as open {
+ id: int32
+}
+
+create dataset DBLP(DBLPType) partitioned by key id;
+
+load dataset DBLP
+using "edu.uci.ics.asterix.external.dataset.adapter.NCFileSystemAdapter"
+(("path"="nc1://data/dblp-small/dblp-small.adm"),("format"="adm"));
+
+write output to nc1:'rttest/fuzzyjoin_dblp-aqlplus_2.adm';
+
+set simthreshold '.5f';
+
+for $dblp in dataset('DBLP')
+for $dblp2 in dataset('DBLP')
+where $dblp.title ~= $dblp2.title and $dblp.id < $dblp2.id
+order by $dblp.id, $dblp2.id
+return {'dblp': $dblp, 'dblp2': $dblp2}
diff --git a/asterix-app/src/test/resources/runtimets/queries/records/field-access-on-open-field.aql b/asterix-app/src/test/resources/runtimets/queries/records/field-access-on-open-field.aql
new file mode 100644
index 0000000..2592c67
--- /dev/null
+++ b/asterix-app/src/test/resources/runtimets/queries/records/field-access-on-open-field.aql
@@ -0,0 +1,24 @@
+/*
+ * Description : Tests whether a field access on an open field (statically of type ANY) succeeds.
+ * Guards against regression to issue 207.
+ * Success : Yes
+ */
+
+drop dataverse test if exists;
+create dataverse test;
+use dataverse test;
+
+create type TestType as open {
+ id : int32,
+ name : string
+}
+
+create dataset testds(TestType) partitioned by key id;
+
+insert into dataset testds({"id": 123, "name": "John Doe", "address": { "zip": 92617} });
+
+write output to nc1:"rttest/records_field-access-on-open-field.adm";
+
+for $l in dataset("testds")
+let $a := $l.address
+return $a.zip
\ No newline at end of file
diff --git a/asterix-app/src/test/resources/runtimets/results/fuzzyjoin/dblp-aqlplus_2.adm b/asterix-app/src/test/resources/runtimets/results/fuzzyjoin/dblp-aqlplus_2.adm
new file mode 100644
index 0000000..e69de29
--- /dev/null
+++ b/asterix-app/src/test/resources/runtimets/results/fuzzyjoin/dblp-aqlplus_2.adm
diff --git a/asterix-app/src/test/resources/runtimets/results/records/field-access-on-open-field.adm b/asterix-app/src/test/resources/runtimets/results/records/field-access-on-open-field.adm
new file mode 100644
index 0000000..d4edf93
--- /dev/null
+++ b/asterix-app/src/test/resources/runtimets/results/records/field-access-on-open-field.adm
@@ -0,0 +1 @@
+92617
\ No newline at end of file
diff --git a/asterix-app/src/test/resources/runtimets/testsuite.xml b/asterix-app/src/test/resources/runtimets/testsuite.xml
index 145ef62..2d740be 100644
--- a/asterix-app/src/test/resources/runtimets/testsuite.xml
+++ b/asterix-app/src/test/resources/runtimets/testsuite.xml
@@ -1474,6 +1474,12 @@
</compilation-unit>
</test-case>
<test-case FilePath="fuzzyjoin">
+ <compilation-unit name="dblp-aqlplus_2">
+ <output-file compare="Text">dblp-aqlplus_2.adm</output-file>
+ <expected-error>edu.uci.ics.asterix.common.exceptions.AsterixException</expected-error>
+ </compilation-unit>
+ </test-case>
+ <test-case FilePath="fuzzyjoin">
<compilation-unit name="dblp-csx-2_1">
<output-file compare="Text">dblp-csx-2_1.adm</output-file>
</compilation-unit>
@@ -2561,6 +2567,11 @@
</compilation-unit>
</test-case>
<test-case FilePath="records">
+ <compilation-unit name="field-access-on-open-field">
+ <output-file compare="Text">field-access-on-open-field.adm</output-file>
+ </compilation-unit>
+ </test-case>
+ <test-case FilePath="records">
<compilation-unit name="open-record-constructor_01">
<output-file compare="Text">open-record-constructor_01.adm</output-file>
</compilation-unit>
diff --git a/asterix-om/src/main/java/edu/uci/ics/asterix/om/typecomputer/impl/FieldAccessByIndexResultType.java b/asterix-om/src/main/java/edu/uci/ics/asterix/om/typecomputer/impl/FieldAccessByIndexResultType.java
index dbe338d..d4b11fa 100644
--- a/asterix-om/src/main/java/edu/uci/ics/asterix/om/typecomputer/impl/FieldAccessByIndexResultType.java
+++ b/asterix-om/src/main/java/edu/uci/ics/asterix/om/typecomputer/impl/FieldAccessByIndexResultType.java
@@ -38,6 +38,9 @@
}
IAType type0 = (IAType) obj;
ARecordType t0 = NonTaggedFieldAccessByNameResultType.getRecordTypeFromType(type0, expression);
+ if (t0 == null) {
+ return BuiltinType.ANY;
+ }
ILogicalExpression arg1 = f.getArguments().get(1).getValue();
if (arg1.getExpressionTag() != LogicalExpressionTag.CONSTANT) {
return BuiltinType.ANY;
diff --git a/asterix-om/src/main/java/edu/uci/ics/asterix/om/typecomputer/impl/NonTaggedFieldAccessByNameResultType.java b/asterix-om/src/main/java/edu/uci/ics/asterix/om/typecomputer/impl/NonTaggedFieldAccessByNameResultType.java
index 4b8b32a..a0c35b6 100644
--- a/asterix-om/src/main/java/edu/uci/ics/asterix/om/typecomputer/impl/NonTaggedFieldAccessByNameResultType.java
+++ b/asterix-om/src/main/java/edu/uci/ics/asterix/om/typecomputer/impl/NonTaggedFieldAccessByNameResultType.java
@@ -9,7 +9,6 @@
import edu.uci.ics.asterix.om.types.BuiltinType;
import edu.uci.ics.asterix.om.types.IAType;
import edu.uci.ics.hyracks.algebricks.common.exceptions.AlgebricksException;
-import edu.uci.ics.hyracks.algebricks.common.exceptions.NotImplementedException;
import edu.uci.ics.hyracks.algebricks.core.algebra.base.ILogicalExpression;
import edu.uci.ics.hyracks.algebricks.core.algebra.base.LogicalExpressionTag;
import edu.uci.ics.hyracks.algebricks.core.algebra.expressions.AbstractFunctionCallExpression;
@@ -36,6 +35,9 @@
}
IAType type0 = (IAType) obj;
ARecordType t0 = getRecordTypeFromType(type0, expression);
+ if (t0 == null) {
+ return BuiltinType.ANY;
+ }
AbstractLogicalExpression arg1 = (AbstractLogicalExpression) f.getArguments().get(1).getValue();
if (arg1.getExpressionTag() != LogicalExpressionTag.CONSTANT) {
@@ -58,7 +60,7 @@
return (ARecordType) type0;
}
case ANY: {
- throw new NotImplementedException();
+ return null;
}
case UNION: {
AUnionType u = (AUnionType) type0;
diff --git a/asterix-runtime/src/main/java/edu/uci/ics/asterix/runtime/evaluators/common/AsterixListAccessor.java b/asterix-runtime/src/main/java/edu/uci/ics/asterix/runtime/evaluators/common/AsterixListAccessor.java
index fd6b691..58cbd07 100644
--- a/asterix-runtime/src/main/java/edu/uci/ics/asterix/runtime/evaluators/common/AsterixListAccessor.java
+++ b/asterix-runtime/src/main/java/edu/uci/ics/asterix/runtime/evaluators/common/AsterixListAccessor.java
@@ -51,7 +51,7 @@
public void reset(byte[] listBytes, int start) throws AsterixException {
this.listBytes = listBytes;
this.start = start;
- listType = EnumDeserializer.ATYPETAGDESERIALIZER.deserialize(listBytes[start]);
+ listType = EnumDeserializer.ATYPETAGDESERIALIZER.deserialize(listBytes[start]);
if (listType != ATypeTag.UNORDEREDLIST && listType != ATypeTag.ORDEREDLIST) {
throw new AsterixException("Unsupported type: " + listType);
}
