illegal start byte 0xf0 about androidsvg

[GoogleCodeExporter]

This looks like the problem is in the XML parser and is probably out of my 
control.

Are you able to provide a test file that contains one of these problematic 
strings?

Original comment by [email protected] on 15 Jan 2014 at 2:02

[GoogleCodeExporter]

Sorry about that, attached to this post is one of the files that threw the 
error.

Original comment by [email protected] on 15 Jan 2014 at 2:35

Attachments:

• 04e45.svg

[GoogleCodeExporter]

I played around with the SVG file today and confirmed that the issue is, as you 
suggested, a problem with the XML parser itself and not AndroidSVG; I had the 
same problem when I tried to use SVG-Android2 to display the file.

I used some Python to strip out the [kvg:element] attribute from the <g> nodes 
in that 05e45.svg file and am pleased to report that the file loaded properly 
with AndroidSVG. I imagine I'll be able to prevent these errors by running the 
other SVG files through the script and strip out other such problematic 
characters.

I suppose you can close this issue since the cause is unrelated to AndroidSVG.

Original comment by [email protected] on 15 Jan 2014 at 10:14

[GoogleCodeExporter]

The problematic sequence of bytes in question is f0 a0 92 8a.

The initial f0 byte indicates that it is a four byte encoding sequence. See 
https://en.wikipedia.org/wiki/UTF-8 .

Unfortunately, it seems that the "Modified UTF8" encoding, that Java 
implements, only supports up to 3 byte sequences.  That's according to 
http://docs.oracle.com/javase/6/docs/api/java/io/DataInput.html#modified-utf-8 .

Unfortunately I can't really intercept these strings before the XML parser 
processes them, so it seems like the simplest solution is to do as you proposed 
and strip the problematic attributes before passing them to AndroidSVG.

Original comment by [email protected] on 16 Jan 2014 at 12:08

[GoogleCodeExporter]

Marking "Won't Fix"

Original comment by [email protected] on 16 Jan 2014 at 12:09

• Changed state: WontFix

[GoogleCodeExporter]

Actually, the SAX parser in Java likes it fine.  I think it may be the one in 
Harmony/Android that doesn't like four byte UTF-8 sequences.  Doing some more 
tests to confirm.

Original comment by [email protected] on 16 Jan 2014 at 4:57

[GoogleCodeExporter]

Android bug report filed for this issue.

https://code.google.com/p/android/issues/detail?id=64892

Original comment by [email protected] on 16 Jan 2014 at 11:55