Projects that need to full text search documents often look at solutions like Solr or Elasticsearch. While these technologies provide a vast amount of features and great scalability, they also demand a large infrastructure and can be difficult to set up, even if only a small set of the available features are used.

While PostgreSQL is widely used as SQL database, its full text search capabilities are lesser known. The Django framework offers an integration that allows to implement web applications with full text search. It is comparably easier to set up and requires fewer system resources. It is particularly well suited for projects that already use an SQL database and want a full text search for about 100.000 documents or less.

In this meetup we look at the code needed to do just that. The format is a mix of presentation and coding dojo. For each topic, there is a short presentation explain the mechanics. After that we apply the knowledge on a Django application to search e-books provided by Project Gutenberg.

Topics covered are:

1. Quick overview on downloading and importing e-books from Project Gutenberg
2. Basics of full text search: token, stop words, stemming
3. PostgreSQL search queries: words, phrases, operators, prefixes
4. Ranking search results
5. Indexing search for improved performance
6. Supporting multiple languages

Interested audience should have a basic understanding of Django models and queries because this is where we will spend most of the time. While we also use a form and some views, they are already provided by the skeleton application and modified only a little.

Sessions

Keine Sessions für dieses Meetup eingetragen.

Notizen, Links, etc.

Github Repo : https://github.com/roskakori/django-search-example/

Siehe Foliensatz und Issues im Repo!