Apache Tika Utility

Apache Tika is a toolkit from the Apache Software Foundation that detects and extracts metadata and text from over a thousand different file types (such as PPT, XLS, and PDF).

In the context of ABCD, Tika is the engine that powers the Full Text Indexing capabilities of the Digital Library module (e.g., DubCore database). It is not part of the original CISIS suite but is distributed as an external Java dependency (tika-app.jar).

Role in ABCD

Unlike standard bibliographic databases where data is typed manually, Digital Libraries (repositories) require the system to "read" the content of uploaded documents.

• Extraction: When a file is uploaded via the Batch Import utility, ABCD executes a system call to the Tika JAR file.
• Conversion: Tika reads the binary format (PDF, Word, etc.) and converts it into a standard HTML or text stream.
• Storage: This extracted text is injected into Tag 99 (Text Word) of the Dublin Core record.
• Indexing: Finally, the WXIS engine indexes this content, making the full document searchable via the OPAC.

Prerequisites

Since Tika is a Java application, the server running ABCD must have:

1. Java Runtime Environment (JRE) installed and accessible via the system PATH.
2. The tika-app.jar file located in the configured utilities folder (often cgi-bin or central/dataentry/wxis).
3. Correct permissions for the web server user (www-data or IUSR) to execute Java.

Usage (Command Line)

Although ABCD handles the execution automatically during imports, administrators can use Tika manually for testing or debugging.

Syntax:

java -jar tika-app.jar [options] [file...]

Common Options

Based on Tika version 2.1.0+:

Option	Description
-x or --xml	Output XHTML content (Default). Preserves some structure.
-h or --html	Output HTML content.
-t or --text	Output plain text content. Useful for pure indexing without markup.
-m or --metadata	Extracts only metadata (Author, Creation Date, etc.).
-l or --language	Detects the language of the file.
--gui	Starts the Apache Tika GUI (requires graphical environment).
--fork	Uses "Fork Mode" for out-of-process extraction (safer for batch stability).

Example: Manual Extraction

To test if Tika is working correctly on your server, try extracting text from a sample PDF:

java -jar tika-app.jar -t sample_document.pdf

If this command outputs the text of the PDF to the console, your Java environment is correctly set up.

Configuration in ABCD

To enable Tika integration, you must configure the path in the central configuration file (config.php or dr_path.def).

• IMPORTPDF=Y: Enables the import toolbar icon.
• Collection Path: Defines where physical files are stored.

Troubleshooting

• "Java command not found": Ensure JRE is installed and added to the environment variables.
• Slow Import: On Linux, starting the JVM for every file can be slow. Using the --fork option or running Tika in server mode (if supported by the script) can improve performance.
• Memory Errors: Large PDF files may require increasing the Java heap size (e.g., java -Xmx1g -jar ...).

References

• Official Site: tika.apache.org
• ABCD Module: Dublin Core Repository