feat: add opt-in strict chunk_index validation

includes/class-wpvdb-rest.php

@@ -912,6 +912,43 @@ public static function handle_metadata( \WP_REST_Request $request ) { // phpcs:i
 	 * @return int|\WP_Error        Row ID, or WP_Error on validation failure or DB insert failure.
 	 */
 	public static function insert_embedding_row( $doc_id, $chunk_id, $chunk_content, $summary, $embedding, $model = '', $doc_type = 'post', $chunk_index = null ) {
+		/**
+		 * Filters whether a missing or invalid chunk_index is a hard error.
+		 *
+		 * Default false preserves the legacy behavior (warn, then default to 0).
+		 * Return true to reject a null, non-integer, negative, or out-of-range
+		 * chunk_index with a WP_Error instead of silently storing 0, so caller
+		 * regressions fail loudly.
+		 *
+		 * @param bool $strict Whether to reject an invalid chunk_index. Default false.
+		 */
+		$strict_chunk_index = (bool) apply_filters( 'wpvdb_strict_chunk_index', false );
+		$valid_chunk_index  = false;
+		if ( is_int( $chunk_index ) ) {
+			$valid_chunk_index = $chunk_index >= 0;
+		} elseif ( is_string( $chunk_index ) && preg_match( '/^\d+$/', $chunk_index ) ) {
+			$normalized_chunk_index = ltrim( $chunk_index, '0' );
+			$max_chunk_index        = (string) PHP_INT_MAX;
+			$valid_chunk_index      = '' === $normalized_chunk_index
+				|| strlen( $normalized_chunk_index ) < strlen( $max_chunk_index )
+				|| (
+					strlen( $normalized_chunk_index ) === strlen( $max_chunk_index )
+					&& strcmp( $normalized_chunk_index, $max_chunk_index ) <= 0
+				);
+		}
+		if ( $strict_chunk_index && ! $valid_chunk_index ) {
+			Logger::error( 'insert_embedding_row rejected invalid chunk_index for doc_id=' . $doc_id );
+			return new \WP_Error(
+				'chunk_index_invalid',
+				'Refused to store an embedding with a missing, non-integer, negative, or out-of-range chunk_index while strict mode is enabled.',
+				array(
+					'doc_id'      => $doc_id,
+					'chunk_index' => $chunk_index,
+					'status'      => 400,
+				)
+			);
+		}
+
 		// Detect callers using the legacy 5/6/7-arg signature; fall back to 0 for backward
 		// compatibility but emit a warning so the regression is visible.
 		if ( null === $chunk_index ) {

phpunit.xml

@@ -17,10 +17,10 @@
         <testsuite name="integration">
             <directory>tests/integration</directory>
         </testsuite>
-        <testsuite name="api">
-            <directory>tests/api</directory>
-        </testsuite>
     </testsuites>
+    <extensions>
+        <extension class="WPVDB_Filter_Reset_Hook" file="tests/filter-reset-hook.php" />
+    </extensions>
     <coverage processUncoveredFiles="true">
         <include>
             <directory suffix=".php">includes</directory>
@@ -38,4 +38,4 @@
         <const name="WP_TESTS_FORCE_KNOWN_BUGS" value="true" />
         <server name="SERVER_NAME" value="http://example.org" />
     </php>
-</phpunit>
+</phpunit>

tests/bootstrap.php

@@ -12,6 +12,9 @@
 echo "Using standalone testing mode.\n";
 
 // Mock essential WordPress functions for testing
+global $_wp_filters;
+$_wp_filters = [];
+
 if ( ! function_exists( 'add_action' ) ) {
     function add_action( $tag, $function_to_add, $priority = 10, $accepted_args = 1 ) {
         return true;
@@ -20,10 +23,45 @@ function add_action( $tag, $function_to_add, $priority = 10, $accepted_args = 1
 
 if ( ! function_exists( 'add_filter' ) ) {
     function add_filter( $tag, $function_to_add, $priority = 10, $accepted_args = 1 ) {
+        global $_wp_filters;
+        if ( ! isset( $_wp_filters[ $tag ] ) ) {
+            $_wp_filters[ $tag ] = [];
+        }
+        if ( ! isset( $_wp_filters[ $tag ][ $priority ] ) ) {
+            $_wp_filters[ $tag ][ $priority ] = [];
+        }
+        $_wp_filters[ $tag ][ $priority ][] = [
+            'function'      => $function_to_add,
+            'accepted_args' => $accepted_args,
+        ];
         return true;
     }
 }
 
+if ( ! function_exists( 'remove_filter' ) ) {
+    function remove_filter( $tag, $function_to_remove, $priority = 10 ) {
+        global $_wp_filters;
+        if ( empty( $_wp_filters[ $tag ][ $priority ] ) ) {
+            return false;
+        }
+
+        foreach ( $_wp_filters[ $tag ][ $priority ] as $index => $filter ) {
+            if ( $filter['function'] === $function_to_remove ) {
+                unset( $_wp_filters[ $tag ][ $priority ][ $index ] );
+                if ( empty( $_wp_filters[ $tag ][ $priority ] ) ) {
+                    unset( $_wp_filters[ $tag ][ $priority ] );
+                }
+                if ( empty( $_wp_filters[ $tag ] ) ) {
+                    unset( $_wp_filters[ $tag ] );
+                }
+                return true;
+            }
+        }
+
+        return false;
+    }
+}
+
 if ( ! function_exists( 'do_action' ) ) {
     function do_action( $hook_name, ...$args ) {
         // Mock implementation
@@ -32,8 +70,22 @@ function do_action( $hook_name, ...$args ) {
 }
 
 if ( ! function_exists( 'apply_filters' ) ) {
-    function apply_filters( $tag, $value ) {
-        // Simple mock that just returns the value unchanged
+    function apply_filters( $tag, $value, ...$args ) {
+        global $_wp_filters;
+        if ( empty( $_wp_filters[ $tag ] ) ) {
+            return $value;
+        }
+
+        ksort( $_wp_filters[ $tag ] );
+        $filter_args = array_merge( [ $value ], $args );
+        foreach ( $_wp_filters[ $tag ] as $filters ) {
+            foreach ( $filters as $filter ) {
+                $accepted_args = max( 0, (int) $filter['accepted_args'] );
+                $value = call_user_func_array( $filter['function'], array_slice( $filter_args, 0, $accepted_args ) );
+                $filter_args[0] = $value;
+            }
+        }
+
         return $value;
     }
 }
@@ -183,7 +235,24 @@ function has_action( $tag, $function_to_check = false ) {
 
 if ( ! function_exists( 'has_filter' ) ) {
     function has_filter( $tag, $function_to_check = false ) {
-        return false; // Mock - always return false
+        global $_wp_filters;
+        if ( empty( $_wp_filters[ $tag ] ) ) {
+            return false;
+        }
+
+        if ( false === $function_to_check ) {
+            return true;
+        }
+
+        foreach ( $_wp_filters[ $tag ] as $priority => $filters ) {
+            foreach ( $filters as $filter ) {
+                if ( $filter['function'] === $function_to_check ) {
+                    return $priority;
+                }
+            }
+        }
+
+        return false;
     }
 }
 
@@ -266,6 +335,18 @@ function stripslashes_from_strings_only( $value ) {
     define( 'REST_REQUEST', false );
 }
 
+if ( ! function_exists( 'wpvdb_is_sqlite' ) ) {
+    function wpvdb_is_sqlite() {
+        return false;
+    }
+}
+
+if ( ! function_exists( 'wpvdb_should_log_to_error_log' ) ) {
+    function wpvdb_should_log_to_error_log( $level = 'debug', $message = '', $context = array() ) {
+        return false;
+    }
+}
+
 if ( ! defined( 'OBJECT' ) ) {
     define( 'OBJECT', 'OBJECT' );
 }
@@ -435,8 +516,10 @@ public function suppress_errors( $suppress = true ) {
 require_once dirname( __DIR__ ) . '/includes/class-wpvdb-providers.php';
 require_once dirname( __DIR__ ) . '/includes/class-wpvdb-models.php';
 require_once dirname( __DIR__ ) . '/includes/class-wpvdb-settings.php';
+require_once dirname( __DIR__ ) . '/includes/class-wpvdb-cache.php';
 require_once dirname( __DIR__ ) . '/includes/class-wpvdb-core.php';
 require_once dirname( __DIR__ ) . '/includes/class-wpvdb-database.php';
+require_once dirname( __DIR__ ) . '/includes/class-wpvdb-rest.php';
 require_once dirname( __DIR__ ) . '/includes/class-wpvdb-embedding-enqueuer.php';
 require_once dirname( __DIR__ ) . '/includes/class-wpvdb-queue.php';
 require_once dirname( __DIR__ ) . '/includes/class-wpvdb-security.php';

tests/filter-reset-hook.php

@@ -0,0 +1,41 @@
+<?php
+/**
+ * PHPUnit hook for resetting standalone WordPress filter mocks.
+ *
+ * @package WPVDB
+ */
+
+use PHPUnit\Runner\AfterTestHook;
+use PHPUnit\Runner\BeforeTestHook;
+
+/**
+ * Reset the mocked WordPress filter registry around every test.
+ */
+class WPVDB_Filter_Reset_Hook implements BeforeTestHook, AfterTestHook {
+
+	/**
+	 * Reset filters before each test starts.
+	 *
+	 * @param string $test Test name.
+	 */
+	public function executeBeforeTest( string $test ): void {
+		$this->reset_filters();
+	}
+
+	/**
+	 * Reset filters after each test finishes.
+	 *
+	 * @param string $test Test name.
+	 * @param float  $time Test runtime.
+	 */
+	public function executeAfterTest( string $test, float $time ): void {
+		$this->reset_filters();
+	}
+
+	/**
+	 * Reset the standalone filter registry.
+	 */
+	private function reset_filters(): void {
+		$GLOBALS['_wp_filters'] = array();
+	}
+}