Ensure to apply default style for paragraphs, to avoid a paragraph between 2 list is mis-guessed

Author: onizet

src/Html2OpenXml/Expressions/BlockElementExpression.cs

@@ -23,11 +23,21 @@ namespace HtmlToOpenXml.Expressions;
 /// Process the parsing of block contents (like <c>p</c>, <c>span</c>, <c>heading</c>).
 /// A block-level element always starts on a new line, and the browsers automatically add some space (a margin) before and after the element.
 /// </summary>
-class BlockElementExpression(IHtmlElement node, params OpenXmlLeafElement[]? styleProperty) : PhrasingElementExpression(node)
+class BlockElementExpression: PhrasingElementExpression
 {
-    private readonly OpenXmlLeafElement[]? defaultStyleProperties = styleProperty;
+    private readonly OpenXmlLeafElement[]? defaultStyleProperties;
     protected readonly ParagraphProperties paraProperties = new();
 
+    public BlockElementExpression(IHtmlElement node, OpenXmlLeafElement? styleProperty) : base(node)
+    {
+        if (styleProperty is not null)
+            defaultStyleProperties = [styleProperty];
+    }
+    public BlockElementExpression(IHtmlElement node, params OpenXmlLeafElement[]? styleProperty) : base(node)
+    {
+        defaultStyleProperties = styleProperty;
+    }
+
 
     /// <inheritdoc/>
     public override IEnumerable<OpenXmlElement> Interpret (ParsingContext context)

src/Html2OpenXml/Expressions/BodyExpression.cs

@@ -23,7 +23,8 @@ namespace HtmlToOpenXml.Expressions;
 /// Top parent expression, processing the <c>body</c> tag,
 /// even if it is not directly specified in the provided Html.
 /// </summary>
-sealed class BodyExpression(IHtmlElement node) : BlockElementExpression(node)
+sealed class BodyExpression(IHtmlElement node, ParagraphStyleId? defaultStyle)
+    : BlockElementExpression(node, defaultStyle)
 {
     private bool shouldRegisterTopBookmark;
 

src/Html2OpenXml/HtmlConverter.cs

@@ -127,8 +127,7 @@ public async Task ParseHeader(string html, HeaderFooterValues? headerType = null
             new ParallelOptions() { CancellationToken = cancellationToken },
             htmlStyles.GetParagraphStyle(htmlStyles.DefaultStyles.HeaderStyle));
 
-        foreach (var p in paragraphs)
-            headerPart.Header.AddChild(p);
+        headerPart.Header.Append(paragraphs);
     }
 
     /// <summary>
@@ -152,8 +151,7 @@ public async Task ParseFooter(string html, HeaderFooterValues? footerType = null
             new ParallelOptions() { CancellationToken = cancellationToken },
             htmlStyles.GetParagraphStyle(htmlStyles.DefaultStyles.FooterStyle));
 
-        foreach (var p in paragraphs)
-            footerPart.Footer.AddChild(p);
+        footerPart.Footer.Append(paragraphs);
     }
 
     /// <summary>
@@ -166,7 +164,8 @@ public async Task ParseBody(string html, CancellationToken cancellationToken = d
     {
         bodyImageLoader ??= new ImagePrefetcher<MainDocumentPart>(mainPart, webRequester);
         var paragraphs = await ParseCoreAsync(html, mainPart, bodyImageLoader,
-            new ParallelOptions() { CancellationToken = cancellationToken });
+            new ParallelOptions() { CancellationToken = cancellationToken },
+            htmlStyles.GetParagraphStyle(htmlStyles.DefaultStyles.Paragraph));
 
         if (!paragraphs.Any())
             return;
@@ -263,11 +262,9 @@ private async Task<IEnumerable<OpenXmlCompositeElement>> ParseCoreAsync(string h
 
         Expressions.HtmlDomExpression expression;
         if (hostingPart is MainDocumentPart)
-            expression = new Expressions.BodyExpression(htmlDocument.Body!);
-        else if (defaultParagraphStyleId?.Val?.HasValue == true)
-            expression = new Expressions.BlockElementExpression(htmlDocument.Body!, defaultParagraphStyleId);
+            expression = new Expressions.BodyExpression(htmlDocument.Body!, defaultParagraphStyleId);
         else
-            expression = new Expressions.BlockElementExpression(htmlDocument.Body!);
+            expression = new Expressions.BlockElementExpression(htmlDocument.Body!, defaultParagraphStyleId);
 
         var parsingContext = new ParsingContext(this, hostingPart, imageLoader);
         var paragraphs = expression.Interpret(parsingContext);

src/Html2OpenXml/PredefinedStyles.cs

@@ -24,6 +24,7 @@ internal class PredefinedStyles
     public const string TableGrid = "TableGrid";
     public const string Header = "Header";
     public const string Footer = "Footer";
+    public const string Paragraph = "Normal";
 
 
 

src/Html2OpenXml/Primitives/DefaultStyles.cs

@@ -101,4 +101,10 @@ public class DefaultStyles
     /// </summary>
     /// <value>Footer</value>
     public string FooterStyle { get; set; } = PredefinedStyles.Footer;
+
+    /// <summary>
+    /// Default style for body paragraph.
+    /// </summary>
+    /// <value>Normal</value>
+    public string Paragraph { get; set; } = PredefinedStyles.Paragraph;
 }

src/Html2OpenXml/WordDocumentStyle.cs

@@ -53,7 +53,8 @@ internal WordDocumentStyle(MainDocumentPart mainPart)
             PredefinedStyles.ListParagraph,
             PredefinedStyles.Quote,
             PredefinedStyles.QuoteChar,
-            PredefinedStyles.TableGrid
+            PredefinedStyles.TableGrid,
+            PredefinedStyles.Paragraph
         ];
         this.mainPart = mainPart;
     }

test/HtmlToOpenXml.Tests/BodyTests.cs

@@ -12,9 +12,11 @@ public class BodyTests : HtmlConverterTestBase
     {
         [TestCase("landscape", ExpectedResult = true)]
         [TestCase("portrait", ExpectedResult = false)]
-        public bool PageOrientation_ReturnsLandscapeDimension(string orientation)
+        public async Task<bool> PageOrientation_ReturnsLandscapeDimension(string orientation)
         {
-            var _ = converter.Parse($@"<body style=""page-orientation:{orientation}""><body>");
+            await converter.ParseBody($@"<body style=""page-orientation:{orientation}""><body>");
+            AssertThatOpenXmlDocumentIsValid();
+
             var sectionProperties = mainPart.Document.Body!.GetFirstChild<SectionProperties>();
             Assert.That(sectionProperties, Is.Not.Null);
             var pageSize = sectionProperties.GetFirstChild<PageSize>();
@@ -24,7 +26,7 @@ public bool PageOrientation_ReturnsLandscapeDimension(string orientation)
 
         [TestCase("portrait", ExpectedResult = true)]
         [TestCase("landscape", ExpectedResult = false)]
-        public bool PageOrientation_OverrideExistingLayout_ReturnsLandscapeDimension(string orientation)
+        public async Task<bool> PageOrientation_OverrideExistingLayout_ReturnsLandscapeDimension(string orientation)
         {
             using var generatedDocument = new MemoryStream();
             using (var buffer = ResourceHelper.GetStream("Resources.DocWithLandscape.docx"))
@@ -35,7 +37,9 @@ public bool PageOrientation_OverrideExistingLayout_ReturnsLandscapeDimension(str
             MainDocumentPart mainPart = package.MainDocumentPart!;
             HtmlConverter converter = new(mainPart);
 
-            var _ = converter.Parse($@"<body style=""page-orientation:{orientation}""><body>");
+            await converter.ParseBody($@"<body style=""page-orientation:{orientation}""><body>");
+            AssertThatOpenXmlDocumentIsValid();
+
             var sectionProperties = mainPart.Document.Body!.GetFirstChild<SectionProperties>();
             Assert.That(sectionProperties, Is.Not.Null);
             var pageSize = sectionProperties.GetFirstChild<PageSize>();

test/HtmlToOpenXml.Tests/HeaderFooterTests.cs

@@ -109,5 +109,48 @@ public async Task WithExistingHeader_Even_ReturnsAnotherHeaderPart()
             });
             AssertThatOpenXmlDocumentIsValid();
         }
+
+        [Test]
+        public async Task Header_ReturnsStyleParagraphs()
+        {
+            await converter.ParseHeader(@"
+                <header>
+                <p>Placeholder
+                    <nav>
+                        <ul><li>Home</li><li>Contact</li></ul>
+                    </nav>
+                </p>
+                </header>
+            ");
+
+            var header = mainPart.HeaderParts.FirstOrDefault()?.Header;
+            Assert.That(header, Is.Not.Null);
+            var paragraphs = header.Elements<Paragraph>();
+            Assert.That(paragraphs.Count(), Is.EqualTo(3));
+            Assert.That(paragraphs.First().ParagraphProperties?.ParagraphStyleId?.Val?.Value,
+                Is.EqualTo(converter.HtmlStyles.DefaultStyles.HeaderStyle));
+            Assert.That(paragraphs.Skip(1).Select(p => p.ParagraphProperties?.ParagraphStyleId?.Val?.Value),
+                Has.All.EqualTo(converter.HtmlStyles.DefaultStyles.ListParagraphStyle));
+        }
+
+        [Test]
+        public async Task Footer_ReturnsStyleParagraphs()
+        {
+            await converter.ParseFooter(@"
+                <footer>
+                <p>
+                    <a rel=""license"" href=""https://creativecommons.org/licenses/by/4.0/"">Copyrighted but you can use what's here as long as you credit me</a>
+                    <small>&copy; Copyright 2058, Company Inc.</small>
+                </p>
+                </footer>
+            ");
+
+            var footer = mainPart.FooterParts.FirstOrDefault()?.Footer;
+            Assert.That(footer, Is.Not.Null);
+            var paragraphs = footer.Elements<Paragraph>();
+            Assert.That(paragraphs.Count(), Is.EqualTo(2));
+            Assert.That(paragraphs.Select(p => p.ParagraphProperties?.ParagraphStyleId?.Val?.Value),
+                Has.All.EqualTo(converter.HtmlStyles.DefaultStyles.FooterStyle));
+        }
     }
 }