This is going to be a tricky one, particularly since I don't have access
to Illustrator. All I know is the eps and pdf output of simple_plot.py
looks fine for me in ghostscript, acroread, xpdf and evince. Even text
select copy and paste works fine in acroread (with both Type 3 and 42
fonts). So my guess is that unfortunately Illustrator is not fully
covering the spec, and we've inadvertently moved to using a feature it
doesn't like.
Now if this really did work in Illustrator in 0.90.x and not now, one
could start to compare the output by hand and see what the root cause
may be. If you set "pdf.compression" to 0, the Pdf file becomes vaguely
readable, and tools such as diff can help to locate differences. I
would take the output of 0.90.1 and 0.91.2 and compare (the SVN trunk
will be too different to be useful), and then start moving things over
piece by piece until it works again. I'm afraid that without
Illustrator, it will be hard for me to do that.
No other obvious thoughts pop to mind at the moment, but I'll keep
thinking about it in a background process and let you know if I think of
anything.
Cheers,
Mike
Rob Hetland wrote:
>
> Off the list because of the figures.....
>
> Here is an example of a pdf and eps produced from the latest svn (with
> the transforms that broke my polygon editors....)
>
> Illustrator (my version is from CS3) looks identical to the screen
> grabs. The text in the eps is not editable (it is the frame of the
> characters), the text in the pdf _is_ editable, but the characters are
> all single, so, for example cool is a series of independent characters.
>
> Note the yaxis text in the pdf is offset. It looks the same in other
> viewers, so it is not Illustrators fault.
>
> -Rob
>
>
> ------------------------------------------------------------------------
>
>
> ------------------------------------------------------------------------
>
>
>
>
>
> On Jan 17, 2008, at 2:55 PM, Michael Droettboom wrote:
>
>> There was a change between 0.90 to 0.91 as to how the TrueType to Type
>> 42 (which is essentially a thin wrapper around a TrueType font)
>> conversion is done. In 0.90 it was done in Python -- in 0.91 when font
>> subsetting was added, this conversion is done in some borrowed C code
>> called "ttconv". That's just a guess as to what may be the specific
>> cause. I don't have a copy of Illustrator, so I haven't done any
>> testing with that. I'm happy to hear of a positive report from Rob.
>> Perhaps you both (Rob and Jordan) are using different versions of
>> Illustrator?
>>
>> Note that ps.useafm essentially overrides the ps.fonttype setting. If
>> ps.useafm is True, you will always only get one of the 14 "builtin"
>> Postscript fonts, which are not embedded in the file because all
>> Postscript interpreters are required to support them. If ps.useafm is
>> False, then the other setting kicks in. Type 42 essentially includes
>> the whole Truetype font somewhat verbatim in the Ps file. Type 3 will
>> include only the used characters, using a format that is more native to
>> Postscript.
>>
>> All of them *should* be able to be editable on a purely theoretical
>> level, though I would imagine Type 3 to be broken if you try to add a
>> character that wasn't embedded in the file. (This is different from how
>> matplotlib does SVG font embedding where the original text really is
>> lost). But I don't really know what Illustrator's abilities and
>> limitations are in that regard.
>>
>> Hopefully Rob's suggestions will work for you. Otherwise, I may want to
>> revisit pulling the old Type 42 code back out of SVN to see if that
>> solves your problem. A good place to start might be if you could send
>> me (off list) two eps files of the same plot -- one created with 0.90.1,
>> and one with 0.91.2.
>>
>> Cheers,
>> Mike
>>
>> Rob Hetland wrote:
>>> There have been quite a few changes to fonts in MPL since 8.x.
>>> Perhaps one of the biggest is mathtext. Real unicode fonts with
>>> mathematics. Since you do what I do, sort of, I am guessing that you
>>> might be happy with the same configuration I use. I looks great, and
>>> can be edited in Illustrator (et al.).
>>>
>>> I really like Arev Sans, but others like the new STIX fonts. This
>>> setup is for Arev Sans. The important stuff is what begins with
>>> mathtext. If you want serif fonts, this should be enough to get you
>>> started on customizing. Just make sure whatever fontset you pick has
>>> all of the math symbols built in, or change fallback_to_cm to True.
>>>
>>> -Rob
>>>
>>>
>>> font.family : sans-serif
>>> font.style : normal
>>> font.variant : normal
>>> font.weight : medium
>>> font.stretch : normal
>>> font.size : 12.0
>>> font.serif : STIX, Bitstream Vera Serif, New Century
>>> Schoolbook, Century Schoolbook L, Utopia, ITC Bookman, Bookman,
>>> Nimbus Roman No9 L, Times New Roman, Times, Palatino, Charter, serif
>>> font.sans-serif : Arev Sans
>>> font.cursive : Zapfino, Apple Chancery, Textile, Sand, cursive
>>> font.fantasy : Chicago, Comic Sans MS, Charcoal, Impact,
>>> Western, fantasy
>>> font.monospace : Bitstream Vera Sans Mono, Andale Mono, Nimbus
>>> Mono L, Courier New, Courier, Fixed, Terminal, monospace
>>>
>>>
>>> text.color : black
>>> text.usetex : False
>>> text.dvipnghack : True
>>>
>>> mathtext.fallback_to_cm : False
>>> mathtext.fontset : custom
>>>
>>> mathtext.cal : Arev Sans:oblique
>>> mathtext.it : Arev Sans:oblique
>>> mathtext.rm : Arev Sans
>>> mathtext.bf : Arev Sans:bold
>>> mathtext.sf : Arev Sans
>>>
>>> pdf.fonttype : 42
>>>
>>>
>>>
>>>
>>>
>>> On Jan 16, 2008, at 10:13 PM, Jordan Dawe wrote:
>>>
>>>> Setting ps.useafm = True while ps.fonttype = 42 allows illustrator to
>>>> open the eps files again and gives me back text editing capability,
>>>> but
>>>> I don't have any control over the typeface anymore, it just
>>>> defaults to
>>>> Helvetica.
>>>>
>>>> Jordan
>>>>
>>>> ----------------------------------------------------------------------
>>>> ---
>>>> This SF.net email is sponsored by: Microsoft
>>>> Defy all challenges. Microsoft(R) Visual Studio 2008.
>>>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>>>> _______________________________________________
>>>> Matplotlib-users mailing list
>>>> Matplotlib-users@...
>>>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>>
>>> ----
>>> Rob Hetland, Associate Professor
>>> Dept. of Oceanography, Texas A&M University
>>> http://pong.tamu.edu/~rob
>>> phone: 979-458-0096, fax: 979-845-6331
>>>
>>>
>>>
>>> -------------------------------------------------------------------------
>>>
>>> This SF.net email is sponsored by: Microsoft
>>> Defy all challenges. Microsoft(R) Visual Studio 2008.
>>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>>> _______________________________________________
>>> Matplotlib-users mailing list
>>> Matplotlib-users@...
>>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>
>> --
>> Michael Droettboom
>> Science Software Branch
>> Operations and Engineering Division
>> Space Telescope Science Institute
>> Operated by AURA for NASA
>>
>> -------------------------------------------------------------------------
>> This SF.net email is sponsored by: Microsoft
>> Defy all challenges. Microsoft(R) Visual Studio 2008.
>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>> _______________________________________________
>> Matplotlib-users mailing list
>> Matplotlib-users@...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
> ----
> Rob Hetland, Associate Professor
> Dept. of Oceanography, Texas A&M University
> http://pong.tamu.edu/~rob
> phone: 979-458-0096, fax: 979-845-6331
>
--
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA
|