Editable header and footer (Feature #94)

Hossam Elgabarty wrote:

It would be nice to patch automultiplechoice.sty to have the possibility of editing the header and footer via e.g. custom commands.
Currently any attempt to edit the header and/or footer overrides the ones created by AMC.

I do have a foot implementation, using the fancyhdr package. Op until now I avoided the head, but that is possible to, but you must avoid the identification boxes at the top.
I create a centre element in the foot like so:

  \cfoot[\textbf{\examcode}{} \surname,\name{} (\id), date   \examdate%
  ]{\textbf{\examcode}{} \surname,\name{} (\id) date \examdate}

The fancyhdr package is already loaded by amc itselves.
This is sufficient for my purpose.

Maybe you should add some advise to the formatting on the footer (or header).
I noticed that the footer could easily be put in the same area in the foot, which the identification binary boxes take in the head, all in relation to the circles to position the paper during scanning. This would make it hard for your software to detect if the form has been scanned up side down and could there for introduce scanning problems that could be avoided without a footer (in the centre as I currently do).

I noticed this when looking at some intentionally unrecognised scans through the unrecognised scans and comparing original scan and preprocessed windows.
There one can see that the software also makes an attempt to find the binary boxes at the bottom, which is then of course off limits for user defined headers (or footers).

I'm sorry I'm afraid I don't understand exactly what you mean. Can you provide a short sample of a subject that will cause some problems?

the software also makes an attempt to find the binary boxes at the bottom

In the following way: AMC first looks at binary boxes at the same place as on the subject, and then, it looks at binary boxes at the place on a subject that is rotated to 180 degrees.

• for ID detection, AMC makes a pre-processing on the scan, trying to remove pencil drawings, printed characters, etc.
• a false ID detection would be done only if the IDs corresponds to an actual page. That means that the IDs computed from the things you wrote in the footer must give an existing student number, an existing page number, and the exact corresponding check number (the thrid number in the +1/12/49+ format).

However, this problem can (hardly ever) arise, so the advice is to avoid including images (that won't be erased with pre-processing) in the footer.

Excellent that your software is robust in this case.