Pig

.travis.yml

@@ -3,8 +3,13 @@ language: java
 jdk:
 - oraclejdk8
 
+install: true
+
+script:
+- mvn verify
+
 after_success:
-- mvn clean cobertura:cobertura coveralls:report
+- mvn coveralls:report
 - test "${TRAVIS_PULL_REQUEST}" == "false" && test "${TRAVIS_TAG}" != "" && mvn deploy --settings travis/settings.xml
 
 cache:

Makefile

@@ -1,13 +1,13 @@
 all: full
 
 full:
-	mvn clean checkstyle:check cobertura:cobertura javadoc:jar package
+	mvn clean javadoc:jar package
 
 clean:
 	mvn clean
 
 test:
-	mvn clean checkstyle:check test
+	mvn clean verify
 
 jar:
 	mvn clean package
@@ -16,13 +16,13 @@ release:
 	mvn -B release:prepare release:clean
 
 coverage:
-	mvn clean cobertura:cobertura
+	mvn clean clover2:setup test clover2:aggregate clover2:clover
 
 doc:
 	mvn clean javadoc:javadoc
 
 see-coverage: coverage
-	cd target/site/cobertura; python -m SimpleHTTPServer
+	cd target/site/clover; python -m SimpleHTTPServer
 
 see-doc: doc
 	cd target/site/apidocs; python -m SimpleHTTPServer

README.md

@@ -2,7 +2,9 @@
 
 [![Build Status](https://travis-ci.org/yahoo/validatar.svg?branch=master)](https://travis-ci.org/yahoo/validatar) [![Coverage Status](https://coveralls.io/repos/yahoo/validatar/badge.svg?branch=master)](https://coveralls.io/r/yahoo/validatar?branch=master) [![Download](https://api.bintray.com/packages/yahoo/maven/validatar/images/download.svg)](https://bintray.com/yahoo/maven/validatar/_latestVersion)
 
-Functional testing framework for Big Data pipelines. Current support is only for Hive, but we are planning support for Pig as well as others.
+Functional testing framework for Big Data pipelines. Currently support querying pipeline results through Hive (HiveServer2) and Pig (PigServer).
+
+Validatar is currently compiled against *Pig-0.14*. Running against an older or newer version may result in issues if interfaces have changed. These are relatively minor from experience and can be fixed with relatively minor fixes to engine code.
 
 ## How to build Validatar
 
@@ -14,40 +16,56 @@ Run:
 
 ## How to run
 
-To run Validatar:
+Use hadoop jar validatar-jar-with-dependencies.jar com.yahoo.validatar.App --help (or -h) for Help
+
+### To run Hive tests in Validatar:
 
     export HADOOP_CLASSPATH="$HADOOP_CLASSPATH:/path/to/hive/jdbc/lib/jars/*"
-    hadoop jar validatar-jar-with-dependencies.jar com.yahoo.validatar.App -s tests/ --report report.xml
+    hadoop jar validatar-jar-with-dependencies.jar com.yahoo.validatar.App -s tests/ --report report.xml --hive-jdbc ...
 
-You will also need the settings specified for the engine you are planning to run.
+Hive needs the JDBC uri of HiveServer2. Note that the DB is in the URI. Do not add it if your queries use
+```
+... FROM DB.TABLE WHERE ...
+
+--hive-jdbc "jdbc:hive2://<URI>/<DB>;<Optional params: E.g. sasl.qop=auth;principal=hive/<PRINCIPAL_URL> etc>
+```
+
+### To run Pig tests in Validatar:
+
+    export HADOOP_CLASSPATH="$HADOOP_CLASSPATH:/path/to/pig/lib/*" (Add other jars here depending on your pig exec type or if hive/hcat is used in Pig)
+    hadoop jar validatar-jar-with-dependencies.jar com.yahoo.validatar.App -s tests/ --report report.xml --pig-exec-type mr --pig-setting 'mapreduce.job.acl-view-job=*' ...
+
+Pig parameters are not supported in the pig query. Instead, use our parameter substitution (see below).
 
 ## Writing Tests
 
 ### Test file format
 
-Test files are written in the YAML format. The schema is as follows:
+Test files are written in the YAML format. See example in src/test/resources/. The schema is as follows:
 
 ```
 name: Test family name : String
 description: Test family description : String
 queries:
    - name: Query name : String : Ex "Analytics"
-     engine: Execution engine : String ("Hive")
+     engine: Execution engine : String Ex "hive" or "pig"
      value: Query : String : Ex "SELECT COUNT(*) AS pv_count FROM page_data"
    ...
 tests:
    - name: Test name : String
      description: Test description : String
      asserts:
-        - Assertion on some query. Query name is prefixed to the value. : Ex: Analytics.pv_count > 10000
+        - Assertion on some query. Query name is prefixed to the value. : Ex Analytics.pv_count > 10000
    ...
 ```
 
 Queries are named, this name is used as a namespace for all the values returned from the query. In the above example, we created a query named "Analytics". It stores the return value "pv_count". We are then able to use this in our later asserts.
 
+Validatar can run a single test file or a folder of test files. Use the --help option to see more details.
+
 ### Assertions
 
-Assertions are quite flexibile, allowing for the following operations:
+Assertions are quite flexible, allowing for the following operations:
 
 ```
                    >  : greater than
@@ -82,10 +100,11 @@ Version | Notes
 0.1.8 | Null types in Hive results fix
 0.1.9 | Empty results handling bug fix
 0.2.0 | Internal switch to Java 8. hive-queue is no longer a setting. Use hive-setting.
+0.3.0 | Pig support added.
 
 ## Members
 
-Akshai Sarma, akshaisarma@gmail.com  
+Akshai Sarma, akshaisarma@gmail.com
 Josh Walters, josh@joshwalters.com
 
 ## Contributors