Ignoring Pages in an Algolia Index

Posted: Thursday, September 17, 2026 at 07:39 PM | Categories: Eleventy

When I built this site, I followed the instructions in Ray Camden's Adding Algolia Search to Eleventy and Netlify post to add an Algolia index to my site and allow visitors to use search to locate specific articles. This worked well for this site for many years although, like with Ray's site, not many people use search. I assume this is because its just too easy to let Google point people to posts anywhere.

As I built out my new Breakfast Quest site, I realized that I wanted to include more than just the site's posts in the search index, I wanted to include some additional pages as well, but not all of the pages.

Ray's post recommended using an algolia.liquid file to generate a permalinked .json file with the index for the site. The file is called algolia.json and 11ty stores the generated file in the site's root folder. Here's the code for the file.

---
permalink: /algolia.json
---

{% assign posts = collections.post | reverse
 %}
[
{% for post in posts %}
  {
  "title": {{ post.data.title |
 json }},
  "date":"{{ post.date }}
",
  "date_timestamp": {{ post
.date | 
date: '%s' }},
  "url":"{{ post.url | prepend
: site.url }}",
  "content":{{ post.templateContent | algExcerpt | json }},
  "tags":[
  {% for tag in post.
data.tags %}
    {{ tag | json }}{% unless 
forloop.last %},{% endunless %}
  {% endfor %}
  ],
  "categories":[
  {% for cat in post.
data.categories %}
    {{ cat | json }}{% unless 
forloop.last %},{% endunless %}
  {% endfor %}
  ]
  }{% unless forloop.
last %},{% 
endunless %}
{% endfor %}
]

As you can see from the first line of liquid code in the file, it generates the Algolia index using just the site's posts:

{% assign posts = collections.post |
 reverse %}

For my site, I wanted more than just those pages. One of the things I could do is use the 11ty collections.all to get all pages for the index:

{%- for item in collections.
all %}

But then that would get all generated files/pages (like the paginated lists of posts which didn't need to be in the index), so I needed to do something else.

The solution I settled on was to set a idxIgnore property in the frontmatter on pages I didn't want included in the index. For example, here's the categories page:

---
layout: generic
title: Article Categories
eleventyNavigation:
  key: Categories
  parent: Articles
  order: 2

idxIgnore: true
---

With that in place and all the skipped pages marked with the property, all I had to do was update indexing algorithm to ignore those pages.

Here's the updated algolia.liquid file:

---
permalink: /algolia.json
idxIgnore: true
---

[
  {%- assign comma = "" 
%}
  {%- for item in collections.
all %}
  {%- unless item.url ==
 page.url 
or item.data.idxIgnore or item.url == 
false %}
    {{ comma }}{
      "objectID": {{ item
.url | 
json }},
      "title": {{ item.data.title |
 default: item.fileSlug | 
json }},
      "date": "{{ item.date }}
",
      "date_timestamp": {{
 item.date |
 date: "%s"
 }},
      "url": {{ item.url | prepend
: site.url | json }},
      "content": {{ item
.templateContent | algExcerpt | json }},
      "tags": {{ item.data.tags |
 default: empty | json
 }},
      "categories": {{ item
.data.categories | default: 
empty | json }}
    }
    {%- assign comma = "," 
%}
  {%- endunless %}
{%- endfor %}]

I didn't write that file, AI gave it to me and I had to study it to understand exactly what it was doing. The comma thing threw me the most, but I finally figured it out.

The use of unless flips the code on its head, I don't know why it was written that way, I assume you could right it the opposite way, but I didn't try since this worked.

The code starts with comma blank/empty, so the index doesn't start with, you know, a , . After that though, and it has to reset the variable in every loop, comma contains a comma, so that separates every index entry.

Header image: Eleventy Project Home Page.