DynaPDF Manual - Page 599

Previous Page 598   Index   Next Page 600

Function Reference

Page 599 of 839

The function parses the content stream of the current open page. The content parser can be used to

extract text, images, and vector graphics from a PDF file. The parameter Stack holds a set of callback

functions which are executed if corresponding operators were found in the content stream. The

parameter Data is a user defined pointer that is passed unchanged to the callback functions.

All callback functions are optional. Which callback functions must be set depends on the kind of

information that should be extracted. For example, an application that extracts images must at least

provide the callback function TInsertImage.

All callback functions which return an integer value can break processing if necessary. A return

value of zero indicates success and processing continues. A return value of 1 of the TBeginTemplate

or TBeginPattern callback functions indicates that the object should be skipped. The corresponding

content streams are not executed in this case. This can be useful when extracting images. Any other

return value breaks processing.

Notice:

It is allowed to write arbitrary objects into the page while the content parser is executed but it is

strongly required to check whether a fatal error occurred when writing something to the page.

The callback function must return a negative value in such a case to break processing. This is

required because the parser doesn't notice when a fatal error occurs. New objects will be

ignored when parsing a page.

ParseContent() is already part of DynaPDF since version 2.0.30 but it was never documented.

Because the function was undocumented, a few important changes were made in DynaPDF 2.5

which do not break backward compatibility. However, an application that uses the following

features must be slightly changed when it is recompiled:

• The flag pfTranslateStrings is no longer defined because the used callback function specifies

already whether strings should be converted to Unicode. The old constant 1 is ignored.

• Two additional callback functions TShowTextA and TShowTextW were defined at the

reserved fields Reserved0001 and Reserved0002. These callback functions are no longer

defined and no longer executed. These functions could be used in combination with

TShowTextArrayA or TShowTextArrayW only. DynaPDF processes now the entire text with

the array versions. So, existing applications will still work as expected.

The function supports several flags which are useful when extracting images from a PDF file. These

flags are meaningful only, if the TInsertImage callback function is set.

Previous topic: ParseContent

Next topic: The Graphics State