For all owners of a source license we've uploaded a new beta version 6.00 here:

http://www.quickpdf.org/forum/forum_posts.asp?TID=855 - http://www.quickpdf.org/forum/forum_posts.asp?TID=855

 

Please keep in mind that it's a beta version ... obviously few of us already use it in production environments ;-)

 

First checks in own (test-)environments with own "problem-pdf-files" would be greatly appreciated. Please post your experiences with the new version.

 

If you have own bugfixes or functions still not found in the sources please post it here or send it to Michel or me... We'll put it together, make a first test and a new upload.

 

If the new version seems to work in a stable manner we'll begin compiling new versions for the owners of the compiled-licenses, too. 

 

This new version is enhanced with bugfixes (mainly for rendering) from pschott and HarryS. They didn't have to do it but they did it and spent a lot of time with testing and put together. I hope here are a lot wanting the same... Keep in mind: One for all and all for one ;-)

 

We're planning further enhancements for the forum:

A section for reference-files (so each new version first must work properly with this critical pdf-files before release).

A section for source posting (bugfixes and new functions).

 

Again: The new QuickPDF-version needs in the project-settings of a delphi-ide the definition "PMAPP" and a second search-path "...\GdiPlus".

 

Have fun with the new version!

 

Thanks to HarryS, pschott for pushing this new version and for Michel offering this forum and infra structure for us!

 

Best regards,

Ingo

 

 

It will be hard !  

 

 

Thanks again to all  contributors.

 

Best regards,

 

Loïc

 

It goes on and on ...

The next beta version 6.01 is out now at:

http://www.quickpdf.org/forum/forum_posts.asp?TID=857 - http://www.quickpdf.org/forum/forum_posts.asp?TID=857

 

Please try...

In the section "Change request" you can post your experiences with this new version. We call it beta but it's already very stable. There're members here working with this version in very large productive environments! So don't hesitate and try it.

 

Thanks a lot to Harald, Percy, Uli, Oliver and Michel!

 

Best regards,

Ingo

 

 

Here the changes for version 6.00:

From member: HarryS

uiSEDQuickPDF.pas

New Parameter CaseSensitive for GetCatalogInformation
According to older Versions, KeySearch must be set CaseSensitive=True by default!
Depending Units uPDFDocument and UPDFStructure

uPDFGDIPlus.pas is no more used for Rendering, exchanged with
complete Mozilla Opensource Api-Wrapper, due to missing functions in uPDFGDIPlus.pas.
Api-Wrapper is Object-Based.

Render-Bug for embedded monochrome Bitmaps. Were drawn reversed.

Implementing basic functions for Patterns. Renderer had no functionality for
Pdf-documents with patterns. Now it has and is rendering at least the documents of my customers.

uPDFPNG.Pas  Renderexception in MakeIDATChunk corrected. ScanLine -
Bytearray is zerobased, but transfer started from 1! Pictures with odd width chrashed
with memory-acces-Error at the last pixel.

uPDFDocument.pas

New Parameter CaseSensitive
According to older Versions, KeySearch must be CaseSensitive=True by default!

uPDFImageDecoder.pas

Let Monochrome Bitmaps always use mtStencil because
then it is drawn with recognicing the Decode_Entry for reverse.

Adding +1 to only one corner (TopRightY) stretches the Image. I dont
think that this effect was wanted. One of my reference documents has
3 pieces of Images which should be positioned as one complete picture.
But in the picture are small lines between the images. So I think
something still goes wrong here. In standard images you wont see this
one pixel ...

Updfpng.pas

Original-Code ... crashed if there's an odd pixel-count ....
I think we can asume that counting begins with zero for three bytes for each
pixel my variant should be the better one ;-)

uPDFRenderer.pas

Actual PatternObjectName when FillColorSpace = 'Pattern'
Actual PatternObject when FillColorSpace = 'Pattern'

New Procedure FillPattern 2007:
written by HSD Harald Seidlein Datentechnik / HarryS
Creates a clipregion for the actual Picasso.Path and
repeats the pattern into this region as often as
nessesary. Restores the state before, so the initial
rendering can continue.

uPDFStructure.pas

New Paramater CaseSensitive for FindIndexByKeyName
According to older Versions, KeySearch must be set CaseSensitive=True by default!
added Windows-Reference in Uses - List!!!

New Paramater CaseSensitive
According to older Versions, KeySearch must be CaseSensitive by default!


From member: pschott

uPDFExtractor.pas

function TPDFExtractor.ExtractText
function NextToken

uPDFFontRasterizer.pas

function TPDFFontCollection.ConvertCMap
function TPDFFontCollection.ProcessCMap

Changes for version 6.01:

From members: ukobsa / grahl

[+] Two new functions to read/check Page-Labels:
    function GetPageLabelDefinition(const Index: Integer; 
      var Start,  Style, Offset: Integer; 
      var Prefix: string): Integer;
 
    Index: 0 up to GetCountPageLabelDefinition-1
    Start: Pagenumber, where this Pagelabeldefinition start working
    Style: D - Decimal-number
           R - Roman Characters (Upper Case)
           r - Roman Characters (Lower Case)
           A - A..Z for page 1..26, AA..ZZ for page 27..52, ...
           a - Like A but lower case characters
    Offset: To move the starting point for labels. 
            e.g. Page 1, Offset 3, Style R will be III
    Prefix: fix String standing before the Labelstring if you want
    

    function GetCountPageLabelDefinition: integer;
    
    Returns with the Page-Label-count

[+] Destinations: Reading the Destinations was standardized 
                  and enhanced for a better handling with
                  the different types: GetDestArrayByID               

[+] Annotations:  Enhanced with new types.

[+] TPDFDocument: TPDFIndObjList. Improve overall performance 
                  by using a sorted index list. This avoids linear
                  probing to find items.  
                  This replaces FIndObjList of TPDFDocument and 
                  removes duplicated code

 TPDFStructure.SmartDecode: Better performance now:
      Copy-compares replaced by direct string-compares.
      Now there's no need anymore copying into the memory.

 TPDFString.Output: Better performance (like TPDFStructure.SmartDecode)       
[-] There were problems with documents having an xref-section with only
one Cross-Reference-Stream. If you open these ones with QuickPDF and 
save them without showing all pages (or for example only call 
RemoveSharedLayer) you can get a wrong xref-section and informations 
would be lost. To avoid this there's a new Property CheckXREFMode on 
TPDFDocument to handle these documents. For this you can use SetCheckXREFMode. 
Because this needs performance and memory the default setting is without/no.

[-] small Bugfixes in case of memory-usage with formfields

[-] TPDFIndObj.PopText: An error in the analyze-/reporting-function was fixed.

[-} TPDFDocument.FirstPageFrom: Ensure cancel-options!

[-] TPDFStructure.SmartDecode: BugFix for Unicode characters

 

Best regards,

Ingo

 

 

I'm out of office 'till thursday and cannot do anything in this case now but anyway ...

take ingo[dot]schmoekel[at]ewetel[dot]net

 

Best regards,

Ingo