meta-search-engine

This is a meta search engine to join the forces of several Schul-Cloud search engines. The meta-search-engine is part of a greater project to scale up the content search and this issue is the reason why it was created.

This is part of the minimal working prototype. This README file must explain,

• how to use the search engine and how to configure it, in the How To Section,
• how to install the search engine, in the Installation Section
• how to work on it, in the project guidelines
• what else it is related to, in the Related Work Section

Related Work

You can view the issue for the creation of this search engine.

How To Use

You can use this search engine using different means:

You can use Docker to run the search engine. Therefore, you can pass the arguments of which searches to join to the docker command.

As an example, you want to see how your search engine performs when it would be used inside the official Schul-Cloud search. Your search engine runs on http://localhost:8080 and the Schul-Cloud search runs at http://search.schul-cloud.org. You can run the docker-container at port 9999.

docker run -p 9999:80 schulcloud/meta-search-engine http://localhost:8080 http://search.schul-cloud.org

Now, if you request a search for Einstein at http://localhost:9999/v1/search?Q=Einstein, you should see your results along the other results.

Likewise, if you install the seach engine locally, you can edit the search-engines.txt file. Each line contains a search engine.

Query

If you run the search engine, you can also specify which search engines should be requested. If a usual query looks like Q=einstein, a query requesting an additional search engine running under http://localhost:8080/v1/search looks like Q=einstein&Search=http://localhost:8080/v1/search.

You can request several additional search engines by adding a number behind the Search parameter:

Q=einstein
Q=einstein&Search1=http://localhost:8080/v1/search
Q=einstein&Search1=http://localhost:8080/v1/search&Search2=http://localhost:8080/v1/search
...

If you do not like to request the default search engines, you can use the Default=false parameter. Then, only the search engines passed with Search query parameters are requested.

• Q=einstein will request all search engines in the search engine listing.
• Q=einstein&Default=false will request no search engine.

Installation

• Install apache and PHP as mentioned in this guide.
  Under Ubuntu, do this:

  sudo apt-get update
  sudo apt-get -y install apache2

  Check that http://localhost returns a webpage.

  sudo apt-get -y install libapache2-mod-php curl \
                         libcurl3 libcurl3-dev php-curl

• Get this repository.
  You can download the sources or clone it usig git.

  git clone https://github.com/schul-cloud/meta-search-engine.git
  cd meta-search-engine
  pwd # /path/to/meta-search-engine

• Create a symlink to your location for http://localhost/v1/search
  Therefore you need the path from pwd.

  sudo mkdir -p /var/www/html/v1
  sudo ln -sT /path/to/meta-search-engine/src /var/www/html/v1/search

  Now, you should be able to access http://localhost/v1/search

Docker

You can install docker.

wget -O- https://get.docker.com | sh
sudo usermod -aG docker $USER

To enable that you do not need sudo to run the docker containers, log in and out.

Once you installed docker, you can create a new schulcloud/meta-search-engine container like this:

docker build -t schulcloud/meta-search-engine .

Then, you can run the container and map the http port to port 8000.

docker run --rm -p 8000:80 schulcloud/meta-search-engine

Now, you should be able to request a search:

curl -i 'http://localhost:8000/?Q=einstein'