From 0199631e02dfcc395990379ca765d9ba44485ea3 Mon Sep 17 00:00:00 2001
From: Dominic Tubach <tubach@systopia.de>
Date: Thu, 20 Jun 2024 09:54:06 +0200
Subject: [PATCH] Fix handling of styles

---
 .../LocalUnoconv/PhpWordTemplateProcessor.php |  32 +--
 Civi/Civioffice/PhpWord/StyleMerger.php       |  95 ++++++++
 ...alUnoconv_PhpWordTemplateProcessorTest.php | 230 +++++++++++++++++-
 .../Civioffice/PhpWord/StyleMergerTest.php    |  64 +++++
 tests/phpunit/bootstrap.php                   |   9 +
 5 files changed, 413 insertions(+), 17 deletions(-)
 create mode 100644 Civi/Civioffice/PhpWord/StyleMerger.php
 create mode 100644 tests/phpunit/Civi/Civioffice/PhpWord/StyleMergerTest.php

diff --git a/CRM/Civioffice/DocumentRendererType/LocalUnoconv/PhpWordTemplateProcessor.php b/CRM/Civioffice/DocumentRendererType/LocalUnoconv/PhpWordTemplateProcessor.php
index 5e533b6..4a4b973 100644
--- a/CRM/Civioffice/DocumentRendererType/LocalUnoconv/PhpWordTemplateProcessor.php
+++ b/CRM/Civioffice/DocumentRendererType/LocalUnoconv/PhpWordTemplateProcessor.php
@@ -15,6 +15,7 @@
 
 declare(strict_types = 1);
 
+use Civi\Civioffice\PhpWord\StyleMerger;
 use CRM_Civioffice_ExtensionUtil as E;
 use PhpOffice\PhpWord;
 
@@ -111,13 +112,6 @@ public function replaceHtmlToken(string $macroVariable, string $renderedTokenMes
           // will not be removed (i.e. leave an empty paragraph).
           $this->setValue($macroVariable, '');
         }
-        elseif (count($elements) === 1 && $elements[0] instanceof PhpWord\Element\Text) {
-          // ... either as plain text (if there is only a single Text
-          // element, which can't have style properties), ...
-          // Note: $rendered_token_message shouldn't be used directly
-          // because it may contain HTML entities.
-          $this->setValue($macroVariable, $elements[0]->getText());
-        }
         else {
           // ... or as HTML: Render all elements and insert in the text
           // run or paragraph containing the macro.
@@ -145,14 +139,14 @@ public function replaceHtmlToken(string $macroVariable, string $renderedTokenMes
    *
    * @param \PhpOffice\PhpWord\Element\AbstractElement[] $elements
    * @param bool $inheritStyle
-   *   If TRUE and an element contains no style, it will be inherited from the
-   *   paragraph/text run the macro is inside.
+   *   If TRUE the style will be inherited from the paragraph/text run the macro
+   *   is inside. If the element already contains styles, they will be merged.
    *
    * @throws \PhpOffice\PhpWord\Exception\Exception
    */
   public function setElementsValue(string $search, array $elements, bool $inheritStyle = FALSE): void {
     $search = static::ensureMacroCompleted($search);
-    $elementsData = '';
+    $elementsDataList = [];
     $hasParagraphs = FALSE;
     foreach ($elements as $element) {
       $elementName = substr(
@@ -169,7 +163,7 @@ public function setElementsValue(string $search, array $elements, bool $inheritS
       /** @var \PhpOffice\PhpWord\Writer\Word2007\Element\AbstractElement $elementWriter */
       $elementWriter = new $objectClass($xmlWriter, $element, !$withParagraph);
       $elementWriter->write();
-      $elementsData .= $xmlWriter->getData();
+      $elementsDataList[] = preg_replace('/>\s+</', '><', $xmlWriter->getData());
     }
     $blockType = $hasParagraphs ? 'w:p' : 'w:r';
     $where = $this->findContainingXmlBlockForMacro($search, $blockType);
@@ -182,10 +176,17 @@ public function setElementsValue(string $search, array $elements, bool $inheritS
         ? $this->splitParagraphIntoParagraphs($block, $paragraphStyle, $textRunStyle)
         : $this->splitTextIntoTexts($block, $textRunStyle);
       if ($inheritStyle) {
-        $elementsData = str_replace(['<w:pPr/>', '<w:rPr/>'], [$paragraphStyle, $textRunStyle], $elementsData);
+        $elementsDataList = preg_replace_callback_array([
+          '#<w:pPr/>#' => fn() => $paragraphStyle,
+          '#<w:pPr.*</w:pPr>#' => fn (array $matches) => StyleMerger::mergeStyles($matches[0], $paragraphStyle),
+          // <w:pPr> may contain <w:rPr> itself so we have to match for <w:rPr> inside of <w:r>
+          '#<w:r><w:rPr/>.*</w:r>#' => fn(array $matches) => str_replace('<w:rPr/>', $textRunStyle, $matches[0]),
+          '#<w:r>.*(<w:rPr.*</w:rPr>).*</w:r>#' => fn (array $matches) =>
+          preg_replace('#<w:rPr.*</w:rPr>#', StyleMerger::mergeStyles($matches[1], $textRunStyle), $matches[0]),
+        ], $elementsDataList);
       }
       $this->replaceXmlBlock($search, $parts, $blockType);
-      $this->replaceXmlBlock($search, $elementsData, $blockType);
+      $this->replaceXmlBlock($search, implode('', $elementsDataList), $blockType);
     }
   }
 
@@ -212,8 +213,9 @@ public function splitParagraphIntoParagraphs(
     preg_match('#<w:pPr.*</w:pPr>#i', $paragraph, $matches);
     $extractedParagraphStyle = $matches[0] ?? '';
 
-    preg_match('#<w:rPr.*</w:rPr>#i', $paragraph, $matches);
-    $extractedTextRunStyle = $matches[0] ?? '';
+    // <w:pPr> may contain <w:rPr> itself so we have to match for <w:rPr> inside of <w:r>
+    preg_match('#<w:r>.*(<w:rPr.*</w:rPr>).*</w:r>#i', $paragraph, $matches);
+    $extractedTextRunStyle = $matches[1] ?? '';
 
     $result = str_replace(
       [
diff --git a/Civi/Civioffice/PhpWord/StyleMerger.php b/Civi/Civioffice/PhpWord/StyleMerger.php
new file mode 100644
index 0000000..25ac1e1
--- /dev/null
+++ b/Civi/Civioffice/PhpWord/StyleMerger.php
@@ -0,0 +1,95 @@
+<?php
+/*
+ * Copyright (C) 2024 SYSTOPIA GmbH
+ *
+ *  This program is free software: you can redistribute it and/or modify
+ *  it under the terms of the GNU Affero General Public License as published by
+ *  the Free Software Foundation in version 3.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU Affero General Public License for more details.
+ *
+ *  You should have received a copy of the GNU Affero General Public License
+ *  along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+declare(strict_types = 1);
+
+namespace Civi\Civioffice\PhpWord;
+
+final class StyleMerger {
+
+  private \DOMElement $styleElement;
+
+  /**
+   * @phpstan-var array<string, \DOMElement>
+   */
+  private array $elements = [];
+
+  public static function mergeStyles(string $style, string ...$styles): string {
+    $styleMerger = new self($style);
+    foreach ($styles as $styleToMerge) {
+      $styleMerger->merge($styleToMerge);
+    }
+
+    return $styleMerger->getStyleString();
+  }
+
+  public function __construct(string $style) {
+    $this->styleElement = $this->createStyleElement($style);
+    foreach ($this->styleElement->childNodes as $node) {
+      if ($node instanceof \DOMElement) {
+        $this->elements[$node->tagName] = $node;
+      }
+    }
+  }
+
+  public function merge(string $style): self {
+    $styleElement = $this->createStyleElement($style);
+    foreach ($styleElement->childNodes as $node) {
+      if ($node instanceof \DOMElement) {
+        // @todo Do we need recursive merging for some elements?
+        if (!isset($this->elements[$node->tagName])) {
+          // @phpstan-ignore-next-line
+          $importedNode = $this->styleElement->ownerDocument->importNode($node, TRUE);
+          if (!$importedNode instanceof \DOMElement) {
+            throw new \RuntimeException('Importing node failed');
+          }
+
+          $this->styleElement->appendChild($importedNode);
+          $this->elements[$node->tagName] = $importedNode;
+        }
+      }
+    }
+
+    return $this;
+  }
+
+  public function getStyleString(): string {
+    // @phpstan-ignore-next-line
+    return $this->styleElement->ownerDocument->saveXML($this->styleElement);
+  }
+
+  private function createStyleElement(string $style): \DOMElement {
+    if (NULL === $style = preg_replace('/>\s+</', '><', $style)) {
+      throw new \RuntimeException('Error processing style');
+    }
+
+    $doc = new \DOMDocument();
+    $doc->loadXML(
+      '<root xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main">' . $style . '</root>'
+    );
+
+    // @phpstan-ignore-next-line
+    foreach ($doc->documentElement->childNodes as $node) {
+      if ($node instanceof \DOMElement) {
+        return $node;
+      }
+    }
+
+    throw new \RuntimeException('Could not create style element');
+  }
+
+}
diff --git a/tests/phpunit/CRM/Civioffice/DocumentRendererType/LocalUnoconv/CRM_Civioffice_DocumentRendererType_LocalUnoconv_PhpWordTemplateProcessorTest.php b/tests/phpunit/CRM/Civioffice/DocumentRendererType/LocalUnoconv/CRM_Civioffice_DocumentRendererType_LocalUnoconv_PhpWordTemplateProcessorTest.php
index b67aa6a..d8bd369 100644
--- a/tests/phpunit/CRM/Civioffice/DocumentRendererType/LocalUnoconv/CRM_Civioffice_DocumentRendererType_LocalUnoconv_PhpWordTemplateProcessorTest.php
+++ b/tests/phpunit/CRM/Civioffice/DocumentRendererType/LocalUnoconv/CRM_Civioffice_DocumentRendererType_LocalUnoconv_PhpWordTemplateProcessorTest.php
@@ -61,7 +61,19 @@ public function testReplaceSimple(): void {
         <w:rPr>
           <w:b w:val="true"/>
         </w:rPr>
-        <w:t>Foo test 123 bar</w:t>
+        <w:t xml:space="preserve">Foo </w:t>
+      </w:r>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="true"/>
+        </w:rPr>
+        <w:t xml:space="preserve">test 123</w:t>
+      </w:r>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="true"/>
+        </w:rPr>
+       <w:t xml:space="preserve"> bar</w:t>
       </w:r>
     </w:p>
   </w:body>
@@ -105,7 +117,19 @@ public function testReplaceSpan(): void {
         <w:rPr>
           <w:b w:val="true"/>
         </w:rPr>
-        <w:t>Foo test 123 bar</w:t>
+        <w:t xml:space="preserve">Foo </w:t>
+      </w:r>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="true"/>
+        </w:rPr>
+        <w:t xml:space="preserve">test 123</w:t>
+      </w:r>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="true"/>
+        </w:rPr>
+        <w:t xml:space="preserve"> bar</w:t>
       </w:r>
     </w:p>
   </w:body>
@@ -360,6 +384,208 @@ public function testReplaceParagraphEnd(): void {
     </w:p>
   </w:body>
 </w:document>
+EOD;
+
+    static::assertXmlStringEqualsXmlString($expectedMainPart, $templateProcessor->getMainPart());
+  }
+
+  public function testStrong(): void {
+    $mainPart = <<<EOD
+<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main">
+  <w:body>
+    <w:p>
+      <w:pPr>
+        <w:pStyle w:val="Normal"/>
+      </w:pPr>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="true"/>
+          <w:color w:val="FF0000"/>
+        </w:rPr>
+        <w:t>Foo {place.holder} bar</w:t>
+      </w:r>
+    </w:p>
+  </w:body>
+</w:document>
+EOD;
+
+    $templateProcessor = new TestablePhpWordTemplateProcessor($mainPart);
+    $templateProcessor->civiTokensToMacros();
+    $templateProcessor->replaceHtmlToken('place.holder', '<strong>bold</strong>');
+
+    $expectedMainPart = <<<EOD
+<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main">
+  <w:body>
+    <w:p>
+      <w:pPr>
+        <w:pStyle w:val="Normal"/>
+      </w:pPr>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="true"/>
+          <w:color w:val="FF0000"/>
+        </w:rPr>
+        <w:t xml:space="preserve">Foo </w:t>
+      </w:r>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="1"/>
+          <w:bCs w:val="1"/>
+          <w:color w:val="FF0000"/>
+        </w:rPr>
+        <w:t xml:space="preserve">bold</w:t>
+      </w:r>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="true"/>
+          <w:color w:val="FF0000"/>
+        </w:rPr>
+        <w:t xml:space="preserve"> bar</w:t>
+      </w:r>
+    </w:p>
+  </w:body>
+</w:document>
+EOD;
+
+    static::assertXmlStringEqualsXmlString($expectedMainPart, $templateProcessor->getMainPart());
+  }
+
+  /**
+   * Tests replace with a paragraph where the paragraph that contains the
+   * placeholder has a paragraph style that has a text run style.
+   */
+  public function testReplaceParagraphWithTextRunStyle(): void {
+    $mainPart = <<<EOD
+<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main">
+  <w:body>
+    <w:p>
+      <w:pPr>
+        <w:pStyle w:val="Normal"/>
+        <w:rPr>
+          <w:rFonts w:ascii="Liberation Sans" w:hAnsi="Liberation Sans"/>
+        </w:rPr>
+      </w:pPr>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="true"/>
+        </w:rPr>
+        <w:t>Foo {place.holder}</w:t>
+      </w:r>
+    </w:p>
+  </w:body>
+</w:document>
+EOD;
+
+    $templateProcessor = new TestablePhpWordTemplateProcessor($mainPart);
+    $templateProcessor->civiTokensToMacros();
+    $templateProcessor->replaceHtmlToken('place.holder', '<p>test 123</p>');
+
+    $expectedMainPart = <<<EOD
+<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main">
+  <w:body>
+    <w:p>
+      <w:pPr>
+        <w:pStyle w:val="Normal"/>
+        <w:rPr>
+          <w:rFonts w:ascii="Liberation Sans" w:hAnsi="Liberation Sans"/>
+        </w:rPr>
+      </w:pPr>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="true"/>
+        </w:rPr>
+        <w:t xml:space="preserve">Foo </w:t>
+      </w:r>
+    </w:p>
+    <w:p>
+      <w:pPr>
+        <w:pStyle w:val="Normal"/>
+        <w:rPr>
+          <w:rFonts w:ascii="Liberation Sans" w:hAnsi="Liberation Sans"/>
+        </w:rPr>
+      </w:pPr>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="true"/>
+        </w:rPr>
+        <w:t xml:space="preserve">test 123</w:t>
+      </w:r>
+    </w:p>
+  </w:body>
+</w:document>
+EOD;
+
+    static::assertXmlStringEqualsXmlString($expectedMainPart, $templateProcessor->getMainPart());
+  }
+
+  /**
+   * Tests replace with a paragraph that has a text run style where the
+   * paragraph that contains the placeholder has a paragraph style that has a
+   * text run style.
+   */
+  public function testReplaceParagraphWithRunStyleAndStrong(): void {
+    $mainPart = <<<EOD
+<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main">
+  <w:body>
+    <w:p>
+      <w:pPr>
+        <w:pStyle w:val="Normal"/>
+        <w:rPr>
+          <w:rFonts w:ascii="Liberation Sans" w:hAnsi="Liberation Sans"/>
+        </w:rPr>
+      </w:pPr>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="true"/>
+          <w:color w:val="FF0000"/>
+        </w:rPr>
+        <w:t>Foo {place.holder}</w:t>
+      </w:r>
+    </w:p>
+  </w:body>
+</w:document>
+EOD;
+
+    $templateProcessor = new TestablePhpWordTemplateProcessor($mainPart);
+    $templateProcessor->civiTokensToMacros();
+    $templateProcessor->replaceHtmlToken('place.holder', '<p><strong>test 123</strong></p>');
+
+    $expectedMainPart = <<<EOD
+<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main">
+  <w:body>
+    <w:p>
+      <w:pPr>
+        <w:pStyle w:val="Normal"/>
+        <w:rPr>
+          <w:rFonts w:ascii="Liberation Sans" w:hAnsi="Liberation Sans"/>
+        </w:rPr>
+      </w:pPr>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="true"/>
+          <w:color w:val="FF0000"/>
+        </w:rPr>
+        <w:t xml:space="preserve">Foo </w:t>
+      </w:r>
+    </w:p>
+    <w:p>
+      <w:pPr>
+        <w:pStyle w:val="Normal"/>
+        <w:rPr>
+          <w:rFonts w:ascii="Liberation Sans" w:hAnsi="Liberation Sans"/>
+        </w:rPr>
+      </w:pPr>
+      <w:r>
+        <w:rPr>
+          <w:b w:val="1"/>
+          <w:bCs w:val="1"/>
+          <w:color w:val="FF0000"/>
+        </w:rPr>
+        <w:t xml:space="preserve">test 123</w:t>
+      </w:r>
+    </w:p>
+  </w:body>
+</w:document>
 EOD;
 
     static::assertXmlStringEqualsXmlString($expectedMainPart, $templateProcessor->getMainPart());
diff --git a/tests/phpunit/Civi/Civioffice/PhpWord/StyleMergerTest.php b/tests/phpunit/Civi/Civioffice/PhpWord/StyleMergerTest.php
new file mode 100644
index 0000000..cc3a9bc
--- /dev/null
+++ b/tests/phpunit/Civi/Civioffice/PhpWord/StyleMergerTest.php
@@ -0,0 +1,64 @@
+<?php
+/*
+ * Copyright (C) 2024 SYSTOPIA GmbH
+ *
+ *  This program is free software: you can redistribute it and/or modify
+ *  it under the terms of the GNU Affero General Public License as published by
+ *  the Free Software Foundation in version 3.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU Affero General Public License for more details.
+ *
+ *  You should have received a copy of the GNU Affero General Public License
+ *  along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+declare(strict_types = 1);
+
+namespace phpunit\Civi\Civioffice\PhpWord;
+
+use Civi\Civioffice\PhpWord\StyleMerger;
+use PHPUnit\Framework\TestCase;
+
+/**
+ * @covers \Civi\Civioffice\PhpWord\StyleMerger
+ */
+final class StyleMergerTest extends TestCase {
+
+  public function testMerge(): void {
+    $style = <<<EOD
+  <w:pPr>
+    <w:pStyle w:val="Normal"/>
+    <w:keeplines/>
+  </w:pPr>
+EOD;
+
+    $styleToMerge = <<<EOD
+  <w:pPr>
+    <w:pStyle w:val="Test"/>
+    <w:numPr>
+      <w:ilvl w:val="0"/>
+      <w:numId w:val="1"/>
+    </w:numPr>
+  </w:pPr>
+EOD;
+
+    $expectedStyle = <<<EOD
+  <w:pPr>
+    <w:pStyle w:val="Normal"/>
+    <w:keeplines/>
+    <w:numPr>
+      <w:ilvl w:val="0"/>
+      <w:numId w:val="1"/>
+    </w:numPr>
+  </w:pPr>
+EOD;
+
+    $styleMerger = new StyleMerger($style);
+    $styleMerger->merge($styleToMerge);
+    static::assertXmlStringEqualsXmlString($expectedStyle, $styleMerger->getStyleString());
+  }
+
+}
diff --git a/tests/phpunit/bootstrap.php b/tests/phpunit/bootstrap.php
index fd5ce30..2e87718 100644
--- a/tests/phpunit/bootstrap.php
+++ b/tests/phpunit/bootstrap.php
@@ -18,6 +18,15 @@
 
 $loader->register();
 
+// Make CRM_Civioffice_ExtensionUtil available.
+require_once __DIR__ . '/../../civioffice.civix.php';
+
+if (!function_exists('ts')) {
+  // Ensure function ts() is available - it's declared in the same file as CRM_Core_I18n in CiviCRM < 5.74.
+  // In later versions the function is registered following the composer conventions.
+  \CRM_Core_I18n::singleton();
+}
+
 /**
  * Call the "cv" command.
  *