Popularity wired through

Signed-off-by: AbstractionFactory <[email protected]>
opentofu · Dec 2, 2024 · 530ec95 · 530ec95
1 parent e7abbd0
commit 530ec95
Show file tree

Hide file tree

Showing 3 changed files with 37 additions and 48 deletions.
diff --git a/backend/internal/providerindex/generator.go b/backend/internal/providerindex/generator.go
@@ -350,6 +350,8 @@ func (d *documentationGenerator) extractRepoInfo(ctx context.Context, addr provi
 	providerData.Description = repoInfo.Description
 	providerData.Popularity = repoInfo.Popularity
 	providerData.ForkCount = repoInfo.ForkCount
+	providerData.UpstreamPopularity = upstreamRepoInfo.Popularity
+	providerData.UpstreamForkCount = upstreamRepoInfo.ForkCount
 
 	forkRepo := repoInfo.ForkOf
 	if forkRepo == nil {
@@ -378,8 +380,6 @@ func (d *documentationGenerator) extractRepoInfo(ctx context.Context, addr provi
 		d.log.Warn(ctx, "Cannot fetch upstream repository info for %s (%v)", forkRepo.String(), err)
 		return
 	}
-	providerData.UpstreamPopularity = upstreamRepoInfo.Popularity
-	providerData.UpstreamForkCount = upstreamRepoInfo.ForkCount
 }
 
 func (d *documentationGenerator) scrapeVersion(ctx context.Context, addr providertypes.ProviderAddr, canonicalAddr provider.Addr, providerDetails *providertypes.Provider, version provider.Version, blocked bool, blockedReason string) (providertypes.ProviderVersion, error) {

diff --git a/backend/internal/providerindex/search.go b/backend/internal/providerindex/search.go
@@ -81,7 +81,9 @@ func (p providerSearch) indexProviderVersion(ctx context.Context, providerAddr p
 					"version":   string(version),
 					"id":        string(docItem.Name),
 				},
-				ParentID: providerItem.ID,
+				ParentID:   providerItem.ID,
+				Popularity: popularity,
+				Warnings:   len(providerDetails.Warnings),
 			}); err != nil {
 				return fmt.Errorf("failed to add resource %s to search index (%w)", docItem.Name, err)
 			}

diff --git a/search/worker/src/query.ts b/search/worker/src/query.ts
@@ -7,51 +7,38 @@ export const query = async (client: Client, queryParam: string): Promise<Entity[
 };
 
 const searchQuery = `
-	WITH search_terms AS (
-		SELECT unnest(regexp_split_to_array($1, '[ /]+')) AS term
-	),
-		 term_matches AS (
-			 SELECT e.*
-			 FROM entities e
-					  INNER JOIN search_terms st
-								 ON e.addr ILIKE '%' || st.term || '%'
-		OR e.description ILIKE '%' || st.term || '%'
-	GROUP BY e.id
-		),
-		ranked_entities AS (
-	SELECT *,
-		/* TODO: remove hard-coded hashicorp/opentofu preferential treatment */
-		CASE WHEN link_variables->>'namespace' = 'hashicorp' THEN 1 WHEN link_variables->>'namespace' = 'opentofu' THEN 0 ELSE 0.5 END AS popularity_fudge,
-		CASE WHEN type = 'provider' THEN 1 ELSE 0 END AS provider_rank_fudge,
-		similarity(tm.addr, $1) AS title_sim,
-		similarity(tm.description, $1) AS description_sim,
-		similarity(link_variables->>'name', $1) AS name_sim
-	FROM term_matches tm
-		),
-		providers AS (
+	WITH search_terms AS (SELECT unnest(regexp_split_to_array($1, '[ /]+')) AS term),
+		 term_matches AS (SELECT e.*
+						  FROM entities e
+								   INNER JOIN search_terms st
+											  ON e.addr ILIKE '%' || st.term || '%'
+												  OR e.description ILIKE '%' || st.term || '%'
+						  GROUP BY id, last_updated, type, addr, version, title, description, link_variables, document, popularity, warnings),
+		 ranked_entities AS (SELECT *,
+								 /* The provider rank fudge ranks providers higher than their resources */
+									CASE WHEN type = 'provider' THEN 1 ELSE 0 END          AS provider_rank_fudge,
+								 /* When warnings are present, rank the provider lower because it's likely deprecated. */
+									CASE WHEN warnings > 0 THEN -1 ELSE 0 END              AS warnings_rank_fudge,
+								 /* Give a slight boost to providers with a higher star rating. */
+									tm.popularity / (SELECT max(popularity) FROM tm) AS popularity_rank,
+								 /* Text similarity rankings, each taking a value from 0 to 1. */
+									similarity(tm.addr, $1)                                AS title_sim,
+									similarity(tm.description, $1)                         AS description_sim,
+									similarity(link_variables ->> 'name', $1)              AS name_sim,
+							 FROM term_matches tm),
+		 providers AS (SELECT *
+					   FROM ranked_entities
+					   WHERE type LIKE 'provider%'
+					   ORDER BY (provider_rank_fudge + warnings_rank_fudge + 1) *(popularity_rank + title_sim + name_sim + description_sim/0.5) DESC
+					   LIMIT 5),
+		 modules AS (SELECT *
+					 FROM ranked_entities
+					 WHERE type LIKE 'module%'
+					 ORDER BY (warnings_rank_fudge + 1) * (popularity_rank + title_sim + name_sim + description_sim/0.5) DESC
+					 LIMIT 5)
 	SELECT *
-	FROM ranked_entities
-	WHERE type LIKE 'provider%'
-	ORDER BY
-		provider_rank_fudge DESC,
-		popularity_fudge DESC,
-		title_sim DESC,
-		name_sim DESC,
-		description_sim DESC
-		LIMIT 5
-		),
-		modules AS (
-	SELECT *
-	FROM ranked_entities
-	WHERE type LIKE 'module%'
-	ORDER BY
-		popularity_fudge DESC,
-		title_sim DESC,
-		name_sim DESC,
-		description_sim DESC
-		LIMIT 5
-		)
-	SELECT * FROM providers
+	FROM providers
 	UNION ALL
-	SELECT * FROM modules;
+	SELECT *
+	FROM modules;
 `;