file type mapping for Office documents [was: Re: ANNOUNCE: amavisd-new-2.11.0-rc1 release candidate is available]

[Thomas Jarosch]

Hi Mark,

On Monday, 21. March 2016 09:08:24 Thomas Jarosch wrote:
> In the previous amavisd version, all Office files types
> were mapped to 'doc' since file(1) reported just "Office document".
> 
> Office 2007+ documents get properly distinguished by file(1).
> 
> We could either do this (I run that locally):
> 
>      [qr/^Microsoft Office Document\b/i  => 'doc'],  # OLE2: doc, ppt,
> xls, ... +    [qr/^Microsoft (Word|Excel|PowerPoint)\b/i  => 'doc'],  #
> Office 2007+: docx, docm, pptx, xlsx, ...
> 
> 
> for the sake of "compatibility"
> 
> *or*
> 
> recognize each file type on it's own to give
> the users a better control about the content type.
> 
> What do you think?

could you revisit this one for 2.11.1? :)

I've opted for this one on my installation:

    [qr/^Microsoft Office Document\b/i     => 'doc'], # OLE2: doc, ppt, xls,...
    [qr/^Microsoft Word\b/i                => 'doc'],
    [qr/^Microsoft Excel\b/i               => 'xls'],
    [qr/^Microsoft PowerPoint\b/i          => 'ppt'],


and also added this as fallback if file(1) is not 100% sure:

    [qr/^Composite Document File\b/i       => 'doc'], # OLE2 container


Best regards,
Thomas