Merge pull request #206 from tigergraph/jdbc-dev-sync

[Automated]JDBC Driver Release 1.3.11
tigergraph · Oct 20, 2023 · f5fe532 · f5fe532
2 parents eb3da5d + 4ef681c
commit f5fe532
Show file tree

Hide file tree

Showing 8 changed files with 220 additions and 177 deletions.
diff --git a/tools/etl/tg-jdbc-driver/CHANGELOG.md b/tools/etl/tg-jdbc-driver/CHANGELOG.md
@@ -1,7 +1,15 @@
+## Release 1.3.11
+* Supported TG Version: 2.4.1+
+* Bugfixes:
+    1. Refine the retry mechanism for connection errors.
+    2. Set default array element type to string for Spark query.
+* Features:
+    2. Provide release package with separated dependency jars.
+
 ## Release 1.3.10
 * Supported TG Version: 2.4.1+
 * Bugfixes:
-    1. Remove single quote from Spark query string
+    1. Remove single quote from Spark query string.
 
 ## Release 1.3.9
 * Supported TG Version: 2.4.1+

diff --git a/tools/etl/tg-jdbc-driver/README.md b/tools/etl/tg-jdbc-driver/README.md
@@ -67,6 +67,7 @@ Please refer to [Versions](#versions) to find the **suitable version**.
 * Use the jar file directly (e.g. include the JDBC driver for Spark):
 
     Go to [maven.org](https://search.maven.org/artifact/com.tigergraph/tigergraph-jdbc-driver) to get the jar file.
+    The `tigergraph-jdbc-driver-<version>.jar` is the uber jar containing all the runtime dependencies, and the `tigergraph-jdbc-driver-<version>.tar.gz` packs all jars together separately.
 
 ## Minimum viable snippet
 
@@ -554,7 +555,8 @@ df.write.mode("append").format("jdbc").options(
     "sep" -> "|", // separator between columns
     "eol" -> "\n", // End Of Line
     "batchsize" -> "10000",
-    "debug" -> "0")).save()
+    "debug" -> "0",
+    "logFilePattern" -> "/tmp/jdbc.log")).save()
 
 ```
 **If your TG version is 3.9.0 or higher, please use the following new features:**
@@ -872,6 +874,7 @@ Tigergraph JDBC Driver supports 4 logging levels: 0 -> ERROR, 1 -> WARN, 2 -> IN
 It supports two logging frameworks:
 - java.util.logging (JUL)
   - To use logger, only need to pass in logging level by `properties.put("debug", "0|1|2|3");`, it will initialize with default logging handler and formatter, which only print logs to console.
+  - To print logs to file, need to explicitly specify the log file path by `properties.put("logFilePattern", "/path/to/log/file");`
   - To customize the JUL configuration, please provide your logging configuration file `logging.properties` and specify the JVM system property **explicitly**: `-Djava.util.logging.config.file=path_to_logging.properties`. Reference: [JUL Documentation](https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html).
 
     For example, create a logging configuration file `logging.properties` with following contents:
@@ -972,6 +975,7 @@ It could be insecure to type in some sensitive properties like `password`, `toke
 | `filename`                | (none)      | The filename defined in the loading job.                     | Yes                                                          | loading job       |
 | `sep`                     | (none)      | Column separator. E.g., `,`.                                 | Yes                                                          | loading job       |
 | `eol`                     | (none)      | Line separator. E.g., `\n`.                                  | Yes                                                          | loading job       |
+| `maxRetryCount`           | 10          | Max retry count for transient loading failures               | Yes                                                          | loading job       |
 | `dbtable`                 | (none)      | The specification of the operation of the form: `operation_type operation_object`.  For loading job: `job JOB_NAME`; E.g. for querying loading statistics: `jobid JOB_ID`. | Yes                                                          | Spark             |
 | `batchsize`               | 1000        | Maximum number of lines per POST request.                    | Yes                                                          | Spark loading job |
 | `partitionColumn`         | (none)      | The column used for partitioning, it has to be numeric or date or timestamp column. | No | Spark partitioning query |

diff --git a/tools/etl/tg-jdbc-driver/tg-jdbc-driver/pom.xml b/tools/etl/tg-jdbc-driver/tg-jdbc-driver/pom.xml
@@ -4,7 +4,7 @@
 
     <groupId>com.tigergraph</groupId>
     <artifactId>tigergraph-jdbc-driver</artifactId>
-    <version>1.3.10</version>
+    <version>1.3.11</version>
     <packaging>jar</packaging>
 
     <name>TigerGraph JDBC Driver Parent</name>
@@ -206,6 +206,23 @@
                             <autoReleaseAfterClose>true</autoReleaseAfterClose>
                         </configuration>
                     </plugin>
+                    <plugin>
+                        <artifactId>maven-assembly-plugin</artifactId>
+                        <configuration>
+                            <appendAssemblyId>false</appendAssemblyId>
+                            <descriptors>
+                                <descriptor>src/assembly/package.xml</descriptor>
+                            </descriptors>
+                        </configuration>
+                        <executions>
+                            <execution>
+                                <phase>package</phase>
+                                <goals>
+                                    <goal>single</goal>
+                                </goals>
+                            </execution>
+                        </executions>
+                    </plugin>
                     <plugin>
                         <groupId>org.apache.maven.plugins</groupId>
                         <artifactId>maven-gpg-plugin</artifactId>

diff --git a/tools/etl/tg-jdbc-driver/tg-jdbc-driver/src/assembly/package.xml b/tools/etl/tg-jdbc-driver/tg-jdbc-driver/src/assembly/package.xml
@@ -0,0 +1,26 @@
+
+<assembly xmlns="http://maven.apache.org/plugins/maven-assembly-plugin/assembly/1.1.2"
+          xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+          xsi:schemaLocation="http://maven.apache.org/plugins/maven-assembly-plugin/assembly/1.1.2 http://maven.apache.org/xsd/assembly-1.1.2.xsd">
+    <id>release</id>
+    <includeBaseDirectory>false</includeBaseDirectory>
+
+    <formats>
+        <format>tar.gz</format>
+    </formats>
+    <files>
+        <file>
+            <source>${project.build.directory}/original-${project.artifactId}-${project.version}.jar</source>
+            <destName>${project.artifactId}-${project.version}.jar</destName>
+            <outputDirectory>/</outputDirectory>
+        </file>
+    </files>
+    <dependencySets>
+        <dependencySet>
+            <outputDirectory>/lib</outputDirectory>
+            <excludes>
+                <exclude>${project.groupId}:${project.artifactId}:jar:*</exclude>
+            </excludes>
+        </dependencySet>
+    </dependencySets>
+</assembly>
diff --git a/...tl/tg-jdbc-driver/tg-jdbc-driver/src/main/java/com/tigergraph/jdbc/TigergraphDialect.java b/...tl/tg-jdbc-driver/tg-jdbc-driver/src/main/java/com/tigergraph/jdbc/TigergraphDialect.java
@@ -24,7 +24,8 @@ public Option<DataType> getCatalystType(
       int sqlType, String typeName, int size, MetadataBuilder md) {
     if (sqlType == Types.ARRAY) {
       // LIST.ELEMENT_TYPE => ELEMENT_TYPE
-      Option<DataType> elementType = toCatalystType(typeName.split("\\.")[1]);
+      String[] tokens = typeName.split("\\.");
+      Option<DataType> elementType = toCatalystType(tokens.length > 1 ? tokens[1] : "");
       if (!elementType.isEmpty()) {
         return Option.apply(new ArrayType(elementType.get(), false));
       }