embulk-filter-calcite 0.1.0 → 0.1.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: fb9b57b9556e4b1ea27783f26a9e4f13275b8b92
-  data.tar.gz: 61227ab14eceb59c0e2d40d04a101f12077cbc11
+  metadata.gz: a72ba31671d8572762457c6fede3f290ab0a81c2
+  data.tar.gz: dde80b10ba7bb64b85174a5469de93db29f89200
 SHA512:
-  metadata.gz: d8a5c2ebc604a0af4ac45ab1c113ce2deedcaa651fd5533098fb9fc05a75f529ab8b48c522469fb71d1a89da89b8c5dd43c8f9b1540735055e074f8c8f47dca3
-  data.tar.gz: adc86d69db46d7b426e3d869dcc687cf7324a278d6f202a0d89ac05d08124c4c0480cf288da14549644f0b7940f494f344ae08024d659000a9745f9d5b798f51
+  metadata.gz: 27bc215b3a74a4011271d3e082744f5eedccaa19405a2adf878e8aad0688d624f1f24b02d21a2beb1ff59e409b770b1f7fa6810ef89389ec91c96afd5df1d1d6
+  data.tar.gz: cc83d78f01ca54a125f39598db233a7acee8eb690e54bd069d290b318c409515fb7678f5b579817747b2ad616b51568530b892ff50ff5e69ce137849aa9e8ad7

data/CHANGELOG.md

@@ -1,3 +1,7 @@
+## 0.1.1 - 2017-03-22
+
+* [bug fix] Use default_timezone option to configure JDBC connection timeZone param [#11](https://github.com/muga/embulk-filter-calcite/pull/11)
+
 ## 0.1.0 - 2017-03-17
 
 The first release

data/README.md

@@ -33,8 +33,8 @@ Data types by Apache Calcite: https://calcite.apache.org/docs/reference.html#dat
 
 ## Configuration
 
-- **query**: SQL to run (string)
-- **default_timezone**: If the sql type of a column is `date`/`time`/`datetime` and the embulk type is `string`, column values are formatted int this default_timezone. You can overwrite timezone for each columns using column_options option. (string, default: `UTC`)
+- **query**: SQL to run (string, required)
+- **default_timezone**: Configure timezone that is used for JDBC connection properties and Calcite engine. This option is one of [JDBC connect parameters](https://calcite.apache.org/docs/adapter.html#jdbc-connect-string-parameters) provided by Apache Calcite. java.util.TimeZone's [AvailableIDs](http://docs.oracle.com/javase/7/docs/api/java/util/TimeZone.html#getAvailableIDs) can be specified. (string, default: 'UTC')
 
 ## Example
 
@@ -52,6 +52,14 @@ filters:
     query: SELECT first_name || last_name AS name, * FROM $PAGES
 ```
 
+Adds the new column by CURRENT_TIMESTAMP function.
+```yaml
+filters:
+  - type: calcite
+    query: SELECT CURRENT_TIMESTAMP, * FROM $PAGES
+    default_timezone: 'America/Los_Angeles'
+```
+
 SQL language provided by Apache Calcite: https://calcite.apache.org/docs/reference.html
 
 ## Build

data/build.gradle

@@ -17,7 +17,7 @@ configurations {
     provided
 }
 
-version = "0.1.0"
+version = "0.1.1"
 
 sourceCompatibility = 1.7
 targetCompatibility = 1.7

data/src/main/java/org/embulk/filter/calcite/CalciteFilterPlugin.java

@@ -84,6 +84,7 @@ public class CalciteFilterPlugin
     {
         PluginTask task = config.loadConfig(PluginTask.class);
         Properties props = System.getProperties(); // TODO should be configured as config option
+        setupProperties(task, props);
 
         // Set input schema in PageSchema
         PageSchema.schema = inputSchema;
@@ -107,6 +108,13 @@ public class CalciteFilterPlugin
         }
     }
 
+    private void setupProperties(PluginTask task, Properties props)
+    {
+        // @see https://calcite.apache.org/docs/adapter.html#jdbc-connect-string-parameters
+        props.setProperty("caseSensitive", "false"); // Relax case-sensitive
+        props.setProperty("timeZone", task.getDefaultTimeZone().getID());
+    }
+
     private PageConverter newPageConverter(PluginTask task, Schema inputSchema)
     {
         return new PageConverter(inputSchema, task.getDefaultTimeZone().toTimeZone());
@@ -116,9 +124,6 @@ public class CalciteFilterPlugin
     {
         String jdbcUrl = buildJdbcUrl();
         try {
-            // Relax case-sensitive
-            // @see https://calcite.apache.org/docs/adapter.html#jdbc-connect-string-parameters
-            props.setProperty("caseSensitive", "false");
             return new Driver().connect(jdbcUrl, props);
         }
         catch (SQLException e) {
@@ -240,6 +245,7 @@ public class CalciteFilterPlugin
         ColumnGetterFactory factory = newColumnGetterFactory(task, Optional.of(pageBuilder));
         List<ColumnGetter> getters = newColumnGetters(factory, task.getQuerySchema());
         Properties props = System.getProperties(); // TODO should be configured as config option
+        setupProperties(task, props);
         return new FilterPageOutput(outputSchema, task.getQuery(), pageBuilder, pageConverter, getters, props);
     }
 

data/src/main/java/org/embulk/filter/calcite/getter/FilterColumnGetterFactory.java

@@ -7,7 +7,6 @@ import org.embulk.input.jdbc.JdbcInputConnection;
 import org.embulk.input.jdbc.getter.ColumnGetter;
 import org.embulk.input.jdbc.getter.ColumnGetterFactory;
 import org.embulk.spi.PageBuilder;
-import org.embulk.spi.time.TimestampFormatter;
 import org.embulk.spi.type.Type;
 import org.joda.time.DateTimeZone;
 
@@ -28,18 +27,10 @@ public class FilterColumnGetterFactory
         String valueType = option.getValueType();
         Type toType = getToType(option);
         if (valueType.equals("coalesce") && sqlTypeToValueType(column, column.getSqlType()).equals("timestamp")) {
-            return new UTCTimestampColumnGetter(to, toType, newTimestampFormatter(option, "%Y-%m-%d"));
+            return new FilterTimestampColumnGetter(to, toType, option.getTimeZone().or(defaultTimeZone));
         }
         else {
             return super.newColumnGetter(con, task, column, option);
         }
     }
-
-    private TimestampFormatter newTimestampFormatter(JdbcColumnOption option, String defaultTimestampFormat)
-    {
-        return new TimestampFormatter(
-                option.getJRuby(),
-                option.getTimestampFormat().isPresent() ? option.getTimestampFormat().get().getFormat() : defaultTimestampFormat,
-                option.getTimeZone().or(defaultTimeZone));
-    }
 }

data/src/main/java/org/embulk/filter/calcite/getter/{UTCTimestampColumnGetter.java → FilterTimestampColumnGetter.java}

@@ -3,8 +3,8 @@ package org.embulk.filter.calcite.getter;
 import org.embulk.input.jdbc.getter.TimestampColumnGetter;
 import org.embulk.spi.PageBuilder;
 import org.embulk.spi.time.Timestamp;
-import org.embulk.spi.time.TimestampFormatter;
 import org.embulk.spi.type.Type;
+import org.joda.time.DateTimeZone;
 
 import java.sql.ResultSet;
 import java.sql.SQLException;
@@ -13,20 +13,15 @@ import java.util.Calendar;
 import static java.util.Calendar.getInstance;
 import static java.util.TimeZone.getTimeZone;
 
-public class UTCTimestampColumnGetter
+public class FilterTimestampColumnGetter
         extends TimestampColumnGetter
 {
-    private static ThreadLocal<Calendar> calendar = new ThreadLocal<Calendar>() {
-        @Override
-        protected Calendar initialValue()
-        {
-            return getInstance(getTimeZone("UTC"));
-        }
-    };
+    private static final ThreadLocal<Calendar> calendar = new ThreadLocal<>();
 
-    public UTCTimestampColumnGetter(PageBuilder to, Type toType, TimestampFormatter timestampFormatter)
+    public FilterTimestampColumnGetter(PageBuilder to, Type toType, DateTimeZone timeZone)
     {
-        super(to, toType, timestampFormatter);
+        super(to, toType, null);
+        calendar.set(getInstance(getTimeZone(timeZone.getID()))); // set TLS here
     }
 
     @Override

data/src/test/resources/org/embulk/filter/calcite/test/test_int_ops_filter.yml

@@ -1,2 +1,3 @@
 type: calcite
-query: 'SELECT id, id + 2, (id + 1) * id FROM $PAGES'
+query: 'SELECT id, id + 2, (id + 1) * id FROM $PAGES'
+default_timezone: 'UTC'

data/src/test/resources/org/embulk/filter/calcite/test/test_simple_filter.yml

@@ -1,2 +1,3 @@
 type: calcite
-query: 'SELECT * FROM $PAGES'
+query: 'SELECT * FROM $PAGES'
+default_timezone: 'UTC'

data/src/test/resources/org/embulk/filter/calcite/test/test_string_ops_filter.yml

@@ -1,2 +1,3 @@
 type: calcite
-query: 'SELECT purchase, purchase || purchase, CHAR_LENGTH(purchase), SUBSTRING(purchase FROM 2), SUBSTRING(purchase FROM 2 FOR 4) FROM $PAGES'
+query: 'SELECT purchase, purchase || purchase, CHAR_LENGTH(purchase), SUBSTRING(purchase FROM 2), SUBSTRING(purchase FROM 2 FOR 4) FROM $PAGES'
+default_timezone: 'UTC'

data/src/test/resources/org/embulk/filter/calcite/test/test_where_int_cond_filter.yml

@@ -1,2 +1,3 @@
 type: calcite
-query: 'SELECT * FROM $PAGES WHERE MOD(id, 2) = 0'
+query: 'SELECT * FROM $PAGES WHERE MOD(id, 2) = 0'
+default_timezone: 'UTC'

data/src/test/resources/org/embulk/filter/calcite/test/test_where_string_cond_filter.yml

@@ -1,2 +1,3 @@
 type: calcite
 query: SELECT * FROM $PAGES WHERE purchase LIKE '%0127'
+default_timezone: 'UTC'

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: embulk-filter-calcite
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
 platform: ruby
 authors:
 - Muga Nishizawa
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2017-03-17 00:00:00.000000000 Z
+date: 2017-03-23 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   requirement: !ruby/object:Gem::Requirement
@@ -65,7 +65,7 @@ files:
 - src/main/java/org/embulk/filter/calcite/adapter/page/PageSchemaFactory.java
 - src/main/java/org/embulk/filter/calcite/adapter/page/PageTable.java
 - src/main/java/org/embulk/filter/calcite/getter/FilterColumnGetterFactory.java
-- src/main/java/org/embulk/filter/calcite/getter/UTCTimestampColumnGetter.java
+- src/main/java/org/embulk/filter/calcite/getter/FilterTimestampColumnGetter.java
 - src/test/java/org/embulk/filter/calcite/TestCalciteFilterPlugin.java
 - src/test/resources/org/embulk/filter/calcite/test/test_int_ops_expected.csv
 - src/test/resources/org/embulk/filter/calcite/test/test_int_ops_filter.yml
@@ -99,7 +99,7 @@ files:
 - classpath/commons-lang3-3.2.jar
 - classpath/commons-logging-1.2.jar
 - classpath/commons-pool-1.5.4.jar
-- classpath/embulk-filter-calcite-0.1.0.jar
+- classpath/embulk-filter-calcite-0.1.1.jar
 - classpath/embulk-input-jdbc-0.8.2.jar
 - classpath/httpclient-4.5.2.jar
 - classpath/httpcore-4.4.4.jar