Reading Web Application Resources

23,037

Solution 1

The right location (and also the common practice) is to place them under your source directory, which will then gets compiled into WEB-INF/classes directory. I'm not sure what you meant by "classes directory is volatile" in your response to @Dave, but this is how most (if not all) Java web apps store things. WEB-INF/classes is not just for Java classes. It's common to see logging properties file (like log4j), Hibernate and Spring XML files stored under source directory and you can safely access the files using something like this:-

// in this case, the business-areas.sql is located right under "source/sql" directory
InputStream is = getClass().getClassLoader().getResourceAsStream("sql/business-areas.sql");
BufferedReader br = new BufferedReader(new InputStreamReader(is));

Some useful information about the use of META-INF: What's the purpose of META-INF?

Solution 2

I had similar concerns as Dave Jarvis about mixing resources with classes and lib, so I did some fiddling and found this solution:

I placed my resource files in WEB-INF/resources. Then, to load them, I used this:

getClass().getClassLoader().getResourceAsStream("../resources/main.xml");

I don't know that using a .. is a much cleaner solution, but my files are at least not mixed with classes or jars.

Share:
23,037
Dave Jarvis
Author by

Dave Jarvis

https://dave.autonoma.ca/blog/

Updated on July 09, 2022

Comments

  • Dave Jarvis
    Dave Jarvis almost 2 years

    Background

    Developing a simple web application (Eclipse + JBoss + Apache Tomcat) to generate XML files.

    Problem

    The "Business Area" list queries against the database, and the "Column Cluster" list queries the database using the selected "Business Area" items. Both of these are unique queries that are stored external text files.

    The files are currently stored in the following locations:

    • WebContent/META-INF/business-areas.sql
    • WebContent/META-INF/column-clusters.sql

    These are then used to seed PreparedStatements.

    Source Code

    The method to read the SQL code might resemble:

      private String getSQL() {
        String result = "";
    
        try {
          BufferedReader br = open( "business-areas.sql" );
          String line = null;
    
          while( (line = br.readLine()) != null ) {
            result += line;
          }
    
          br.close();
        }
        catch( Exception e ) {
          e.printStackTrace();
        }
    
        return result;
      }
    

    Questions

    I would like to know:

    1. What are the best practices for storing such assets for deployment as part of a web app? (That is, is META-INF a good location, or is META-INF/resources preferred?)
    2. What APIs would you recommend for reading the file content? (That is, how do I write the open method so that it finds the files to open?)

    I already have JNDI in place to establish the database connection, but would rather not use JNDI to obtain handles to the files, if possible.

    Related Sites

    Thank you!