Re: eps file

[David Wright]

On Thu 02 Sep 2021 at 14:14:12 (-0400), ming tsang wrote:

> I must have included the wrong screen print.
> Here is the right one.

Ah, I can replicate that:

$ lily /tmp/ming.
/home/david/lilypond-2.23.3-1.linux-64/bin/lilypond -dno-point-and-click 
--include=/home/david/LilyLib/ --include=/home/david/LilyLib/margins/ 
--include=/home/david/LilyLib/parts-midis/ ming.ly
GNU LilyPond 2.23.3
Processing `ming.ly'
Parsing...
ming.ly:1: warning: no \version statement found, please add

\version "2.23.3"

for future compatibility
/home/david/lilypond-2.23.3-1.linux-64/lilypond/usr/share/lilypond/current/scm/lily/stencil.scm:986:26:
 In procedure list-ref in expression (list-ref bbox 3):
/home/david/lilypond-2.23.3-1.linux-64/lilypond/usr/share/lilypond/current/scm/lily/stencil.scm:986:26:
 Wrong type argument in position 1: #f

ERROR

$ 

(I'm guessing there's something in the Windows version that loses a
line of stencil.scm somewhere.)

I took your original MWE and substituted some EPSs of "my own", a
perfect one, and one that is oddly malformed. The malformation makes
some programs refuse to display it, other groan but cooperate, but
ghostview opens it without a murmer. Here's a hex dump of the start
of the file:

$ hexdump -C ave/tex/logos/xmas018.eps | head -n 16
00000000  c5 d0 d3 c6 1e 00 00 00  45 6d 00 00 00 00 00 00  |........Em......|
00000010  00 00 00 00 63 6d 00 00  90 4c 03 00 ff ff 25 21  |....cm...L....%!|
00000020  50 53 2d 41 64 6f 62 65  2d 33 2e 30 20 45 50 53  |PS-Adobe-3.0 EPS|
00000030  46 2d 33 2e 30 0d 0a 25  25 42 6f 75 6e 64 69 6e  |F-3.0..%%Boundin|
00000040  67 42 6f 78 3a 20 33 30  20 31 39 38 20 35 38 38  |gBox: 30 198 588|
00000050  20 35 38 35 20 0d 0a 25  25 4c 61 6e 67 75 61 67  | 585 ..%%Languag|
00000060  65 4c 65 76 65 6c 3a 20  31 0d 0a 25 25 43 72 65  |eLevel: 1..%%Cre|
00000070  61 74 6f 72 3a 20 43 6f  72 65 6c 44 52 41 57 20  |ator: CorelDRAW |
00000080  31 32 0d 0a 25 25 54 69  74 6c 65 3a 20 58 4d 41  |12..%%Title: XMA|
00000090  53 30 31 38 2e 65 70 73  0d 0a 25 25 43 72 65 61  |S018.eps..%%Crea|
000000a0  74 69 6f 6e 44 61 74 65  3a 20 53 61 74 20 4e 6f  |tionDate: Sat No|
000000b0  76 20 32 35 20 32 30 3a  34 34 3a 35 36 20 32 30  |v 25 20:44:56 20|
000000c0  30 36 0d 0a 25 25 44 6f  63 75 6d 65 6e 74 50 72  |06..%%DocumentPr|
000000d0  6f 63 65 73 73 43 6f 6c  6f 72 73 3a 20 43 79 61  |ocessColors: Cya|
000000e0  6e 20 4d 61 67 65 6e 74  61 20 59 65 6c 6c 6f 77  |n Magenta Yellow|
000000f0  20 42 6c 61 63 6b 20 0d  0a 25 25 44 6f 63 75 6d  | Black ..%%Docum|
$ 

So the file starts with 30 bytes of "garbage", which could be some
sort of header for some program of other. (The EPS is a piece of
clipart.)

Ah (again), I've just looked in more detail at the file. Buried under
the EPS layer is actually a TIFF image. So the program that groaned
waded through the PS part until it hit the II decimal42 TIFF header,
groaned some more about tags that it didn't understand, but happily
displayed the TIFF.

Upshot: at least try and find some test EPS file where the first
characters are something like:
%!PS-Adobe-3.0 EPSF-3.0
%%BoundingBox: 37 160 568 701
and see if that works.

Cheers,
David.