This repository has been archived by the owner on Nov 17, 2020. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 1
/
README
37 lines (31 loc) · 1.94 KB
/
README
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Auto-rotate filter for cups
---------------------------
Applications might expect that mixed-orientation pdfs print correctly.
As cups, and especially OS X (e.g. when using a remote print queue on OS X via IPP),
expects print-jobs to have homogenous orientations, this filter normalizes them
(also taking 'landscape'/'orientation-requested' into account).
The filter can be controlled by the "pdfAutoRotate" option.
The current default (might change in the future) is enabled.
E.g. setting 'pdfAutoRotate=off' will disable it.
Unfortunately I don't know enought about the cups filter chain creation process
to say, how it can be included there.
A few thing have to be considered, though:
* The filter has to be run even when a remote queue is used.
* All print dialogs should provide a way to switch off the 'pdfAutoRotate'-option,
as the user might want to resolve the print data to media size mismatch differenty,
e.g. by using 'fit-to-page'.
In the future the pdftopdf filter from cups-filters (>1.0.18) could be extended
to handles the auto-rotation internally, saving one temporary file.
BUT this filter still has to be used in all filter chains where a such extended
pdftopdf is not run (but e.g. cgpdftopdf from OS X).
For example some pdftoXY-filters might be called without pdftopdf in front of them
('PDF-job > cgpdftops > pstops > ...' on OS X comes to mind).
Don't expect to be safe, just because your local setup always calls a "good"
pdftopdf, as the IPP printing protocol can forward unfiltered
(or partially filtered) jobs (together with the options set for the job)
when printing to remote queues, and the local pdftopdf is skipped.
In this cases pdfautorotate MUST be called locally for autorotation
to work. Note that some pdftopdf might still be called, so we can't run the
full pdftopdf (or: the options corresponding to the now already applied filters
[like number-up] have to be removed, otherwise the remote end will apply them
again).