/PLSQL-Clone-Detection

Token based and tree based clone detection tools for PL/SQL-queries. This is a bachelor thesis.

Primary LanguageJava

PL/SQL Clone Detection

Clone Detection with PMD (token based) and Clone Digger (tree based).

Contents

  1. PMD
  2. Clone Digger
  3. Typical usage pattern
  4. Helper tools

PMD

Usage

GUI
$ cd tool_pmd/bin
$ ./run.sh cpdgui

In the GUI select the source folder containing the query-files. Select 'PL/SQL' as language. After getting the results, they can be saved as an XML-file by selecting 'File' > 'Save as XML'. Note: because of a bug the selected source folder is wrongly displayed. The last part of the path is shown twice. Simply delete the last part of the path to make it work.

CLI
$ cd tool_pmd/bin
$ ./run.sh cpd --minimum-tokens 75 --files Path/To/PLSQL-Queries --language plsql --format xml > output_PMD.xml

Generates output_PMD.xml with clones.

More CLI options: http://pmd.sourceforge.net/snapshot/usage/cpd-usage.html

Clone Digger

Installation

$ cd tool_clonedigger/clonedigger/plsql_antlr
$ javac -cp antlr-4.6-complete.jar:. TreeProducer.java
$ jar -cfe TreeProducer.jar TreeProducer  *.class org/
$ rm *.class

Now copy the generated TreeProducer.jar to your working directory.

$ cd tool_clonedigger
$ sudo python setup.py install --user

Will install in default location ~/Library/Python/2.7/site-packages/ and add 'clonedigger' command to /usr/local/bin.

Usage

Now from the working directory (directory with the TreeProducer.jar), execute the following command.

$ clonedigger -l plsql --cpd-output [Path/To/PLSQL-Queries]

[Path/To/PLSQL-Queries] is the path to the folder with all SQL-files. This will generate a clones.xml file. Clone Digger may take a while to complete.

Typical use pattern

A typical use pattern:

  1. Convert the single XML-file with all the exported queries to separate SQL-files. See 'helpertool_create_query_files'.
  2. Use PMD or Clone Digger on the SQL-files to generate an clone.xml file with all found clones.
  3. If Clone Digger has been used, convert the output to clone classes. See 'helpertool_combine_clone_instances'.
  4. Categorize the foundclones based on some heuristics. See 'helpertool_categorize_clones'.
  5. Create an HTML-file to easily view all found clones. See 'helpertool_clone_viewer'.

Categorizing clones, displaying clones (HTML), combining clone instances

Some helpertools are created:

  • Create Query Files - To create single SQL-files from the exported query list by CALI.
  • Combine Query instances - Clone Digger outputs clones is pairs instead of classes. This tool can be used to create classes from those pairs.
  • Categorize Clones - Create three categories of clones.
  • Easely view clones - Use this tool to generate an interactive HTML-file to easily view the clones.

The usage and installation of these tools is explained in the README file in each corresponding directory.