How to: Retrieve Paragraphs from an Office Open XML Document

This topic presents an example that opens an Office Open XML document, and retrieves a collection of all of the paragraphs in the document.

For more information on Office Open XML, see www.openxmldeveloper.org.

Example

This example opens an Office Open XML package, uses the relationships within the Open XML package to find the document and the style parts. It then queries the document, projecting a collection of an anonymous type that contains the paragraph XElement node, the style name of each paragraph, and the text of each paragraph.

The example uses an extension method named StringConcatenate, which is also supplied in the example.

For a detailed tutorial that explains how this example works, see Pure Functional Transformations of XML.

This example uses classes found in the WindowsBase assembly. It uses types in the System.IO.Packaging namespace.

public static class LocalExtensions
{
    public static string StringConcatenate(this IEnumerable<string> source)
    {
        StringBuilder sb = new StringBuilder();
        foreach (string s in source)
            sb.Append(s);
        return sb.ToString();
    }

    public static string StringConcatenate<T>(this IEnumerable<T> source,
        Func<T, string> func)
    {
        StringBuilder sb = new StringBuilder();
        foreach (T item in source)
            sb.Append(func(item));
        return sb.ToString();
    }

    public static string StringConcatenate(this IEnumerable<string> source, string separator)
    {
        StringBuilder sb = new StringBuilder();
        foreach (string s in source)
            sb.Append(s).Append(separator);
        return sb.ToString();
    }

    public static string StringConcatenate<T>(this IEnumerable<T> source,
        Func<T, string> func, string separator)
    {
        StringBuilder sb = new StringBuilder();
        foreach (T item in source)
            sb.Append(func(item)).Append(separator);
        return sb.ToString();
    }
}

class Program
{
    public static string ParagraphText(XElement e)
    {
        XNamespace w = e.Name.Namespace;
        return e
               .Elements(w + "r")
               .Elements(w + "t")
               .StringConcatenate(element => (string)element);
    }

    static void Main(string[] args)
    {
        const string fileName = "SampleDoc.docx";

        const string documentRelationshipType =
          "https://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument";
        const string stylesRelationshipType =
          "https://schemas.openxmlformats.org/officeDocument/2006/relationships/styles";
        const string wordmlNamespace =
          "https://schemas.openxmlformats.org/wordprocessingml/2006/main";
        XNamespace w = wordmlNamespace;

        XDocument xDoc = null;
        XDocument styleDoc = null;

        using (Package wdPackage = Package.Open(fileName, FileMode.Open, FileAccess.Read))
        {
            PackageRelationship docPackageRelationship =
              wdPackage
              .GetRelationshipsByType(documentRelationshipType)
              .FirstOrDefault();
            if (docPackageRelationship != null)
            {
                Uri documentUri =
                    PackUriHelper
                    .ResolvePartUri(
                       new Uri("/", UriKind.Relative),
                             docPackageRelationship.TargetUri);
                PackagePart documentPart =
                    wdPackage.GetPart(documentUri);

                //  Load the document XML in the part into an XDocument instance.
                xDoc = XDocument.Load(XmlReader.Create(documentPart.GetStream()));

                //  Find the styles part. There will only be one.
                PackageRelationship styleRelation =
                  documentPart.GetRelationshipsByType(stylesRelationshipType)
                  .FirstOrDefault();
                if (styleRelation != null)
                {
                    Uri styleUri = PackUriHelper.ResolvePartUri(documentUri, styleRelation.TargetUri);
                    PackagePart stylePart = wdPackage.GetPart(styleUri);

                    //  Load the style XML in the part into an XDocument instance.
                    styleDoc = XDocument.Load(XmlReader.Create(stylePart.GetStream()));
                }
            }
        }

        string defaultStyle =
            (string)(
                from style in styleDoc.Root.Elements(w + "style")
                where (string)style.Attribute(w + "type") == "paragraph" &&
                      (string)style.Attribute(w + "default") == "1"
                select style
            ).First().Attribute(w + "styleId");

        // Find all paragraphs in the document.
        var paragraphs =
            from para in xDoc
                         .Root
                         .Element(w + "body")
                         .Descendants(w + "p")
            let styleNode = para
                            .Elements(w + "pPr")
                            .Elements(w + "pStyle")
                            .FirstOrDefault()
            select new
            {
                ParagraphNode = para,
                StyleName = styleNode != null ?
                    (string)styleNode.Attribute(w + "val") :
                    defaultStyle
            };

        // Retrieve the text of each paragraph.
        var paraWithText =
            from para in paragraphs
            select new
            {
                ParagraphNode = para.ParagraphNode,
                StyleName = para.StyleName,
                Text = ParagraphText(para.ParagraphNode)
            };

        foreach (var p in paraWithText)
            Console.WriteLine("StyleName:{0} >{1}<", p.StyleName, p.Text);
    }
}

Imports <xmlns:w="https://schemas.openxmlformats.org/wordprocessingml/2006/main">

Module Module1
    <System.Runtime.CompilerServices.Extension()> _
    Public Function StringConcatenate(ByVal source As IEnumerable(Of String)) As String
        Dim sb As StringBuilder = New StringBuilder()
        For Each s As String In source
            sb.Append(s)
        Next
        Return sb.ToString()
    End Function

    <System.Runtime.CompilerServices.Extension()> _
    Public Function StringConcatenate(Of T)(ByVal source As IEnumerable(Of T), _
    ByVal func As Func(Of T, String)) As String
        Dim sb As StringBuilder = New StringBuilder()
        For Each item As T In source
            sb.Append(func(item))
        Next
        Return sb.ToString()
    End Function

    <System.Runtime.CompilerServices.Extension()> _
    Public Function StringConcatenate(Of T)(ByVal source As IEnumerable(Of T), _
    ByVal separator As String) As String
        Dim sb As StringBuilder = New StringBuilder()
        For Each s As T In source
            sb.Append(s).Append(separator)
        Next
        Return sb.ToString()
    End Function

    <System.Runtime.CompilerServices.Extension()> _
    Public Function StringConcatenate(Of T)(ByVal source As IEnumerable(Of T), _
    ByVal func As Func(Of T, String), ByVal separator As String) As String
        Dim sb As StringBuilder = New StringBuilder()
        For Each item As T In source
            sb.Append(func(item)).Append(separator)
        Next
        Return sb.ToString()
    End Function

    Public Function ParagraphText(ByVal e As XElement) As String
        Dim w As XNamespace = e.Name.Namespace
        Return (e.<w:r>.<w:t>).StringConcatenate(Function(element) CStr(element))
    End Function

    ' Following function is required because VB does not support short circuit evaluation
    Private Function GetStyleOfParagraph(ByVal styleNode As XElement, ByVal defaultStyle As String) _
                As String
        If (styleNode Is Nothing) Then
            Return defaultStyle
        Else
            Return styleNode.@w:val
        End If
    End Function

    Sub Main()
        Dim fileName = "SampleDoc.docx"

        Dim documentRelationshipType = _
          "https://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument"
        Dim stylesRelationshipType = _
          "https://schemas.openxmlformats.org/officeDocument/2006/relationships/styles"
        Dim wordmlNamespace = _
          "https://schemas.openxmlformats.org/wordprocessingml/2006/main"

        Dim xDoc As XDocument = Nothing
        Dim styleDoc As XDocument = Nothing
        Using wdPackage As Package = Package.Open(fileName, FileMode.Open, FileAccess.Read)
            Dim docPackageRelationship As PackageRelationship = _
              wdPackage _
              .GetRelationshipsByType(documentRelationshipType) _
              .FirstOrDefault()
            If (docPackageRelationship IsNot Nothing) Then
                Dim documentUri As Uri = _
                  PackUriHelper _
                  .ResolvePartUri(New Uri("/", UriKind.Relative), docPackageRelationship.TargetUri)
                Dim documentPart As PackagePart = wdPackage.GetPart(documentUri)

                '  Load the document XML in the part into an XDocument instance.
                xDoc = XDocument.Load(XmlReader.Create(documentPart.GetStream()))

                '  Find the styles part. There will only be one.
                Dim styleRelation As PackageRelationship = documentPart.GetRelationshipsByType(stylesRelationshipType).FirstOrDefault()
                If (styleRelation IsNot Nothing) Then
                    Dim styleUri As Uri = PackUriHelper.ResolvePartUri(documentUri, styleRelation.TargetUri)
                    Dim stylePart As PackagePart = wdPackage.GetPart(styleUri)

                    '  Load the style XML in the part into an XDocument instance.
                    styleDoc = XDocument.Load(XmlReader.Create(stylePart.GetStream()))
                End If
            End If
        End Using

        Dim defaultStyle As String = _
            ( _
                From style In styleDoc.Root.<w:style> _
                Where style.@w:type = "paragraph" And _
                      style.@w:default = "1" _
                Select style _
            ).First().@w:styleId

        ' Find all paragraphs in the document.
        Dim paragraphs = _
            From para In xDoc.Root.<w:body>...<w:p> _
        Let styleNode As XElement = para.<w:pPr>.<w:pStyle>.FirstOrDefault _
        Select New With { _
            .ParagraphNode = para, _
            .StyleName = GetStyleOfParagraph(styleNode, defaultStyle) _
        }

        ' Retrieve the text of each paragraph.
        Dim paraWithText = _
            From para In paragraphs _
            Select New With { _
                .ParagraphNode = para.ParagraphNode, _
                .StyleName = para.StyleName, _
                .Text = ParagraphText(para.ParagraphNode) _
            }

        For Each p In paraWithText
            Console.WriteLine("StyleName:{0} >{1}<", p.StyleName, p.Text)
        Next
    End Sub
End Module

When run with the sample Open XML document described in Creating the Source Office Open XML Document, this example produces the following output:

StyleName:Heading1 >Parsing WordprocessingML with LINQ to XML<
StyleName:Normal ><
StyleName:Normal >The following example prints to the console.<
StyleName:Normal ><
StyleName:Code >using System;<
StyleName:Code ><
StyleName:Code >class Program {<
StyleName:Code >    public static void (string[] args) {<
StyleName:Code >        Console.WriteLine("Hello World");<
StyleName:Code >    }<
StyleName:Code >}<
StyleName:Normal ><
StyleName:Normal >This example produces the following output:<
StyleName:Normal ><
StyleName:Code >Hello World<

See Also

Concepts

Advanced Query Techniques (LINQ to XML)

Build Date:

2012-08-02