Multiple Algolia search index? #8467
Replies: 1 comment 5 replies
-
|
Hi We assign metadata tags to each page for the search crawler. For example your regular docs: https://docs.typebot.io/ Your API ref docs: https://docs.typebot.io/api <meta name="docusaurus_tag" content="default" data-rh="true">The plugin you use for the API docs does not populate Now we have the concept of "contextual facet filters" which permit to craft an Algolia query based on the current path. // Translate search-engine agnostic search filters to Algolia search filters
export function useAlgoliaContextualFacetFilters(): [string, string[]] {
const {locale, tags} = useContextualSearchFilters();
// Seems safe to convert locale->language, see AlgoliaSearchMetadata comment
const languageFilter = `language:${locale}`;
const tagsFilter = tags.map((tag) => `docusaurus_tag:${tag}`);
// [] = AND query
return [languageFilter, tagsFilter];
}Unfortunately I don't think you can swizzle that file anymore but you can probably create a Webpack alias to override it with your own implementation, and use the // Translate search-engine agnostic search filters to Algolia search filters
export function useAlgoliaContextualFacetFilters(): [string, string[]] {
const {pathname} = useLocation()
const {locale, tags} = useContextualSearchFilters();
if (pathname.startsWith("/api/") {
return ["docusaurus_tag:my-api-ref-tag"]
}
// Seems safe to convert locale->language, see AlgoliaSearchMetadata comment
const languageFilter = `language:${locale}`;
const tagsFilter = tags.map((tag) => `docusaurus_tag:${tag}`);
// [] = AND query
return [languageFilter, tagsFilter];
}Not sure we provide any better way to do that currently. Maybe @shortcuts has another idea? |
Beta Was this translation helpful? Give feedback.
-
|
Hey! For simplicity I'd avise you to do a single index with filters, since it works out of the box for both DocSearch and Docusaurus. My idea is similar, but it's also possible to do the filtering part on the Algolia side. Inside your Crawler configuration, you can leverage the recordExtractor: ({ $, helpers, url }) => {
const filter = url.pathname.includes("...") ? "a" : "b"
// ... rest of the logic
}and then follow https://docsearch.algolia.com/docs/record-extractor/#indexing-content-for-faceting You'd still however to make it dynamic on the contextual search for Docusaurus |
Beta Was this translation helpful? Give feedback.
-
|
Thank you so much for the info! 🙏 https://docs.typebot.io/ has Contextual search should work properly since What am I missing? |
Beta Was this translation helpful? Give feedback.
-
|
Here is my crawler config: new Crawler({
rateLimit: 8,
startUrls: ["https://docs.typebot.io/"],
renderJavaScript: false,
sitemaps: ["https://docs.typebot.io//sitemap.xml"],
ignoreCanonicalTo: true,
discoveryPatterns: ["https://docs.typebot.io/**"],
exclusionPatterns: ["https://docs.typebot.io/self-hosting/configuration/**"],
schedule: "at 3:19 PM on Monday",
actions: [
{
indexName: "typebot",
pathsToMatch: ["https://docs.typebot.io/**"],
recordExtractor: ({ $, helpers }) => {
// priority order: deepest active sub list header -> navbar active item -> 'Documentation'
const lvl0 =
$(
".menu__link.menu__link--sublist.menu__link--active, .navbar__item.navbar__link--active"
)
.last()
.text() || "Documentation";
return helpers.docsearch({
recordProps: {
lvl0: {
selectors: "",
defaultValue: lvl0,
},
lvl1: ["header h1", "article h1"],
lvl2: "article h2",
lvl3: "article h3",
lvl4: "article h4",
lvl5: "article h5, article td:first-child",
lvl6: "article h6",
content: "article p, article li, article td:last-child",
},
aggregateContent: true,
recordVersion: "v3",
});
},
},
],
initialIndexSettings: {
typebot: {
attributesForFaceting: [
"type",
"lang",
"language",
"version",
"docusaurus_tag",
],
attributesToRetrieve: [
"hierarchy",
"content",
"anchor",
"url",
"url_without_anchor",
"type",
],
attributesToHighlight: ["hierarchy", "hierarchy_camel", "content"],
attributesToSnippet: ["content:10"],
camelCaseAttributes: ["hierarchy", "hierarchy_radio", "content"],
searchableAttributes: [
"unordered(hierarchy_radio_camel.lvl0)",
"unordered(hierarchy_radio.lvl0)",
"unordered(hierarchy_radio_camel.lvl1)",
"unordered(hierarchy_radio.lvl1)",
"unordered(hierarchy_radio_camel.lvl2)",
"unordered(hierarchy_radio.lvl2)",
"unordered(hierarchy_radio_camel.lvl3)",
"unordered(hierarchy_radio.lvl3)",
"unordered(hierarchy_radio_camel.lvl4)",
"unordered(hierarchy_radio.lvl4)",
"unordered(hierarchy_radio_camel.lvl5)",
"unordered(hierarchy_radio.lvl5)",
"unordered(hierarchy_radio_camel.lvl6)",
"unordered(hierarchy_radio.lvl6)",
"unordered(hierarchy_camel.lvl0)",
"unordered(hierarchy.lvl0)",
"unordered(hierarchy_camel.lvl1)",
"unordered(hierarchy.lvl1)",
"unordered(hierarchy_camel.lvl2)",
"unordered(hierarchy.lvl2)",
"unordered(hierarchy_camel.lvl3)",
"unordered(hierarchy.lvl3)",
"unordered(hierarchy_camel.lvl4)",
"unordered(hierarchy.lvl4)",
"unordered(hierarchy_camel.lvl5)",
"unordered(hierarchy.lvl5)",
"unordered(hierarchy_camel.lvl6)",
"unordered(hierarchy.lvl6)",
"content",
],
distinct: true,
attributeForDistinct: "url",
customRanking: [
"desc(weight.pageRank)",
"desc(weight.level)",
"asc(weight.position)",
],
ranking: [
"words",
"filters",
"typo",
"attribute",
"proximity",
"exact",
"custom",
],
highlightPreTag: '<span class="algolia-docsearch-suggestion--highlight">',
highlightPostTag: "</span>",
minWordSizefor1Typo: 3,
minWordSizefor2Typos: 7,
allowTyposOnNumericTokens: false,
minProximity: 1,
ignorePlurals: true,
advancedSyntax: true,
attributeCriteriaComputedByMinProximity: true,
removeWordsIfNoResults: "allOptional",
},
},
appId: "...",
apiKey: "...",
indexPrefix: "",
}); |
Beta Was this translation helpful? Give feedback.
-
|
Is this a bug? Should I fill an issue? @slorber |
Beta Was this translation helpful? Give feedback.
-
|
First, I'm surprised that your API section has It is likely that you changed something that lead to this default value, or do not use the latest Docusaurus version or whatever but it's not normal. If I can't see your code/config, I can't help. Edit: oh this is because you use a third-party OpenAPI plugin. Each plugin is responsible to declare the docusaurus search tag it wants for each individual page. Second, Contextual search is meant to apply some dynamic filters for you automatically. The main goals are:
Imagine a site where you have iOS v1 + v2 (current), and Android v1 + v2 (current). It means you will search in any of:
Contextual search is not a way to "sandbox" the search to a single part of your site but rather a way to avoid displaying duplicate search results due to localization or versioning. Contextual search always search inside all pages that have the "default" docusaurus_tag, that is set by your OpenAPI plugin. |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
I have a doc with a sub path
/apithat points to the API doc.I'd like to change the Algolia search index based on this:
If user in documentation (every page other than
/api/**) then don't show search results from Typebot API indexOtherwise, use the API index
How can I achieve this?
Beta Was this translation helpful? Give feedback.
All reactions