Optimized support for ignoring faulty structures in page tree

The parsing of the page tree is stopped now as soon as the page count is reached. By doing this we can ignore following faulty objects in the Kids structure.

Author: JanSlabon

composer.lock

@@ -202,7 +202,8 @@ protected function readPages($readAll = false)
             return;
         }
 
-        $readPages = function ($kids, $count) use (&$readPages, $readAll) {
+        $expectedPageCount = $this->getPageCount();
+        $readPages = function ($kids, $count) use (&$readPages, $readAll, $expectedPageCount) {
             $kids = PdfArray::ensure($kids);
             $isLeaf = ($count->value === \count($kids->value));
 
@@ -222,6 +223,11 @@ protected function readPages($readAll = false)
                 } else {
                     $this->pages[] = $object;
                 }
+
+                // stop if all pages are read - faulty documents exists with additional entries with invalid data.
+                if (count($this->pages) === $expectedPageCount) {
+                    break;
+                }
             }
         };
 

src/PdfReader/PdfReader.php

@@ -49,6 +49,11 @@ public function getPageCountProvider()
             13
         ];
 
+        $data[] = [
+            $path . '/specials/page-trees/PageTreeWithInvalidTypeAndMisslLeadingCount.pdf',
+            1
+        ];
+
         return $data;
     }
 
@@ -835,6 +840,27 @@ public function getPageProvider()
             ]
         ];
 
+        $data[] = [
+            $path . '/specials/page-trees/PageTreeWithInvalidTypeAndMisslLeadingCount.pdf',
+            [
+                1 => PdfIndirectObject::create(
+                    1,
+                    0,
+                    PdfDictionary::create([
+                        'Type' => PdfName::create('Page'),
+                        'MediaBox' => PdfArray::create([
+                            PdfNumeric::create(0),
+                            PdfNumeric::create(0),
+                            PdfNumeric::create(595.28),
+                            PdfNumeric::create(841.89)
+                        ]),
+                        'Resources' => PdfDictionary::create(),
+                        'Parent' => PdfIndirectObjectReference::create(3, 0),
+                    ])
+                )
+            ]
+        ];
+
         return $data;
     }