[poppler] pdftohtml : enhancing it to use embedded fonts

[mpsuzuki at hiroshima-u.ac.jp]

On Mon, 6 Jun 2011 00:17:56 +0100
Albert Astals Cid <aacid at kde.org> wrote:
>> If I only use the extracted font to display the original document in it's
>> original form, and not to draw additional glyphs in any document, I
>> believe I'll be in compliance with "fair use" and digital copyright rules
>> for the font.
>
>Copyright law can be funny, but given we are not doing anything, it is the 
>user that does it, i don't think we should be worried at all.

A flag to indicate the permitted use of the font is
included in an OpenType font:

http://www.microsoft.com/typography/otspec/os2.htm#fst

It indicates "can be or cannot be embedded", "can be
or cannot be subsetted", "can be or cannot be modified",
"bitmap or outline", but there is not indication whether
the font can be extracted from the document which the
font is embedded to. It is reasonable (at least when
the spec was designed) because the flag is for the font,
not for the document including the font.

On the side of PDF, the permission to use document are
classified in the encryption dictionary, the permissions
relating with the embedded font handling would be
"enable printing of the document" and
"enable copying of text and graphics from the document".
They can be defined in parallel, so if one (or both) of
them is set to disable, disabling the font extraction
would be safer.

Anyway, it is my personal opinion. There might be some
objection saying that putting such conditional is not
intuitive and letting the users to check the permission
and decide by themselves would be better.

Regards,
mpsuzuki