Fix missing pages in search indexes (#212)

* Search indexes: Fix missing search entries

This handles duplicated ids - but requires changes on web-php to use new indexes

Author: AllenJB

phpdotnet/phd/IndexRepository.php

@@ -162,6 +162,16 @@ protected function SQLiteIndex($context, $index, $id, $filename, $parent, $sdesc
         ];
     }
 
+    public function getIndexesWithDuplicates(): array
+    {
+        $results =  $this->db->query('SELECT docbook_id, filename, parent_id, sdesc, ldesc, element, previous, next, chunk FROM ids');
+        $indexes = [];
+        while ($row = $results->fetchArray(SQLITE3_ASSOC)) {
+            $indexes[] = $row;
+        }
+        return $indexes;
+    }
+
     private static function SQLiteFinal($context): mixed {
         return $context;
     }

phpdotnet/phd/Package/PHP/Web.php

@@ -249,17 +249,40 @@ protected function writeJsonIndex() {
             json_encode($descriptions)
         );
         $this->outputHandler->v("Index written", VERBOSE_FORMAT_RENDERING);
+
+        $entries = $this->processCombinedJsonIndex();
+        file_put_contents(
+            $this->getOutputDir() . "search-combined.json",
+            json_encode($entries)
+        );
+        $entries = 'var localSearchIndexes = '. json_encode($entries) .';';
+        file_put_contents(
+            $this->getOutputDir() . "search-combined.js",
+            $entries
+        );
+        $this->outputHandler->v("Combined Index written", VERBOSE_FORMAT_RENDERING);
     }
 
     /**
      * Processes the index to extract entries and descriptions. These are
      * used to generate the search index and the descriptions JSON files.
      */
     private function processJsonIndex(): array {
+        $alwaysIncludeElements = [
+            'refentry',
+            'stream_wrapper',
+            'phpdoc:classref',
+            'phpdoc:exceptionref',
+            'phpdoc:varentry',
+        ];
+
         $entries = [];
         $descriptions = [];
         foreach($this->indexes as $id => $index) {
-            if (!$index["chunk"]) {
+            if (
+                (! $index['chunk'])
+                && (! in_array($index['element'], $alwaysIncludeElements, true))
+            ) {
                 continue;
             }
 
@@ -281,6 +304,80 @@ private function processJsonIndex(): array {
         return [$entries, $descriptions];
     }
 
+    private function processCombinedJsonIndex(): array
+    {
+        $alwaysIncludeElements = [
+            'refentry',
+            'stream_wrapper',
+            'phpdoc:classref',
+            'phpdoc:exceptionref',
+            'phpdoc:varentry',
+        ];
+
+        $entries = [];
+        $indexes = $this->indexRepository->getIndexesWithDuplicates();
+        foreach ($indexes as $index) {
+            if (
+                (! $index['chunk'])
+                && (! in_array($index['element'], $alwaysIncludeElements, true))
+            ) {
+                continue;
+            }
+
+            if ($index["sdesc"] === "" && $index["ldesc"] !== "") {
+                $index["sdesc"] = $index["ldesc"];
+                $bookOrSet = $this->findParentBookOrSet($index['parent_id']);
+                if ($bookOrSet) {
+                    $index["ldesc"] = Format::getLongDescription(
+                        $bookOrSet['docbook_id']
+                    );
+                }
+            }
+
+            $nameParts = explode('::', $index['sdesc']);
+            $methodName = array_pop($nameParts);
+
+            if (str_contains('wrapper', $index['filename'])) {
+                print "Combined index: adding " . $index['filename'] . " :: " . $index['sdesc'] . "\n";
+            }
+
+            $type = 'General';
+            switch ($index['element']) {
+                case "phpdoc:varentry":
+                    $type = "Variable";
+                    break;
+
+                case "refentry":
+                    $type = "Function";
+                    break;
+
+                case "phpdoc:exceptionref":
+                    $type = "Exception";
+                    break;
+
+                case "phpdoc:classref":
+                    $type = "Class";
+                    break;
+
+                case "set":
+                case "book":
+                case "reference":
+                    $type = "Extension";
+                    break;
+            }
+
+            $entries[] = [
+                'id' => $index['filename'],
+                'name' => $index['sdesc'],
+                'description' => html_entity_decode($index['ldesc']),
+                'tag' => $index['element'],
+                'type' => $type,
+                'methodName' => $methodName,
+            ];
+        }
+        return $entries;
+    }
+
     /**
      * Finds the closest parent book or set in the index hierarchy.
      */

tests/render/render_001.phpt

@@ -16,4 +16,5 @@ require_once __DIR__ . "/../../render.php";
 %s[%d:%d:%d - Rendering Format      ]%s Starting PHP-Web rendering
 %s[%d:%d:%d - Rendering Format      ]%s Writing search indexes..
 %s[%d:%d:%d - Rendering Format      ]%s Index written
+%s[%d:%d:%d - Rendering Format      ]%s Combined Index written
 %s[%d:%d:%d - Rendering Format      ]%s Finished rendering