Html Reader Process Titles as Headings Not Paragraphs

composer.json

@@ -122,7 +122,7 @@
         "phpmd/phpmd": "^2.13",
         "phpstan/phpstan": "^0.12.88 || ^1.0.0",
         "phpstan/phpstan-phpunit": "^1.0 || ^2.0",
-        "phpunit/phpunit": ">=7.0",
+        "phpunit/phpunit": "^7.0 || ^8.0 || ^9.0 || ^10.0",
         "symfony/process": "^4.4 || ^5.0",
         "tecnickcom/tcpdf": "^6.5"
     },

src/PhpWord/Shared/Html.php

@@ -345,18 +345,22 @@ protected static function parseInput($node, $element, &$styles): void
     /**
      * Parse heading node.
      *
-     * @param string $argument1 Name of heading style
-     *
      * @todo Think of a clever way of defining header styles, now it is only based on the assumption, that
      * Heading1 - Heading6 are already defined somewhere
      */
-    protected static function parseHeading(DOMNode $node, AbstractContainer $element, array &$styles, string $argument1): TextRun
+    protected static function parseHeading(DOMNode $node, AbstractContainer $element, array &$styles, string $headingStyle): TextRun
     {
         $style = new Paragraph();
-        $style->setStyleName($argument1);
+        $style->setStyleName($headingStyle);
         $style->setStyleByArray(self::parseInlineStyle($node, $styles['paragraph']));
+        $textRun = new TextRun($style);
+
+        // Create a title with level corresponding to number in heading style
+        // (Eg, Heading1 = 1)
+        $element->addTitle($textRun, (int) ltrim($headingStyle, 'Heading'));
 
-        return $element->addTextRun($style);
+        // Return TextRun so children are parsed
+        return $textRun;
     }
 
     /**

src/PhpWord/Writer/HTML/Element/Title.php

@@ -18,7 +18,11 @@
 
 namespace PhpOffice\PhpWord\Writer\HTML\Element;
 
+use PhpOffice\PhpWord\Element\Title as ElementTitle;
+use PhpOffice\PhpWord\Style;
 use PhpOffice\PhpWord\Writer\HTML;
+use PhpOffice\PhpWord\Writer\HTML\Style\Font;
+use PhpOffice\PhpWord\Writer\HTML\Style\Paragraph;
 
 /**
  * TextRun element HTML writer.
@@ -34,21 +38,41 @@ class Title extends AbstractElement
      */
     public function write()
     {
-        if (!$this->element instanceof \PhpOffice\PhpWord\Element\Title) {
+        if (!$this->element instanceof ElementTitle) {
             return '';
         }
 
         $tag = 'h' . $this->element->getDepth();
 
         $text = $this->element->getText();
+        $paragraphStyle = null;
         if (is_string($text)) {
             $text = $this->parentWriter->escapeHTML($text);
         } else {
+            $paragraphStyle = $text->getParagraphStyle();
             $writer = new Container($this->parentWriter, $text);
             $text = $writer->write();
         }
+        $css = '';
+        $write1 = $write2 = $write3 = '';
+        $style = Style::getStyle('Heading_' . $this->element->getDepth());
+        if ($style !== null) {
+            $styleWriter = new Font($style);
+            $write1 = $styleWriter->write();
+        }
+        if (is_object($paragraphStyle)) {
+            $styleWriter = new Paragraph($paragraphStyle);
+            $write3 = $styleWriter->write();
+            if ($write1 !== '' && $write3 !== '') {
+                $write2 = ' ';
+            }
+        }
+        $css = "$write1$write2$write3";
+        if ($css !== '') {
+            $css = " style=\"$css\"";
+        }
 
-        $content = "<{$tag}>{$text}</{$tag}>" . PHP_EOL;
+        $content = "<{$tag}{$css}>{$text}</{$tag}>" . PHP_EOL;
 
         return $content;
     }

src/PhpWord/Writer/HTML/Part/Head.php

@@ -92,17 +92,15 @@ private function writeStyles(): string
             'font-size' => Settings::getDefaultFontSize() . 'pt',
             'color' => "#{$defaultFontColor}",
         ];
-        // Mpdf sometimes needs separate tag for body; doesn't harm others.
-        $bodyarray = $astarray;
 
         $defaultWhiteSpace = $this->getParentWriter()->getDefaultWhiteSpace();
         if ($defaultWhiteSpace) {
             $astarray['white-space'] = $defaultWhiteSpace;
         }
+        $bodyarray = $astarray;
 
         foreach ([
             'body' => $bodyarray,
-            '*' => $astarray,
             'a.NoteRef' => [
                 'text-decoration' => 'none',
             ],
@@ -135,12 +133,13 @@ private function writeStyles(): string
                     $styleWriter = new FontStyleWriter($style);
                     if ($style->getStyleType() == 'title') {
                         $name = str_replace('Heading_', 'h', $name);
+                        $css .= "{$name} {" . $styleWriter->write() . '}' . PHP_EOL;
                         $styleParagraph = $style->getParagraph();
                         $style = $styleParagraph;
                     } else {
                         $name = '.' . $name;
+                        $css .= "{$name} {" . $styleWriter->write() . '}' . PHP_EOL;
                     }
-                    $css .= "{$name} {" . $styleWriter->write() . '}' . PHP_EOL;
                 }
                 if ($style instanceof Paragraph) {
                     $styleWriter = new ParagraphStyleWriter($style);

tests/PhpWordTests/Shared/HtmlHeadingsTest.php

@@ -0,0 +1,76 @@
+<?php
+
+/**
+ * This file is part of PHPWord - A pure PHP library for reading and writing
+ * word processing documents.
+ *
+ * PHPWord is free software distributed under the terms of the GNU Lesser
+ * General Public License version 3 as published by the Free Software Foundation.
+ *
+ * For the full copyright and license information, please read the LICENSE
+ * file that was distributed with this source code. For the full list of
+ * contributors, visit https://github.com/PHPOffice/PHPWord/contributors.
+ *
+ * @see         https://github.com/PHPOffice/PHPWord
+ *
+ * @license     http://www.gnu.org/licenses/lgpl.txt LGPL version 3
+ */
+
+namespace PhpOffice\PhpWordTests\Shared;
+
+use PhpOffice\PhpWord\Element\TextRun;
+use PhpOffice\PhpWord\PhpWord;
+use PhpOffice\PhpWord\Settings;
+use PhpOffice\PhpWord\Shared\Html as SharedHtml;
+use PhpOffice\PhpWord\Writer\HTML as HtmlWriter;
+use PHPUnit\Framework\TestCase;
+
+/**
+ * Test class for PhpOffice\PhpWord\Shared\Html.
+ *
+ * @coversDefaultClass \PhpOffice\PhpWord\Shared\Html
+ */
+class HtmlHeadingsTest extends TestCase
+{
+    public function testRoundTripHeadings(): void
+    {
+        Settings::setOutputEscapingEnabled(true);
+        $originalDoc = new PhpWord();
+        $originalDoc->addTitleStyle(1, ['size' => 20]);
+        $section = $originalDoc->addSection();
+        $expectedStrings = [];
+        $section->addTitle('Title 1', 1);
+        $expectedStrings[] = '<h1 style="font-size: 20pt;">Title 1</h1>';
+        for ($i = 2; $i <= 6; ++$i) {
+            $textRun = new TextRun();
+            $textRun->addText('Title ');
+            $textRun->addText("$i", ['italic' => true]);
+            $section->addTitle($textRun, $i);
+            $expectedStrings[] = "<h$i>Title <span style=\"font-style: italic;\">$i</span></h$i>";
+        }
+        $writer = new HtmlWriter($originalDoc);
+        $content = $writer->getContent();
+        foreach ($expectedStrings as $expectedString) {
+            self::assertStringContainsString($expectedString, $content);
+        }
+
+        $newDoc = new PhpWord();
+        $newSection = $newDoc->addSection();
+        SharedHtml::addHtml($newSection, $content, true);
+        $newWriter = new HtmlWriter($newDoc);
+        $newContent = $newWriter->getContent();
+        // Reader does not yet support h1 declaration in css.
+        $content = str_replace('h1 {font-size: 20pt;}' . PHP_EOL, '', $content);
+
+        // Reader transforms Text to TextRun,
+        //  but result is functionally the same.
+        self::assertSame(
+            $newContent,
+            str_replace(
+                '<h1 style="font-size: 20pt;">Title 1</h1>',
+                '<h1><span style="font-size: 20pt;">Title 1</span></h1>',
+                $content
+            )
+        );
+    }
+}

tests/PhpWordTests/Shared/HtmlTest.php

@@ -24,6 +24,7 @@
 use PhpOffice\PhpWord\Element\Table;
 use PhpOffice\PhpWord\Element\Text;
 use PhpOffice\PhpWord\Element\TextRun;
+use PhpOffice\PhpWord\Element\Title;
 use PhpOffice\PhpWord\PhpWord;
 use PhpOffice\PhpWord\Shared\Converter;
 use PhpOffice\PhpWord\Shared\Html;
@@ -116,6 +117,8 @@ public function testParseHeader(): void
 
         self::assertCount(1, $section->getElements());
         $element = $section->getElement(0);
+        self::assertInstanceOf(Title::class, $element);
+        $element = $element->getText();
         self::assertInstanceOf(TextRun::class, $element);
         self::assertInstanceOf(Paragraph::class, $element->getParagraphStyle());
         self::assertEquals('Heading1', $element->getParagraphStyle()->getStyleName());
@@ -137,6 +140,8 @@ public function testParseHeaderStyle(): void
 
         self::assertCount(1, $section->getElements());
         $element = $section->getElement(0);
+        self::assertInstanceOf(Title::class, $element);
+        $element = $element->getText();
         self::assertInstanceOf(TextRun::class, $element);
         self::assertInstanceOf(Paragraph::class, $element->getParagraphStyle());
         self::assertEquals('Heading1', $element->getParagraphStyle()->getStyleName());

tests/PhpWordTests/Writer/HTML/FontTest.php

@@ -121,29 +121,24 @@ public function testFontNames1(): void
         self::assertEquals('style5', Helper::getTextContent($xpath, '/html/body/div/p[6]/span', 'class'));
 
         $style = Helper::getTextContent($xpath, '/html/head/style');
-        $prg = preg_match('/^[*][^\\r\\n]*/m', $style, $matches);
+        $prg = preg_match('/^body[^\\r\\n]*/m', $style, $matches);
         self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
-        self::assertEquals('* {font-family: \'Courier New\'; font-size: 12pt; color: #000000;}', $matches[0]);
+        self::assertSame(1, $prg);
+        self::assertEquals('body {font-family: \'Courier New\'; font-size: 12pt; color: #000000;}', $matches[0]);
         $prg = preg_match('/^[.]style1[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style1 {font-family: \'Tahoma\'; font-size: 10pt; color: #1B2232; font-weight: bold;}', $matches[0]);
         $prg = preg_match('/^[.]style2[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style2 {font-family: \'Arial\'; font-size: 10pt;}', $matches[0]);
         $prg = preg_match('/^[.]style3[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style3 {font-family: \'hack attempt'}; display:none\'; font-size: 10pt;}', $matches[0]);
         $prg = preg_match('/^[.]style4[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style4 {font-family: \'padmaa 1.1\'; font-size: 10pt; font-weight: bold;}', $matches[0]);
         $prg = preg_match('/^[.]style5[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style5 {font-family: \'MingLiU-ExtB\'; font-size: 10pt; font-weight: bold;}', $matches[0]);
     }
 
@@ -177,25 +172,21 @@ public function testFontNames2(): void
         self::assertEquals('style4', Helper::getTextContent($xpath, '/html/body/div/p[5]/span', 'class'));
 
         $style = Helper::getTextContent($xpath, '/html/head/style');
-        $prg = preg_match('/^[*][^\\r\\n]*/m', $style, $matches);
+        $prg = preg_match('/^body[^\\r\\n]*/m', $style, $matches);
         self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
-        self::assertEquals('* {font-family: \'Courier New\'; font-size: 12pt; color: #000000;}', $matches[0]);
+        self::assertSame(1, $prg);
+        self::assertEquals('body {font-family: \'Courier New\'; font-size: 12pt; color: #000000;}', $matches[0]);
         $prg = preg_match('/^[.]style1[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style1 {font-family: \'Tahoma\'; font-size: 10pt; color: #1B2232; font-weight: bold;}', $matches[0]);
         $prg = preg_match('/^[.]style2[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style2 {font-family: \'Arial\', sans-serif; font-size: 10pt;}', $matches[0]);
         $prg = preg_match('/^[.]style3[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style3 {font-family: \'DejaVu Sans Monospace\', monospace; font-size: 10pt;}', $matches[0]);
         $prg = preg_match('/^[.]style4[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style4 {font-family: \'Arial\'; font-size: 10pt;}', $matches[0]);
     }
 
@@ -229,25 +220,21 @@ public function testFontNames3(): void
         self::assertEquals('style4', Helper::getTextContent($xpath, '/html/body/div/p[5]/span', 'class'));
 
         $style = Helper::getTextContent($xpath, '/html/head/style');
-        $prg = preg_match('/^[*][^\\r\\n]*/m', $style, $matches);
+        $prg = preg_match('/^body[^\\r\\n]*/m', $style, $matches);
         self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
-        self::assertEquals('* {font-family: \'Courier New\', monospace; font-size: 12pt; color: #000000;}', $matches[0]);
+        self::assertSame(1, $prg);
+        self::assertEquals('body {font-family: \'Courier New\', monospace; font-size: 12pt; color: #000000;}', $matches[0]);
         $prg = preg_match('/^[.]style1[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style1 {font-family: \'Tahoma\'; font-size: 10pt; color: #1B2232; font-weight: bold;}', $matches[0]);
         $prg = preg_match('/^[.]style2[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style2 {font-family: \'Arial\', sans-serif; font-size: 10pt;}', $matches[0]);
         $prg = preg_match('/^[.]style3[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style3 {font-family: \'DejaVu Sans Monospace\', monospace; font-size: 10pt;}', $matches[0]);
         $prg = preg_match('/^[.]style4[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style4 {font-family: \'Arial\'; font-size: 10pt;}', $matches[0]);
     }
 
@@ -274,23 +261,19 @@ public function testWhiteSpace(): void
         $xpath = new DOMXPath($dom);
 
         $style = Helper::getTextContent($xpath, '/html/head/style');
-        self::assertNotFalse(preg_match('/^[*][^\\r\\n]*/m', $style, $matches));
-        self::assertEquals('* {font-family: \'Arial\'; font-size: 12pt; color: #000000; white-space: pre-wrap;}', $matches[0]);
+        self::assertNotFalse(preg_match('/^body[^\\r\\n]*/m', $style, $matches));
+        self::assertEquals('body {font-family: \'Arial\'; font-size: 12pt; color: #000000; white-space: pre-wrap;}', $matches[0]);
         $prg = preg_match('/^[.]style1[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style1 {font-family: \'Courier New\'; font-size: 10pt; white-space: pre-wrap;}', $matches[0]);
         $prg = preg_match('/^[.]style2[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style2 {font-family: \'Courier New\'; font-size: 10pt;}', $matches[0]);
         $prg = preg_match('/^[.]style3[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style3 {font-family: \'Courier New\'; font-size: 10pt; white-space: normal;}', $matches[0]);
         $prg = preg_match('/^[.]style4[^\\r\\n]*/m', $style, $matches);
-        self::assertNotEmpty($matches);
-        self::assertNotFalse($prg);
+        self::assertSame(1, $prg);
         self::assertEquals('.style4 {font-family: \'Courier New\'; font-size: 10pt;}', $matches[0]);
     }
 

tests/PhpWordTests/Writer/HTML/Helper.php

@@ -67,7 +67,7 @@ public static function getNamedItem(DOMXPath $xpath, string $query, string $name
             if ($item2 === null) {
                 self::fail('Unexpected null return requesting item');
             } else {
-                $returnValue = $item2->attributes->getNamedItem($namedItem);
+                $returnVal = $item2->attributes->getNamedItem($namedItem);
             }
         }
 
@@ -125,4 +125,13 @@ public static function getAsHTML(PhpWord $phpWord, string $defaultWhiteSpace = '
 
         return $dom;
     }
+
+    public static function getHtmlString(PhpWord $phpWord, string $defaultWhiteSpace = '', string $defaultGenericFont = ''): string
+    {
+        $htmlWriter = new HTML($phpWord);
+        $htmlWriter->setDefaultWhiteSpace($defaultWhiteSpace);
+        $htmlWriter->setDefaultGenericFont($defaultGenericFont);
+
+        return $htmlWriter->getContent();
+    }
 }

tests/PhpWordTests/Writer/HTML/PartTest.php

@@ -185,5 +185,9 @@ public function testTitleStyles(): void
         self::assertNotFalse(strpos($style, 'h2 {margin-top: 0.25pt; margin-bottom: 0.25pt;}'));
         self::assertEquals(1, Helper::getLength($xpath, '/html/body/div/h1'));
         self::assertEquals(2, Helper::getLength($xpath, '/html/body/div/h2'));
+        $html = Helper::getHtmlString($phpWord);
+        self::assertStringContainsString('<h1 style="font-family: \'Calibri\'; font-weight: bold;">Header 1 #1</h1>', $html);
+        self::assertStringContainsString('<h2 style="font-family: \'Times New Roman\'; font-style: italic;">Header 2 #1</h2>', $html);
+        self::assertStringContainsString('<h2 style="font-family: \'Times New Roman\'; font-style: italic;">Header 2 #2</h2>', $html);
     }
 }