Full text search on html or pdf files

hibernator_11 · **Joined:** Thu Jun 16, 2011 12:03 pm **Posts:** 94

Hi all,

I would like to know if it possible to create several Lucene documents from one record in the database. I have the url file in the database and I want to index every paragraph associated with its anchor (<a name="example"></a>) in order to perfom the search and paragraph position.

So for example if I have the record in my database with the :

File database table
FileName
example1.html

with the following text:

Code:

....
the book is fine <a name="ex1"></a>
i love trees <a name="ex2"></a>
play with me <a name="ex3"></a>
.....

I would like index the following:
url ----------------------text
example1.html#ex1------------the book is fine
example1.html#ex2------------i love trees
example1.html#ex3------------play with me

I have all the process implemented in order to create the paragraphs but I have to figure out how to index the three documents from the one database record.

I am thinking about using natively Lucene:

Code:

org.apache.lucene.store.Directory directory = searchFactory.getDirectoryProviders(FileEntity.class)[0].getDirectory();

Any ideas?

Thanks!

sanne.grinovero · **Posted:** Thu Oct 10, 2013 2:21 pm

Hi,
did you consider implementing a FieldBridge?

A FieldBridge is normally a small function, but technically there is no limit to what you do in this function: you're free to open the external files, load them, extract the text and write it to the Lucene Document.