From 55f7f1d446e6280541ea525490ba36fe1db2d7ab Mon Sep 17 00:00:00 2001
From: Pete Bacon Darwin <pete@bacondarwin.com>
Date: Mon, 29 Mar 2021 22:02:54 +0100
Subject: [PATCH] feat(docs-infra): improve search query processing (#41368)

This commit tries to improve the search results by processing
the query and attempting progressively less restrictive searches
until a non-zero set of pages is matched.

The new procesing includes:

* stripping off quote marks, which were causing searches to fail
* first attempting to match pages where ALL the query terms exist
* second attempting to match pages where ANY of the query terms exist
* third attempting to match pages where the title contains partial word matches

The first query attempt approximates, quite well, the idea of searching
for multi-word phrases. This is given the technical nature of the terms
and the fairly small size of the corpus.

PR Close #41368
---
 aio/src/app/search/search.worker.ts           | 21 ++++++++++++++-----
 .../deployment/e2e/smoke-tests.e2e-spec.ts    |  5 ++---
 aio/tests/e2e/src/app.e2e-spec.ts             |  5 ++---
 3 files changed, 20 insertions(+), 11 deletions(-)

diff --git a/aio/src/app/search/search.worker.ts b/aio/src/app/search/search.worker.ts
index 153416a680..5c15cd1c6b 100644
--- a/aio/src/app/search/search.worker.ts
+++ b/aio/src/app/search/search.worker.ts
@@ -84,22 +84,33 @@ function loadIndex(pagesData: PageInfo[]): IndexLoader {
 
 // Query the index and return the processed results
 function queryIndex(query: string): PageInfo[] {
+  // Strip off quotes
+  query = query.replace(/^["']|['"]$/g, '');
   try {
     if (query.length) {
-      let results = index.search(query);
+      // First try a query where every term must be present
+      const queryAll = query.replace(/(^|\s)([^\s]+)/g, '$1+$2');
+      let results = index.search(queryAll);
+
+      // If that was too restrictive just query for any term to be present
       if (results.length === 0) {
-        // Add a relaxed search in the title for the first word in the query
-        // E.g. if the search is "ngCont guide" then we search for "ngCont guide titleWords:ngCont*"
-        const titleQuery = 'titleWords:*' + query.split(' ', 1)[0] + '*';
+        results = index.search(query);
+      }
+
+      // If that is still too restrictive then search in the title for the first word in the query
+      if (results.length === 0) {
+        // E.g. if the search is "ngCont guide" then we search for "ngCont guide title:ngCont*"
+        const titleQuery = 'title:*' + query.split(' ', 1)[0] + '*';
         results = index.search(query + ' ' + titleQuery);
       }
+
       // Map the hits into info about each page to be returned as results
       return results.map(hit => pages[hit.ref]);
     }
   } catch (e) {
     // If the search query cannot be parsed the index throws an error
     // Log it and recover
-    console.log(e);
+    console.error(e);
   }
   return [];
 }
diff --git a/aio/tests/deployment/e2e/smoke-tests.e2e-spec.ts b/aio/tests/deployment/e2e/smoke-tests.e2e-spec.ts
index e724deea86..2d58484fc6 100644
--- a/aio/tests/deployment/e2e/smoke-tests.e2e-spec.ts
+++ b/aio/tests/deployment/e2e/smoke-tests.e2e-spec.ts
@@ -103,11 +103,10 @@ describe(browser.baseUrl, () => {
     });
 
     it('should show relevant results on 404', async () => {
-      await page.goTo('http/router');
+      await page.goTo('common/http');
       const results = await page.getSearchResults();
 
-      expect(results).toContain('HttpClient');
-      expect(results).toContain('Router');
+      expect(results).toContain('common/http package');
     });
   });
 });
diff --git a/aio/tests/e2e/src/app.e2e-spec.ts b/aio/tests/e2e/src/app.e2e-spec.ts
index 442c840ff0..e341324858 100644
--- a/aio/tests/e2e/src/app.e2e-spec.ts
+++ b/aio/tests/e2e/src/app.e2e-spec.ts
@@ -217,11 +217,10 @@ describe('site App', () => {
     });
 
     it('should search the index for words found in the url', async () => {
-      await page.navigateTo('http/router');
+      await page.navigateTo('common/http');
       const results = await page.getSearchResults();
 
-      expect(results).toContain('HttpRequest');
-      expect(results).toContain('Router');
+      expect(results).toContain('common/http package');
     });
   });