From f15331fb3f1f1d679b459fcb33e44fbf3657e295 Mon Sep 17 00:00:00 2001
From: Joseph Bielawski <stloyd@gmail.com>
Date: Sat, 11 Jul 2026 12:49:07 +0200
Subject: [PATCH] Reduce complexity of XML reading & writing

---
 .../Flow/ETL/Adapter/XML/Loader/XMLLoader.php | 15 ++----
 .../ETL/Adapter/XML/XMLParserExtractor.php    | 54 +++++--------------
 2 files changed, 17 insertions(+), 52 deletions(-)

diff --git a/src/adapter/etl-adapter-xml/src/Flow/ETL/Adapter/XML/Loader/XMLLoader.php b/src/adapter/etl-adapter-xml/src/Flow/ETL/Adapter/XML/Loader/XMLLoader.php
index b6e895a00b..74d441432b 100644
--- a/src/adapter/etl-adapter-xml/src/Flow/ETL/Adapter/XML/Loader/XMLLoader.php
+++ b/src/adapter/etl-adapter-xml/src/Flow/ETL/Adapter/XML/Loader/XMLLoader.php
@@ -22,11 +22,7 @@
 use Throwable;
 
 use function array_key_exists;
-use function array_keys;
-use function array_map;
-use function array_values;
 use function count;
-use function implode;
 
 final class XMLLoader implements Closure, FileLoader, Loader
 {
@@ -194,13 +190,12 @@ public function write(Rows $nextRows, array $partitions, FlowContext $context, R
                 $this->writes[$stream->path()->path()] = 0;
             }
 
-            $xmlAttributes = implode(' ', array_map(
-                static fn(string $key, string $value) => $key . '="' . $value . '"',
-                array_keys($this->xmlAttributes),
-                array_values($this->xmlAttributes),
-            ));
+            $xmlAttributes = '';
+            foreach ($this->xmlAttributes as $key => $value) {
+                $xmlAttributes .= $key . '="' . $value . '" ';
+            }
 
-            $stream->append('<?xml ' . $xmlAttributes . "?>\n<" . $this->rootElementName . ">\n");
+            $stream->append('<?xml ' . trim($xmlAttributes) . "?>\n<" . $this->rootElementName . ">\n");
         } else {
             $stream = $streams->writeTo($this->path, $partitions);
         }
diff --git a/src/adapter/etl-adapter-xml/src/Flow/ETL/Adapter/XML/XMLParserExtractor.php b/src/adapter/etl-adapter-xml/src/Flow/ETL/Adapter/XML/XMLParserExtractor.php
index 61d02d07cf..ebade62727 100644
--- a/src/adapter/etl-adapter-xml/src/Flow/ETL/Adapter/XML/XMLParserExtractor.php
+++ b/src/adapter/etl-adapter-xml/src/Flow/ETL/Adapter/XML/XMLParserExtractor.php
@@ -60,7 +60,7 @@ final class XMLParserExtractor implements Extractor, FileExtractor, LimitableExt
     private string $xmlNodePath = '';
 
     /**
-     * In order to iterate only over <element> nodes use `$loader->withXMLNodePath('root/elements/element')`.
+     * To iterate only over <element> nodes, use `$loader->withXMLNodePath('root/elements/element')`.
      *
      * <root>
      *   <elements>
@@ -124,51 +124,18 @@ public function extract(FlowContext $context): Generator
                     ));
                 }
 
-                if (count($this->elements)) {
-                    foreach ($this->elements as $element) {
-                        if ($shouldPutInputIntoRows) {
-                            $rowData = [
-                                'node' => $this->createDOMNode($element),
-                                '_input_file_uri' => $uri,
-                            ];
-                        } else {
-                            $rowData = ['node' => $this->createDOMNode($element)];
-                        }
-
-                        $signal = yield array_to_rows(
-                            $rowData,
-                            $context->entryFactory(),
-                            $stream->path()->partitions(),
-                            $this->schema,
-                        );
-
-                        $this->incrementReturnedRows();
-
-                        if ($signal === Signal::STOP || $this->reachedLimit()) {
-                            $context->streams()->closeStreams($this->path);
-                            $this->freeParser();
-
-                            return;
-                        }
-                    }
-                    $this->elements = [];
-                }
-            }
-
-            xml_parse($this->parser(), '', true);
-
-            if (count($this->elements)) {
                 foreach ($this->elements as $element) {
+                    $rowData = ['node' => $this->createDOMNode($element)];
                     if ($shouldPutInputIntoRows) {
-                        $rowData = [
-                            'node' => $this->createDOMNode($element),
-                            '_input_file_uri' => $uri,
-                        ];
-                    } else {
-                        $rowData = ['node' => $this->createDOMNode($element)];
+                        $rowData['_input_file_uri'] = $uri;
                     }
 
-                    $signal = yield array_to_rows([$rowData], $context->entryFactory(), $stream->path()->partitions());
+                    $signal = yield array_to_rows(
+                        $rowData,
+                        $context->entryFactory(),
+                        $stream->path()->partitions(),
+                        $this->schema,
+                    );
 
                     $this->incrementReturnedRows();
 
@@ -179,9 +146,12 @@ public function extract(FlowContext $context): Generator
                         return;
                     }
                 }
+
                 $this->elements = [];
             }
 
+            xml_parse($this->parser(), '', true);
+
             $this->freeParser();
         }
     }