Py-hiverunner

Added
- Support shell-specific `source` (`hive`) and ``!run`` (`beeline`) commands. These commands allow one to import and execute the contents of external files in statements or scripts.

Changed
- Default supported Hive version is now 2.3.3 (was 1.2.1).
- Default supported Tez version is now 0.9.1 (was 0.7.0).
- Supported Java version is 8 (was 7).
- In-memory DB used by HiveRunner is now Derby (was HSQLDB).
- Log4J configuration file removed from jar artifact.
- System property to configure command shell emulation mode renamed to `commandShellEmulator` (was `commandShellEmulation`).

Changed
- Fixed issue where if case of column name in a file was different to case in table definition they would be treated as different [73](https://github.com/klarna/HiveRunner/issues/73).
- The way of setting writable permissions on JUnit temporary folder changed to make it compatible with Windows [63](https://github.com/klarna/HiveRunner/issues/63).

Added
- Added functionality for headers in TSV parser. This way you can dynamically add TSV files declaring a subset of columns using insertInto.

Added
- Added debug logging of result set. Enable by setting log4j.logger.com.klarna.hiverunner.HiveServerContainer=DEBUG in log4j.properties.

Added
- Added methods to the shell that allow statements contained in files to be executed and their results gathered. These are particularly useful for HQL scripts that generate no table based data and instead write results to STDOUT. In practice we've seen these scripts used in data processing job orchestration scripts (e.g `bash`) to check for new data, calculate processing boundaries, etc. These values are then used to appropriately configure and launch some downstream job.
- Support abstract base class [48](https://github.com/klarna/HiveRunner/issues/48).

Changed
- Upgraded to Hive 1.2.1 (Note: new major release with backwards incompatibility issues). As of Hive 1.2 there are a number of new reserved keywords, see [DDL manual](https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDLLanguageManualDDL-Keywords,Non-reservedKeywordsandReservedKeywords) for more information. If you happen to have one of these as an identifier, you could either backtick quote them (e.g. \`date\`, \`timestamp\` or \`update\`) or set hive.support.sql11.reserved.keywords=false.
- Users of Hive version 0.14 or older are recommended to use HiveRunner version 2.6.0.
- Removed the custom HiveConf hive.vs. Use hadoop.tmp.dir instead.