Implement and launch the string search context-enhancement solution as Pages website

Following the demo provided by @mchrastek-ext, the next steps towards implementation and launch are:

• Complete support for Ruby HAML files
• Fix highlighting issues for multiline strings
• Combine two separate projects (astro frontend and Code occurrences extraction) into one (combined into the astro)
• Create a new project under the localization group to host the newly-combined project (code moved in gitlab-string-search!1 (merged))
• Set up testing environment and test CI pipeline after combining projects
• Deploy to production 🚀
• Set up Crowdin webhooks so that the JSON files are rebuilt (and the changes are deploy to GitLab Pages) each time there are any updates to the strings in Crowdin
• Check if we need to set the site and base config values in Astro after the MR gitlab-string-search!1 (merged) is merged and the website has been published to GitLab Pages
• Enable the new Crowdin webhook after the MR gitlab-string-search!1 (merged) is merged and the website has been published to GitLab Pages
• Update automation that injects context links for new strings (done in crowdin-automation!5 (merged))
• Replace existing context links in Crowdin for existing strings (updated mostly locally, then in https://gitlab.com/gitlab-com/localization/crowdin-automation/-/jobs/8874076460 and https://gitlab.com/gitlab-com/localization/crowdin-automation/-/jobs/8874161251)
• Go through translator comments mentioning issues with the code search being broken and reply that the issue has been fixed