desktop-packages team mailing list archive

[Arlo James Barnes <arlo.barnes@xxxxxxxxx>]

Public bug reported:

'pdffonts' (https://en.wikipedia.org/wiki/Poppler_(software)#poppler-
utils) is a command-line utility that reports 0) what fonts are embedded
or referenced in a PDF file, and for each font:

1) If a subset is used, and if so, what subset (using CID-keying for reference)
2) The font type (OTF, 1C, et cetera)
3) The character encoding (UTF8, WinAnsi, custom, et cetera)
4) Whether the font is embedded or referenced
5) Whether the characters in the font are Unicode-mappable
and
6) The unique two-part numerical Object ID for each font.

<<In this case, "font" is used in a similar way to the classical sense,
minus specifying the size; that is, unlike that modern usage which
equates "font" to the older "type", this is analog to "typeface": name
of the family (sometimes including foundry), weight, and obliquity
(among other features) are all specified. This means there can be
multiple objects for multiple variations in a face.>>

The downsides of using 'pdffonts' versus the requested featureset in Evince:
1) Is a separate application from that used to view PDFs.
2) Does not sort the objects by font family.
3) Does not and cannot offer a preview of or icon for each fontface.
4) Is neither sortable nor searchable.
5) Once the ASCII 'table' is generated, is static and cannot be dynamically resized (and looks rather ugly or at least plain).
6) Meaning of the CID key cannot be interactively unpacked as to the specifics
7) Font cannot be extracted (perhaps you would consider this an anti-piracy "feature"?), or removed and replaced with a referenced font (this would probably not be appropriate for either application since they are geared towards reading, not editing, PDF files).
8) It is not possible to see what text uses which font object. In fact, sometimes a PDF embeds a font but never uses it, and this would be very useful to see if the user planned to later optimize the PDF.

This functionality is an obvious addition to the current Evince feature
set, as it is often the case the user wonders "Oh, what font could that
be?" and should be a (relatively, I know these things tend to become a
little harder than expected) straightforward and easy feature to
implement since in one sense the functionality is already explicated in
the code of 'pdffonts'.

Thank you for reading.

** Affects: evince (Ubuntu)
     Importance: Undecided
         Status: New


** Tags: cid font pdffonts poppler postscript unicode wishlist

-- 
You received this bug notification because you are a member of Desktop
Packages, which is subscribed to evince in Ubuntu.
https://bugs.launchpad.net/bugs/1546415

Title:
  [wishlist] Add features from 'pdffonts'

Status in evince package in Ubuntu:
  New

Bug description:
  'pdffonts' (https://en.wikipedia.org/wiki/Poppler_(software)#poppler-
  utils) is a command-line utility that reports 0) what fonts are
  embedded or referenced in a PDF file, and for each font:

  1) If a subset is used, and if so, what subset (using CID-keying for reference)
  2) The font type (OTF, 1C, et cetera)
  3) The character encoding (UTF8, WinAnsi, custom, et cetera)
  4) Whether the font is embedded or referenced
  5) Whether the characters in the font are Unicode-mappable
  and
  6) The unique two-part numerical Object ID for each font.

  <<In this case, "font" is used in a similar way to the classical
  sense, minus specifying the size; that is, unlike that modern usage
  which equates "font" to the older "type", this is analog to
  "typeface": name of the family (sometimes including foundry), weight,
  and obliquity (among other features) are all specified. This means
  there can be multiple objects for multiple variations in a face.>>

  The downsides of using 'pdffonts' versus the requested featureset in Evince:
  1) Is a separate application from that used to view PDFs.
  2) Does not sort the objects by font family.
  3) Does not and cannot offer a preview of or icon for each fontface.
  4) Is neither sortable nor searchable.
  5) Once the ASCII 'table' is generated, is static and cannot be dynamically resized (and looks rather ugly or at least plain).
  6) Meaning of the CID key cannot be interactively unpacked as to the specifics
  7) Font cannot be extracted (perhaps you would consider this an anti-piracy "feature"?), or removed and replaced with a referenced font (this would probably not be appropriate for either application since they are geared towards reading, not editing, PDF files).
  8) It is not possible to see what text uses which font object. In fact, sometimes a PDF embeds a font but never uses it, and this would be very useful to see if the user planned to later optimize the PDF.

  This functionality is an obvious addition to the current Evince
  feature set, as it is often the case the user wonders "Oh, what font
  could that be?" and should be a (relatively, I know these things tend
  to become a little harder than expected) straightforward and easy
  feature to implement since in one sense the functionality is already
  explicated in the code of 'pdffonts'.

  Thank you for reading.

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/evince/+bug/1546415/+subscriptions