简单的Django queryset __icontains搜索将首先找到大多数特定元素

Question

I'm doing a queryset filter for "foo" in a CharField using __ictonains and I want to find the most specific search matches first.

Dataset in database for a certain field named "description":

fooal;skdjfkasdgh;alskdjrf
foobar
foo-nstastical
foobariffic
foo-ntastic
foo

When I search for:

MyModel.objects.filter(description__icontains="foo")

The queryset returns the results in some order according to id number (or other factors?):

fooal;skdjfkasdgh;alskdjrf
foobar
foo-nstastical
foobariffic
foo-ntastic

If I only show the top 5 results I will leave out "foo" which is actually the best match that I want to show first. How can I give higher weight to better matches in a simple way without implementing a "real heavy duty" search engine like Lucene? I want an easier queryset filter hack.

This is a crude search engine for a text dataset, the reason it is more important to show foo first when only showing the top N results is that if the person was looking for foobariffic, they could type in more letters. But there is no way to get "foo" by typing more letters if it is displaced by longer entries.

Answer 1

If it's a small field, you could potentially sort by field length and get good results.

Given a match, the shortest result is the hardest and most significant match. Wouldn't work for full text of course...

I would be comfortable doing this for an autocomplete username or tag field.

foo
bofoo
foobar
barfood

Python sort

# python sort
x = [results]
x.sort(key=len)

Or in SQL

MyModel.objects.extra(select={'myfield_length':'Length(myfield)'}).order_by('myfield_length')