public IActionResult WordToHtml()
{
var file = Request.Form.Files.First();
try
{
using (MemoryStream memoryStream = new MemoryStream())
{
file.CopyTo(memoryStream);
using (WordprocessingDocument doc =
WordprocessingDocument.Open(memoryStream, true))
{
HtmlConverterSettings settings = new HtmlConverterSettings()
{
PageTitle = file.Name
};
XElement html = HtmlConverter.ConvertToHtml(doc, settings);
return(Json(new { success = true, data = html.ToStringNewLineOnAttributes() }));
}
}
}
catch (Exception ex)
{
return(Json(new { success = false, data = ex.Message }));
}
}
public bool WordToHTML()
{
bool res = false;
try
{
byte[] byteArray = File.ReadAllBytes("C:/Users/lsamaniego/Documents/Sprint Backlog.docx");
using (MemoryStream memoryStream = new MemoryStream())
{
memoryStream.Write(byteArray, 0, byteArray.Length);
using (WordprocessingDocument doc = WordprocessingDocument.Open(memoryStream, true))
{
HtmlConverterSettings settings = new HtmlConverterSettings()
{
PageTitle = "My Page Title"
};
XElement html = HtmlConverter.ConvertToHtml(doc, settings);
File.WriteAllText("C:/Users/lsamaniego/Documents/Sprint Backlog.html", html.ToStringNewLineOnAttributes());
}
}
res = true;
}
catch (Exception ex)
{
throw;
}
return(res);
}
private string ConvertToHtml(IFormFile file)
{
using var ms = file.OpenReadStream();
using var fs = new FileStream(Path.Combine(Environment.GetEnvironmentVariable("temp") ??
"upload", file.FileName), FileMode.OpenOrCreate, FileAccess.ReadWrite);
ms.CopyTo(fs);
using var doc = WordprocessingDocument.Open(fs, true);
var pageTitle = file.FileName;
var part = doc.CoreFilePropertiesPart;
if (part != null)
{
pageTitle ??= (string)part.GetXDocument().Descendants(DC.title).FirstOrDefault();
}
var settings = new HtmlConverterSettings()
{
PageTitle = pageTitle,
FabricateCssClasses = false,
RestrictToSupportedLanguages = false,
RestrictToSupportedNumberingFormats = false,
ImageHandler = imageInfo =>
{
var stream = new MemoryStream();
imageInfo.Bitmap.Save(stream, imageInfo.Bitmap.RawFormat);
var base64String = Convert.ToBase64String(stream.ToArray());
return(new XElement(Xhtml.img, new XAttribute(NoNamespace.src, $"data:{imageInfo.ContentType};base64," + base64String), imageInfo.ImgStyleAttribute, imageInfo.AltText != null ? new XAttribute(NoNamespace.alt, imageInfo.AltText) : null));
}
};
var htmlElement = HtmlConverter.ConvertToHtml(doc, settings);
var html = new XDocument(new XDocumentType("html", null, null, null), htmlElement);
var htmlString = html.ToString(SaveOptions.DisableFormatting);
return(htmlString);
}
public byte[] HtmlToPdf(byte[] bytes)
{
MemoryStream origin = new MemoryStream(bytes, true);
MemoryStream destination = new MemoryStream();
origin.CopyToAsync(destination);
origin.Close();
using (MemoryStream memoryStream = new MemoryStream())
{
using (WordprocessingDocument doc = WordprocessingDocument.Open(destination, true))
{
HtmlConverterSettings settings = new HtmlConverterSettings()
{
FabricateCssClasses = true,
};
XElement html = HtmlConverter.ConvertToHtml(doc, settings);
MemoryStream result = new MemoryStream();
html.Save(result);
result.Position = 0;
return(result.ToArray());
}
}
}
public string WordToHtml(MemoryStream ms)
{
using (WordprocessingDocument doc = WordprocessingDocument.Open(ms, true))
{
HtmlConverterSettings settings = new HtmlConverterSettings()
{
};
XElement html = HtmlConverter.ConvertToHtml(doc, settings);
return(html.ToStringNewLineOnAttributes());
}
}
public string ConvertToHTML()
{
var wmldoc = new WmlDocument("tempDoc", MediaFile.GetMediaStream);
var settings = new HtmlConverterSettings()
{
PageTitle = "HTML Converted Page",
};
var html = HtmlConverter.ConvertToHtml(wmldoc, settings);
}
public async ValueTask <ManagedIdDto> RenderWordAsync(Stream stream, string title, TUserBrief user, Guid id)
{
var anytask = AnyAsync(a => a.ArticleId == id);
var article = Activator.CreateInstance <TArticle>();
using WordprocessingDocument doc = WordprocessingDocument.Open(stream, true);
HtmlConverterSettings settings = new HtmlConverterSettings()
{
PageTitle = "My Page Title",
ImageHandler = WordImageHandler
};
XElement html = HtmlConverter.ConvertToHtml(doc, settings);
article.AuthorBrief = user;
var santilizetask = Santilize(html.ToStringNewLineOnAttributes());
article.Title = title;
article.IsDraft = true;
article.Visibility = ArticleVisibility.everyone;
article.Saved = true;
article.DraftOrPublishId = Guid.Empty;
Guid aid = Guid.Empty;
article.Content = await santilizetask;
article.Content = string.Concat(article.Content, @"
<hr>
<p>本文章由word文档转换而成</p>
");
var ud = Builders <TArticle> .Update
.Set(a => a.Content, article.Content)
.Set(a => a.Saved, true);
if (id != Guid.Empty && await anytask)
{
await UpDateAsync(a => a.ArticleId == id && a.IsDraft == true, ud);
aid = id;
}
else
{
article.ArticleId = Guid.NewGuid();
article.DraftOrPublishId = Guid.Empty;
await AddArticleAsync(article);
aid = article.ArticleId;
}
return(new ManagedIdDto(await GetManagedIdAsync(aid), aid));
}
public static XElement ConvertToHtml(WordprocessingDocument wordDocument)
{
HtmlConverterSettings settings = new HtmlConverterSettings()
{
PageTitle = "My Page Title",
CssClassPrefix = "",
Css = "",
ConvertFormatting = false
};
XElement html = HtmlConverter.ConvertToHtml(wordDocument, settings, null);
return(html);
}
private static async Task <string> ConvertToHtml(IFormFile file)
{
var docfile = Path.Combine(Environment.GetEnvironmentVariable("temp") ?? "upload", file.FileName);
try
{
await using var ms = file.OpenReadStream();
await using var fs = System.IO.File.Create(docfile, 1024, FileOptions.DeleteOnClose);
await ms.CopyToAsync(fs);
using var doc = WordprocessingDocument.Open(fs, true);
var pageTitle = file.FileName;
var part = doc.CoreFilePropertiesPart;
if (part != null)
{
pageTitle ??= (string)part.GetXDocument().Descendants(DC.title).FirstOrDefault();
}
var settings = new HtmlConverterSettings()
{
PageTitle = pageTitle,
FabricateCssClasses = false,
RestrictToSupportedLanguages = false,
RestrictToSupportedNumberingFormats = false,
ImageHandler = imageInfo =>
{
var stream = new MemoryStream();
imageInfo.Bitmap.Save(stream, imageInfo.Bitmap.RawFormat);
var base64String = Convert.ToBase64String(stream.ToArray());
return(new XElement(Xhtml.img, new XAttribute(NoNamespace.src, $"data:{imageInfo.ContentType};base64," + base64String), imageInfo.ImgStyleAttribute, imageInfo.AltText != null ? new XAttribute(NoNamespace.alt, imageInfo.AltText) : null));
}
};
var htmlElement = HtmlConverter.ConvertToHtml(doc, settings);
var html = new XDocument(new XDocumentType("html", null, null, null), htmlElement);
var htmlString = html.ToString(SaveOptions.DisableFormatting);
return(htmlString);
}
finally
{
if (System.IO.File.Exists(docfile))
{
Policy.Handle <IOException>().WaitAndRetry(5, i => TimeSpan.FromSeconds(1)).Execute(() => System.IO.File.Delete(docfile));
}
}
}
public StringBuilder getWordDoc(string docName, string version)
{
var str = GetUploadedDoc(docName, version);
Stream stream = str.Result;
StringBuilder word = new StringBuilder();
using (WordprocessingDocument doc = WordprocessingDocument.Open(stream, true))
{
HtmlConverterSettings settings = new HtmlConverterSettings()
{
PageTitle = docName
};
XElement html = HtmlConverter.ConvertToHtml(doc, settings);
word.Append(html);
}
return(word);
}
public IActionResult DocToPdf()
{
//Это если нужно конвертировать меньше 3 страниц
// Document document = new Document(Path.Combine(Environment.CurrentDirectory, "SignData_and_VerifyData.docx"), FileFormat.Docx);
// document.SaveToFile(Path.Combine(Environment.CurrentDirectory, "docToPdfTest3.pdf"), FileFormat.PDF);
// var source = Package.Open(Path.Combine(Environment.CurrentDirectory, "SignData_and_VerifyData.docx"));
var source = Package.Open(@"Table.docx");
var document = WordprocessingDocument.Open(source);
HtmlConverterSettings settings = new HtmlConverterSettings();
XElement html = HtmlConverter.ConvertToHtml(document, settings);
string htmlStr = html.ToString();
var doc = new HtmlToPdfDocument()
{
GlobalSettings =
{
ColorMode = ColorMode.Color,
Orientation = Orientation.Landscape,
PaperSize = PaperKind.A4Plus,
Out = Path.Combine(Environment.CurrentDirectory, "test23.pdf"),
},
Objects =
{
new ObjectSettings()
{
PagesCount = true,
HtmlContent = htmlStr,
WebSettings = { DefaultEncoding = "utf-8" },
HeaderSettings ={ FontSize =9, Right = "Страница [page] из [toPage]", Line = true, Spacing = 2.812 },
FooterSettings ={ FontSize =9, Right = "Страница [page] из [toPage]" }
}
}
};
_converter.Convert(doc);
return(View());
}
private static string ConverWordToHtml(string wordFilePath)
{
byte[] byteArray = File.ReadAllBytes(wordFilePath);
using (MemoryStream memoryStream = new MemoryStream())
{
memoryStream.Write(byteArray, 0, byteArray.Length);
using (WordprocessingDocument doc = WordprocessingDocument.Open(memoryStream, true))
{
HtmlConverterSettings settings = new HtmlConverterSettings()
{
PageTitle = "Daily Status"
};
XElement html = HtmlConverter.ConvertToHtml(doc, settings);
return(html.ToStringNewLineOnAttributes());
}
}
}
/// <summary>
/// Convert word to html
/// </summary>
/// <param name="input"></param>
/// <param name="mediaPath"></param>
/// <returns></returns>
public HtmlResult ToHtml(byte[] input, string mediaPath)
{
var result = new HtmlResult();
var imageHandler = mediaPath == null ? new Base64ImageHandler() : new FileImageHandler(mediaPath) as IImageHandler;
using (var stream = new MemoryStream(input))
{
using (var doc = WordprocessingDocument.Open(stream, true))
{
var settings = new HtmlConverterSettings
{
PageTitle = ""
};
var htmlConverterService = HostContainer.GetInstance <IHtmlConverterService>();
var html = htmlConverterService.ConvertToHtml(doc, settings, imageHandler.Handle);
result.Html = html.ToStringNewLineOnAttributes();
}
}
result.Files = imageHandler.Files;
return(result);
}
public FileModel WordToPdf(byte[] data, string fileName = null)
{
byte[] byteArray = data;
using (MemoryStream memoryStream = new MemoryStream()) {
memoryStream.Write(byteArray, 0, byteArray.Length);
using (WordprocessingDocument doc =
WordprocessingDocument.Open(memoryStream, true)) {
HtmlConverterSettings settings = new HtmlConverterSettings()
{
//PageTitle = "My Page Title"
AdditionalCss = "td { border : 1px solid #000 !important; }"
};
XElement html = HtmlConverter.ConvertToHtml(doc, settings);
SelectPdf.HtmlToPdf htmlToPdf = new SelectPdf.HtmlToPdf();
return(_pdfSvc.GetPdfFromHtmlString(html.ToStringNewLineOnAttributes(), Path.GetFileNameWithoutExtension(fileName) ?? "Document" + ".pdf"));
//File.writeal(@"Test.html", html.ToStringNewLineOnAttributes());
}
}
}
public static string Convert(string inputPath)
{
var fi = new FileInfo(inputPath);
byte[] byteArray = File.ReadAllBytes(fi.FullName);
using (MemoryStream memoryStream = new MemoryStream()) {
memoryStream.Write(byteArray, 0, byteArray.Length);
using (WordprocessingDocument wDoc = WordprocessingDocument.Open(memoryStream, true)) {
HtmlConverterSettings settings = new HtmlConverterSettings()
{
FabricateCssClasses = true,
RestrictToSupportedLanguages = false,
RestrictToSupportedNumberingFormats = false
};
XElement html = HtmlConverter.ConvertToHtml(wDoc, settings);
return(html.ToString(SaveOptions.DisableFormatting));
}
}
}
private async Task <IActionResult> ConvertDocxToHtmlWithoutImageAsync(string fileName, byte[] fileBytes)
{
try
{
var docFilePath = Path.Combine(_tempPath, fileName);
System.IO.File.WriteAllBytes(docFilePath, fileBytes);
var source = Package.Open(docFilePath);
var document = WordprocessingDocument.Open(source);
HtmlConverterSettings settings = new HtmlConverterSettings();
XElement html = HtmlConverter.ConvertToHtml(document, settings);
var outfile = Path.Combine(_tempPath, ("htmldata" + ".html"));
var writer = System.IO.File.CreateText(outfile);
writer.WriteLine(html.ToString());
writer.Dispose();
source.Close();
System.IO.File.Delete(docFilePath);
return(File(await System.IO.File.ReadAllBytesAsync(outfile), "application/octet-stream", (Path.GetFileNameWithoutExtension(fileName) + ".html")));
}
catch (Exception ex)
{
return(StatusCode(StatusCodes.Status500InternalServerError));
}
}
public static void ConvertToHtml(string file, string outputDirectory)
{
var fi = new FileInfo(file);
Console.WriteLine(fi.Name);
byte[] byteArray = File.ReadAllBytes(fi.FullName);
using (MemoryStream memoryStream = new MemoryStream())
{
memoryStream.Write(byteArray, 0, byteArray.Length);
using (WordprocessingDocument wDoc = WordprocessingDocument.Open(memoryStream, true))
{
var destFileName = new FileInfo(fi.Name.Replace(".docx", ".html"));
if (outputDirectory != null && outputDirectory != string.Empty)
{
DirectoryInfo di = new DirectoryInfo(outputDirectory);
if (!di.Exists)
{
throw new OpenXmlPowerToolsException("Output directory does not exist");
}
destFileName = new FileInfo(Path.Combine(di.FullName, destFileName.Name));
}
var imageDirectoryName = destFileName.FullName.Substring(0, destFileName.FullName.Length - 5) + "_files";
int imageCounter = 0;
var pageTitle = fi.FullName;
var part = wDoc.CoreFilePropertiesPart;
if (part != null)
{
pageTitle = (string)part.GetXDocument().Descendants(DC.title).FirstOrDefault() ?? fi.FullName;
}
// TODO: Determine max-width from size of content area.
HtmlConverterSettings settings = new HtmlConverterSettings()
{
AdditionalCss = "body { margin: 1cm auto; max-width: 20cm; padding: 0; }",
PageTitle = pageTitle,
FabricateCssClasses = true,
CssClassPrefix = "pt-",
RestrictToSupportedLanguages = false,
RestrictToSupportedNumberingFormats = false,
ImageHandler = imageInfo =>
{
DirectoryInfo localDirInfo = new DirectoryInfo(imageDirectoryName);
if (!localDirInfo.Exists)
{
localDirInfo.Create();
}
++imageCounter;
string extension = imageInfo.ContentType.Split('/')[1].ToLower();
ImageFormat imageFormat = null;
if (extension == "png")
{
imageFormat = ImageFormat.Png;
}
else if (extension == "gif")
{
imageFormat = ImageFormat.Gif;
}
else if (extension == "bmp")
{
imageFormat = ImageFormat.Bmp;
}
else if (extension == "jpeg")
{
imageFormat = ImageFormat.Jpeg;
}
else if (extension == "tiff")
{
// Convert tiff to gif.
extension = "gif";
imageFormat = ImageFormat.Gif;
}
else if (extension == "x-wmf")
{
extension = "wmf";
imageFormat = ImageFormat.Wmf;
}
// If the image format isn't one that we expect, ignore it,
// and don't return markup for the link.
if (imageFormat == null)
{
return(null);
}
string imageFileName = imageDirectoryName + "/image" +
imageCounter.ToString() + "." + extension;
try
{
imageInfo.Bitmap.Save(imageFileName, imageFormat);
}
catch (System.Runtime.InteropServices.ExternalException)
{
return(null);
}
string imageSource = localDirInfo.Name + "/image" +
imageCounter.ToString() + "." + extension;
XElement img = new XElement(Xhtml.img,
new XAttribute(NoNamespace.src, imageSource),
imageInfo.ImgStyleAttribute,
imageInfo.AltText != null ?
new XAttribute(NoNamespace.alt, imageInfo.AltText) : null);
return(img);
}
};
XElement htmlElement = HtmlConverter.ConvertToHtml(wDoc, settings);
// Produce HTML document with <!DOCTYPE html > declaration to tell the browser
// we are using HTML5.
var html = new XDocument(
new XDocumentType("html", null, null, null),
htmlElement);
// Note: the xhtml returned by ConvertToHtmlTransform contains objects of type
// XEntity. PtOpenXmlUtil.cs define the XEntity class. See
// http://blogs.msdn.com/ericwhite/archive/2010/01/21/writing-entity-references-using-linq-to-xml.aspx
// for detailed explanation.
//
// If you further transform the XML tree returned by ConvertToHtmlTransform, you
// must do it correctly, or entities will not be serialized properly.
var htmlString = html.ToString(SaveOptions.DisableFormatting);
File.WriteAllText(destFileName.FullName, htmlString, Encoding.UTF8);
}
}
}
public static void ConvertToHtmlNoCssClasses(FileInfo sourceDocx, FileInfo destFileName)
{
byte[] byteArray = File.ReadAllBytes(sourceDocx.FullName);
using (MemoryStream memoryStream = new MemoryStream())
{
memoryStream.Write(byteArray, 0, byteArray.Length);
using (WordprocessingDocument wDoc = WordprocessingDocument.Open(memoryStream, true))
{
var outputDirectory = destFileName.Directory;
destFileName = new FileInfo(Path.Combine(outputDirectory.FullName, destFileName.Name));
var imageDirectoryName = destFileName.FullName.Substring(0, destFileName.FullName.Length - 5) + "_files";
int imageCounter = 0;
var pageTitle = (string)wDoc.CoreFilePropertiesPart.GetXDocument().Descendants(DC.title).FirstOrDefault();
if (pageTitle == null)
{
pageTitle = sourceDocx.FullName;
}
HtmlConverterSettings settings = new HtmlConverterSettings()
{
PageTitle = pageTitle,
FabricateCssClasses = false,
RestrictToSupportedLanguages = false,
RestrictToSupportedNumberingFormats = false,
ImageHandler = imageInfo =>
{
DirectoryInfo localDirInfo = new DirectoryInfo(imageDirectoryName);
if (!localDirInfo.Exists)
{
localDirInfo.Create();
}
++imageCounter;
string extension = imageInfo.ContentType.Split('/')[1].ToLower();
ImageFormat imageFormat = null;
if (extension == "png")
{
// Convert png to jpeg.
extension = "gif";
imageFormat = ImageFormat.Gif;
}
else if (extension == "gif")
{
imageFormat = ImageFormat.Gif;
}
else if (extension == "bmp")
{
imageFormat = ImageFormat.Bmp;
}
else if (extension == "jpeg")
{
imageFormat = ImageFormat.Jpeg;
}
else if (extension == "tiff")
{
// Convert tiff to gif.
extension = "gif";
imageFormat = ImageFormat.Gif;
}
else if (extension == "x-wmf")
{
extension = "wmf";
imageFormat = ImageFormat.Wmf;
}
// If the image format isn't one that we expect, ignore it,
// and don't return markup for the link.
if (imageFormat == null)
{
return(null);
}
string imageFileName = imageDirectoryName + "/image" +
imageCounter.ToString() + "." + extension;
try
{
imageInfo.Bitmap.Save(imageFileName, imageFormat);
}
catch (System.Runtime.InteropServices.ExternalException)
{
return(null);
}
XElement img = new XElement(Xhtml.img,
new XAttribute(NoNamespace.src, imageFileName),
imageInfo.ImgStyleAttribute,
imageInfo.AltText != null ?
new XAttribute(NoNamespace.alt, imageInfo.AltText) : null);
return(img);
}
};
XElement html = HtmlConverter.ConvertToHtml(wDoc, settings);
// Note: the xhtml returned by ConvertToHtmlTransform contains objects of type
// XEntity. PtOpenXmlUtil.cs define the XEntity class. See
// http://blogs.msdn.com/ericwhite/archive/2010/01/21/writing-entity-references-using-linq-to-xml.aspx
// for detailed explanation.
//
// If you further transform the XML tree returned by ConvertToHtmlTransform, you
// must do it correctly, or entities will not be serialized properly.
var htmlString = html.ToString(SaveOptions.DisableFormatting);
File.WriteAllText(destFileName.FullName, htmlString, Encoding.UTF8);
}
}
}
private void button1_Click(object sender, EventArgs e)
{
Thread.CurrentThread.CurrentCulture = new System.Globalization.CultureInfo("pl-PL");
using (DocX document = DocX.Load(@"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\Mój1.docx"))
{
// Check if all the replace patterns are used in the loaded document.
document.ReplaceText("@PrzykładowyTekstDoZmiany0", "PierwszaWstawkaBezSPacji_Znakmi!#$%^)&()", false, System.Text.RegularExpressions.RegexOptions.None, new Formatting()
{
Bold = true
});
document.ReplaceText("@PrzykładowyTekstDoZmiany1", "Zwykły tekst ", false, System.Text.RegularExpressions.RegexOptions.None, new Formatting()
{
Bold = true
});
document.ReplaceText("@PrzykładowyTekstDoZmiany2", "Zwykły tekst, dodany przez aplikację ( ) ", false, System.Text.RegularExpressions.RegexOptions.None, new Formatting()
{
Bold = true
});
var p = document.InsertParagraph();
p.Append("Przykładowy czerwony tekst dodany na końcu,")
.Font(new Font("Arial"))
.FontSize(25)
.Color(System.Drawing.Color.Red)
.Bold()
.Append(" zawierający dodatkowe inne niebieskie formatowanie").Font(new Font("Times New Roman")).Color(System.Drawing.Color.Blue).Italic()
.SpacingAfter(40);
// Save this document to disk.
document.SaveAs(@"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\Mój2.docx");
using (DocX document1 = DocX.Load(@"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\Mój1.docx"))
{
using (DocX document3 = DocX.Load(@"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\Mój2.docx"))
{
document1.InsertDocument(document3, true);
document1.SaveAs(@"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\Mój3.docx");
}
}
var n = DateTime.Now;
var tempDi = new DirectoryInfo(string.Format("ExampleOutput-{0:00}-{1:00}-{2:00}-{3:00}{4:00}{5:00}", n.Year - 2000, n.Month, n.Day, n.Hour, n.Minute, n.Second));
tempDi.Create();
/*
* This example loads each document into a byte array, then into a memory stream, so that the document can be opened for writing without
* modifying the source document.
*/
//foreach (var file in Directory.GetFiles("../../", "*.docx"))
//{
// ConvertToHtml(file, tempDi.FullName);
//}
byte[] byteArray = File.ReadAllBytes(@"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\Mój2.docx");
using (MemoryStream memoryStream = new MemoryStream())
{
memoryStream.Write(byteArray, 0, byteArray.Length);
using (WordprocessingDocument doc = WordprocessingDocument.Open(memoryStream, true))
{
HtmlConverterSettings settings = new HtmlConverterSettings()
{
PageTitle = "My Page Title"
};
XElement html = HtmlConverter.ConvertToHtml(doc, settings);
File.WriteAllText(@"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\Mój2.html", html.ToStringNewLineOnAttributes());
}
}
string test = File.ReadAllText(@"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\Mój2.hmtl");
//Byte[] res = null;
//using (MemoryStream ms = new MemoryStream())
//{
// var pdf = TheArtOfDev.HtmlRenderer.PdfSharp.PdfGenerator.GeneratePdf(test, PdfSharp.PageSize.A4);
// pdf.Save(@"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\Mój2.pdf");
// res = ms.ToArray();
//}
//Load Document
Spire.Doc.Document document2 = new Spire.Doc.Document();
document2.LoadFromFile(@"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\Mój2.docx");
//Convert Word to PDF
document2.SaveToFile("toPDF.PDF", FileFormat.PDF);
//Launch Document
System.Diagnostics.Process.Start("toPDF.PDF");
string[] pdfFiles = { @"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\bin\Debug\toPDF.PDF", @"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\bin\Debug\toPDF.PDF", @"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\bin\Debug\toPDF.PDF", @"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\bin\Debug\toPDF.PDF", @"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\bin\Debug\toPDF.PDF", @"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\bin\Debug\toPDF.PDF" };
PdfDocument outputPDFDocument = new PdfDocument();
foreach (string pdfFile in pdfFiles)
{
PdfDocument inputPDFDocument = PdfReader.Open(pdfFile, PdfDocumentOpenMode.Import);
outputPDFDocument.Version = inputPDFDocument.Version;
foreach (PdfPage page in inputPDFDocument.Pages)
{
outputPDFDocument.AddPage(page);
}
}
outputPDFDocument.Save(@"C:\Users\Ernest\source\repos\Word_Pdf\Word_Pdf\bin\Debug\toPDF2.PDF");
}
}
public static HtmlConverterSettings Settings(string htmlPath, string pageTitle)
{
string imageDir = Path.Combine(Path.GetDirectoryName(htmlPath), "_files");
var dirInfo = new DirectoryInfo(imageDir);
if (dirInfo.Exists)
{
foreach (var f in dirInfo.GetFiles())
{
f.Delete();
}
dirInfo.Delete();
}
var imageCounter = 0;
HtmlConverterSettings settings = new HtmlConverterSettings()
{
AdditionalCss = "body { margin: 1cm auto; max-width: 20cm; padding: 0; }",
PageTitle = pageTitle,
FabricateCssClasses = true,
CssClassPrefix = "pt-",
RestrictToSupportedLanguages = false,
RestrictToSupportedNumberingFormats = false,
ImageHandler = imageInfo =>
{
DirectoryInfo localDirInfo = new DirectoryInfo(imageDir);
if (!localDirInfo.Exists)
{
localDirInfo.Create();
}
++imageCounter;
string extension = imageInfo.ContentType.Split('/')[1].ToLower();
ImageFormat imageFormat = null;
if (extension == "png")
{
imageFormat = ImageFormat.Png;
}
else if (extension == "gif")
{
imageFormat = ImageFormat.Gif;
}
else if (extension == "bmp")
{
imageFormat = ImageFormat.Bmp;
}
else if (extension == "jpeg")
{
imageFormat = ImageFormat.Jpeg;
}
else if (extension == "tiff")
{
// Convert tiff to gif.
extension = "gif";
imageFormat = ImageFormat.Gif;
}
else if (extension == "x-wmf")
{
extension = "wmf";
imageFormat = ImageFormat.Wmf;
}
if (imageFormat == null)
{
return(null);
}
string imageFileName = imageDir + "/image" +
imageCounter.ToString() + "." + extension;
try
{
imageInfo.Bitmap.Save(imageFileName, imageFormat);
}
catch (System.Runtime.InteropServices.ExternalException)
{
return(null);
}
string imageSource = localDirInfo.Name + "/image" +
imageCounter.ToString() + "." + extension;
XElement img = new XElement(Xhtml.img,
new XAttribute(NoNamespace.src, imageSource),
imageInfo.ImgStyleAttribute,
imageInfo.AltText != null ?
new XAttribute(NoNamespace.alt, imageInfo.AltText) : null);
return(img);
}
};
return(settings);
}
public static (string title, string content) ConvertToHtml(string docxFilePath, ConverterSettings settings)
{
string title;
string content;
var fi = new FileInfo(docxFilePath);
var byteArray = File.ReadAllBytes(fi.FullName);
using (var memoryStream = new MemoryStream())
{
memoryStream.Write(byteArray, 0, byteArray.Length);
using (var wDoc = WordprocessingDocument.Open(memoryStream, true))
{
title = fi.FullName;
var part = wDoc.CoreFilePropertiesPart;
if (part != null)
{
title = (string)part.GetXDocument().Descendants(DC.title).FirstOrDefault() ?? fi.FullName;
}
title = PathUtils.GetFileNameWithoutExtension(title);
// TODO: Determine max-width from size of content area.
var htmlSettings = new HtmlConverterSettings
{
// AdditionalCss = "body { margin: 1cm auto; max-width: 20cm; padding: 0; }",
PageTitle = title,
FabricateCssClasses = true,
CssClassPrefix = "pt-",
RestrictToSupportedLanguages = false,
RestrictToSupportedNumberingFormats = false,
ImageHandler = imageInfo =>
{
if (settings.IsClearImages || string.IsNullOrEmpty(settings.ImageDirectoryPath))
{
return(null);
}
DirectoryUtils.CreateDirectoryIfNotExists(settings.ImageDirectoryPath);
var extension = imageInfo.ContentType.Split('/')[1].ToLower();
ImageFormat imageFormat = null;
if (extension == "png")
{
imageFormat = ImageFormat.Png;
}
else if (extension == "gif")
{
imageFormat = ImageFormat.Gif;
}
else if (extension == "bmp")
{
imageFormat = ImageFormat.Bmp;
}
else if (extension == "jpeg")
{
imageFormat = ImageFormat.Jpeg;
}
else if (extension == "tiff")
{
// Convert tiff to gif.
extension = "gif";
imageFormat = ImageFormat.Gif;
}
else if (extension == "x-wmf")
{
extension = "wmf";
imageFormat = ImageFormat.Wmf;
}
// If the image format isn't one that we expect, ignore it,
// and don't return markup for the link.
if (imageFormat == null)
{
return(null);
}
var imageFileName = StringUtils.GetShortGuid(false) + "." + extension;
var imageFilePath = PathUtils.Combine(settings.ImageDirectoryPath, imageFileName);
try
{
imageInfo.Bitmap.Save(imageFilePath, imageFormat);
}
catch (System.Runtime.InteropServices.ExternalException)
{
return(null);
}
var imageSource = PageUtils.Combine(settings.ImageDirectoryUrl, imageFileName);
var img = new XElement(Xhtml.img,
new XAttribute(NoNamespace.src, imageSource),
imageInfo.ImgStyleAttribute,
imageInfo.AltText != null ?
new XAttribute(NoNamespace.alt, imageInfo.AltText) : null);
return(img);
}
};
var htmlElement = HtmlConverter.ConvertToHtml(wDoc, htmlSettings);
// Produce HTML document with <!DOCTYPE html > declaration to tell the browser
// we are using HTML5.
var html = new XDocument(
new XDocumentType("html", null, null, null),
htmlElement);
// Note: the xhtml returned by ConvertToHtmlTransform contains objects of type
// XEntity. PtOpenXmlUtil.cs define the XEntity class. See
// http://blogs.msdn.com/ericwhite/archive/2010/01/21/writing-entity-references-using-linq-to-xml.aspx
// for detailed explanation.
//
// If you further transform the XML tree returned by ConvertToHtmlTransform, you
// must do it correctly, or entities will not be serialized properly.
var htmlString = html.ToString(SaveOptions.DisableFormatting);
var htmlDoc = new HtmlDocument();
htmlDoc.LoadHtml(htmlString);
var style = htmlDoc.DocumentNode.SelectSingleNode("//style").OuterHtml;
var body = htmlDoc.DocumentNode.SelectSingleNode("//body").InnerHtml;
// var style = HtmlToWmlConverter.CleanUpCss((string)htmlElement.Descendants().FirstOrDefault(d => d.Name.LocalName.ToLower() == "style"));
content = $"{style}{Environment.NewLine}{body}";
if (settings.IsSaveHtml && !string.IsNullOrEmpty(settings.HtmlDirectoryPath) && DirectoryUtils.IsDirectoryExists(settings.HtmlDirectoryPath))
{
var htmlFilePath = PathUtils.Combine(settings.HtmlDirectoryPath, PathUtils.GetFileNameWithoutExtension(docxFilePath) + ".html");
File.WriteAllText(htmlFilePath, htmlString, Encoding.UTF8);
}
}
}
if (settings.IsFirstLineTitle)
{
var contentTitle = RegexUtils.GetInnerContent("p", content);
contentTitle = StringUtils.StripTags(contentTitle);
if (!string.IsNullOrEmpty(contentTitle) && settings.IsFirstLineRemove)
{
content = StringUtils.ReplaceFirst(contentTitle, content, string.Empty);
}
if (!string.IsNullOrEmpty(contentTitle))
{
contentTitle = contentTitle.Trim();
contentTitle = contentTitle.Trim(' ', ' ');
contentTitle = StringUtils.StripEntities(contentTitle);
}
if (!string.IsNullOrEmpty(contentTitle))
{
title = contentTitle;
}
}
if (settings.IsClearFormat)
{
content = HtmlClearUtils.ClearFormat(content);
}
if (settings.IsFirstLineIndent)
{
content = HtmlClearUtils.FirstLineIndent(content);
}
if (settings.IsClearFontSize)
{
content = HtmlClearUtils.ClearFontSize(content);
}
if (settings.IsClearFontFamily)
{
content = HtmlClearUtils.ClearFontFamily(content);
}
if (settings.IsFirstLineRemove)
{
content = StringUtils.ReplaceFirst(title, content, string.Empty);
}
if (string.IsNullOrEmpty(title))
{
title = PathUtils.GetFileNameWithoutExtension(docxFilePath);
}
return(title, content);
}
public static string Convert(FileInfo fileInfo, string outputDir)
{
var byteArray = File.ReadAllBytes(fileInfo.FullName);
using var memoryStream = new MemoryStream();
memoryStream.Write(byteArray, offset: 0, byteArray.Length);
using var wordDoc = WordprocessingDocument.Open(memoryStream, isEditable: true);
var htmlFileName = new FileInfo(fileInfo.Name.Replace(".docx", ".html"));
if (!string.IsNullOrEmpty(outputDir))
{
var dirInfo = new DirectoryInfo(outputDir);
if (!dirInfo.Exists)
{
throw new Exception("Output directory does not exist");
}
htmlFileName = new FileInfo(Path.Combine(dirInfo.FullName, htmlFileName.Name));
}
var imageDir = htmlFileName.FullName.Substring(startIndex: 0, htmlFileName.FullName.Length - 5) + "_images";
var imageCounter = 0;
var pageTitle = fileInfo.FullName;
var part = wordDoc.CoreFilePropertiesPart;
if (part != null)
{
pageTitle = (string)part.GetXDocument().Descendants(DC.title).FirstOrDefault() ?? fileInfo.FullName;
}
var settings = new HtmlConverterSettings
{
AdditionalCss = "body { margin: 1cm auto; max-width: 20cm; padding: 0; }",
PageTitle = pageTitle,
FabricateCssClasses = true,
CssClassPrefix = "pt-",
RestrictToSupportedLanguages = false,
RestrictToSupportedNumberingFormats = false,
ImageHandler = imageInfo =>
{
var localDirInfo = new DirectoryInfo(imageDir);
if (!localDirInfo.Exists)
{
localDirInfo.Create();
}
++imageCounter;
var extension = imageInfo.ContentType.Split(separator: '/')[1].ToLower();
ImageFormat imageFormat = null;
switch (extension)
{
case "png":
imageFormat = ImageFormat.Png;
break;
case "gif":
imageFormat = ImageFormat.Gif;
break;
case "bmp":
imageFormat = ImageFormat.Bmp;
break;
case "jpeg":
imageFormat = ImageFormat.Jpeg;
break;
case "tiff":
// Convert tiff to gif.
extension = "gif";
imageFormat = ImageFormat.Gif;
break;
case "x-wmf":
extension = "wmf";
imageFormat = ImageFormat.Wmf;
break;
}
// If the image format isn't one that we expect, ignore it,
// and don't return markup for the link.
if (imageFormat == null)
{
return(null);
}
var imageFileName = imageDir + "/image" +
imageCounter + "." + extension;
try
{
imageInfo.Bitmap.Save(imageFileName, imageFormat);
}
catch (ExternalException)
{
return(null);
}
var imageSource = localDirInfo.Name + "/image" +
imageCounter + "." + extension;
var img = new XElement(Xhtml.img,
new XAttribute(NoNamespace.src, imageSource),
imageInfo.ImgStyleAttribute,
imageInfo.AltText != null ? new XAttribute(NoNamespace.alt, imageInfo.AltText) : null);
return(img);
}
};
var htmlElement = HtmlConverter.ConvertToHtml(wordDoc, settings);
// Produce HTML document with <!DOCTYPE html > declaration to tell the browser
// we are using HTML5.
var html = new XDocument(
new XDocumentType("html", publicId: null, systemId: null, internalSubset: null),
htmlElement);
var htmlString = html.ToString(SaveOptions.DisableFormatting);
File.WriteAllText(htmlFileName.FullName, htmlString, Encoding.UTF8);
return(htmlFileName.FullName);
}
public static FileInfo ToHtml(string docxPath, string htmlPath)
{
var docxFileInfo = new FileInfo(docxPath);
byte[] docxBytes = File.ReadAllBytes(docxFileInfo.FullName);
var htmlFileInfo = new FileInfo(htmlPath);
htmlFileInfo.Directory.Create();
if (docxBytes.Length <= 0)
{
// Create file with empty content
if (!File.Exists(htmlFileInfo.FullName))
{
File.Create(htmlFileInfo.FullName);
}
return(htmlFileInfo);
}
try
{
using (MemoryStream memoryStream = new MemoryStream())
{
memoryStream.Write(docxBytes, 0, docxBytes.Length);
using (WordprocessingDocument wDoc = WordprocessingDocument.Open(memoryStream, true))
{
var imageDirectoryName = Path.Combine(htmlFileInfo.DirectoryName, $"{Path.GetFileNameWithoutExtension(htmlFileInfo.FullName)}_files");
int imageCounter = 0;
var pageTitle = Path.GetFileNameWithoutExtension(docxFileInfo.Name);
// TODO: Determine max-width from size of content area.
HtmlConverterSettings settings = new HtmlConverterSettings
{
AdditionalCss = "body { margin: 1cm auto; max-width: 20cm; padding: 0; }",
PageTitle = pageTitle,
FabricateCssClasses = true,
CssClassPrefix = "pt-",
RestrictToSupportedLanguages = false,
RestrictToSupportedNumberingFormats = false,
ImageHandler = imageInfo =>
{
++imageCounter;
string extension = imageInfo.ContentType.Split('/')[1].ToLower();
ImageFormat imageFormat = null;
if (extension == "png")
{
imageFormat = ImageFormat.Png;
}
else if (extension == "gif")
{
imageFormat = ImageFormat.Gif;
}
else if (extension == "bmp")
{
imageFormat = ImageFormat.Bmp;
}
else if (extension == "jpeg")
{
imageFormat = ImageFormat.Jpeg;
}
else if (extension == "tiff")
{
// Convert tiff to gif.
extension = "gif";
imageFormat = ImageFormat.Gif;
}
else if (extension == "x-wmf")
{
extension = "wmf";
imageFormat = ImageFormat.Wmf;
}
// If the image format isn't one that we expect, ignore it, and don't
// return markup for the link.
if (imageFormat == null)
{
return(null);
}
string imageFileName = $"{imageDirectoryName}/image_{imageCounter.ToString()}.{extension}";
try
{
imageInfo.Bitmap.Save(imageFileName, imageFormat);
}
catch (System.Runtime.InteropServices.ExternalException)
{
return(null);
}
string imageSource = $"{htmlFileInfo.Directory.Name}/image_{imageCounter.ToString()}.{extension}";
XElement img =
new XElement(
Xhtml.img,
new XAttribute(NoNamespace.src, imageSource),
imageInfo.ImgStyleAttribute,
imageInfo.AltText != null
? new XAttribute(NoNamespace.alt, imageInfo.AltText)
: null);
return(img);
}
};
XElement htmlElement = HtmlConverter.ConvertToHtml(wDoc, settings);
// Produce HTML document with <!DOCTYPE html > declaration to tell the
// browser we are using HTML5.
var html = new XDocument(new XDocumentType("html", null, null, null), htmlElement);
// Note: the xhtml returned by ConvertToHtmlTransform contains objects of type
// XEntity. PtOpenXmlUtil.cs define the XEntity class. See
// http://blogs.msdn.com/ericwhite/archive/2010/01/21/writing-entity-references-using-linq-to-xml.aspx
// for detailed explanation. // If you further transform the XML tree
// returned by ConvertToHtmlTransform, you must do it correctly, or entities
// will not be serialized properly.
var htmlString = html.ToString(SaveOptions.DisableFormatting);
File.WriteAllText(htmlFileInfo.FullName, htmlString, Encoding.UTF8);
}
}
}
catch
{
// Ignore
}
return(htmlFileInfo);
}
public async Task ConvertWordToRazorViewAsync(IFormFile BlogBody, Blog Blog)
{
fileName = ContentDispositionHeaderValue.Parse(BlogBody.ContentDisposition).FileName.Trim('"');
if (fileName.EndsWith(".docx"))
{
var oldBlogFileUploads = new DirectoryInfo($"{_environment.WebRootPath}\\BlogFiles");
if (oldBlogFileUploads.Exists)
{
if (oldBlogFileUploads.GetFiles() != null)
{
foreach (var file in oldBlogFileUploads.GetFiles())
{
file.Delete();
}
}
oldBlogFileUploads.Delete();
}
var BlogFiles = Directory.CreateDirectory("BlogFiles");
var BlogFileUploads = Path.Combine(_environment.WebRootPath, BlogFiles.FullName);
var newFile = fileName;
var newHtmldoc = ChangeFileExtension(".docx", ".cshtml", BlogBody);
var newHtmldocPath = Path.Combine(BlogFileUploads, newHtmldoc);
var newPath = Path.Combine(BlogFileUploads, newFile);
await BlogBody.SaveAsAsync(newPath);
using (WordprocessingDocument doc = WordprocessingDocument.Open(newPath, true))
{
var settings = new HtmlConverterSettings
{
PageTitle = Blog.BlogTitle
};
var htmlDocument = HtmlConverter.ConvertToHtml(doc, settings);
File.WriteAllText(newHtmldocPath, htmlDocument.ToStringNewLineOnAttributes());
//get folder: "BlogFiles"
var BlogUploadsdir = new DirectoryInfo(newHtmldocPath).Parent.ToString();
//get the parent folder full path, i.e the application name.
var parentDir = new DirectoryInfo(BlogUploadsdir).Parent.FullName.ToString();
//get the "Views" folder of the application.
var viewsDir = GetViewsDirectory(parentDir);
//get a collection of directories under "Views" folder and create "BlogUploads" folder.
var ViewsDirs = new DirectoryInfo(viewsDir).EnumerateDirectories();
var blogUploadsDir = ReturnHomeDir(ViewsDirs).CreateSubdirectory("BlogUploads");
//create new path for the newly created Razor view file and move it there.
var viewHtmldocPath = Path.Combine(blogUploadsDir.FullName, newHtmldoc);
File.Move(newHtmldocPath, viewHtmldocPath);
//persist the path to the newly created Razor View in the database.
Blog.BlogBody = $"{ConvertToRelativePath(viewHtmldocPath)}";
}
}
}
public override string ConvertToHtml(FileInfo file, EpubWriter writer)
{
this.LastStatus = $"Processing File: {file.Name}";
byte[] byteArray = File.ReadAllBytes(file.FullName);
using (MemoryStream memoryStream = new MemoryStream())
{
memoryStream.Write(byteArray, 0, byteArray.Length);
using (WordprocessingDocument wDoc = WordprocessingDocument.Open(memoryStream, true))
{
var pageTitle = file.FullName;
var part = wDoc.CoreFilePropertiesPart;
if (part != null)
{
pageTitle = (string)part.GetXDocument().Descendants(DC.title).FirstOrDefault() ?? file.FullName;
}
HtmlConverterSettings settings = new HtmlConverterSettings()
{
AdditionalCss = "body { margin: 1cm auto; max-width: 20cm; padding: 0; }",
PageTitle = pageTitle,
FabricateCssClasses = true,
CssClassPrefix = "pt-",
RestrictToSupportedLanguages = false,
RestrictToSupportedNumberingFormats = false,
ImageHandler = imageInfo =>
{
string extension = imageInfo.ContentType.Split('/')[1].ToLower();
SysImg.ImageFormat imageFormat = null;
if (extension == "png")
{
imageFormat = SysImg.ImageFormat.Png;
}
else if (extension == "gif")
{
imageFormat = SysImg.ImageFormat.Gif;
}
else if (extension == "bmp")
{
imageFormat = SysImg.ImageFormat.Bmp;
}
else if (extension == "jpeg")
{
imageFormat = SysImg.ImageFormat.Jpeg;
}
else if (extension == "tiff")
{
// Convert tiff to gif.
extension = "gif";
imageFormat = SysImg.ImageFormat.Gif;
}
else if (extension == "x-wmf")
{
extension = "wmf";
imageFormat = SysImg.ImageFormat.Wmf;
}
// If the image format isn't one that we expect, ignore it,
// and don't return markup for the link.
if (imageFormat == null)
{
return(null);
}
string fileName = $"{Guid.NewGuid()}.{extension}";
string filePath = $"images/{fileName}";
using (var ms = new MemoryStream())
{
imageInfo.Bitmap.Save(ms, SysImg.ImageFormat.Png);
writer.AddFile(filePath, ms.ToArray(), EpubSharp.Format.EpubContentType.ImagePng);
}
XElement ximg = new XElement(Xhtml.img,
new XAttribute(NoNamespace.src, filePath),
imageInfo.ImgStyleAttribute,
imageInfo.AltText != null ?
new XAttribute(NoNamespace.alt, imageInfo.AltText) : null);
return(ximg);
}
};
XElement htmlElement = HtmlConverter.ConvertToHtml(wDoc, settings);
var html = new XDocument(
new XDocumentType("html", null, null, null),
htmlElement);
var htmlString = html.ToString(SaveOptions.DisableFormatting);
return(htmlString);
}
}
}
protected override void ProcessRecord()
{
foreach (var document in AllDocuments("Export-OpenXmlToHtml"))
{
try
{
if (!(document is WmlDocument))
{
throw new PowerToolsDocumentException("Not a wordprocessing document.");
}
FileInfo info = null;
if (document.FileName != null)
{
info = new FileInfo(document.FileName);
}
string htmlFileName;
if (OutputPath != null)
{
htmlFileName = System.IO.Path.Combine(SessionState.Path.CurrentLocation.Path, OutputPath);
}
else
{
if (info == null)
{
throw new ArgumentException("No output file name available.");
}
htmlFileName = System.IO.Path.Combine(info.DirectoryName, info.Name.Substring(0, info.Name.Length - info.Extension.Length) + ".html");
}
string imageFolder;
FileInfo outputInfo = new FileInfo(htmlFileName);
if (ImageFolder != null)
{
imageFolder = SessionState.Path.GetResolvedPSPathFromPSPath(ImageFolder).First().Path;
}
else
{
imageFolder = System.IO.Path.Combine(outputInfo.DirectoryName, "images");
}
DirectoryInfo imageFolderInfo = new DirectoryInfo(imageFolder);
if (!imageFolderInfo.Exists)
{
imageFolderInfo.Create();
}
HtmlConverterSettings settings = new HtmlConverterSettings();
settings.PageTitle = PageTitle;
if (CssPath != null)
{
settings.CssClassPrefix = CssClassPrefix;
// Read CSS file into a string
string cssFileName = SessionState.Path.GetResolvedPSPathFromPSPath(CssPath).First().Path;
settings.Css = File.ReadAllText(cssFileName);
}
int imageCounter = 0;
XElement html = HtmlConverter.ConvertToHtml((WmlDocument)document, settings,
imageInfo =>
{
++imageCounter;
string extension = imageInfo.ContentType.Split('/')[1].ToLower();
ImageFormat imageFormat = null;
if (extension == "png")
{
// Convert png to jpeg.
extension = "jpeg";
imageFormat = ImageFormat.Jpeg;
}
else if (extension == "bmp")
{
imageFormat = ImageFormat.Bmp;
}
else if (extension == "jpeg")
{
imageFormat = ImageFormat.Jpeg;
}
else if (extension == "tiff")
{
imageFormat = ImageFormat.Tiff;
}
// If the image format isn't one that we expect, ignore it,
// and don't return markup for the link.
if (imageFormat == null)
{
return(null);
}
string imageFileName = System.IO.Path.Combine(imageFolder, "image" + imageCounter.ToString() + "." + extension);
try
{
imageInfo.Bitmap.Save(imageFileName, imageFormat);
}
catch (System.Runtime.InteropServices.ExternalException)
{
return(null);
}
XElement img = new XElement(Xhtml.img,
new XAttribute(NoNamespace.src, imageFileName),
imageInfo.ImgStyleAttribute,
imageInfo.AltText != null ?
new XAttribute(NoNamespace.alt, imageInfo.AltText) : null);
return(img);
});
File.WriteAllText(htmlFileName, html.ToStringNewLineOnAttributes());
}
catch (Exception e)
{
WriteError(PowerToolsExceptionHandling.GetExceptionErrorRecord(e, document));
}
}
}
private static void ConvertToHtml(byte[] byteArray, DirectoryInfo destDirectory,
string htmlFileName)
{
FileInfo fiHtml = new FileInfo(Path.Combine(destDirectory.FullName, htmlFileName));
using (MemoryStream memoryStream = new MemoryStream())
{
memoryStream.Write(byteArray, 0, byteArray.Length);
using (WordprocessingDocument wDoc =
WordprocessingDocument.Open(memoryStream, true))
{
var imageDirecotryFullName =
fiHtml.FullName.Substring(0, fiHtml.FullName.Length - fiHtml.Extension.Length) + "_files";
var imageDirectoryRelativeName =
fiHtml.Name.Substring(0, fiHtml.Name.Length - fiHtml.Extension.Length) + "_files";
int imageCounter = 0;
//var pageTitle = (string)wDoc
// .CoreFilePropertiesPart
// .GetXDocument()
// .Descendants(DC.title)
// .FirstOrDefault();
HtmlConverterSettings settings = new HtmlConverterSettings()
{
PageTitle = "File",
FabricateCssClasses = true,
CssClassPrefix = "pt-",
RestrictToSupportedLanguages = false,
RestrictToSupportedNumberingFormats = false,
ImageHandler = imageInfo =>
{
DirectoryInfo localDirInfo = new DirectoryInfo(imageDirecotryFullName);
if (!localDirInfo.Exists)
{
localDirInfo.Create();
}
++imageCounter;
string extension = imageInfo.ContentType.Split('/')[1].ToLower();
ImageFormat imageFormat = null;
if (extension == "png")
{
extension = "gif";
imageFormat = ImageFormat.Gif;
}
else if (extension == "gif")
{
imageFormat = ImageFormat.Gif;
}
else if (extension == "bmp")
{
imageFormat = ImageFormat.Bmp;
}
else if (extension == "jpeg")
{
imageFormat = ImageFormat.Jpeg;
}
else if (extension == "tiff")
{
extension = "gif";
imageFormat = ImageFormat.Gif;
}
else if (extension == "x-wmf")
{
extension = "wmf";
imageFormat = ImageFormat.Wmf;
}
if (imageFormat == null)
{
return(null);
}
FileInfo imageFileName = new FileInfo(imageDirecotryFullName + "/image" +
imageCounter.ToString() + "." + extension);
try
{
imageInfo.Bitmap.Save(imageFileName.FullName, imageFormat);
}
catch (System.Runtime.InteropServices.ExternalException)
{
return(null);
}
XElement img = new XElement(Xhtml.img,
new XAttribute(NoNamespace.src, imageDirectoryRelativeName + "/" + imageFileName.Name),
imageInfo.ImgStyleAttribute,
imageInfo.AltText != null ?
new XAttribute(NoNamespace.alt, imageInfo.AltText) : null);
return(img);
}
};
// XElement html = HtmlConverter.ConvertToHtml(wDoc, settings);
XElement html = HtmlConverter.ConvertToHtml(wDoc, settings);
var body = html.Descendants(Xhtml.body).First();
//body.AddFirst(
// new XElement(Xhtml.p,
// new XElement(Xhtml.a,
// new XAttribute("href", "WebForm1.aspx"), "Go back to upload page")));
var htmlString = html.ToString(SaveOptions.DisableFormatting);
File.WriteAllText(fiHtml.FullName, htmlString, Encoding.UTF8);
}
}
}
public static void WordToHtml()
{
byte[] byteArray = File.ReadAllBytes(@"D:\OfficeDev\Word\201701\Test.docx");
string css = @"
p.PtNormal
{margin-bottom:10.0pt;
font-size:11.0pt;
font-family:""Times"";}
span.PtDefaultParagraphFont
{margin-top:24.0pt;
font-size:14.0pt;
font-family:""Helvetica"";
color:yellow;}
h1.PtHeading1
{margin-top:24.0pt;
font-size:14.0pt;
font-family:""Helvetica"";
color:blue;}
h2.PtHeading2
{margin-top:10.0pt;
font-size:13.0pt;
font-family:""Helvetica"";
color:blue;}";
using (MemoryStream memoryStream = new MemoryStream())
{
memoryStream.Write(byteArray, 0, byteArray.Length);
using (WordprocessingDocument doc = WordprocessingDocument.Open(memoryStream, true))
{
int imageCounter = 0;
HtmlConverterSettings settings = new HtmlConverterSettings()
{
PageTitle = "My Page Title",
CssClassPrefix = "Pt",
AdditionalCss = css,
ImageHandler = imageInfo =>
{
DirectoryInfo localDirInfo = new DirectoryInfo(@"D:\OfficeDev\Word\201701\img");
if (!localDirInfo.Exists)
{
localDirInfo.Create();
}
++imageCounter;
string extension = imageInfo.ContentType.Split('/')[1].ToLower();
ImageFormat imageFormat = null;
if (extension == "png")
{
extension = "gif";
imageFormat = ImageFormat.Gif;
}
else if (extension == "gif")
{
imageFormat = ImageFormat.Gif;
}
else if (extension == "bmp")
{
imageFormat = ImageFormat.Bmp;
}
else if (extension == "jpeg")
{
imageFormat = ImageFormat.Jpeg;
}
else if (extension == "tiff")
{
extension = "gif";
imageFormat = ImageFormat.Gif;
}
else if (extension == "x-wmf")
{
extension = "wmf";
imageFormat = ImageFormat.Wmf;
}
if (imageFormat == null)
{
return(null);
}
string imageFileName = @"D:\OfficeDev\Word\201701\image" +
imageCounter.ToString() + "." + extension;
try
{
imageInfo.Bitmap.Save(imageFileName, imageFormat);
}
catch (System.Runtime.InteropServices.ExternalException)
{
return(null);
}
XElement img = new XElement(Xhtml.img,
new XAttribute(NoNamespace.src, imageFileName),
imageInfo.ImgStyleAttribute,
imageInfo.AltText != null ?
new XAttribute(NoNamespace.alt, imageInfo.AltText) : null);
return(img);
}
};
XElement html = HtmlConverter.ConvertToHtml(doc, settings);
File.WriteAllText(@"D:\OfficeDev\Word\201701\WordHtml.html", html.ToStringNewLineOnAttributes());
};
}
}
public XElement ConvertToHTML(byte[] byteArray)
{
XElement html;
using (MemoryStream memoryStream = new MemoryStream())
{
memoryStream.Write(byteArray, 0, byteArray.Length);
using (WordprocessingDocument doc = WordprocessingDocument.Open(memoryStream, true))
{
int imageCounter = 0;
HtmlConverterSettings settings = new HtmlConverterSettings()
{
PageTitle = "My Page Title",
FabricateCssClasses = true,
AdditionalCss = "",
GeneralCss = "",
CssClassPrefix = "",
RestrictToSupportedLanguages = false,
RestrictToSupportedNumberingFormats = false,
ImageHandler = imageInfo =>
{
DirectoryInfo localDirInfo = new DirectoryInfo("img");
if (!localDirInfo.Exists)
{
localDirInfo.Create();
}
++imageCounter;
string extension = imageInfo.ContentType.Split('/')[1].ToLower();
ImageFormat imageFormat = null;
if (extension == "png")
{
extension = "gif";
imageFormat = ImageFormat.Gif;
}
else if (extension == "gif")
{
imageFormat = ImageFormat.Gif;
}
else if (extension == "bmp")
{
imageFormat = ImageFormat.Bmp;
}
else if (extension == "jpeg")
{
imageFormat = ImageFormat.Jpeg;
}
else if (extension == "tiff")
{
extension = "gif";
imageFormat = ImageFormat.Gif;
}
else if (extension == "x-wmf")
{
extension = "wmf";
imageFormat = ImageFormat.Wmf;
}
if (imageFormat == null)
{
return(null);
}
string imageFileName = "img/image" +
imageCounter.ToString() + "." + extension;
try
{
imageInfo.Bitmap.Save(imageFileName, imageFormat);
}
catch (System.Runtime.InteropServices.ExternalException)
{
return(null);
}
XElement img = new XElement(Xhtml.img,
new XAttribute(NoNamespace.src, imageFileName),
imageInfo.ImgStyleAttribute,
imageInfo.AltText != null ?
new XAttribute(NoNamespace.alt, imageInfo.AltText) : null);
return(img);
}
};
html = HtmlConverter.ConvertToHtml(doc, settings);
//System.IO.File.WriteAllText("kk.html", html.ToStringNewLineOnAttributes());
};
}
return(html);
}
protected override void ProcessRecord()
{
foreach (var document in AllDocuments("Export-OpenXmlToHtml"))
{
try
{
if (!(document is WmlDocument))
throw new PowerToolsDocumentException("Not a wordprocessing document.");
FileInfo info = null;
if (document.FileName != null)
info = new FileInfo(document.FileName);
string htmlFileName;
if (OutputPath != null)
htmlFileName = System.IO.Path.Combine(SessionState.Path.CurrentLocation.Path, OutputPath);
else
{
if (info == null)
throw new ArgumentException("No output file name available.");
htmlFileName = System.IO.Path.Combine(info.DirectoryName, info.Name.Substring(0, info.Name.Length - info.Extension.Length) + ".html");
}
string imageFolder;
FileInfo outputInfo = new FileInfo(htmlFileName);
if (ImageFolder != null)
imageFolder = SessionState.Path.GetResolvedPSPathFromPSPath(ImageFolder).First().Path;
else
imageFolder = System.IO.Path.Combine(outputInfo.DirectoryName, "images");
DirectoryInfo imageFolderInfo = new DirectoryInfo(imageFolder);
if (!imageFolderInfo.Exists)
imageFolderInfo.Create();
HtmlConverterSettings settings = new HtmlConverterSettings();
settings.PageTitle = PageTitle;
if (CssPath != null)
{
settings.CssClassPrefix = CssClassPrefix;
// Read CSS file into a string
string cssFileName = SessionState.Path.GetResolvedPSPathFromPSPath(CssPath).First().Path;
settings.Css = File.ReadAllText(cssFileName);
}
int imageCounter = 0;
XElement html = HtmlConverter.ConvertToHtml((WmlDocument)document, settings,
imageInfo =>
{
++imageCounter;
string extension = imageInfo.ContentType.Split('/')[1].ToLower();
ImageFormat imageFormat = null;
if (extension == "png")
{
// Convert png to jpeg.
extension = "jpeg";
imageFormat = ImageFormat.Jpeg;
}
else if (extension == "bmp")
imageFormat = ImageFormat.Bmp;
else if (extension == "jpeg")
imageFormat = ImageFormat.Jpeg;
else if (extension == "tiff")
imageFormat = ImageFormat.Tiff;
// If the image format isn't one that we expect, ignore it,
// and don't return markup for the link.
if (imageFormat == null)
return null;
string imageFileName = System.IO.Path.Combine(imageFolder, "image" + imageCounter.ToString() + "." + extension);
try
{
imageInfo.Bitmap.Save(imageFileName, imageFormat);
}
catch (System.Runtime.InteropServices.ExternalException)
{
return null;
}
XElement img = new XElement(Xhtml.img,
new XAttribute(NoNamespace.src, imageFileName),
imageInfo.ImgStyleAttribute,
imageInfo.AltText != null ?
new XAttribute(NoNamespace.alt, imageInfo.AltText) : null);
return img;
});
File.WriteAllText(htmlFileName, html.ToStringNewLineOnAttributes());
}
catch (Exception e)
{
WriteError(PowerToolsExceptionHandling.GetExceptionErrorRecord(e, document));
}
}
}
