Merge pull request #378 from Roave/replace-code-blocks-with-confluenc…

…e-code-widgets Added Confluence writer method to replace code blocks with Confluence code widgets
Roave · Apr 23, 2024 · 42accd4 · 42accd4
2 parents 07eb4db + 81a596a
commit 42accd4
Show file tree

Hide file tree

Showing 2 changed files with 176 additions and 0 deletions.
diff --git a/src/Writer/ConfluenceWriter.php b/src/Writer/ConfluenceWriter.php
@@ -20,6 +20,7 @@
 use function array_merge;
 use function dirname;
 use function hash_equals;
+use function html_entity_decode;
 use function in_array;
 use function md5;
 use function preg_replace_callback;
@@ -35,6 +36,37 @@
 /** @psalm-type ListOfExtractedImageData = list<array{hashFilename: string, data: string}> */
 final class ConfluenceWriter implements OutputWriter
 {
+    /** @link https://confluence.atlassian.com/doc/code-block-macro-139390.html */
+    private const ALLOWED_CONFLUENCE_CODE_FORMATS = [
+        'actionscript',
+        'applescript',
+        'bash',
+        'csharp',
+        'coldfusion',
+        'cpp',
+        'css',
+        'delphi',
+        'diff',
+        'erlang',
+        'groovy',
+        'html',
+        'java',
+        'javafx',
+        'javascript',
+        'none',
+        'perl',
+        'php',
+        'powershell',
+        'python',
+        'ruby',
+        'sass',
+        'scala',
+        'sql',
+        'xml',
+        'vb',
+        'yaml',
+    ];
+
     private const CONFLUENCE_HEADER = '<p><strong style="color: #ff0000;">NOTE: This documentation is auto generated, do not edit this directly in Confluence, as your changes will be overwritten!</strong></p>';
 
     private readonly string $confluenceContentApiUrl;
@@ -87,6 +119,7 @@ public function __invoke(array $docbookPages): void
             );
 
             $confluenceContent = $this->replaceLocalMarkdownLinks($page, $mapPathsToConfluencePageIds, $confluenceContent);
+            $confluenceContent = $this->replaceCodeBlocks($confluenceContent);
 
             $hashUpdateMethod  = 'POST';
             $latestContentHash = md5($confluenceContent);
@@ -280,6 +313,38 @@ function (array $m) use ($currentPagePath, $mapPathsToConfluencePageIds): string
         );
     }
 
+    private function replaceCodeBlocks(string $renderedContent): string
+    {
+        return (string) preg_replace_callback(
+            '/<pre><code(?: class="lang-([^"]+)"|)>([^<]+)<\/code><\/pre>/',
+            static function (array $m): string {
+                /** @var array{1: string, 2: string} $m */
+                $confluenceCodeLanguage = match ($m[1]) {
+                    'json', 'js' => 'javascript',
+                    'shell' => 'bash',
+                    default => $m[1]
+                };
+
+                if (! in_array($confluenceCodeLanguage, self::ALLOWED_CONFLUENCE_CODE_FORMATS, true)) {
+                    $confluenceCodeLanguage = 'none';
+                }
+
+                return sprintf(
+                    <<<'XML'
+<ac:structured-macro ac:name="code" ac:schema-version="1">
+  <ac:parameter ac:name="language">%s</ac:parameter>
+  <ac:plain-text-body><![CDATA[%s]]>
+  </ac:plain-text-body>
+</ac:structured-macro>
+XML,
+                    $confluenceCodeLanguage,
+                    html_entity_decode($m[2]), // Since this is rendered in CDATA, we should not escape HTML entities
+                );
+            },
+            $renderedContent,
+        );
+    }
+
     /**
      * @param array<array-key, mixed>|null                                                        $bodyContent
      * @param array<string, string>                                                               $overrideHeaders

diff --git a/test/unit/Writer/ConfluenceWriterTest.php b/test/unit/Writer/ConfluenceWriterTest.php
@@ -379,4 +379,115 @@ public function testConfluenceUploadWithLinksToOtherPages(): void
             $this->testLogger->logMessages,
         );
     }
+
+    public function testConfluenceUploadReplacesCodeBlocks(): void
+    {
+        $guzzleLog = [];
+
+        $handlerStack = HandlerStack::create(new MockHandler([
+            // GET /{pageId}
+            0 => new Response(200, [], json_encode([
+                'id' => '123456789',
+                'type' => 'page type',
+                'title' => 'page title',
+                'space' => ['key' => 'space key'],
+                'version' => ['number' => '1'],
+            ], JSON_THROW_ON_ERROR)),
+            // GET /{pageId}/child/attachment
+            1 => new Response(200, [], json_encode([
+                'results' => [
+                    ['title' => 'attachment'],
+                ],
+            ], JSON_THROW_ON_ERROR)),
+            // PUT /{pageId}
+            2 => new Response(200, [], json_encode([], JSON_THROW_ON_ERROR)),
+        ]));
+        $handlerStack->push(Middleware::history($guzzleLog));
+
+        $confluence = new ConfluenceWriter(
+            new Client(['handler' => $handlerStack]),
+            'https://fake-confluence-url',
+            'Something',
+            $this->testLogger,
+            true,
+        );
+
+        $confluence->__invoke([
+            DocbookPage::fromSlugAndContent(
+                '/fake/path',
+                'page-slug',
+                <<<'HTML'
+<p>Regular text before.</p>
+<pre><code>just some plaintext code</code></pre>
+<pre><code class="lang-json">{
+    "some": true,
+    "json": 123
+}
+</code></pre>
+<pre><code class="lang-shell">make build</code></pre>
+<pre><code class="lang-gobbledygook">this gobbledygook language is not a supported Confluence language</code></pre>
+<pre><code class="lang-html">&lt;html&gt;
+&lt;body&gt;
+&lt;pre&gt;&lt;code&gt;Example of some HTML to trip up the regex&lt;/code&gt;&lt;/pre&gt;
+&lt;/body&gt;
+&lt;/html&gt;
+</code></pre>
+<p>Regular text after.</p>
+HTML,
+            )->withFrontMatter(['confluencePageId' => 123456789]),
+        ]);
+
+        /** @psalm-var array<self::*,array{request:RequestInterface}> $guzzleLog */
+
+        $postedPageContent = $guzzleLog[2]['request'];
+        assert($postedPageContent instanceof RequestInterface);
+        $this->assertPostContentRequestWasCorrect(
+            $postedPageContent,
+            123456789,
+            <<<'HTML'
+<p>Regular text before.</p>
+<ac:structured-macro ac:name="code" ac:schema-version="1">
+  <ac:parameter ac:name="language">none</ac:parameter>
+  <ac:plain-text-body><![CDATA[just some plaintext code]]>
+  </ac:plain-text-body>
+</ac:structured-macro>
+<ac:structured-macro ac:name="code" ac:schema-version="1">
+  <ac:parameter ac:name="language">javascript</ac:parameter>
+  <ac:plain-text-body><![CDATA[{
+    "some": true,
+    "json": 123
+}
+]]>
+  </ac:plain-text-body>
+</ac:structured-macro>
+<ac:structured-macro ac:name="code" ac:schema-version="1">
+  <ac:parameter ac:name="language">bash</ac:parameter>
+  <ac:plain-text-body><![CDATA[make build]]>
+  </ac:plain-text-body>
+</ac:structured-macro>
+<ac:structured-macro ac:name="code" ac:schema-version="1">
+  <ac:parameter ac:name="language">none</ac:parameter>
+  <ac:plain-text-body><![CDATA[this gobbledygook language is not a supported Confluence language]]>
+  </ac:plain-text-body>
+</ac:structured-macro>
+<ac:structured-macro ac:name="code" ac:schema-version="1">
+  <ac:parameter ac:name="language">html</ac:parameter>
+  <ac:plain-text-body><![CDATA[<html>
+<body>
+<pre><code>Example of some HTML to trip up the regex</code></pre>
+</body>
+</html>
+]]>
+  </ac:plain-text-body>
+</ac:structured-macro>
+<p>Regular text after.</p>
+HTML,
+            2,
+        );
+
+        self::assertContains(
+            sprintf('[%s] - OK! Successfully updated confluence page 123456789 with page-slug ...', ConfluenceWriter::class),
+            $this->testLogger->logMessages,
+        );
+    }
 }