Fix BOM handling #586

Author: nyamsprod

src/Reader.php

@@ -533,10 +533,9 @@ public function selectAllExcept(string|int ...$columns): TabularDataReader
      */
     public function getRecords(array $header = []): Iterator
     {
-        return $this->combineHeader(
-            $this->prepareRecords(),
-            $this->prepareHeader($header)
-        );
+        $foundHeaders = $this->prepareHeader($header);
+
+        return $this->combineHeader($this->prepareRecords(), $foundHeaders);
     }
 
     /**
@@ -568,12 +567,12 @@ public function getRecordsAsObject(string $className, array $header = []): Itera
     protected function prepareRecords(): Iterator
     {
         $normalized = fn ($record): bool => is_array($record) && ($this->is_empty_records_included || $record !== [null]);
-        $bom = null;
+        $document = $this->getDocument();
         if (!$this->is_input_bom_included) {
-            $bom = Bom::tryFrom($this->getInputBOM());
+            $document = SkipBomIterator::fromDocument($document, $this->delimiter, $this->enclosure, $this->escape);
         }
 
-        $records = $this->stripBOM(new CallbackFilterIterator($this->getDocument(), $normalized), $bom);
+        $records = new CallbackFilterIterator($document, $normalized);
         if (null !== $this->header_offset) {
             $records = new CallbackFilterIterator($records, fn (array $record, int $offset): bool => $offset !== $this->header_offset);
         }
@@ -587,6 +586,7 @@ protected function prepareRecords(): Iterator
 
     /**
      * Strips the BOM sequence from the returned records if necessary.
+     * @deprecated
      */
     protected function stripBOM(Iterator $iterator, ?Bom $bom): Iterator
     {

src/ReaderTest.php

@@ -21,6 +21,7 @@
 
 use function array_keys;
 use function count;
+use function dump;
 use function fclose;
 use function fopen;
 use function fputcsv;
@@ -652,4 +653,18 @@ public function testStreamWithFiltersDestructsGracefully(): void
         // An explicitly closed file handle makes the stream filter resources invalid
         fclose($fp);
     }
+
+    public function test_handline_bom_expression_issue_586(): void
+    {
+        $bom = Bom::Utf8->value;
+        $contents = [
+            '"a,b",c' => ["a,b", "c"],
+            '"a'."\n".'b",c' => ['a'."\n".'b', "c"],
+        ];
+
+        foreach ($contents as $content => $expected) {
+            self::assertSame($expected, Reader::fromString($content)->first());
+            self::assertSame($expected, Reader::fromString($bom.$content)->first());
+        }
+    }
 }

src/SkipBomIterator.php

@@ -0,0 +1,84 @@
+<?php
+
+/**
+ * League.Csv (https://csv.thephpleague.com)
+ *
+ * (c) Ignace Nyamagana Butera <nyamsprod@gmail.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+declare(strict_types=1);
+
+namespace League\Csv;
+
+use Iterator;
+use SplFileObject;
+use const SEEK_SET;
+
+/**
+ * @internal
+ */
+final class SkipBomIterator implements Iterator
+{
+    public static function fromDocument(
+        SplFileObject|Stream $document,
+        string $delimiter = ',',
+        string $enclosure = '"',
+        string $escape = '\\',
+    ): self {
+
+        return new self(
+            $document,
+            Bom::tryFromSequence($document)?->length() ?? 0,
+            $delimiter,
+            $enclosure,
+            $escape,
+        );
+    }
+
+    public function __construct(
+        private readonly SplFileObject|Stream $file,
+        private readonly int $offset,
+        private readonly string $delimiter = ',',
+        private readonly string $enclosure = '"',
+        private readonly string $escape = '\\',
+    ) {
+        $this->file->setFlags(SplFileObject::READ_CSV);
+        $this->file->setCsvControl($this->delimiter, $this->enclosure, $this->escape);
+    }
+
+    public function current(): mixed
+    {
+        return $this->file->current();
+    }
+
+    public function fseek(int $offset, int $whence): int
+    {
+        return $this->file->fseek($offset, $whence);
+    }
+
+    public function key(): int
+    {
+        return $this->file->key();
+    }
+
+    public function next(): void
+    {
+        $this->file->next();
+    }
+
+    public function rewind(): void
+    {
+        $this->file->rewind();
+        if (0 !== $this->offset) {
+            $this->fseek($this->offset, SEEK_SET);
+        }
+    }
+
+    public function valid(): bool
+    {
+        return !$this->file->eof();
+    }
+}