Antelope Search

"antelope" is a super fast search engine written in C/C++. Initially it was developed by Andrew Trotman who teaches in University of Otago.

File Structure

The files that are different with the upstream

• antelope
• version.h

Build Antelope

cd [path/to/antelope]
./autogen.sh
mkdir -p build/{release,debug}

## DEBUG BUILD WITH g++
## In a new terminal
cd build/debug
source ../../scripts/setdebugenv.sh
../../configure
make
sudo make install

## RELEASE BUILD
## In a new terminal
cd build/release
../../configure
make
sudo make install

## DEBUG BUILD WITH clang++
## In a new terminal
cd build/debug
source ../../scripts/setdebugenv.clang
../../configure CC=clang++ CXX=clang++
make
sudo make install

Usage Examples

At the moment, antelope supports indexing documents in structural formats such as csv, xml, wiki, etc.

Indexing / Search with TREC data

Zettair search engine provides a pre-compiled text of novel - "Moby Dick" in TREC format.

To index documents in TREC format:

cd [path/to/zettair/txt]
index -rtrec moby.txt

To index Wiki style documents as as Wikipedia articles:

For the xml documents, we need to specify the root tag of a document, then the title tag.

cd [path/to/wikipedia/xml/files]
index -rtrec:tag:page:title file1 file2 file3 ...

Or, index all documents under a folder:

cd [path/to/wikipedia/xml/files]
index -rtrec:tag:page:title folder1 ...

To search:

cd [path/to/zettair/txt]
antelope