public static string Parse(int siteId, string filePath, bool isClearFormat, bool isFirstLineIndent, bool isClearFontSize, bool isClearFontFamily, bool isClearImages)
{
if (string.IsNullOrEmpty(filePath))
{
return(string.Empty);
}
var filename = PathUtils.GetFileNameWithoutExtension(filePath);
//被转换的html文档保存的位置
try
{
var saveFilePath = PathUtils.GetTemporaryFilesPath(filename + ".html");
FileUtils.DeleteFileIfExists(saveFilePath);
WordDntb.buildWord(filePath, saveFilePath);
var parsedContent = FileUtils.ReadText(saveFilePath, System.Text.Encoding.Default);
parsedContent = RegexUtils.GetInnerContent("body", parsedContent);
//try
//{
// parsedContent = HtmlClearUtils.ClearElementAttributes(parsedContent, "p");
//}
//catch { }
if (isClearFormat)
{
parsedContent = HtmlClearUtils.ClearFormat(parsedContent);
}
if (isFirstLineIndent)
{
parsedContent = HtmlClearUtils.FirstLineIndent(parsedContent);
}
if (isClearFontSize)
{
parsedContent = HtmlClearUtils.ClearFontSize(parsedContent);
}
if (isClearFontFamily)
{
parsedContent = HtmlClearUtils.ClearFontFamily(parsedContent);
}
if (isClearImages)
{
parsedContent = StringUtils.StripTags(parsedContent, "img");
}
else
{
var siteInfo = SiteManager.GetSiteInfo(siteId);
var imageFileNameArrayList = RegexUtils.GetOriginalImageSrcs(parsedContent);
if (imageFileNameArrayList != null && imageFileNameArrayList.Count > 0)
{
foreach (var imageFileName in imageFileNameArrayList)
{
var imageFilePath = PathUtils.GetTemporaryFilesPath(imageFileName);
var fileExtension = PathUtils.GetExtension(imageFilePath);
var uploadDirectoryPath = PathUtility.GetUploadDirectoryPath(siteInfo, fileExtension);
var uploadDirectoryUrl = PageUtility.GetSiteUrlByPhysicalPath(siteInfo, uploadDirectoryPath, true);
if (!FileUtils.IsFileExists(imageFilePath))
{
continue;
}
var uploadFileName = PathUtility.GetUploadFileName(siteInfo, imageFilePath);
var destFilePath = PathUtils.Combine(uploadDirectoryPath, uploadFileName);
FileUtils.MoveFile(imageFilePath, destFilePath, false);
parsedContent = parsedContent.Replace(imageFileName, PageUtils.Combine(uploadDirectoryUrl, uploadFileName));
FileUtils.DeleteFileIfExists(imageFilePath);
}
}
}
FileUtils.DeleteFileIfExists(filePath);
FileUtils.DeleteFileIfExists(saveFilePath);
return(parsedContent.Trim());
}
catch (Exception ex)
{
LogUtils.AddErrorLog(ex);
return(string.Empty);
}
}
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);
}
