Mixing text and images causes incorrect vertical positioning - itext

Using iTextSharp version 5.5.8 (same bug existed in 5.5.7), there's an unpleasant bug when you add images to Chapters and Sections - the images and the section headings start out OK but quickly become offset relative to each other.
The PDF generated from the following code starts out correctly, it says "Section 1" and below it is the image. The next section ("Section 2") has a little of the image overlapping the section text, the next section is even worse, etc. I think it's the text that's mal-positioned, not the image.
Is this a known iTextSharp bug?
static Document m_doc = null;
static BaseFont m_helvetica = null;
static Font m_font = null;
static PdfWriter m_writer = null;
static Image m_image = null;
static void Main(string[] args)
{
m_doc = new Document(PageSize.LETTER);
m_helvetica = BaseFont.CreateFont(BaseFont.HELVETICA, BaseFont.CP1252, BaseFont.NOT_EMBEDDED);
m_font = new Font(m_helvetica, 10.0f);
m_writer = PdfWriter.GetInstance(m_doc, new FileStream("Output.pdf", FileMode.Create));
m_writer.StrictImageSequence = true;
m_doc.Open();
m_doc.Add(new Chunk("Created by iTextSharp version " + new iTextSharp.text.Version().GetVersion, m_font));
Chapter chapter = new Chapter("Chapter 1", 1);
chapter.TriggerNewPage = false;
if (m_image == null)
{
m_image = Image.GetInstance(new Uri("https://pbs.twimg.com/profile_images/2002307628/Captura_de_pantalla_2012-03-17_a_la_s__22.14.48.png"));
m_image.ScaleAbsolute(100, 100);
}
for (int i = 0; i < 5; i++)
{
Section section = chapter.AddSection(18, "Section " + (i + 1));
section.Add(new Chunk(" ", m_font));
section.Add(m_image);
}
m_doc.Add(chapter);
m_doc.Close();
}

From the documentation for the Java version:
A Section is a part of a Document containing other Sections, Paragraphs, List and/or Tables.
Further looking at the Add() method in the C# source we see:
Adds a Paragraph, List, Table or another Section
Basically, instead of a Chunk use a Paragraph. So instead of this
section.Add(new Chunk(" ", m_font));
Use this:
section.Add(new Paragraph(new Chunk(" ", m_font)));
Or even just this:
section.Add(new Paragraph(" ", m_font));

Related

LayoutResult more one page in Itext7

i ask this: Remove the first and last lines properties in the paper Itext7
and if i do it:
PdfWriter pdfWriter = new PdfWriter(dest);
PdfDocument pdfDoc = new PdfDocument(pdfWriter);
Div div = new Div();
Document doc = new Document(pdfDoc, PageSize.A5);
doc.setMargins(0,0,0,36);
for (int i = 0; i <50 ; i++) {
ListItem listItem = new ListItem();
String s= "hello "+i;
Paragraph p = new Paragraph();
for (int j = 0; j <s.length() ; j++) {
p.add("HELLO " +I);
}
LayoutResult result = div.createRendererSubTree().setParent(doc.getRenderer()).layout(new LayoutContext(new LayoutArea(0,PageSize.A5)));
List<IRenderer> childRendererParagraph = result.getSplitRenderer().getChildRenderers();
childRendererParagraph contain Paragraphs only from first page.And i don't know how many pages well be in pdf
As I mentioned in the answer to your previous question,
split renderer represent the part of the content which iText can place on the area, overflow - the content which overflows.
So if you want to layout the rest of the content, you can perform the same operation (layout) on your overflowRenderer.
The code is as follows:
LayoutResult firstPageResult = div.createRendererSubTree().setParent(doc.getRenderer()).layout(new LayoutContext(new LayoutArea(0, PageSize.A5)));
LayoutResult secondPageResult = firstPageResult.getOverflowRenderer().layout(new LayoutContext(new LayoutArea(1, PageSize.A5)));
Once the content has been fully placed, the overflowRenderer will be null.
pdf.setDefaultPageSize(new PageSize(595, 14400));
Resize the page to display all content on one page

How do you underline text with dashedline in ITEXT PDF

I have underlined "Some text" by
var par = new Paragraph();
par.Add(new Chunk("Some text", CreateFont(12, Font.UNDERLINE)));
document.Add(par);
It is possible underline just "Some text" with dashed line (not the paragraph)?
Thanks
This answer tells you how to do it but unfortunately doesn't provide any code so I've provided it below.
To the best on my knowledge there isn't a direct way to achieve this in iTextSharp by just setting a simple property. Instead, you need to use a PageEvent and manually draw the line yourself.
First you need to subclass PdfPageEventHelper:
private class UnderlineMaker : PdfPageEventHelper {
}
You then want to override the OnGenericTag() method:
public override void OnGenericTag(PdfWriter writer, Document document, iTextSharp.text.Rectangle rect, string text) {
}
The text variable will hold an arbitrary value that you set later on. Here's a full working example with comments:
private class UnderlineMaker : PdfPageEventHelper {
public override void OnGenericTag(PdfWriter writer, Document document, iTextSharp.text.Rectangle rect, string text) {
switch (text) {
case "dashed":
//Grab the canvas underneath the content
var cb = writer.DirectContentUnder;
//Save the current state so that we don't affect future canvas drawings
cb.SaveState();
//Set a line color
cb.SetColorStroke(BaseColor.BLUE);
//Set a dashes
//See this for more details: http://api.itextpdf.com/itext/com/itextpdf/text/pdf/PdfContentByte.html#setLineDash(float)
cb.SetLineDash(3, 2);
//Move the cursor to the left edge with the "pen up"
cb.MoveTo(rect.Left, rect.Bottom);
//Move to cursor to the right edge with the "pen down"
cb.LineTo(rect.Right, rect.Bottom);
//Actually draw the lines (the "pen downs")
cb.Stroke();
//Reset the drawing states to what it was before we started messing with it
cb.RestoreState();
break;
}
//There isn't actually anything here but I always just make it a habit to call my base
base.OnGenericTag(writer, document, rect, text);
}
}
And below is an implementation of it:
//Create a test file on the desktop
var testFile = Path.Combine(Environment.GetFolderPath(Environment.SpecialFolder.Desktop), "test.pdf");
//Normal iTextSharp boilerplate, nothing special here
using (var fs = new FileStream(testFile, FileMode.Create, FileAccess.Write, FileShare.None)) {
using (var doc = new Document()) {
using (var writer = PdfWriter.GetInstance(doc, fs)) {
doc.Open();
//Bind our helper class
writer.PageEvent = new UnderlineMaker();
//Create a new paragraph
var p = new Paragraph();
//Add a normal chunk
p.Add(new Chunk("This is a "));
//Create another chunk with an arbitrary tag associated with it and add to the paragraph
var c = new Chunk("dashed underline test");
c.SetGenericTag("dashed");
p.Add(c);
//Add the paragraph to the document
doc.Add(p);
doc.Close();
}
}
}
If you wanted to get fancy you could pass a delimited string to SetGenericTag() like dashed-black-2x4 or something and parse that out in the OnGenericTag event.

How do I change the weight of a simulated bold font using itext

I am using the iText library to generate text. I am loading the Arial Unicode MS font which does not contain a bold style so iText is simulating the bold. This works fine, but the weight of the bold font appears too heavy compared with text generated using the Java API or even using Microsoft Word.
I tried to get the weight from the FontDescriptor, but the value returned is always 0.0
float weight = font.getBaseFont().getFontDescriptor(BaseFont.FONT_WEIGHT, fontSize);
Is there a way I can change the weight of a simulated bold font?
As an addendum to #Chris' answer: You do not need to construct those Object[]s as there is a Chunk convenience method:
BaseFont arialUnicodeMs = BaseFont.createFont("c:\\Windows\\Fonts\\ARIALUNI.TTF", BaseFont.WINANSI, BaseFont.EMBEDDED);
Font arial12 = new Font(arialUnicodeMs, 12);
Paragraph p = new Paragraph();
for (int i = 1; i < 100; i++)
{
Chunk chunk = new Chunk(String.valueOf(i) + " ", arial12);
chunk.setTextRenderMode(PdfContentByte.TEXT_RENDER_MODE_FILL_STROKE, i/100f, null);
p.add(chunk);
}
document.add(p);
results in
EDIT
Sorry, I just realized after posting this that you're using iText but my answer is for iTextSharp. You should, however, be able to use most of the code below. I've updated the source code link to reference the appropriate Java source.
Bold simulation (faux bold) is done by drawing the text with a stroke. When iText is asked to draw bold text with a non-bold font it defaults to applying a stroke with a width of of the font's size divided by 30. You can see this in the current source code here. The magic part is setting the chunk's text rendering mode to a stroke of your choice:
//.Net code
myChunk.Attributes[Chunk.TEXTRENDERMODE] = new Object[] { PdfContentByte.TEXT_RENDER_MODE_FILL_STROKE, MAGIC_NUMBER_HERE, null };
//Java code
myChunk.attributes.put(Chunk.TEXTRENDERMODE, new Object[]{Integer.valueOf(PdfContentByte.TEXT_RENDER_MODE_FILL_STROKE), MAGIC_NUMBER_HERE, null});
Knowing that you can just apply the same logic but using your weight preference. The sample below creates four chunks, the first normal, the second faux-bold, the third ultra-heavy faux-bold and the fourth ultra-lite faux-bold.
//.Net code below but should be fairly easy to convert to Java
//Path to our PDF
var testFile = Path.Combine(Environment.GetFolderPath(Environment.SpecialFolder.Desktop), "test.pdf");
//Path to our font
var ff = Path.Combine(Environment.GetFolderPath(Environment.SpecialFolder.Fonts), "ARIALUNI.TTF");
//Normal document setup, nothing special here
using (var fs = new FileStream(testFile, FileMode.Create, FileAccess.Write, FileShare.None)) {
using (var doc = new Document()) {
using (var writer = PdfWriter.GetInstance(doc, fs)) {
doc.Open();
//Register our font
FontFactory.Register(ff, "Arial Unicode MS");
//Declare a size to use throughout the demo
var size = 20;
//Get a normal and a faux-bold version of the font
var f = FontFactory.GetFont("Arial Unicode MS", BaseFont.IDENTITY_H, BaseFont.EMBEDDED, size, iTextSharp.text.Font.NORMAL);
var fb = FontFactory.GetFont("Arial Unicode MS", BaseFont.IDENTITY_H, BaseFont.EMBEDDED, size, iTextSharp.text.Font.BOLD);
//Create a normal chunk
var cNormal = new Chunk("Hello ", f);
//Create a faux-bold chunk
var cFauxBold = new Chunk("Hello ", fb);
//Create an ultra heavy faux-bold
var cHeavy = new Chunk("Hello ", f);
cHeavy.Attributes = new Dictionary<string, object>();
cHeavy.Attributes[Chunk.TEXTRENDERMODE] = new Object[] { PdfContentByte.TEXT_RENDER_MODE_FILL_STROKE, size / 10f, null };
//Create a lite faux-bold
var cLite = new Chunk("Hello ", f);
cLite.Attributes = new Dictionary<string, object>();
cLite.Attributes[Chunk.TEXTRENDERMODE] = new Object[] { PdfContentByte.TEXT_RENDER_MODE_FILL_STROKE, size / 50f, null };
//Add to document
var p = new Paragraph();
p.Add(cNormal);
p.Add(cFauxBold);
p.Add(cHeavy);
p.Add(cLite);
doc.Add(p);
doc.Close();
}
}
}

iTextSharp does not render header/footer when using element generated by XmlWorkerHelper

I am trying to add header/footer to a PDF whose content is otherwise generated by XMLWorkerHelper. Not sure if it's a placement issue but I can't see the header/footer.
This is in an ASP.NET MVC app using iTextSharp and XmlWorker packages ver 5.4.4 from Nuget.
Code to generate PDF is as follows:
private byte[] ParseXmlToPdf(string html)
{
XhtmlToListHelper xhtmlHelper = new XhtmlToListHelper();
Document document = new Document(PageSize.A4, 30, 30, 90, 90);
MemoryStream msOutput = new MemoryStream();
PdfWriter writer = PdfWriter.GetInstance(document, msOutput);
writer.PageEvent = new TextSharpPageEventHelper();
document.Open();
var htmlContext = new HtmlPipelineContext(null);
htmlContext.SetTagFactory(Tags.GetHtmlTagProcessorFactory());
var cssResolver = XMLWorkerHelper.GetInstance().GetDefaultCssResolver(false);
cssResolver.AddCssFile(HttpContext.Server.MapPath("~/Content/themes/mytheme.css"), true);
var pipeline = new CssResolverPipeline(cssResolver, new HtmlPipeline(htmlContext, new PdfWriterPipeline(document, writer)));
var worker = new XMLWorker(pipeline, true);
var parser = new XMLParser();
parser.AddListener(worker);
using (TextReader sr = new StringReader(html))
{
parser.Parse(sr);
}
//string text = "Some Random Text";
//for (int k = 0; k < 8; ++k)
//{
// text += " " + text;
// Paragraph p = new Paragraph(text);
// p.SpacingBefore = 8f;
// document.Add(p);
//}
worker.Close();
document.Close();
return msOutput.ToArray();
}
Now instead of using these three lines
using (TextReader sr = new StringReader(html))
{
parser.Parse(sr);
}
if I comment them out and uncomment the code to add a random Paragraph of text (commented in above sample), I see the header/footer along with the random text.
What am I doing wrong?
The EventHandler is as follows:
public class TextSharpPageEventHelper : PdfPageEventHelper
{
public Image ImageHeader { get; set; }
public override void OnEndPage(PdfWriter writer, Document document)
{
float cellHeight = document.TopMargin;
Rectangle page = document.PageSize;
PdfPTable head = new PdfPTable(2);
head.TotalWidth = page.Width;
PdfPCell c = new PdfPCell(ImageHeader, true);
c.HorizontalAlignment = Element.ALIGN_RIGHT;
c.FixedHeight = cellHeight;
c.Border = PdfPCell.NO_BORDER;
head.AddCell(c);
c = new PdfPCell(new Phrase(
DateTime.UtcNow.ToString("yyyy-MM-dd HH:mm:ss") + " GMT",
new Font(Font.FontFamily.COURIER, 8)
));
c.Border = PdfPCell.TOP_BORDER | PdfPCell.RIGHT_BORDER | PdfPCell.BOTTOM_BORDER | PdfPCell.LEFT_BORDER;
c.VerticalAlignment = Element.ALIGN_BOTTOM;
c.FixedHeight = cellHeight;
head.AddCell(c);
head.WriteSelectedRows(
0, -1, // first/last row; -1 flags all write all rows
0, // left offset
// ** bottom** yPos of the table
page.Height - cellHeight + head.TotalHeight,
writer.DirectContent
);
}
}
Feeling angry and stupid for having lost more than two hours of my life to Windows 8.1! Apparently the built in PDF Reader app isn't competent enough to show 'Headers'/'Footers'. Installed Foxit Reader and opened the output and it shows the Headers allright!
I guess I should try with Adobe Reader too!!!
UPDATES:
I tried opening the generated PDF in Adobe Acrobat reader and finally saw some errors. This made me look at the HTML that I was sending to the XMLWorkerHelper more closely. The structure of the HTML had <div>s inside <table>s and that was tripping up the generated PDF. I cleaned up the HTML to ensure <table> inside <table> and thereafter the PDF came out clean in all readers.
Long story short, the code above is fine, but if you are testing for PDF's correctness, have Adobe Acrobat Reader handy at a minimum. Other readers behave unpredictably as in either not reporting errors or incorrectly rendering the PDF.

iText - Strange column- / page changing behaviour with ColumnText

I am quite new to iText and trying to accomplish the following:
read a list of text files from local hd
arrange the texts of the files in a 2-column layout pdf file
add a consecutively numbered index before each text
I started with the MovieColumns1 example (http://itextpdf.com/examples/iia.php?id=64) and ended up with the following code:
final float[][] COLUMNS_COORDS = { { 36, 36, 296, 806 }, { 299, 36, 559, 806 } };
Document document = new Document(PageSize.A4);
PdfWriter writer = PdfWriter.getInstance(document, resultFile);
document.open();
ColumnText ct = new ColumnText(writer.getDirectContent());
ct.setSimpleColumn(COLUMNS_COORDS[0][0], COLUMNS_COORDS[0][1],
COLUMNS_COORDS[0][2], COLUMNS_COORDS[0][3]);
File textDir = new File("c:/Users/raddatz/Desktop/123/texts/");
File[] files = textDir.listFiles();
int i = 1;
int column = 0;
for (File file : files) {
String text = FileUtils.readFileToString(file, "UTF-8");
float yLine = ct.getYLine();
System.out.println("adding '" + file.getName() + "'");
PdfPCell theText = new PdfPCell(new Phrase(text, new Font(Font.HELVETICA, 10)));
theText.setBorder(Rectangle.NO_BORDER);
theText.setPaddingBottom(10);
PdfPCell runningNumber = new PdfPCell(new Phrase(new DecimalFormat("00").format(i++), new Font(
Font.HELVETICA, 14, Font.BOLDITALIC,
new Color(0.7f, 0.7f, 0.7f))));
runningNumber.setBorder(Rectangle.NO_BORDER);
runningNumber.setPaddingBottom(10);
PdfPTable table = new PdfPTable(2);
table.setWidths(new int[] { 12, 100 });
table.addCell(runningNumber);
table.addCell(theText);
ct.addElement(table);
int status = ct.go(true);
if (ColumnText.hasMoreText(status)) {
column = Math.abs(column - 1);
if (column == 0) {
document.newPage();
System.out.println("inserting new page with size :" + document.getPageSize());
}
ct.setSimpleColumn(
COLUMNS_COORDS[column][0], COLUMNS_COORDS[column][1],
COLUMNS_COORDS[column][2], COLUMNS_COORDS[column][3]);
yLine = COLUMNS_COORDS[column][3];
System.out.println("correcting yLine to: " + yLine);
} else {
ct.addElement(table);
}
ct.setYLine(yLine);
System.out.println("before adding: " + ct.getYLine());
status = ct.go(false);
System.out.println("after adding: " + ct.getYLine());
System.out.println("--------------------------------");
}
document.close();
Here you can see the result:
http://d.pr/f/NEmx
Looking at the first page of the resulting PDF I assumed everything was working out fine.
But on second page you can see the problem(s):
text #31 is not displayed completely (first line + index is cut / not in visible area)
text #46 is not displayed completely (first three lines + index is cut / not in visible area)
On page 3 everything seems to be ok again. I am really lost here.
-- UPDATE (2013-03-14) --
I have analyzed the contents of the PDF now. The problem is not that content is show in non-visible areas but that the content is not present in the pdf at all. The missing part of the content is exactly the one which would have fit in the previous column / page. So it seems like ColumnText.go(true) is manipulating the object passed by addElement() before. Can someone confirm this? If so: what can I do about it?
-- end UPDATE (2013-03-14) --
Looking forward to your reply
regards,
sven
Solved! As soon as ColumnText indicates a table will not fit the current column I reinitialize ct with a new instance of ColumnText and add the table again.
In other words:
Each instance of ColumnText is exactly dealing one column of my document.
if (ColumnText.hasMoreText(status) || mediaPoolSwitch) {
column = Math.abs(column - 1);
if (mediaPoolSwitch) {
currentMediaPool = mediapool;
column = 0;
}
if (column == 0) {
document.newPage();
writeTitle(writer.getDirectContent(), mediapool.getName());
}
ct = new ColumnText(writer.getDirectContent());
ct.addElement(table);
ct.setSimpleColumn(
COLUMNS_COORDS[column][0], COLUMNS_COORDS[column][1],
COLUMNS_COORDS[column][2], COLUMNS_COORDS[column][3]);
yLine = COLUMNS_COORDS[column][3];
LOG.debug("correcting yLine to: " + yLine);
} else {
ct.addElement(table);
}
ct.setYLine(yLine);
ct.go();