Develop Reference

OpenXML Word Document - Unreadable content found after altchunk

I build a small application using .NET 6 that is replacing values inside a word document and saving a copy.
Some keys are replaced with other files content using an AltChunk.
Using file A, in which I merge AltChunk1, the output is working fine.
Using file B with same AltChunk1, the output produce the error "found unreadable content" when opening with Word.
Using file B and a different AltChunk file (even the same after I trimmed it) can, in some cases, work.
I don't have any clue what the issue might be.
I tried comparing files using OpenXML productivity tool however:
File A and File B have a lot of differences, it is really hard to find anything that would explain this behavior
They are identical in the place the AltChunk is put.
Tried comparing the not working result with what word is creating with a repair but word is not keeping the AltChunk, it completely merges content of AltChunk with File B making any comparison almost impossible with my non-working result.
Here is the code Is use:
First method is creating the AltChunk from file, then calls methods used to replace "keys" with the wanted value (including case where the key is split accross various runs)
internal static void MergeOutSideDocument(string key, string filePath, IEnumerable<string> outsideDocs)
{
if (string.IsNullOrEmpty(key)) throw new ArgumentException("Cannot replace empty key.");
if (!File.Exists(filePath) || outsideDocs.Any(path => !File.Exists(path))) throw new FileNotFoundException();
using WordprocessingDocument doc = WordprocessingDocument.Open(filePath, true);
List<OpenXmlElement> altChunks = new();
foreach (var outsideDoc in outsideDocs)
{
var existingIds = doc.MainDocumentPart.Document.Body.Descendants<AltChunk>();
string altChunkId = "AltChunkId" + DateTime.Now.Ticks.ToString();
MainDocumentPart mainPart = doc.MainDocumentPart;
AlternativeFormatImportPart chunk = mainPart.AddAlternativeFormatImportPart(
AlternativeFormatImportPartType.WordprocessingML, altChunkId);
using (FileStream fileStream = File.Open(outsideDoc, FileMode.Open))
chunk.FeedData(fileStream);
altChunks.Add(new AltChunk()
{
Id = altChunkId
});
inMemoryAltChunkIds.Add(altChunkId);
}
var body = doc.MainDocumentPart.Document.Body;
SetElementForKey(key, altChunks,
body.Descendants<Paragraph>().First(par => par.Contains(key)),
body);
}
private static void SetElementForKey(string key, List<OpenXmlElement> replacements, OpenXmlElement el, Body body)
{
List<Run> previousRuns = new();
if (el?.InnerText.Contains(key) != true) return;
for (int i = 0; i <= el.Descendants<Run>().Count(); i++)
{
var innerText = string.Join("", previousRuns.Select(r => r.InnerText));
if (innerText.Contains(key))
{
var usedRuns = GetRequiredRunsForText(previousRuns, key);
var firstRun = usedRuns.First();
MergeRunsWithKey(key, usedRuns, firstRun);
var usedRun = usedRuns.First();
var firstPart = usedRun.InnerText.IndexOf(key) != -1 ? usedRun.InnerText[..usedRun.InnerText.IndexOf(key)] : "";
ReplaceText(key, "", usedRun);
foreach (var replacement in replacements)
el.Parent.InsertAfter(replacement, el);
if (string.IsNullOrEmpty(usedRun.InnerText)) usedRun.Remove();
if (string.IsNullOrEmpty(el.InnerText)) el.Remove();
break;
}
else
{
previousRuns.Add(el.Descendants<Run>().ElementAt(i));
}
}
}
private static void MergeRunsWithKey(string key, List<Run> usedRuns, Run firstRun)
{
while (!usedRuns.First().InnerText.Contains(key))
{
AddText(usedRuns.Skip(1).First().InnerText, firstRun);
usedRuns.Skip(1).First().Remove();
usedRuns.RemoveAt(1);
}
}
private static void AddText(string newText, Run run)
{
Text text = run.Elements<Text>().LastOrDefault();
if (text == null)
{
run.Append(new Text());
text = run.Elements<Text>().Last();
}
text.Text += newText;
if (text.Text.StartsWith(" ") || text.Text.EndsWith(" "))
text.Space = SpaceProcessingModeValues.Preserve;
}
What can I do to understand where the problem lies?
I tried replacing some values I don't understand in File B with the ones from File A (header and footer have rectangles with different gfxdata values, the "recovered" from word was setting the same values as File A).
I tried a different way of generating the AltChunkIds and storing a global list for the file.
I tried comparing various parts of the documents (File A and B or Fil B's result and its recovered version). There are differences, but too many and none seem to be relevant.

Text extraction using itext7: garbage characters for some pdf documents

I have a problem extracting text from pdf documents using iText7. For documents coming from a specific source textRenderInfo.GetText() returns only garbage chars (0xfdff) in the event handler of my extraction strategy:
internal class CustomExtractionStrategy : ITextExtractionStrategy
{
public virtual void EventOccurred(IEventData data, EventType type)
{
if (!type.Equals((object)EventType.RENDER_TEXT))
{
return;
}
var textRenderInfo = (TextRenderInfo)data;
bool currentResultEmpty = _result.Length == 0;
bool isInNewLine = false;
var baseline = textRenderInfo.GetBaseline();
var startPoint = baseline.GetStartPoint();
var endPoint = baseline.GetEndPoint();
var currentText = textRenderInfo.GetText(); // returns garbage for specific pdfs
// further processing below
...
}
}
I'm not very familiar with the way text/glyph encoding words in PDF but I try to give some details when comparing the problematic pdfs with an example where extraction works. For the pdfs with issues:
textRenderInfo.gs.font is MS-UIGothic
textRenderInfo.gs.font.fontProgram.codeToGlyph contains only mapping (key: 0 to a Glyph with width 1000, unicode -1, code 0)
textRenderInfo.gs.font.fontProgram.unicodeToGlyph contains no records
These are the most obvious discrepancies. If there's any thing else I should look out for please let me know. I would have provided an example of the PDF in question but it might have sensitive information that I must not disclose.
Note: the PDFs can be correctly read in Acrobat Reader and I can copy text from the reader into notepad. Other libraries (pdfium based or ports of PDFBox) can properly extract text from the document. So I think the document as such is "valid".
If this is a known issue for iText7, is there any workaround (other than using a different library altogether)?
Update
With the link provided in the comment and the following code (in addition to the custom extraction strategy snippet shown above) I get garbage chars see VS screenshot:
internal class PdfExtractor
{
internal void ExtractFromPath(string path)
{
PdfReader reader = new PdfReader(path);
var document = new iText.Kernel.Pdf.PdfDocument(reader);
for (int pageNum = 1; pageNum <= document.GetNumberOfPages(); pageNum++)
{
var page = document.GetPage(pageNum);
string text = PdfTextExtractor.GetTextFromPage(page, new CustomExtractionStrategy());
}
}
}

Merged document using AltChunks has innertext empty

I am trying to merge multiple documents into a single one and then open the result document and process it further.
The "ChunkId" is a property that is increased every time this method is called in order to get a unique id. I followed the example from this site.
This is the code used to merge multiple documents (using altchunks):
`
private void MergeDocument(string mergePath, bool appendPageBreak)
{
if (!File.Exists(mergePath))
{
Log.Warn(string.Format("Document: \"{0}\" was not found.", mergePath));
return;
}
ChunkId++;
var altChunkId = "AltChunkId" + ChunkId;
var mainDocPart = DestinationDocument.MainDocumentPart;
if (mainDocPart == null)
{
DestinationDocument.AddMainDocumentPart();
mainDocPart = DestinationDocument.MainDocumentPart;
if (mainDocPart.Document == null)
mainDocPart.Document = new Document { Body = new Body() };
}
try
{
var chunk = mainDocPart.AddAlternativeFormatImportPart(
AlternativeFormatImportPartType.WordprocessingML, altChunkId);
if (chunk != null)
using (var ms = new FileStream(mergePath, FileMode.Open))
{
chunk.FeedData(ms);
}
else
{
Log.Error(string.Format("Merge - Failed to create chunk document based on \"{0}\".", mergePath));
return; // failed to create chunk document, return from merge method
}
}
catch (Exception e)
{
Log.Error(string.Format("Merge - Failed to insert chunk document based on \"{0}\".", mergePath));
return; // failed to create chunk document, return from merge method
}
var altChunk = new AltChunk { Id = altChunkId };
//append the page break
if (appendPageBreak)
try
{
AppendPageBreak(mainDocPart);
Log.Info(string.Format("Successfully appended page break."));
}
catch (Exception ex)
{
Log.Error(string.Format("Eror appending page break. Message: \"{0}\".", ex.Message));
return; // return if page break insertion failed
}
// insert the document
var last = mainDocPart.Document
.Body
.Elements()
.LastOrDefault(e => e is Paragraph || e is AltChunk);
try
{
if (last == null)
mainDocPart.Document.Body.InsertAt(altChunk, 0);
else
last.InsertAfterSelf(altChunk);
Log.Info(string.Format("Successfully inserted new doc \"{0}\" into destination.", mergePath));
}
catch (Exception ex)
{
Log.Error(string.Format("Error merging document \"{0}\". Message: \"{1}\".", mergePath, ex.Message));
return; // return if the merge was not successfull
}
try
{
mainDocPart.Document.Save();
}
catch (Exception ex)
{
Log.Error(string.Format("Error saving document \"{0}\". Message: \"{1}\".", mergePath, ex.Message));
}
}`
If I open the merged document with Word I can see its content (tables, text, paragraphs..), but if I open if from code again it says that inner text is "" (empty string). I need that inner text to reflect what the document contains because I have to replace some placeholders like "##name##" with another text and I can't if the inner text is empty.
This is the innerxml of the merged document,
This is how I open the merged document:
DestinationDocument = WordprocessingDocument.Open(Path.GetFullPath(destinationPath), true);
How can I read the inner text of the document? Or how can I merge these documents into a single one so that this problem would not occur anymore?

When documents merged with AltChunks it is like embedded attachments to the original word document. The client (MS Word) handles the rendering of the altchunk sections. Hence the resulting document won't have the openxml markup of the merged documents.
If you want to use the resulting document for further programmatic post-processing use Openxml Power Tools. pelase refer to my answer here
Openxml powertools - https://github.com/OfficeDev/Open-Xml-PowerTools

The problem is that the documents are not really merged (per se), the altChunk element only defines a place where the alternative content should be placed in the document and it has a reference to that alternative content.
When you open this document in MS Word then it will actually merge all those alternative contents automatically for you. So when you resave that document with MS Word you'll no longer have altChunk elements.
Nevertheless what you can do is actually manipulate with those altChunk DOCX files (the child DOCX documents) just like you do with the main DOCX file (the parent document).
For example:
string destinationPath = "Sample.docx";
string search = "##name##";
string replace ="John Doe";
using (var parent = WordprocessingDocument.Open(Path.GetFullPath(destinationPath), true))
{
foreach (var altChunk in parent.MainDocumentPart.GetPartsOfType<AlternativeFormatImportPart>())
{
if (Path.GetExtension(altChunk.Uri.OriginalString) != ".docx")
continue;
using (var child = WordprocessingDocument.Open(altChunk.GetStream(), true))
{
var foundText = child.MainDocumentPart.Document.Body
.Descendants<Text>()
.Where(t => t.Text.Contains(search))
.FirstOrDefault();
if (foundText != null)
{
foundText.Text = foundText.Text.Replace(search, replace);
break;
}
}
}
}
Alternatively you'll need to use some approach to merge those documents for real. One solution is mentioned by Flowerking, another one that you could try is with GemBox.Document library. It will merge those alternative contents for you on loading (just as MS Word does when opening).
For example:
string destinationPath = "Sample.docx";
string search = "##name##";
string replace = "John Doe";
DocumentModel document = DocumentModel.Load(destinationPath);
ContentRange foundText = document.Content.Find(search).FirstOrDefault();
if (foundText != null)
foundText.LoadText(replace);
document.Save(destinationPath);

how can I put a content in a mergefield in docx

I'm developing a web application with asp.net and I have a file called Template.docx that works like a template to generate other reports. Inside this Template.docx I have some MergeFields (Title, CustomerName, Content, Footer, etc) to replace for some dynamic content in C#.
I would like to know, how can I put a content in a mergefield in docx ?
I don't know if MergeFields is the right way to do this or if there is another way. If you can suggest me, I appreciate!
PS: I have openxml referenced in my web application.
Edits:
private MemoryStream LoadFileIntoStream(string fileName)
{
MemoryStream memoryStream = new MemoryStream();
using (FileStream fileStream = File.OpenRead(fileName))
{
memoryStream.SetLength(fileStream.Length);
fileStream.Read(memoryStream.GetBuffer(), 0, (int) fileStream.Length);
memoryStream.Flush();
fileStream.Close();
}
return memoryStream;
}
public MemoryStream GenerateWord()
{
string templateDoc = "C:\\temp\\template.docx";
string reportFileName = "C:\\temp\\result.docx";
var reportStream = LoadFileIntoStream(templateDoc);
// Copy a new file name from template file
//File.Copy(templateDoc, reportFileName, true);
// Open the new Package
Package pkg = Package.Open(reportStream, FileMode.Open, FileAccess.ReadWrite);
// Specify the URI of the part to be read
Uri uri = new Uri("/word/document.xml", UriKind.Relative);
PackagePart part = pkg.GetPart(uri);
XmlDocument xmlMainXMLDoc = new XmlDocument();
xmlMainXMLDoc.Load(part.GetStream(FileMode.Open, FileAccess.Read));
// replace some keys inside xml (it will come from database, it's just a test)
xmlMainXMLDoc.InnerXml = xmlMainXMLDoc.InnerXml.Replace("field_customer", "My Customer Name");
xmlMainXMLDoc.InnerXml = xmlMainXMLDoc.InnerXml.Replace("field_title", "Report of Documents");
xmlMainXMLDoc.InnerXml = xmlMainXMLDoc.InnerXml.Replace("field_content", "Content of Document");
// Open the stream to write document
StreamWriter partWrt = new StreamWriter(part.GetStream(FileMode.Open, FileAccess.Write));
//doc.Save(partWrt);
xmlMainXMLDoc.Save(partWrt);
partWrt.Flush();
partWrt.Close();
reportStream.Flush();
pkg.Close();
return reportStream;
}
PS: When I convert MemoryStream to a file, I got a corrupted file. Thanks!

I know this is an old post, but I could not get the accepted answer to work for me. The project linked would not even compile (which someone has already commented in that link). Also, it seems to use other Nuget packages like WPFToolkit.
So I'm adding my answer here in case someone finds it useful. This only uses the OpenXML SDK 2.5 and also the WindowsBase v4. This works on MS Word 2010 and later.
string sourceFile = #"C:\Template.docx";
string targetFile = #"C:\Result.docx";
File.Copy(sourceFile, targetFile, true);
using (WordprocessingDocument document = WordprocessingDocument.Open(targetFile, true))
{
// If your sourceFile is a different type (e.g., .DOTX), you will need to change the target type like so:
document.ChangeDocumentType(WordprocessingDocumentType.Document);
// Get the MainPart of the document
MainDocumentPart mainPart = document.MainDocumentPart;
var mergeFields = mainPart.RootElement.Descendants<FieldCode>();
var mergeFieldName = "SenderFullName";
var replacementText = "John Smith";
ReplaceMergeFieldWithText(mergeFields, mergeFieldName, replacementText);
// Save the document
mainPart.Document.Save();
}
private void ReplaceMergeFieldWithText(IEnumerable<FieldCode> fields, string mergeFieldName, string replacementText)
{
var field = fields
.Where(f => f.InnerText.Contains(mergeFieldName))
.FirstOrDefault();
if (field != null)
{
// Get the Run that contains our FieldCode
// Then get the parent container of this Run
Run rFldCode = (Run)field.Parent;
// Get the three (3) other Runs that make up our merge field
Run rBegin = rFldCode.PreviousSibling<Run>();
Run rSep = rFldCode.NextSibling<Run>();
Run rText = rSep.NextSibling<Run>();
Run rEnd = rText.NextSibling<Run>();
// Get the Run that holds the Text element for our merge field
// Get the Text element and replace the text content
Text t = rText.GetFirstChild<Text>();
t.Text = replacementText;
// Remove all the four (4) Runs for our merge field
rFldCode.Remove();
rBegin.Remove();
rSep.Remove();
rEnd.Remove();
}
}
What the code above does is basically this:
Identify the 4 Runs that make up the merge field named "SenderFullName".
Identify the Run that contains the Text element for our merge field.
Remove the 4 Runs.
Update the text property of the Text element for our merge field.
UPDATE
For anyone interested, here is a simple static class I used to help me with replacing merge fields.

Frank Fajardo's answer was 99% of the way there for me, but it is important to note that MERGEFIELDS can be SimpleFields or FieldCodes.
In the case of SimpleFields, the text runs displayed to the user in the document are children of the SimpleField.
In the case of FieldCodes, the text runs shown to the user are between the runs containing FieldChars with the Separate and the End FieldCharValues. Occasionally, several text containing runs exist between the Separate and End Elements.
The code below deals with these problems. Further details of how to get all the MERGEFIELDS from the document, including the header and footer is available in a GitHub repository at https://github.com/mcshaz/SimPlanner/blob/master/SP.DTOs/Utilities/OpenXmlExtensions.cs
private static Run CreateSimpleTextRun(string text)
{
Run returnVar = new Run();
RunProperties runProp = new RunProperties();
runProp.Append(new NoProof());
returnVar.Append(runProp);
returnVar.Append(new Text() { Text = text });
return returnVar;
}
private static void InsertMergeFieldText(OpenXmlElement field, string replacementText)
{
var sf = field as SimpleField;
if (sf != null)
{
var textChildren = sf.Descendants<Text>();
textChildren.First().Text = replacementText;
foreach (var others in textChildren.Skip(1))
{
others.Remove();
}
}
else
{
var runs = GetAssociatedRuns((FieldCode)field);
var rEnd = runs[runs.Count - 1];
foreach (var r in runs
.SkipWhile(r => !r.ContainsCharType(FieldCharValues.Separate))
.Skip(1)
.TakeWhile(r=>r!= rEnd))
{
r.Remove();
}
rEnd.InsertBeforeSelf(CreateSimpleTextRun(replacementText));
}
}
private static IList<Run> GetAssociatedRuns(FieldCode fieldCode)
{
Run rFieldCode = (Run)fieldCode.Parent;
Run rBegin = rFieldCode.PreviousSibling<Run>();
Run rCurrent = rFieldCode.NextSibling<Run>();
var runs = new List<Run>(new[] { rBegin, rCurrent });
while (!rCurrent.ContainsCharType(FieldCharValues.End))
{
rCurrent = rCurrent.NextSibling<Run>();
runs.Add(rCurrent);
};
return runs;
}
private static bool ContainsCharType(this Run run, FieldCharValues fieldCharType)
{
var fc = run.GetFirstChild<FieldChar>();
return fc == null
? false
: fc.FieldCharType.Value == fieldCharType;
}

You could try http://www.codeproject.com/KB/office/Fill_Mergefields.aspx which uses the Open XML SDK to do this.

OpenXml Sdk - Copy Sections of docx into another docx

I am trying the following code. It takes a fileName (docx file with many sections) and I try to iterate through each section getting the section name. The problem is that I end up with unreadable docx files. It does not error, but I think I am doing something wrong with getting the elements in the section.
public void Split(string fileName) {
using (WordprocessingDocument myDoc =
WordprocessingDocument.Open(fileName, true)) {
string curCliCode = "";
MainDocumentPart mdp = myDoc.MainDocumentPart;
foreach (var element in mdp.Document.Body.ChildElements) {
if (element.Descendants().OfType<SectionProperties>().Count() == 1) {
//get the name of the section from the footer
var footer = (FooterPart) mdp.GetPartById(
element.Descendants().OfType<SectionProperties>().First().OfType
<FooterReference>().First().
Id.Value);
foreach (Paragraph p in footer.Footer.ChildElements.OfType<Paragraph>()) {
if (p.InnerText != "") {
curCliCode = p.InnerText;
}
}
if (curCliCode != "") {
var forFile = new List<OpenXmlElement>();
var els = element.ElementsBefore();
if (els != null) {
foreach (var e in els) {
if (e != null) {
forFile.Add(e);
}
}
for (int i = 0; i < els.Count(); i++) {
els.ElementAt(i).Remove();
}
}
Create(curCliCode, forFile);
}
}
}
}
}
private void Create(string cliCode,IEnumerable<OpenXmlElement> docParts) {
var parts = from e in docParts select e.Clone();
const string template = #"\Test\toSplit\blank.docx";
string destination = string.Format(#"\Test\{0}.docx", cliCode);
File.Copy(template, destination,true);
/* Create the package and main document part */
using (WordprocessingDocument myDoc =
WordprocessingDocument.Open(destination, true)) {
MainDocumentPart mainPart = myDoc.MainDocumentPart;
/* Create the contents */
foreach(var part in parts) {
mainPart.Document.Body.Append((OpenXmlElement)part);
}
/* Save the results and close */
mainPart.Document.Save();
myDoc.Close();
}
}
Does anyone know what the problem could be (or how to properly copy a section from one document to another)?

I've done some work in this area, and what I have found invaluable is diffing a known good file with a prospective file; the error is usually fairly obvious.
What I would do is take a file that you know works, and copy all of the sections into the template. Theoretically, the two files should be identical. Run a diff on them the document.xml inside the docx file, and you'll see the difference.
BTW, I'm assuming that you know that the docx is actually a zip; change the extension to "zip", and you'll be able to get at the actual xml files which compose the format.
As far as diff tools, I use Beyond Compare from Scooter Software.

An approach along the lines of what you are doing will work only for simple documents (ie those not containing images, hyperlinks, comments etc). To handle these more complex documents, take a look at http://blogs.msdn.com/b/ericwhite/archive/2009/02/05/move-insert-delete-paragraphs-in-word-processing-documents-using-the-open-xml-sdk.aspx and the resulting DocumentBuilder API (part of the PowerTools for Open XML project on CodePlex).
In order to split a docx into sections using DocumentBuilder, you'll still need to first find the index of the paragraphs containing sectPr elements.