Fixed the feature for inserting temporal types from an external dataset.
- updated nontagged/AdmLexer.jj for the latest temporal type formats.
- added test cases for inserting temporal types from external datasets.
- fixed a bug on incorrect parsing at most 3 digits the millisecond fields: string index out of bound if the temporal string is too short to reach the millisecond fields.
git-svn-id: https://asterixdb.googlecode.com/svn/branches/asterix_stabilization_temporal_functionality@705 eaa15691-b419-025a-1212-ee371bd00084
diff --git a/asterix-app/data/temporal/temporalData.json b/asterix-app/data/temporal/temporalData.json
new file mode 100644
index 0000000..ae100f7
--- /dev/null
+++ b/asterix-app/data/temporal/temporalData.json
@@ -0,0 +1,2 @@
+{"id": "001", "dateField": -2012-12-12, "dateFieldPlus": 0990-01-01, "timeField": 23:49:12.39Z, "timeFieldPlus": 03:23:12.2, "datetimeField": 2012-12-12T00:00:00.001, "datetimeFieldPlus": -00130810T221015398, "durationField": P20Y19DT3H74M23.34S, "durationFieldPlus": -P2MT4M300.68S}
+{"id": "002", "datetimeField": 19201220T232918478 }
\ No newline at end of file
diff --git a/asterix-app/src/test/resources/runtimets/queries/accessors/temporal_accessors.aql b/asterix-app/src/test/resources/runtimets/queries/temp/accessors.aql
similarity index 94%
rename from asterix-app/src/test/resources/runtimets/queries/accessors/temporal_accessors.aql
rename to asterix-app/src/test/resources/runtimets/queries/temp/accessors.aql
index 2a3ba93..dbf16ba 100644
--- a/asterix-app/src/test/resources/runtimets/queries/accessors/temporal_accessors.aql
+++ b/asterix-app/src/test/resources/runtimets/queries/temp/accessors.aql
@@ -8,7 +8,7 @@
create dataverse test;
use dataverse test;
-write output to nc1:"rttest/accessors_temporal_accessors.adm";
+write output to nc1:"rttest/temp_accessors.adm";
let $c1 := date("2010-10-30")
let $c2 := datetime("1987-11-19T23:49:23.938")
diff --git a/asterix-app/src/test/resources/runtimets/queries/temp/insert_from_ext_ds.aql b/asterix-app/src/test/resources/runtimets/queries/temp/insert_from_ext_ds.aql
new file mode 100644
index 0000000..c816047
--- /dev/null
+++ b/asterix-app/src/test/resources/runtimets/queries/temp/insert_from_ext_ds.aql
@@ -0,0 +1,26 @@
+/*
+ * Test case name: date-insert.aql
+ * Description: verify insertion operation for date type
+ * Expected result: success
+ */
+
+drop dataverse testdvt if exists;
+create dataverse testdvt;
+use dataverse testdvt;
+
+create type testtype as open {
+ id: string,
+ dateField: date?,
+ timeField: time?,
+ datetimeField: datetime?,
+ durationField: duration?
+}
+
+write output to nc1:"rttest/temp_insert_from_ext_ds.adm";
+
+create external dataset testds(testtype)
+using "edu.uci.ics.asterix.external.dataset.adapter.NCFileSystemAdapter"
+(("path"="nc1://data/temporal/temporalData.json"),("format"="adm"));
+
+for $r in dataset("testds")
+return {"date": $r.dateField, "time": $r.timeField, "datetime": $r.datetimeField, "duration": $r.durationField}
\ No newline at end of file
diff --git a/asterix-app/src/test/resources/runtimets/results/accessors/temporal_accessors.adm b/asterix-app/src/test/resources/runtimets/results/temp/accessors.adm
similarity index 100%
rename from asterix-app/src/test/resources/runtimets/results/accessors/temporal_accessors.adm
rename to asterix-app/src/test/resources/runtimets/results/temp/accessors.adm
diff --git a/asterix-app/src/test/resources/runtimets/results/temp/insert_from_ext_ds.adm b/asterix-app/src/test/resources/runtimets/results/temp/insert_from_ext_ds.adm
new file mode 100644
index 0000000..00bfc1b
--- /dev/null
+++ b/asterix-app/src/test/resources/runtimets/results/temp/insert_from_ext_ds.adm
@@ -0,0 +1,2 @@
+{ "date": date("-2012-12-12"), "time": time("23:49:12.390Z"), "datetime": datetime("2012-12-12T00:00:00.001Z"), "duration": duration("P20Y19DT4H14M23.34S") }
+{ "date": null, "time": null, "datetime": datetime("1920-12-20T23:29:18.478Z"), "duration": null }
\ No newline at end of file
diff --git a/asterix-om/src/main/java/edu/uci/ics/asterix/om/base/temporal/ADateAndTimeParser.java b/asterix-om/src/main/java/edu/uci/ics/asterix/om/base/temporal/ADateAndTimeParser.java
index c26125f..636b831 100644
--- a/asterix-om/src/main/java/edu/uci/ics/asterix/om/base/temporal/ADateAndTimeParser.java
+++ b/asterix-om/src/main/java/edu/uci/ics/asterix/om/base/temporal/ADateAndTimeParser.java
@@ -206,7 +206,8 @@
}
// error is thrown if more than three digits are seen for the millisecond part
- if (charAccessor.getCharAt(offset) >= '0' && charAccessor.getCharAt(offset) <= '9') {
+ if (charAccessor.getLength() > offset && charAccessor.getCharAt(offset) >= '0'
+ && charAccessor.getCharAt(offset) <= '9') {
throw new Exception("Wrong format of time instance: too many fields for millisecond.");
}
}
diff --git a/asterix-om/src/main/java/edu/uci/ics/asterix/om/base/temporal/ByteArrayCharSequenceAccessor.java b/asterix-om/src/main/java/edu/uci/ics/asterix/om/base/temporal/ByteArrayCharSequenceAccessor.java
index e1a2135..50d4d1a 100644
--- a/asterix-om/src/main/java/edu/uci/ics/asterix/om/base/temporal/ByteArrayCharSequenceAccessor.java
+++ b/asterix-om/src/main/java/edu/uci/ics/asterix/om/base/temporal/ByteArrayCharSequenceAccessor.java
@@ -25,7 +25,16 @@
return (char) (string[index + offset + beginOffset]);
}
- /* The offset is the position of the first letter in the byte array */
+ /**
+ * Reset the wrapped byte array.
+ *
+ * @param obj
+ * The byte array to be wrapped
+ * @param beginOffset
+ * The offset of the string stored in the byte array.
+ * @param offset
+ * The offset of the substring of the string stored (offset from the beginOffset).
+ */
public void reset(byte[] obj, int beginOffset, int offset) {
string = obj;
this.offset = offset;
diff --git a/asterix-om/src/main/java/edu/uci/ics/asterix/om/base/temporal/ICharSequenceAccessor.java b/asterix-om/src/main/java/edu/uci/ics/asterix/om/base/temporal/ICharSequenceAccessor.java
index 6b4e898..c475523 100644
--- a/asterix-om/src/main/java/edu/uci/ics/asterix/om/base/temporal/ICharSequenceAccessor.java
+++ b/asterix-om/src/main/java/edu/uci/ics/asterix/om/base/temporal/ICharSequenceAccessor.java
@@ -16,8 +16,19 @@
public interface ICharSequenceAccessor<T> {
+ /**
+ * Return the character in the wrapped char sequence at the given index.
+ *
+ * @param index
+ * @return
+ */
public char getCharAt(int index);
+ /**
+ * Get the length of the wrapped char sequence.
+ *
+ * @return
+ */
public int getLength();
}
diff --git a/asterix-runtime/pom.xml b/asterix-runtime/pom.xml
index 3b5b392..da3d5d0 100644
--- a/asterix-runtime/pom.xml
+++ b/asterix-runtime/pom.xml
@@ -1,4 +1,5 @@
-<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
+<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+ xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
<modelVersion>4.0.0</modelVersion>
<parent>
<artifactId>asterix</artifactId>
@@ -55,26 +56,26 @@
</dependency>
<dependency>
<groupId>edu.uci.ics.hyracks</groupId>
- <artifactId>hyracks-storage-am-btree</artifactId>
+ <artifactId>hyracks-storage-am-btree</artifactId>
<version>0.2.1-SNAPSHOT</version>
</dependency>
<dependency>
- <groupId>edu.uci.ics.asterix</groupId>
- <artifactId>asterix-transactions</artifactId>
+ <groupId>edu.uci.ics.asterix</groupId>
+ <artifactId>asterix-transactions</artifactId>
<version>0.0.4-SNAPSHOT</version>
<scope>compile</scope>
</dependency>
<dependency>
- <groupId>org.twitter4j</groupId>
- <artifactId>twitter4j-core</artifactId>
- <version>2.2.3</version>
+ <groupId>org.twitter4j</groupId>
+ <artifactId>twitter4j-core</artifactId>
+ <version>2.2.3</version>
</dependency>
<dependency>
- <groupId>org.apache.hadoop</groupId>
- <artifactId>hadoop-core</artifactId>
- <version>0.20.2</version>
- <type>jar</type>
- <scope>compile</scope>
+ <groupId>org.apache.hadoop</groupId>
+ <artifactId>hadoop-core</artifactId>
+ <version>0.20.2</version>
+ <type>jar</type>
+ <scope>compile</scope>
</dependency>
</dependencies>
diff --git a/asterix-runtime/src/main/java/edu/uci/ics/asterix/runtime/operators/file/AdmSchemafullRecordParserFactory.java b/asterix-runtime/src/main/java/edu/uci/ics/asterix/runtime/operators/file/AdmSchemafullRecordParserFactory.java
index c8511d2..d7714dc 100644
--- a/asterix-runtime/src/main/java/edu/uci/ics/asterix/runtime/operators/file/AdmSchemafullRecordParserFactory.java
+++ b/asterix-runtime/src/main/java/edu/uci/ics/asterix/runtime/operators/file/AdmSchemafullRecordParserFactory.java
@@ -294,18 +294,34 @@
parseConstructor(ATypeTag.DATE, objectType, out);
break;
}
+ case AdmLexerConstants.DATE_LITERAL: {
+ parseDate(token.image, out);
+ break;
+ }
case AdmLexerConstants.TIME_CONS: {
parseConstructor(ATypeTag.TIME, objectType, out);
break;
}
+ case AdmLexerConstants.TIME_LITERAL: {
+ parseTime(token.image, out);
+ break;
+ }
case AdmLexerConstants.DATETIME_CONS: {
parseConstructor(ATypeTag.DATETIME, objectType, out);
break;
}
+ case AdmLexerConstants.DATETIME_LITERAL: {
+ parseDatetime(token.image, out);
+ break;
+ }
case AdmLexerConstants.DURATION_CONS: {
parseConstructor(ATypeTag.DURATION, objectType, out);
break;
}
+ case AdmLexerConstants.DURATION_LITERAL: {
+ parseDuration(token.image, out);
+ break;
+ }
case AdmLexerConstants.POINT_CONS: {
parseConstructor(ATypeTag.POINT, objectType, out);
break;
diff --git a/asterix-runtime/src/main/javacc/nontagged/AdmLexer.jj b/asterix-runtime/src/main/javacc/nontagged/AdmLexer.jj
index f556d54..9d9f3ff 100644
--- a/asterix-runtime/src/main/javacc/nontagged/AdmLexer.jj
+++ b/asterix-runtime/src/main/javacc/nontagged/AdmLexer.jj
@@ -144,13 +144,13 @@
<DEFAULT>
TOKEN :
{
- <TIMEZONE_LITERAL : (("+"|"-")<INTEGER>(":")<INTEGER>) | ("Z") >
+ <TIMEZONE_LITERAL : (("+"|"-")<DIGIT><DIGIT>(":")<DIGIT><DIGIT>) | (("+"|"-")<DIGIT><DIGIT><DIGIT><DIGIT>) | ("Z") >
}
<DEFAULT>
TOKEN :
{
- <DATE_LITERAL : ("-")?<INTEGER>("-")<INTEGER>("-")<INTEGER> (<TIMEZONE_LITERAL>)? >
+ <DATE_LITERAL : (("-")?<DIGIT><DIGIT><DIGIT><DIGIT>("-")<DIGIT><DIGIT>("-")<DIGIT><DIGIT>) | (("-")?<DIGIT><DIGIT><DIGIT><DIGIT><DIGIT><DIGIT><DIGIT><DIGIT>) >
}
<DEFAULT>
@@ -162,7 +162,7 @@
<DEFAULT>
TOKEN :
{
- <TIME_LITERAL : <INTEGER>(":")<INTEGER>(":")<INTEGER> ( (":")<INTEGER>)? (<TIMEZONE_LITERAL>)? >
+ <TIME_LITERAL : (<DIGIT><DIGIT>(":")<DIGIT><DIGIT>(":")<DIGIT><DIGIT> ( (".")<DIGIT>(<DIGIT>((<DIGIT>)?))?)? ((("+"|"-")<DIGIT><DIGIT>(":")<DIGIT><DIGIT>) | ("Z"))?) | (<DIGIT><DIGIT><DIGIT><DIGIT><DIGIT><DIGIT> (<DIGIT>(<DIGIT>((<DIGIT>)?))?)? ((("+"|"-")<DIGIT><DIGIT><DIGIT><DIGIT>) | ("Z"))?) >
}
<DEFAULT>
@@ -174,7 +174,7 @@
<DEFAULT>
TOKEN :
{
- <DATETIME_LITERAL : ("-")? <INTEGER>("-")<INTEGER>("-")<INTEGER>("T")<INTEGER>(":")<INTEGER>(":")<INTEGER> ( (":")<INTEGER>)? (<TIMEZONE_LITERAL>)?>
+ <DATETIME_LITERAL : (("-")?<DIGIT><DIGIT><DIGIT><DIGIT>("-")<DIGIT><DIGIT>("-")<DIGIT><DIGIT>("T")<DIGIT><DIGIT>(":")<DIGIT><DIGIT>(":")<DIGIT><DIGIT> ( (".")<DIGIT>(<DIGIT>((<DIGIT>)?))?)? ((("+"|"-")<DIGIT><DIGIT>(":")<DIGIT><DIGIT>) | ("Z"))?) | (("-")?<DIGIT><DIGIT><DIGIT><DIGIT><DIGIT><DIGIT><DIGIT><DIGIT>("T")<DIGIT><DIGIT><DIGIT><DIGIT><DIGIT><DIGIT> (<DIGIT>(<DIGIT>((<DIGIT>)?))?)? ((("+"|"-")<DIGIT><DIGIT><DIGIT><DIGIT>) | ("Z"))?)>
}
<DEFAULT>
@@ -186,7 +186,7 @@
<DEFAULT>
TOKEN :
{
- <DURATION_LITERAL : ("-")? ("D")(<INTEGER>("Y"))?(<INTEGER>("M"))?(<INTEGER>("D"))?(("T")(<INTEGER>("H"))?(<INTEGER>("M"))?(<INTEGER>("S"))?)?>
+ <DURATION_LITERAL : ("-")? ("P")(<INTEGER>("Y"))?(<INTEGER>("M"))?(<INTEGER>("D"))?(("T")(((<INTEGER>("H"))(<INTEGER>("M"))?(<INTEGER>((".")<DIGIT>(<DIGIT>(<DIGIT>)?)?)?("S"))?) | ((<INTEGER>("M"))(<INTEGER>((".")<DIGIT>(<DIGIT>(<DIGIT>)?)?)?("S"))?) | ((<INTEGER>((".")<DIGIT>(<DIGIT>(<DIGIT>)?)?)?("S")))))?>
}
<DEFAULT>