From 7897585d9887a6f7e58d1630d421b01fd81de4a5 Mon Sep 17 00:00:00 2001
From: "J. King" <jking@jkingweb.ca>
Date: Sat, 16 Jan 2021 17:58:31 -0500
Subject: [PATCH] Test scraping

Text search should also match scraped content when appropriate
---
 lib/Database.php                       | 16 +++++++++---
 tests/cases/Database/SeriesArticle.php | 35 +++++++++++++++++++++++---
 2 files changed, 43 insertions(+), 8 deletions(-)

diff --git a/lib/Database.php b/lib/Database.php
index a78ba370..ea70d953 100644
--- a/lib/Database.php
+++ b/lib/Database.php
@@ -1711,10 +1711,10 @@ class Database {
         }
         // handle text-matching context options
         $options = [
-            "titleTerms"      => ["arsse_articles.title"],
-            "searchTerms"     => ["arsse_articles.title", "arsse_articles.content"],
-            "authorTerms"     => ["arsse_articles.author"],
-            "annotationTerms" => ["arsse_marks.note"],
+            "titleTerms"      => ["title"],
+            "searchTerms"     => ["title", "content"],
+            "authorTerms"     => ["author"],
+            "annotationTerms" => ["note"],
         ];
         foreach ($options as $m => $columns) {
             if (!$context->$m()) {
@@ -1722,6 +1722,10 @@ class Database {
             } elseif (!$context->$m) {
                 throw new Db\ExceptionInput("tooShort", ['field' => $m, 'action' => $this->caller(), 'min' => 1]); // must have at least one array element
             }
+            $columns = array_map(function ($c) use ($colDefs) {
+                assert(isset($colDefs[$c]), new Exception("constantUnknown", $c));
+                return $colDefs[$c];
+            }, $columns);
             $q->setWhere(...$this->generateSearch($context->$m, $columns));
         }
         // further handle exclusionary text-matching context options
@@ -1729,6 +1733,10 @@ class Database {
             if (!$context->not->$m() || !$context->not->$m) {
                 continue;
             }
+            $columns = array_map(function ($c) use ($colDefs) {
+                assert(isset($colDefs[$c]), new Exception("constantUnknown", $c));
+                return $colDefs[$c];
+            }, $columns);
             $q->setWhereNot(...$this->generateSearch($context->not->$m, $columns, true));
         }
         // return the query
diff --git a/tests/cases/Database/SeriesArticle.php b/tests/cases/Database/SeriesArticle.php
index 09342c9a..6302f5df 100644
--- a/tests/cases/Database/SeriesArticle.php
+++ b/tests/cases/Database/SeriesArticle.php
@@ -22,10 +22,11 @@ trait SeriesArticle {
                     'num'      => 'int',
                 ],
                 'rows' => [
-                    ["jane.doe@example.com", "",1],
-                    ["john.doe@example.com", "",2],
-                    ["john.doe@example.org", "",3],
-                    ["john.doe@example.net", "",4],
+                    ["jane.doe@example.com", "", 1],
+                    ["john.doe@example.com", "", 2],
+                    ["john.doe@example.org", "", 3],
+                    ["john.doe@example.net", "", 4],
+                    ["jill.doe@example.com", "", 5],
                 ],
             ],
             'arsse_feeds' => [
@@ -110,6 +111,7 @@ trait SeriesArticle {
                     [12,"john.doe@example.net",2,    9,null,0],
                     [13,"john.doe@example.net",3,    8,"Subscription 13",0],
                     [14,"john.doe@example.net",4,    7,null,0],
+                    [15,"jill.doe@example.com",11,null,null,1],
                 ],
             ],
             'arsse_tag_members' => [
@@ -1149,4 +1151,29 @@ trait SeriesArticle {
         $state = $this->primeExpectations($this->data, $this->checkTables);
         $this->compareExpectations(static::$drv, $state);
     }
+
+    public function testSelectScrapedContent(): void {
+        $exp = [
+            ['id' => 101, 'content' => "<p>Article content 1</p>"],
+            ['id' => 102, 'content' => "<p>Article content 2</p>"],
+        ];
+        $this->assertResult($exp, Arsse::$db->articleList("john.doe@example.org", (new Context)->subscription(8), ["id", "content"]));
+        $exp = [
+            ['id' => 101, 'content' => "<p>Scraped content 1</p>"],
+            ['id' => 102, 'content' => "<p>Article content 2</p>"],
+        ];
+        $this->assertResult($exp, Arsse::$db->articleList("jill.doe@example.com", (new Context)->subscription(15), ["id", "content"]));
+    }
+
+    public function testSearchScrapedContent(): void {
+        $exp = [
+            ['id' => 101, 'content' => "<p>Scraped content 1</p>"],
+            ['id' => 102, 'content' => "<p>Article content 2</p>"],
+        ];
+        $this->assertResult($exp, Arsse::$db->articleList("jill.doe@example.com", (new Context)->subscription(15)->searchTerms(["article"]), ["id", "content"]));
+        $exp = [
+            ['id' => 101, 'content' => "<p>Scraped content 1</p>"],
+        ];
+        $this->assertResult($exp, Arsse::$db->articleList("jill.doe@example.com", (new Context)->subscription(15)->searchTerms(["scraped"]), ["id", "content"]));
+    }
 }