Skip to main content
P
PS-Interp
Known Participant
February 4, 2017
Question

illegal FontBBox

  • February 4, 2017
  • 5 replies
  • 1153 views

I have an illegal /FontBBox in a truetype font in a postscript job. I send this file to a printer, the printer flushes the job with "illegalfont" error message.

I have a composite font with a CMap and the base font is the truetype font with the illegal /FontBBox. I send this file to a printer, the printer happily prints the job.

Any expert can shed any lights on this?

This topic has been closed for replies.

5 replies

T
Test Screen Name
Legend
February 6, 2017

So you have nothing in the PS Language Reference that says that THIS error SHALL be detected and reported. So I have nothing more to add.

T
Test Screen Name
Legend
February 6, 2017

Please give a reference to a page in the PD Language Reference. Which supports your claim that all errors must be reported. I believe I understand your description fully so there is no need for examples.

P
PS-InterpAuthor
Known Participant
February 6, 2017

Is the section " 3.11 Errors" on page 114 are you referring to?

Whether or not to report an error based on the error handler is a different issue from what I was referring to. I am searching for the reason why the same exact error is being reported on one kind of font but not on the other kind. The printer error handler is the same and the job does not change the error handling behavior.

 

T
Test Screen Name
Legend
February 6, 2017

Please give a specific reference. Until you do I can only say this is normal and unsurprising.

P
PS-InterpAuthor
Known Participant
February 6, 2017

:

11 dict begin/FontType 42 def

/FontName /SampleFont def

/PaintType 0 def

/FontMatrix def

/FontBBox {500 2 div 400 100 add 100 300 100 sub} def

::

/MyFont currentdict end definefont setfont

1 0 0 setrgbcolor

100 100 moveto

<01020304> show

Send this job to an Adobe printer, you get

ERROR: invalidfontOFFENDING COMMAND: show

 

T
Test Screen Name
Legend
February 5, 2017

No that's not at all what I said. Not a bug, probably not in the reference manual: you just cannot expect that an error will be reported. Different font, different code, different day, different programmer.

P
PS-InterpAuthor
Known Participant
February 6, 2017

The spec states that if there is an error it will be reported.  As in the case of the illegal FontBBox in the font dictionary,  an illegalfont error is reported. I expect that the same illegalfont error be reported from the same illegal FontBBox in the CID-Keyed font but the printer interpreter is not following its own spec in this case.

发送自我的三星设备

T
Test Screen Name
Legend
February 6, 2017

Please give an exact reference fir this very surprising statement in your first sentence.

T
Test Screen Name
Legend
February 5, 2017

Perhaps you missed my reply, which I think covers this all... A RIP is NOT a validator. If a file contains an error the RIP is not required to do anything in particular (with certain specific exceptions in the PS Language). If an error is detected it might produce a PostScript error. Or the value might be ignored as if it isn't there. Or forced in range. Or something weird and unpredictable might happen. It is the responsibility of the PostScript to be correct !

P
PS-InterpAuthor
Known Participant
February 5, 2017

Thanks so much for the replay. I am sorry for not making my question more clear and precise.

I was perplexed by the fact that why the Adobe printer interpreter treats the same error differently.  

I was searching the Postscript Language Reference 3rd edition and could not find an answer.

When the illegal /FontBBox is in a simple truetype font, the printer flush the job with an error message"illegalfont from offending command show".

When the same illegal /FontBBox is in a CID-Keyed composite font, /Fonttype 0 with /FMapType 9, the same printer does not respond to the same illegal /FontBBox and print the job.

Possibly this is an Adobe bug but most likely it is something that somewhere in the reference book that states that /FontBBox is ignored for CID-Keyed fonts. I just have not found this "somewhere" and I hope that someone out there would know this.

Terms & ConditionsAccessibility statement
Using the community Terms of use Privacy Cookie preferences Do not sell or share my personal information ad choicesAdChoices