Opening org.si4t.org in Eclipse

Question

To better understand, and perhaps customise SI4T we have tried opening the org.si4t.solr project in Ecplise.

I don't have a great deal of Java experience, but it looks like we might be missing a JAR file?

UPDATE

What the original question didn't make clear is that we've added all the Tridion 2013 and Solr 4.4.0 libs dependencies but that didn't work.

The Tridion 2011 and Solr < 4.4.0 libs contain the lucene-analyzers-common-4.0.0.jar which does contain the HTMLStripCharFilter class.

However the HTMLStripCharFilter read method is expecting 3 parameters:

public int read(char[] cbuf, int off, int len)

Where the SI4T SolrIndexer.java is only passing a single parameter to it.

HTMLStripCharFilter html = new HTMLStripCharFilter(strReader);
char[] cbuf = new char[1024 * 10];
while (true)
{
    int count = html.read(cbuf);
    if (count == -1)
        break; // end of stream mark is -1
    if (count > 0)
        out.append(cbuf, 0, count);
}

We've changed the code to set the offset and length required by the read method.

int count = html.read(cbuf,0,10240);

Is this the correct thing to do, or should we be using a different JAR file?

Answer 1

If you follow Dominic's documentation, you'll see that HTMLStripCharFilter has an overloaded read method, so it can take either one or three parameters :o).

Further, the latest release of SI4T does not make use anymore of the HTMLStripCharFilter, because it is wrong to strip chars in that place. It's better to let the Solr indexing mechanism on the server side handle it. Do note the @SuppressWarnings("unused") annotation on that method.

If you really want to use the method, then you have to add lucene-analyzers-common-4.4.0.jar to the class path, in case you use Solr 4.4.0.

Answer 2

org.apache.lucene.analysis.charfilter.HTMLStripCharFilter is part of lucene. The documentation can be found here

I would suggest double-checking the documentation on how to set up pre-requisites and configuration for SI4T.