Stefan,

Am 27.07.2011 17:31, schrieb stefan.decuypere:
> I am scratching my head on how to use the "PDF" Blocks type (as opposed to
text or image block types).
> In the design, the user creates a block of type PdfBlock and gives it a name
"externalpage"
> When merging the pdfblocks input file with the variable data, I have to
somehow be able to get two data items for this field to be filled (with anything
other than the default pdf and pagenumber)
> One dataitem is the pdf filename, the other one is the pagenumber.
>
> This seems to preclude a one to one mapping of PDFBlocks and data items.
> To be explicit: when I iterate over the pdfblock, and I encounter field
"externalpage", I have to correlate two data items with this: externalfilename,
externalpagenumber

yes and this is necessary, because only with this two information,
you can address a PDF page.

To address the page, you have to open the document first and then
open the page and apply this page handle to the block.

I do not see here any problem. How do you expect to address a Page
in a PDF with only one data?

Rainer

--
Rainer Plöckl        rp@...        http://www.pdflib.com
   PDFlib GmbH, Franziska-Bilek-Weg 9, 80339 München,  Germany
   Amtsgericht München HRB 129497, Geschäftsführer Thomas Merz
---------------------------------------------------------------
  PDFlib 8: Complex script shaping and bidirectional formatting
for Arabic, Hebrew, Thai, Hindi, and many other writing systems
_______PDFlib - a library for generating PDF on the fly________

Stefan,

> I have a question regarding the cookbook example: linked textblock
> How would one handle the case where I want the text "Visit us at our Web site
at www. kraxi.com!" to come below the body text, which is  variable and may be
from 1 to 6 blocks long?
> If I am not mistaken, the bottom text is now always in a fixed position, so
there could be a huge gap between the body text and the "Visit us... " line.
>
> Another case where this may happen is after a variable length body text, where
you would put: "Sincerly Yours, ... "
> Another case is an address that has one or two address lines, but if the
second one is empty you want to move up the town and zip fields so that there is
no empty space.

when I understand you correct, there is a simple way to get all
this information.
Similar as you retrieve the error reason for a textflow block,
you can use info_textflow() to determine the end positions of
the placed text. Based on this information, you can place the
"Visit us..." or the "Sincerly Yours, ...." in the following
process.

Rainer
--
Rainer Plöckl        rp@...        http://www.pdflib.com
   PDFlib GmbH, Franziska-Bilek-Weg 9, 80339 München,  Germany
   Amtsgericht München HRB 129497, Geschäftsführer Thomas Merz
---------------------------------------------------------------
  PDFlib 8: Complex script shaping and bidirectional formatting
for Arabic, Hebrew, Thai, Hindi, and many other writing systems
_______PDFlib - a library for generating PDF on the fly________

I have been working on some C code using PDFlib.
I have set errorpolicy to "return", and have been using function return
values to detect errors.

I have recently encountered error conditions in functions returning
void, which appear to use the PDFlib exception mechanism regardless of
the errorpolicy parameter.  Of course this isn't a big surprise, since
these functions have no other means to report errors.

I have attached a short program demonstrating this.  So far I have
noticed this kind of error with PDF_set_parameter and PDF_begin_page_ext.

If no PDF_TRY/PDF_CATCH block is used, the program will be terminated
with a non-zero exit status and an error message to standard error.

My question is:
If I wish my program to handle these errors differently, do I have any
options other than using the PDF_TRY/PDF_CATCH exception mechanism?

On 01.08.2011 08:50, BSM | Barry Downes wrote:

> If I wish my program to handle these errors differently, do I have any
> options other than using the PDF_TRY/PDF_CATCH exception mechanism?

No. Actually, it is a severe programming error to
call PDFlib API functions without a corresponding
TRY/CATCH clause. PDFlib detects this case and as a last
resort emits a message to stdout before exiting, but
this is only to tell the API user that something is
severely wrong in his program.

Functions without any return value generally don't fail
if used correctly. When designing the API we followed the
paradigm "an exception is an exception": for example, running
out of memory is considered an exceptional situation which
happens rarely, and cannot be recovered by the client.
As another example, if you have a typo in a parameter or
option name an exception will be thrown so that you can see
early in the development cycle that something is severely
wrong.

On the other hand, some operations may fail more often,
e.g. involving disk files: if you load a font or image
from file, the file may be unavailable, damaged or
of unsuitable type for the intended purpose. This kind
of error can be detected with an error return value,
and details are available with PDF_get_errmsg().

Thomas

_______________________________________________________________
Thomas Merz             tm@...            www.pdflib.com
    PDFlib GmbH, Franziska-Bilek-Weg 9, 80339 München,  Germany
    Amtsgericht München HRB 129497, Geschäftsführer Thomas Merz
---------------------------------------------------------------
   PDFlib 8: Complex script shaping and bidirectional formatting
for Arabic, Hebrew, Thai, Hindi, and many other writing systems
_______PDFlib - a library for generating PDF on the fly________

Thanks for the explanation.
If I have errorpolicy=return, and I'm calling a non-void function, is it
possible there are error conditions for which it will throw a PDFlib
exception rather than returning an error code?

I was previously unsure whether errorpolicy=return would allow me to
avoid the exception system entirely. It seems it won't.
As the code I am working on is a library with a number of functions, I
won't have the option to simply wrap the entire program with a single
try/catch block. Probably I'll wrap most PDFlib calls individually.

On 2011-08-01 17:51, Thomas Merz wrote:
>
> On 01.08.2011 08:50, BSM | Barry Downes wrote:
>
> > If I wish my program to handle these errors differently, do I have any
> > options other than using the PDF_TRY/PDF_CATCH exception mechanism?
>
> No. Actually, it is a severe programming error to
> call PDFlib API functions without a corresponding
> TRY/CATCH clause. PDFlib detects this case and as a last
> resort emits a message to stdout before exiting, but
> this is only to tell the API user that something is
> severely wrong in his program.
>
> Functions without any return value generally don't fail
> if used correctly. When designing the API we followed the
> paradigm "an exception is an exception": for example, running
> out of memory is considered an exceptional situation which
> happens rarely, and cannot be recovered by the client.
> As another example, if you have a typo in a parameter or
> option name an exception will be thrown so that you can see
> early in the development cycle that something is severely
> wrong.
>
> On the other hand, some operations may fail more often,
> e.g. involving disk files: if you load a font or image
> from file, the file may be unavailable, damaged or
> of unsuitable type for the intended purpose. This kind
> of error can be detected with an error return value,
> and details are available with PDF_get_errmsg().
>
> Thomas
>
> __________________________________________________________
> Thomas Merz tm@... <mailto:tm%40pdflib.com> www.pdflib.com
> PDFlib GmbH, Franziska-Bilek-Weg 9, 80339 München, Germany
> Amtsgericht München HRB 129497, Geschäftsführer Thomas Merz
> ----------------------------------------------------------
> PDFlib 8: Complex script shaping and bidirectional formatting
> for Arabic, Hebrew, Thai, Hindi, and many other writing systems
> _______PDFlib - a library for generating PDF on the fly________
>
>

On 02.08.2011 03:34, BSM | Barry Downes wrote:
> If I have errorpolicy=return, and I'm calling a non-void function, is it
> possible there are error conditions for which it will throw a PDFlib
> exception rather than returning an error code?

Yes.

> I was previously unsure whether errorpolicy=return would allow me to
> avoid the exception system entirely. It seems it won't.

No. Keep in mind that some error situations cannot be rectified,
neither inside PDFlib nor in client code. Think of out-of-memory,
disk full, and other fatal problems. In such situations PDFlib
will throw an exception regardless of the errorpolicy.

> As the code I am working on is a library with a number of functions, I
> won't have the option to simply wrap the entire program with a single
> try/catch block. Probably I'll wrap most PDFlib calls individually.

Keep in mind that after an exception the current PDF document
is lost; you cannot continue the same document after an
exception but must give it up.

Thomas

_______________________________________________________________
Thomas Merz             tm@...            www.pdflib.com
    PDFlib GmbH, Franziska-Bilek-Weg 9, 80339 München,  Germany
    Amtsgericht München HRB 129497, Geschäftsführer Thomas Merz
---------------------------------------------------------------
   PDFlib 8: Complex script shaping and bidirectional formatting
for Arabic, Hebrew, Thai, Hindi, and many other writing systems
_______PDFlib - a library for generating PDF on the fly________

This is not a bug but more a limitation of the pdfblock v4.3 plugin:

There is no way to see the difference in the preview between a block filled in
from the defaultxxx attribute and a real datavalue, as it would come from a
database and replaced with fill_XXX call.

To be more specific: I have a text block called firstname that has a default
value "Stefan"
If the text block has status active, the preview will show "Stefan"
If the text block has status static, the preview will show "Stefan"
If the text block has status ignore, the preview will be empty.
If the text block has status ignoredefault, the preview will be empty

Would it make sense to add new text attribute previewtext to textblock and
textline blocks,
a new imageblock attribute previewimage and
2 new pdfblock attributes previewpdf previewpdfpage ?

if the block status is ignoredefault, then the preview would take the field
value from these attributes instead of from the default attribute. if the block
status is active, the preview would be a combination of the preview and default
attributes
This has the advantage that one can simulate more accurately the actual output.


To reuse my example from above, if the text block firstname has previewtext
"John" then:
If the text block has status active, the preview will show "John/Stefan"
If the text block has status static, the preview will show "Stefan"
If the text block has status ignore, the preview will not show anything.
If the text block has status ignoredefault, the preview will show "John"


As an alternative to the above suggestion but still trying to achieve the same
result:
when the status of the block is active could the preview maybe show in some way
that the filled in value came from defaultxxx field? [ different
color/underlined/.. whatever ]

Thanks for considering this,
Stefan  Decuypere

I would like to know does Pdflib supports HTML. I have used the <h1> and other
html tags but none of them is supported. Please tell me the sample file in which
it has been used.

Waiting for your earliest response.

Hello,

Am 05.08.2011 06:03, schrieb khaleel:
> I would like to know does Pdflib supports HTML. I have used the <h1> and other
html tags but none of them is supported. Please tell me the sample file in which
it has been used.

no, PDFlib is not a converter and so you can't apply HTML tags
directly.
When working with textflow, you can specify macros, so you might
do inline formatting in a similar way as you know it from HTML.
Please see PDFlib 8 Tutorial, chapter 8.2 (8.2.3 ) for further
details.

Rainer
--
Rainer Plöckl        rp@...        http://www.pdflib.com
   PDFlib GmbH, Franziska-Bilek-Weg 9, 80339 München,  Germany
   Amtsgericht München HRB 129497, Geschäftsführer Thomas Merz
---------------------------------------------------------------
  PDFlib 8: Complex script shaping and bidirectional formatting
for Arabic, Hebrew, Thai, Hindi, and many other writing systems
_______PDFlib - a library for generating PDF on the fly________

When I was experimenting with creating hyperlinks with PDFlib, I found
something weird.
Based on my understanding from what I read, I thought it was a
limitation of PDF that the box for an annotation would always be a
rectangle with sides parallel to the page edges.  If the "usematchbox"
option were used to specify the annotation area, I expected it would use
the smallest such rectangle that enclosed the matchbox.

But in my experiments with diagonal hyperlinks, I found that sometimes
the hyperlink area exactly enclosed the diagonal link.  This was nice,
but unfortunately it didn't always work like that.  Small and
apparently-insignificant changes in the transformation matrix could
affect whether the hyperlink area was precise or used an axis-aligned
rectangle.

It's probably best explained with the attached sample code in C.
It outputs 2 pages with the same hyperlinked text, differing only in the
transformations applied.
For me, the link on the first page was "precise", but the one on the
second page was "imprecise".  By moving my mouse cursor around the
hyperlink and seeing it change, I can see that the link on the second
page is bounded with an axis-aligned rectangle.

If possible, I'd like them to always work like on the first page, but I
have no idea what really causes the difference.

Hi Barry,

Sorry, but this is a bug in Acrobat :-(
The bug has nothing to do with the page where the annotation was
created. You can verify this to create single documents with 1 page
and with different translation values. Documents with translation
values < 150 approximately show the desired behavior: The sensitive
area for the link is the matchbox's rectangle. In the case of values
  >= 150, the sensitve area is the bounding box of matchbox's rectangle.
I didn't examine the precise threshold value. The single documents
differ only in the - correct - coordinates of the annotation's
rectangle and the sensitive area, expressed by a so-called QuadPoints
entry. Incredibly, but it is the truth.

Kurt

On 05.08.2011 11:51, BSM | Barry Downes wrote:
>
> [Attachment(s) <#TopText> from BSM | Barry Downes included below]
>
> When I was experimenting with creating hyperlinks with PDFlib, I found
> something weird.
> Based on my understanding from what I read, I thought it was a
> limitation of PDF that the box for an annotation would always be a
> rectangle with sides parallel to the page edges. If the "usematchbox"
> option were used to specify the annotation area, I expected it would use
> the smallest such rectangle that enclosed the matchbox.
>
> But in my experiments with diagonal hyperlinks, I found that sometimes
> the hyperlink area exactly enclosed the diagonal link. This was nice,
> but unfortunately it didn't always work like that. Small and
> apparently-insignificant changes in the transformation matrix could
> affect whether the hyperlink area was precise or used an axis-aligned
> rectangle.
>
> It's probably best explained with the attached sample code in C.
> It outputs 2 pages with the same hyperlinked text, differing only in the
> transformations applied.
> For me, the link on the first page was "precise", but the one on the
> second page was "imprecise". By moving my mouse cursor around the
> hyperlink and seeing it change, I can see that the link on the second
> page is bounded with an axis-aligned rectangle.
>
> If possible, I'd like them to always work like on the first page, but I
> have no idea what really causes the difference.
>
>

--
_______________________________________________________________
Kurt Stützer           kurt@...           www.pdflib.com
       phone +49 89 452 33 84-15      fax +49 89 452 33 84-99
    PDFlib GmbH, Franziska-Bilek-Weg 9, 80339 München,  Germany
         Court of registry/Amtsgericht München HRB 129497
   Managing Directors/Geschäftsführer: Thomas Merz, Petra Porst
---------------------------------------------------------------
   PDFlib 8: Complex script shaping and bidirectional formatting
for Arabic, Hebrew, Thai, Hindi, and many other writing systems
_______PDFlib - a library for generating PDF on the fly________

Thanks Kurt.

Based on your feedback I tried the document in Foxit Reader, and found
the sensitive area to be the matchbox rectangle in both cases.  It's
nice to know that it's not a limitation in the PDF format itself.

I probably should have made it clear that the 2 pages were just a way of
demonstrating the 2 cases in a single document, created after I'd
already determined various changes that could trigger the problem.  I
initially triggered it with a rotation, but used the translation in my
example because I found it more suprising, and more likely to result in
a properly-considered response.

On 2011-08-06 03:14, Kurt Stützer wrote:
> Hi Barry,
>
> Sorry, but this is a bug in Acrobat :-(
> The bug has nothing to do with the page where the annotation was
> created. You can verify this to create single documents with 1 page
> and with different translation values. Documents with translation
> values<  150 approximately show the desired behavior: The sensitive
> area for the link is the matchbox's rectangle. In the case of values
>   >= 150, the sensitve area is the bounding box of matchbox's rectangle.
> I didn't examine the precise threshold value. The single documents
> differ only in the - correct - coordinates of the annotation's
> rectangle and the sensitive area, expressed by a so-called QuadPoints
> entry. Incredibly, but it is the truth.
>
> Kurt
>
> On 05.08.2011 11:51, BSM | Barry Downes wrote:
>> [Attachment(s)<#TopText>  from BSM | Barry Downes included below]
>>
>> When I was experimenting with creating hyperlinks with PDFlib, I found
>> something weird.
>> Based on my understanding from what I read, I thought it was a
>> limitation of PDF that the box for an annotation would always be a
>> rectangle with sides parallel to the page edges. If the "usematchbox"
>> option were used to specify the annotation area, I expected it would use
>> the smallest such rectangle that enclosed the matchbox.
>>
>> But in my experiments with diagonal hyperlinks, I found that sometimes
>> the hyperlink area exactly enclosed the diagonal link. This was nice,
>> but unfortunately it didn't always work like that. Small and
>> apparently-insignificant changes in the transformation matrix could
>> affect whether the hyperlink area was precise or used an axis-aligned
>> rectangle.
>>
>> It's probably best explained with the attached sample code in C.
>> It outputs 2 pages with the same hyperlinked text, differing only in the
>> transformations applied.
>> For me, the link on the first page was "precise", but the one on the
>> second page was "imprecise". By moving my mouse cursor around the
>> hyperlink and seeing it change, I can see that the link on the second
>> page is bounded with an axis-aligned rectangle.
>>
>> If possible, I'd like them to always work like on the first page, but I
>> have no idea what really causes the difference.
>>
>>

Hi,

I attached some C code where I attempt to fetch a pCOS path that does
not exist.  Specifically, it's a document info field, though I think the
question here applies similarly to any pCOS path that is not guaranteed
to be present.

Ideally I would like to grab a value from the path if it exists, or
receive some indication that it doesn't exist, if it doesn't.

As far as I can tell, I have an option of fetching the value by name, in
which case an invalid path (ie. a missing document info field) will be
signalled by a PDFlib exception.  (This seems to be the case based on my
testing - I don't think the API documentation actually makes this clear.)
Or I may choose to enumerate the keys under "/Info".  This will allow me
to determine the presence of the desired key without risk of an
exception, but doesn't seem the ideal way to obtain a specific set of
keys I'm interested in.

I'd like to go with fetching the value by name, and would be happy
enough to receive the missing path indication as an exception.  For
functions like PDF_pcos_get_number, which have no out-of-band value
suitable for signalling errors reliably, this makes a lot of sense.  But
I gather that after an exception, an output PDF will be unusable and the
entire PDF generation must be aborted.
Is this true even for pCOS functions?
What about if I am composing an output document with the same PDF * handle?
And if I'm not composing an output document, but just using the PDF *
for pCOS input functions?
Am I forced to enumerate the /Info keys (accessing them by number) to
obtain the values safely?

- Barry

As I want to create a chinese bookmark with the function create_bookmark(),but
the chinese cannot be displayed in the right way. When I input "µÚ1Ò³"
,"¦Ì¨²1¨°3" is displayed.
    Can you give me an example how to solve the  problem?

On 8/25/2011 11:10 AM, Vedat Aral wrote:
> Is there a way to insert/append a pdf file onto *existing* pdf file?
>
> It is easy to create a new file with begin_document() and add pages from
> two other files but It is not what I'm looking for.

Not with PDFlib.

-Lance

Thank you Lance,

Do you know if there is a plan to implement it in future versions?


--- In pdflib@yahoogroups.com, Lance Cotton <lcotton+pdflib@...> wrote:
>
> On 8/25/2011 11:10 AM, Vedat Aral wrote:
> > Is there a way to insert/append a pdf file onto *existing* pdf file?
> >
> > It is easy to create a new file with begin_document() and add pages from
> > two other files but It is not what I'm looking for.
>
> Not with PDFlib.
>
> -Lance
>

Hi,

please show us a small program snippet by which we can see what you are
doing.

Kurt

On 25.08.2011 09:56, zjljlj wrote:
>
>
> As I want to create a chinese bookmark with the function
> create_bookmark(),but the chinese cannot be displayed in the right way.
> When I input "µÚ1Ò³" ,"¦Ì¨²1¨°3" is displayed.
> Can you give me an example how to solve the problem?
>
>

--
_______________________________________________________________
Kurt Stützer           kurt@...           www.pdflib.com
       phone +49 89 452 33 84-15      fax +49 89 452 33 84-99
    PDFlib GmbH, Franziska-Bilek-Weg 9, 80339 München,  Germany
         Court of registry/Amtsgericht München HRB 129497
   Managing Directors/Geschäftsführer: Thomas Merz, Petra Porst
---------------------------------------------------------------
   PDFlib 8: Complex script shaping and bidirectional formatting
for Arabic, Hebrew, Thai, Hindi, and many other writing systems
_______PDFlib - a library for generating PDF on the fly________

Hello,


>     Following is the code i used to render the barcode.
>
>       my $optlist = "barcode={symbology=DataMatrix dataprep=0 ecc=0
xsymwidth=10 } "
>                   . " font=" . $id_automation_font_handle;
>
>       $p->create_field( 50,
>                         70,
>                         100,
>                         110,
>                         "0000000 Test",
>                         "textfield",
>                         $optlist );
>
>     But some reason barcode is not rendered on the pdf. Does create_field
function require encrypted string to place the DataMatix barcode? Can anybody
provide perl code for rendering barcode on pdf?
>

do you really want to place the barcode within a form field?
Displaying and printing the barcode fields will only work with
full Acrobat version 9 or newer, but not with Adobe Reader.

When you have an IDAutomationDMatrix font, I think you will
simple place the code as text.

Similar to our (Java) cookbook example:
http://www.pdflib.com/pdflib-cookbook/fonts/barcode-font/

- load the font with $p->load_font()
- place the text with $p->fit_textline():
   $p->fit_textline("*ABC123*", 10, 75, "font=" .
       $id_automation_font_handle ." fontsize=20");

kind regards,
Rainer

--
_______________________________________________________________
Rainer Plöckl        rp@...        http://www.pdflib.com
   PDFlib GmbH, Franziska-Bilek-Weg 9, 80339 München,  Germany
        Court of registry/Amtsgericht München HRB 129497
  Managing Directors/Geschäftsführer: Thomas Merz, Petra Porst
---------------------------------------------------------------
     PDFlib: powerful toolkits for PDF developers since 1997
_______ See www.pdflib.com/products for product details________

On 8/27/2011 5:10 AM, venki wrote:
> Thanks for your code. The above function fit_textline working fine for
> 3of9barcode bur for DataMatrix barcode the function fit_textline showing
> text instead of barcode. Please see the my perl code and output of the
> program
[...]
> my $barcode_font_handle = $p->load_font("IDAutomationDMatrix",
> "unicode", "embedding");
[...]
> $p->begin_page_ext (0, 0, "width=a4.width height=a4.height" );
> $p->fit_textline("*00000001 Test*",
> 10,
> 275,
> $barcode_options );

DataMatrix barcodes that use a font have to be "encoded", so you can't
simply place the text that you want to encode in the barcode with the
font. What usually accompanies DataMatrix fonts is a program or plug-in
tool that transforms your desired plaintext into a series of characters
that map to the right glyphs in the DataMatrix font.

Typically, a DM font will have only 16 real barcode glyphs, arranged as
the 16 permutations of on/off for a 4x4 grid.

Since a DM code is 2-d, you won't be able to fit_textline() because the
encoding tool returns multiple lines so that the barcode is a square or
whatever 2-D shape you have configured it to be.

-Lance

> Am I able to use PDFLib 7.0 with Visual Studio 2010?

there are no know problems with this combination. When you
open the C/C++ project files with Visual Studio 2010, the
projects will be adjusted and then you can compile the examples.

Rainer

--
_______________________________________________________________
Rainer Plöckl        rp@...        http://www.pdflib.com
   PDFlib GmbH, Franziska-Bilek-Weg 9, 80339 München,  Germany
        Court of registry/Amtsgericht München HRB 129497
  Managing Directors/Geschäftsführer: Thomas Merz, Petra Porst
---------------------------------------------------------------
     PDFlib: powerful toolkits for PDF developers since 1997
_______ See www.pdflib.com/products for product details________

Great news - thank you Rainer.

Can I assume this would be true for .NET with vb & c#, also?

:luke

--- In pdflib@yahoogroups.com, Rainer Plöckl <rp@...> wrote:
>
> > Am I able to use PDFLib 7.0 with Visual Studio 2010?
>
> there are no know problems with this combination. When you
> open the C/C++ project files with Visual Studio 2010, the
> projects will be adjusted and then you can compile the examples.
>
> Rainer
>

I should have also included/asked:

Will it work with ASP.NET and C# or VB coding?

Sorry about that.

Thank you,

:luke

--- In pdflib@yahoogroups.com, "lwgracie" <lukegracie@...> wrote:
>
> Great news - thank you Rainer.
>
> Can I assume this would be true for .NET with vb & c#, also?
>
> :luke
>
> --- In pdflib@yahoogroups.com, Rainer Plöckl <rp@> wrote:
> >
> > > Am I able to use PDFLib 7.0 with Visual Studio 2010?
> >
> > there are no know problems with this combination. When you
> > open the C/C++ project files with Visual Studio 2010, the
> > projects will be adjusted and then you can compile the examples.
> >
> > Rainer
> >
>

Hello,

Am 30.08.2011 17:03, schrieb lwgracie:
> I should have also included/asked:
>
> Will it work with ASP.NET and C# or VB coding?

sure, just check the .NET package to get the related projects.

One additional note, when you now start with PDFlib coding, please
do not start with PDFlib 7. PDFlib 7 will reach end of life time
end of October 2011, and will not longer supported.
http://www.pdflib.com/support-policy/product-lifetime/

So it's highly recommended to use the latest version for new
projects.

kind regards,
Rainer


--
_______________________________________________________________
Rainer Plöckl        rp@...        http://www.pdflib.com
   PDFlib GmbH, Franziska-Bilek-Weg 9, 80339 München,  Germany
        Court of registry/Amtsgericht München HRB 129497
  Managing Directors/Geschäftsführer: Thomas Merz, Petra Porst
---------------------------------------------------------------
     PDFlib: powerful toolkits for PDF developers since 1997
_______ See www.pdflib.com/products for product details________

When you call
p.set_parameter("logging", "filename
={\\\\vboxsrv\SharedDocuments\temp\pdflib.log} remove");
you get the error
Error '1012' Call PDF_set_parameter: Couldn't open log file
'\vboxsrv\SharedDocuments\temp\pdflib.log' for writing

Pdflib does not recognize UNC pathnames \\server\volume\pathname as an absolute
path (like x:\temp\pdflib.log)

To check that a filename is absolute in windows, it starts with either x:\ or \\

I have not verified other path-related calls, but I know that p.begin_document
works for unc filenames


Stefan Decuypere

Stefan,

> p.set_parameter("logging", "filename
={\\\\vboxsrv\SharedDocuments\temp\pdflib.log} remove");
> you get the error
> Error '1012' Call PDF_set_parameter: Couldn't open log file
'\vboxsrv\SharedDocuments\temp\pdflib.log' for writing
>
> Pdflib does not recognize UNC pathnames \\server\volume\pathname as an
absolute path (like x:\temp\pdflib.log)
>
> To check that a filename is absolute in windows, it starts with either x:\ or
\\

might it be possible, that you do not work with the last PDFlib
version? In PDFlib 8.0.3 we fixed a bug, which is related to the
description you sent:
[snip]
- 2011-03-29 (bug #3282)
   An even number of consecutive backslash characters in option lists was
   not treated correctly: one of the backslashes was missing after
   optlist parsing.
[snip]

When you think, you have a different problem, please open a support
case: http://www.pdflib.com/en/support-policy/opening-a-support-case/
add the exact used PDFlib version you use as well an example code
to reproduce the issue.

Rainer

--
_______________________________________________________________
Rainer Plöckl        rp@...        http://www.pdflib.com
   PDFlib GmbH, Franziska-Bilek-Weg 9, 80339 München,  Germany
        Court of registry/Amtsgericht München HRB 129497
  Managing Directors/Geschäftsführer: Thomas Merz, Petra Porst
---------------------------------------------------------------
     PDFlib: powerful toolkits for PDF developers since 1997
_______ See www.pdflib.com/products for product details________

Yup, I was using 8.0.1
I will upgrade
Stefan