Page MenuHomePhabricator

Page titles missing from search
Closed, ResolvedPublic

Description

Expected behavior

Dev Portal search searches page titles in addition to page content.

Observed behavior

Page titles appear to not be included in the search index, although I wasn't able to find the search index to verify this.

Steps to reproduce

  1. From the Dev Portal, search for get started
  2. See that the page titled Get started /get-started does not appear in the results

Event Timeline

I wonder if this is in any way affected by the fact that there is more than one page with "Get started" as it's title? The first hit for me is https://developer-portal.wmcloud.org/build-tools/get-started/

Screen Shot 2022-05-31 at 1.35.58 PM.png (608×507 px, 75 KB)

I wasn't able to find the search index to verify this.

https://developer-portal.wmcloud.org/search/search_index.json

In the version of this file I just fetched, there are multiple entries with "title": "Get started":

$ curl -s https://developer-portal.wmcloud.org/search/search_index.json | jq '.docs[] | select(.title=="Get started")'
{
  "location": "build-tools/#get-started",
  "text": "Create your first tool, learn about running a bot, and get best practices.",
  "title": "Get started"
}
{
  "location": "build-tools/get-started/",
  "text": "Get started ¶ Create your first tool, learn about running a bot, and get best practices. Build your first tool ¶ Learn how to edit wiki pages using Python and the Pywikibot framework. This tutorial helps you get started quickly using only your browser. Read more on mediawiki.org Learn about creating and running bots ¶ Bots help maintain wikis by performing repetitive tasks. Get started creating and running bots by learning about best practices. Read more on mediawiki.org Best practices for open source tools ¶ New to open source? Get tips to help you develop successful tools and bots, including choosing a license, publishing the code, and writing docs. Read more on Wikitech",
  "title": "Get started"
}
{
  "location": "build-tools/get-started/#get-started",
  "text": "Create your first tool, learn about running a bot, and get best practices.",
  "title": "Get started"
}
{
  "location": "get-started/",
  "text": "Learn about Wikimedia technology ¶ Explore the open source software and technical components behind Wikimedia projects. Understand the development process ¶ Learn how Wikimedia software projects work and how to make technical contributions. Learn with tutorials ¶ Follow coding exercises with step-by-step guidance. Browse by programming language ¶ Find client libraries, example apps, and open source projects in your preferred coding languages.",
  "title": "Get started"
}

Thanks! This is helpful. I think seeing that titles are included in the search index is enough to call this resolved