Fix uncaught OverflowException: The buffer size is too big for the de…

…fined file size limit
samdark · Nov 1, 2023 · cf51475 · cf51475
1 parent a7ba091
commit cf51475
Show file tree

Hide file tree

Showing 2 changed files with 177 additions and 5 deletions.
diff --git a/Sitemap.php b/Sitemap.php
@@ -181,7 +181,7 @@ private function createNewFile()
          * elements that did not fit into the previous file. (See self::flush)
          */
         $this->writer->text("\n");
-        $this->flush(true);
+        $this->flush();
     }
 
     /**
@@ -209,17 +209,22 @@ private function finishFile()
      */
     public function write()
     {
-        $this->finishFile();
+        if ($this->writer !== null) {
+            $this->flush();
+            $this->finishFile();
+        }
     }
 
     /**
      * Flushes buffer into file
      *
      * @param int $footSize Size of the remaining closing tags
+     * @return bool is new file created
      * @throws \OverflowException
      */
     private function flush($footSize = 10)
     {
+        $isNewFileCreated = false;
         $data = $this->writer->flush(true);
         $dataSize = mb_strlen($data, '8bit');
 
@@ -235,10 +240,13 @@ private function flush($footSize = 10)
             }
             $this->finishFile();
             $this->createNewFile();
+            $isNewFileCreated = true;
         }
 
         $this->writerBackend->append($data);
         $this->byteCount += $dataSize;
+
+        return $isNewFileCreated;
     }
 
     /**
@@ -268,8 +276,11 @@ protected function validateLocation($location) {
      */
     public function addItem($location, $lastModified = null, $changeFrequency = null, $priority = null)
     {
-        if ($this->urlsCount >= $this->maxUrls) {
-            $this->finishFile();
+        if ($this->urlsCount >= $this->maxUrls && $this->writer !== null) {
+            $isNewFileCreated = $this->flush();
+            if (!$isNewFileCreated) {
+                $this->finishFile();
+            }
         }
 
         if ($this->writerBackend === null) {
@@ -517,4 +528,4 @@ public function setStylesheet($stylesheetUrl)
             $this->stylesheet = $stylesheetUrl;
         }
     }
-}
+}
diff --git a/tests/SitemapTest.php b/tests/SitemapTest.php
@@ -7,6 +7,10 @@
 
 class SitemapTest extends \PHPUnit_Framework_TestCase
 {
+    const HEADER_LENGTH = 100;
+    const FOOTER_LENGTH = 10;
+    const ELEMENT_LENGTH_WITHOUT_URL = 137;
+
     /**
      * Asserts validity of simtemap according to XSD schema
      * @param string $fileName
@@ -365,4 +369,161 @@ public function testBufferSizeImpact()
 
         $this->assertLessThan($times[0] * 1.2, $times[1]);
     }
+
+    public function testBufferSizeIsNotTooBigOnFinishFileInWrite()
+    {
+        $time = 100;
+        $urlLength = 13;
+        $urlsQty = 4;
+
+        $sitemapPath = __DIR__ . '/sitemap.xml';
+        $sitemap = new Sitemap($sitemapPath);
+        $sitemap->setBufferSize(3);
+        $sitemap->setMaxUrls(4);
+        $sitemap->setMaxBytes(
+            self::HEADER_LENGTH + self::FOOTER_LENGTH + self::ELEMENT_LENGTH_WITHOUT_URL * $urlsQty
+                + $urlLength * $urlsQty - 1
+        );
+
+        for ($i = 0; $i < $urlsQty; $i++) {
+            $sitemap->addItem(
+                // url 13 bytes
+                "https://a.b/{$i}",
+                $time,
+                Sitemap::WEEKLY,
+                1
+            );
+        }
+        $sitemap->write();
+
+        $expectedFiles = array(
+            __DIR__ . '/sitemap.xml',
+            __DIR__ . '/sitemap_2.xml',
+        );
+        $expected[] = <<<EOF
+<?xml version="1.0" encoding="UTF-8"?>
+<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
+ <url>
+  <loc>https://a.b/0</loc>
+  <lastmod>1970-01-01T00:01:40+00:00</lastmod>
+  <changefreq>weekly</changefreq>
+  <priority>1.0</priority>
+ </url>
+ <url>
+  <loc>https://a.b/1</loc>
+  <lastmod>1970-01-01T00:01:40+00:00</lastmod>
+  <changefreq>weekly</changefreq>
+  <priority>1.0</priority>
+ </url>
+ <url>
+  <loc>https://a.b/2</loc>
+  <lastmod>1970-01-01T00:01:40+00:00</lastmod>
+  <changefreq>weekly</changefreq>
+  <priority>1.0</priority>
+ </url>
+</urlset>
+EOF;
+        $expected[] = <<<EOF
+<?xml version="1.0" encoding="UTF-8"?>
+<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
+ <url>
+  <loc>https://a.b/3</loc>
+  <lastmod>1970-01-01T00:01:40+00:00</lastmod>
+  <changefreq>weekly</changefreq>
+  <priority>1.0</priority>
+ </url>
+</urlset>
+EOF;
+        foreach ($expectedFiles as $expectedFileNumber => $expectedFile) {
+            $this->assertTrue(file_exists($expectedFile), "$expectedFile does not exist!");
+            $this->assertIsValidSitemap($expectedFile);
+
+            $actual = trim(file_get_contents($expectedFile));
+            $this->assertEquals($expected[$expectedFileNumber], $actual);
+
+            unlink($expectedFile);
+        }
+    }
+
+    public function testBufferSizeIsNotTooBigOnFinishFileInAddItem()
+    {
+        $time = 100;
+        $urlLength = 13;
+        $urlsQty = 5;
+
+        $sitemapPath = __DIR__ . '/sitemap.xml';
+        $sitemap = new Sitemap($sitemapPath);
+        $sitemap->setBufferSize(3);
+        $sitemap->setMaxUrls(4);
+        $sitemap->setMaxBytes(
+            // 100 + 10 + 137 * 4
+            self::HEADER_LENGTH + self::FOOTER_LENGTH + self::ELEMENT_LENGTH_WITHOUT_URL * 4
+                + $urlLength * 4 - 1
+        );
+
+        for ($i = 0; $i < $urlsQty; $i++) {
+            $sitemap->addItem(
+                // url 13 bytes
+                "https://a.b/{$i}",
+                $time,
+                Sitemap::WEEKLY,
+                1
+            );
+        }
+        $sitemap->write();
+
+        $expectedFiles = array(
+            __DIR__ . '/sitemap.xml',
+            __DIR__ . '/sitemap_2.xml',
+        );
+        $expected[] = <<<EOF
+<?xml version="1.0" encoding="UTF-8"?>
+<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
+ <url>
+  <loc>https://a.b/0</loc>
+  <lastmod>1970-01-01T00:01:40+00:00</lastmod>
+  <changefreq>weekly</changefreq>
+  <priority>1.0</priority>
+ </url>
+ <url>
+  <loc>https://a.b/1</loc>
+  <lastmod>1970-01-01T00:01:40+00:00</lastmod>
+  <changefreq>weekly</changefreq>
+  <priority>1.0</priority>
+ </url>
+ <url>
+  <loc>https://a.b/2</loc>
+  <lastmod>1970-01-01T00:01:40+00:00</lastmod>
+  <changefreq>weekly</changefreq>
+  <priority>1.0</priority>
+ </url>
+</urlset>
+EOF;
+        $expected[] = <<<EOF
+<?xml version="1.0" encoding="UTF-8"?>
+<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
+ <url>
+  <loc>https://a.b/3</loc>
+  <lastmod>1970-01-01T00:01:40+00:00</lastmod>
+  <changefreq>weekly</changefreq>
+  <priority>1.0</priority>
+ </url>
+ <url>
+  <loc>https://a.b/4</loc>
+  <lastmod>1970-01-01T00:01:40+00:00</lastmod>
+  <changefreq>weekly</changefreq>
+  <priority>1.0</priority>
+ </url>
+</urlset>
+EOF;
+        foreach ($expectedFiles as $expectedFileNumber => $expectedFile) {
+            $this->assertTrue(file_exists($expectedFile), "$expectedFile does not exist!");
+            $this->assertIsValidSitemap($expectedFile);
+
+            $actual = trim(file_get_contents($expectedFile));
+            $this->assertEquals($expected[$expectedFileNumber], $actual);
+
+            unlink($expectedFile);
+        }
+    }
 }