Web based PDF viewer using PDFTron.

Q: Do you have any plans to make a web based PDF viewer? For example
PDF to HTML5 (like scrib project).

A: PDFTron already offers a range of products that can be used for
viewing PDF within the browser.

SilverDox (http://www.pdftron.com/silverdox/) is a solution to view,
markup and manipulate PDF and other formats in Silverlight. By far,
SilverDox offers the better performance, features, quality, and
reliability compared to any HTML5, SVG, or AJAX offering. The only
potential downside is the Silverlight requirement.

In case the Silverlight is not an option, the second best choice would
be to use SVG (wrapped in HTML). This is a browser (W3C) friendly
solution and the mapping from PDF and graphics quality and is better
than HTML5. Unfortunately SVG support is not part of many existing
browsers and the implementation in the ones that do support SVG is
sketchy and inconsistent. In case you do want to investigate this
option you could use ‘pdftron.PDF.Convert.ToSvg()’ method in PDFNet
Convert Add-on or PDF2SVG (http://www.pdftron.com/pdf2svg/index.html).

There is lots of talk (mostly buzz and misconception) about HTML5. It
seems that there are many people who believe that HTML5 can be used as
replacement for PDF graphics model. In theory web fonts are used to
render the typography, and canvas is used for drawing operations. In
practice HTML5 graphics model can’t really be used to reliably render
PDF graphics and this why there are so many documents that don't
render correctly/accurately on scribd). There are many reasons why
this would not work (starting from fonts, custom encodings,
unsupported transparency and blend modes, soft masks, color management
etc). Add to this that HTML5 is support is missing or is partial/
inconsistent in all current browsers and the perceived benefits of
HTML5 (in the context of replacement for PDF) quickly fade away.

In case you are looking for best possible results and portability with
the current browsers the only option is to render PDF to images and
use JavaScript for text search and selection (similar to Google PDF


In case you are looking for solution along these lines, PDFTron is
about to release an AJAX PDF Viewer add-on for PDFNet SDK. If you are
interested in this option please let us know and we will provide you
with more information.

To wrap up the above discussion it is also possible to use PDFNet SDK
to convert to plain old HTML – where PDF graphics (e.g. vector art and
images) are stored as a background image and text is using CSS. This
approach suffers from the same type of issues present in HTML5 but is
better supported on older browsers. PDFNet SDK comes with a full
source code for a PDF to HTML converter (in C#
http://www.pdftron.com/pdfnet/samplecode/Pdf2Html.cs) that you could
use as a starting point for your development.