Table of Contents

• Overview
• Prerequisite
• Tracing
• Syntax highlighting
• Column
• Search
• Parse
• Unrecognized columns
• Appendix: Regular Expression

Overview

SuperLogReader is a tracer and a reader for LimpidLog, AuditLog and J2EE logging system.

• It provides source code level tracing facility: a single mouse click will show the source code and the line in question will be marked and highlighted.
• It provides syntax highlighting for easy reading.
• It support LimpidLog, AuditLog and native server logs.
• It provides both simple search and Regular Expression search, for both search down and up.
• It provides filter to select the log records you want.

Examples:

Prerequisite

See Prerequisite in Installation for general prerequisite.

Tracing

When you click a file name in a log messages on the Trace Panel, the Source Code Panel will display the source code. The line of the source file in question will be marked and highlighted. The file name must be in the correct format. It can be generated by Java stack trace.

Syntax highlighting

SuperLogReader displays log information by log records. Each record consists of many columns. The syntax highlighting shows different color for different column. The colors can be set by mouse clicks

Column

LimpidLog and AuditLog store log messages in relational database. So it is very easy to parse log information. Each record is naturally divided into columns.

For native log file from J2EE servers. Log files are flat text files. When a log messages is generated, it is written "segment" by "segment". The "segment" is similar to column in relational databases. SuperLogReader calls "segment" as column. 

Filter

The Filter filters unwanted log messages using WHERE clause of SQL. To be user friendly, SuperLogReader provides a GUI input panel for the time period part of the WHERE clause.

When you write WHERE clause, you need to know the column names of LimpidLog and AuditLog message table. See Columns of LimpidLog message table for detailed information.

Search

Search is based on the current position of the caret. If a search is successful, the found text will be highlighted and the caret will be placed on a new position as described below: 

• for Down-ward search: the end of the found text.
• for Up-ward search: the beginning of the found text.

The search facility searches record by record and column by column. The delimiters of columns and records are not part of log messages, so they are not searchable.

The search facility provides the following algorithms: 

• Simple, case insensitive: Simple search, ignore cases.
• Simple, case sensitive: Simple search, matches cases.
• Regular Expression: Regular Expression search. If you are not familiar with it,  See Regular Expression for more information.

Parse

For LimpidLog and AuditLog, log records are stored in a relational database. So, parse is easy and always correct.

But, for flat log files, there is no guarantee. The following is some common issues:

• If a part of an application message falls to the pattern of the start of a record, SuperLogReader may be confused. For example, Weblogic uses "####" as start symbol for each record. If a message from your application contains multi-lines and a line starts with "####", SuperLogReader may or may not parse it correctly.
• Log files may be corrupted by many reasons. 
• A log file may grow too big. Your J2EE server may rotate log files to avoid that. If a record spans to two files during rotation, SuperLogReader will meet difficulties. We have such experience with Websphere 6. 

These problems are rare. SuperLogReader tries its best to solve these issues. 

Unrecognized columns

If a column is unrecognized, it will be displayed as "????".

Appendix: Regular Expression

Regular expression is a powerful search mechanism. If you are not familiar with it, Google for "regular expression". You will find a lot of tutorials. Many of the tutorials are for specific software. You still can read them even you are not interested in the specific software. The general rules are the same among software.