Defining Potential Code Anomalies

You can view and modify existing definitions of potential code anomalies (other than range overlaps and range jumps) in the Interactive Analysis advanced search criteria in the Coding Standards folder.

To define a new code anomaly, you must define an advanced search criterion for the anomaly and a matching entry in the file <EA Home>\Data\CodeDefects.xml. The entry has the form:

<DEFECT Id="name"
Internal="True|False"
Enabled="True|False"
Caption="display name"
ListName="list name"
Criterion="path of criterion"
Model="Model Type"/>

where:

• Id is a unique name identifying the code anomaly in the product.
• Internal specifies whether the anomaly is implemented internally in program code (True), or externally in an advanced search criterion (False).
  Note: You must specify False. Code anomalies with an Internal value of True cannot be modified.
• Enabled specifies whether the code anomaly is displayed in the Executive Report.
• Caption is the display name for the anomaly in the Executive Report.
• ListName is the name of the list of anomalous code constructs displayed in the Executive Report category of the Interactive Analysis Code Search tool.
• Criterion is the full path name of the criterion in the Interactive Analysis Advanced Search tool, including the tab name (General) and any folder names. For example, General:Coding Standards\MOVE Statements\Possible Data Padding.
• Model is the Interactive Analysis model for the source files to be searched, e.g., "COBOL".
  Some of the commonly used model types are:

  • COBOL - COBOL source files
  • JAVAFILE - Java source files
  • JCL - JCL source files
  • NATURAL - NATURAL source files
  • PLI - PL/I source files
  Note: Not all possible model types are listed above. Some model types like PL/SQL files do not have level 2 support, so cannot have Code Search queries run against them. For more information, see Supported Features by Language.

  For some languages, such as C and C++, there are no pre-defined Code Search queries, so they have not been listed.

  Note: You can find the full list of models in the Enterprise Analyzer installation directory, under \Model\HyperCode. You can find the exact name of the model to use in the CodeDefects.xml in the model file itself. For example, you can find the model name for COBOL in the NAME attribute in the following line in the Cobol.xml file:

  <HYPERCODEMODEL NAME="COBOL" DESCRIPTION="COBOL Programs" VERSION="1.1">

The following are examples of different defect entries:

<DEFECT Criterion="General:Coding Standards\MOVE Statements\Possible Sign Loss" 
ListName="Possible Sign Loss" 
Caption="Possible Sign Loss" 
Group="Cobol Coding Standards" 
Enabled="True" 
Internal="False" 
Id="Sign" 
Model="COBOL"/>

<DEFECT Criterion="General:Coding Standards\Unaligned Assignments" 
ListName="Unaligned Assignment" 
Caption="Unaligned Assignment" 
Group="Pli Coding Standards" 
Enabled="True" 
Internal="False" 
Id="Unaligned" 
Model="PLI"/>

<DEFECT Criterion="General:Java Best Practices\Basic\Empty If Statements" 
ListName="Empty If Stats" 
Caption="Empty If Statements" 
Group="Java Results" Enabled="True" 
Internal="False" 
Id="EmpIfStnts" 
Model="JAVAFILE"/>

You can display the anomaly caption in Japanese or Korean in the Executive Report by creating an entry for the anomaly in the file <EA Home>\Language\[Jpn|Kor]\CodeDefects.xrc. The entry has the form:

<String name="name"
listname="list name"
caption="translated display name"
description="description"/>

where:

• name is the unique name of the code anomaly in the product (Id attribute of CodeDefects.xml entry).
• listname is the name of the list of anomalous code constructs displayed in the Executive Report category of the Interactive Analysis Code Searchr tool (ListName attribute of CodeDefects.xml entry).
• caption is the translated display name for the anomaly in the Executive Report.
• description contains a description of the entry.