Hi,
I think it is a realy good idea. Currently, preflight inherite from the PDFBox parser to check some stuff during the parsing like the max length of a String, max number of element in an array or the presence of EOL character after the 'obj' keyword. Having a 'profile' that represents the PDF version/standard should provide more flexibility : - possibility to add callback process for a type of COS element. (on read / write) - link a set of high level controls that must be process for the validation of a PDF to check if it is consisent with the underlying standard. PDFBox could provide a default parser/writer that doesn't care of this profile and a sub class that is "aware of the profile" that will manage additionnal processes. BR, Eric 2013/8/7 Maruan Sahyoun <[email protected]> > Hi, > > working on the parser/lexer (unfortunately very slowly) there are some > potential changes to the COS model. As part of that I would like to discuss > if we should look for adding support for PDF versions and PDF standards > such as PDF/A and PDF/X into PDFBox. E.g. one could enable a 'profile' when > generating a PDF and dependent on that we compress streams, choose the Xref > type and maybe complain if certain criteria are not met such as > transparency used in PDF/A-1. > > WDYT > > Maruan Sahyoun > >
