#29107

We publish a version, which fixes the issue, later today or tomorrow.

#29113

You need to determine how the ASP.Net application propagates session ID. It could be a session ID embedded to URL or (most probably) a cookie.

Cookie can be added to HTTP request this way:

pd4ml.setCookie( "JSESSIONID", cookieValue );
where cookieValue is something like that: “9034657927465;path=/”

JSESSIONID keyword is specific for Java, .Net uses anything else.

#26778

Hi,
I’m new to PD4ML and I need your help with a little issue.
Is there a way in PD4ML to set the pdf page margins? I can’t find it.
I use the following HTML:

<!doctype html>
<html lang="en">
<head>
<meta charset="UTF-8">
<title>Export</title>
<style>
body {margin:0;padding:0;}
</style>
</head>
<body>
<table width="8.125in" cellpadding="0" cellspacing="0" border="0" align="center">
<tr><td>test</td></tr>
</table>
</body>
</html>

Thanks in advance!

#29115

The following info should help:
https://pd4ml.com/cookbook/pdf_page_formatting.htm

#27432

Have there been any updates/changes with regard to this? I’m considering using a charting library that generates SVG charts. I could use Batik to convert those to JPEG/PNG, but it would be great if I could instead embed the SVG in the PDF.

Thanks,
Matt

#26779

When an certain snippet of HTML exists in a document, it will make the renderer enter some kind of death spiral where it will fill up the heap until the machine stops responding.

The snippet seems to be

<SPAN style="white-space:pre" class="Apple-tab-span"> </SPAN>

In context of a very simple example:

<table>
<tr>
<td>
TEXT
<!--<SPAN style="white-space:pre" class="Apple-tab-span"> </SPAN>-->
MORE TEXT
<SPAN style="white-space:pre" class="Apple-tab-span">   </SPAN>
FINAL TEXT
</td>
</tr>
</table>

Here’s a histogram of my heap after trying to render this HTML:

num     #instances         #bytes  class name
<hr class="bbcode_rule" />
1:      12253005      784192320  org.zefer.html.doc.view.b
2:           517       55412344  [Ljava.lang.Object;
3:         26309        3785624  <constMethodKlass>
4:         26309        3590760  <methodKlass>
5:          2671        2474760  [B
6:          1786        2400848  <constantPoolKlass>
7:          1786        1494536  <instanceKlassKlass>
8:          1504        1396576  <constantPoolCacheKlass>
9:         18028        1279584  [C
10:          8536         603344  [I
11:         18388         588416  java.lang.String
12:         12400         396800  java.util.HashMap$Entry
13:           457         321504  <methodDataKlass>
14:          1966         242192  java.lang.Class
15:          2781         204224  [S
16:          2765         173064  [[I
17:          6991         111856  java.lang.Integer
18:           304         108560  [Ljava.util.HashMap$Entry;
19:           163          95192  <objArrayKlassKlass>
20:          1761          56352  java.util.concurrent.ConcurrentHashMap$HashEntry

And the following is a very short Java program to demonstrate the issue:

String html = "<table><tr><td>TEXT<SPAN style="white-space:pre" class="Apple-tab-span"> </SPAN>MORE TEXT<SPAN style="white-space:pre" class="Apple-tab-span">   </SPAN>FINAL TEXT</td></tr></table>";
StringReader sr = new StringReader(html);
OutputStream os = new ByteArrayOutputStream();

PD4ML pd4ml = new PD4ML();
pd4ml.enableDebugInfo();
pd4ml.useTTF( "java:", true );
pd4ml.setHtmlWidth(1050);
pd4ml.enableSmartTableBreaks(true);

try {
pd4ml.render(sr, os);
} catch (Exception e) { }

sop(((ByteArrayOutputStream)os).toByteArray());

#26780

My colleague wrote two JAVA programs, one is JAVA agent in Lotus Notes that will convert email into HTML. Another one is a web service, which will receive the HTML and convert it into PDF by pd4ml.
When the email got a attachment which filename is non-English (In my test case, it is Traditional Chinese characters.), the PDF converted from it cannot show the correct filename of the attachment while the filenames as a text in the content of the email can be showed correctly.

I have checked that the “description” of the attachment still got the correct filename and the HTML header had already specified the encoding in UTF-8. Also, the JAVA program for changing the HTML to PDF has set the TTF and the property file had been created for the fonts.

I have also tried the latest version PD4ML Java v3.8.0fx7.

What should I do?
Thank you very much.

#29116

We reproduced the problem and work on it.

#29097

Did you try to add debug=”true” attribute to (I suppose you use PD4ML JSP taglib) and inspect the server log? Are there any traces it tries to load the image and fails?

#26781

I am using PD4ML 380fx6 taglibs which generates a PDF. This works fine in Chrome, Firefox and IE9 +.

But when I do it in IE8 I get the attached error.

This is the transform tag I am using

<pd4ml:transform
screenWidth="1100"
pageFormat="A4"
pageOrientation="portrait"
pageInsets="15,15,15,15,points"
inline="false"
generateHyperlinks="false"
debug="true"
outline="headings"
fileName="Report.pdf">
<%@ taglib uri="http://tiles.apache.org/tags-tiles" prefix="tiles"%>

Http Response headers

Cache-Control:private
Cache-Control:no-cache
Connection:close
Content-disposition:attachment; filename=Report.pdf
Content-Length:10646
Content-Type:application/pdf
Date:Wed, 19 Sep 2012 16:07:16 GMT
Expires:0
Pragma:no-cache
Pragma:cache
Server:Apache-Coyote/1.1

Anyone have any ideas?

#29123

I think this may be related to it.

http://blogs.msdn.com/b/ieinternals/archive/2009/10/02/internet-explorer-cannot-download-over-https-when-no-cache.aspx

Any idea how to fix this using the PD4ML TagLib?

#29118

Here are more details of the problem.

The original filenames of the attachments in the email:

The filenames of the attachments in the rendered PDF, please note that the filenames in the content are correct:

Thanks.

#26782

In the latest version (3.80fx7), when using background-color: transparent the rendering can be incorrect.

For the sample HTML given:

<table style='background-color: white; color: black; border: 2px solid red;'>
<tr>
<td style='background-color: transparent'>TEXT MORE TEXT FINAL TEXT</td>
</tr>
<tr>
<td>2 TEXT MORE TEXT FINAL TEXT</td>
</tr>
<tr>
<td style='background-color: transparent; color: yellow'>TEXT MORE TEXT FINAL TEXT</td>
</tr>
</table>

I would expect a table with a 2 pixel border, white background and black text. It should have three rows, each with a white background, and the first two rows black text while the third has yellow text. Instead, what I get is the rows that have background-color: transparent are getting BLACK background – I’m guessing it has something to do with how the tables are rendered using layering in the Java graphics package.

#29119

It looks like the file names are UTF-8-encoded, but treated as Latin by PD4ML.

Please make sure you use the most recent dxl4pd4ml.xsl and pd4ml(_demo).jar. The actual stylesheet can be obtained from https://pd4ml.com/command-line-dxl-to-pdf-converter.htm

Also it is a good idea to dump the document source as DXL and to try to convert it offline with the command line tool. It simplifies debug a lot. -dumphtml switch should help you to analyse if the names in HTML are correct, before the doc passed to PDF converter.

#29124

It looks like the problem is caused by a caching of documents, downloaded by HTTPS. Please correct if I am wrong.
It was a known issue of older PD4ML JSP taglib versions, until we found a safe HTTP header directives set.

Currently PD4ML taglib sets only the following HTTP cache control directives:

((HttpServletResponse) sr).setHeader("Pragma", "cache");
((HttpServletResponse) sr).setHeader("Expires", "0");
((HttpServletResponse) sr).setHeader("Cache-control", "private");

In your case there are some other cache controls (including conflicting with the above) – probably the hosting application is configured to add them to each reply. Try to switch it off for PDF generating JSPs.

[Author]

Posts