[sane-devel] sane_read() and padding bytes

Olaf Meeuwissen olaf.meeuwissen at avasys.jp
Wed Jun 13 11:51:36 UTC 2012


Hi,

I'm lazy and didn't feel like removing the padding bytes from the image
data that I pass to the frontend via sane_read().  The SANE API
specification for sane_get_parameters() [1], is very clear about what
SANE frontends should be capable of WRT padding bytes:

  Note that the number of bytes per line can be larger than the minimum
  value imposed by the right side of this equation. A frontend must be
  able to properly cope with such "padded" image formats.

  [1] http://sane.alioth.debian.org/html/doc012.html#s4.3.8

That is, the bytes_per_pixel value of SANE_Parameters only has a lower
limit that is determined by pixels_per_line, format and depth.  There is
no upper limit (other than whatever the maximum of SANE_Int is).  Let's
say we scan a 1 pixel wide image in SANE_FORMAT_GRAY with 8 bits to the
sample.  Then the lower bound on bytes_per_pixel is 1, the upper bound
is 2^31 - 1.  If it pleases me to use the latter (for whatever devious
reasons), the frontend should cope.

Much to my chagrin, I found out that scanimage and xsane do NOT cope
properly with such padded image formats.  Much to my surprise, however,
simple-scan does (but has a bunch of problems of in own in terms of
being a SANE frontend).  I haven't checked any further.

Any chance that these (and other non-compliant) SANE frontends get fixed
and, if so, anytime soon?  It's a bit embarrasing that even scanimage
does not live up to the expectations of the SANE standard ... :-(

In the mean time, though, I guess I have little choice but to not be
lazy and rip out the padding bytes in the backend ... :-((

Hope this helps,
-- 
Olaf Meeuwissen, LPIC-2           FLOSS Engineer -- AVASYS CORPORATION
FSF Associate Member #1962               Help support software freedom
                 http://www.fsf.org/jf?referrer=1962



More information about the sane-devel mailing list