DynaPDF Manual - Page 561

Previous Page 560 Index Next Page 562

Function Reference

Page 561 of 787

A correct handling of the graphics state is always required if the application needs to compute

coordinates of graphic objects.

Note also that the graphics state can be saved and restored very often. The maximum allowed

nesting level of save/restore graphics state operators in PDF is 28. The application should not

depend on this limit but it can be used for optimizations, e.g. to create a graphics state cache.

Coordinate Spaces

The coordinates of a page extend from left to right and from bottom to top (subject to alteration by

the orientation of the page (see GetOrientation()). Note that SetPageCoords() does not affect the

coordinates returned by the content parser.

The length of a unit in default user space is 1/72 inch. The visible area of a page is defined by the

crop box of the page, or if absent, by the media box (see GetBBox() for further information). The

media box represents the coordinate system in which the page is defined.

In addition to user space, PDF uses a variety of other coordinate spaces for specialized purposes:

• Text coordinates are defined in text space. The transformation from text space to user space

is defined by a text matrix which is provided in the text callback functions. PDF defines also

several text related positioning parameters in the graphics state. However, to make the usage

easier DynaPDF pre-computes the text transformation matrix so that no further calculations

are required to transform the text space to user space. Note that the current leading, which

specifies the distance between two text lines, is already considered in the text matrix. The

text matrix is always provided. To transform text space to user space the text matrix must be

multiplied with the one of the current graphics state.

• Images are defined in image space. An image in PDF is always one unit wide and one unit

high in user space, regardless of the number of samples in the image. The mapping to the

visible region in user space is achieved by temporarily altering the current transformation

matrix. The image position, as well as the width and height must be computed from the

current transformation matrix (take a look at the description of the TInsertImage callback

function for further information).

• Templates (Form XObjects in PDF terms) are defined in form space. The transformation from

form space to user space is defined by a form matrix that is provided in the TBeginTemplate

callback function. However, the matrix parameter of the callback function is optional and can

be set to NULL. Null means the matrix is set to the identity matrix. To transform form space

to user space the form matrix must be multiplied with the one of the current graphics state.

• Patterns are defined in pattern space. The transformation from pattern space to the initial

default user space of the object in which it is used is provided in the TBeginPattern callback

function. However, the matrix parameter of the callback function is optional and can be set

to NULL. Null means the matrix is set to the identity matrix. Note that patterns are treated

like a color. The TBeginPattern callback function starts the definition of the pattern and

Previous topic: The Graphics State

Next topic: Working with Transformation Matrices