SVG group loading failing #5412
Comments
|
What it is the purpose of that image tag? |
|
I'm not sure what the purpose is, but should Fabric be able to parse this correctly, since Chrome does? Perhaps it should strip out |
|
Maybe you should have something like xlink:href='data:image/JPEG;base64,...' |
|
Your Chrome does parse because he can reach the image file that it is outside the svg. However, anyone who tries to see that svg in chrome cannot see that image |
|
Please try http://freeonlinetools24.com/base64-image |
|
The file doesn't actually exist; it's a bug inside the SVG from a user-submitted upload. I think fabric should filter out image tags on image values that it can't reach? |
|
I don't kown, sorry |
|
yes fabric should not die and give you a blank image object. |
|
Yeh I’m stripping it out to prevent the fabric error. But I don’t think the
error should happen in the first place.
On Sat, Dec 8, 2018 at 9:37 AM Andrea Bogazzi ***@***.***> wrote:
yes fabric should not die and give you a blank image object.
Not filter it out. But if you know that this may happen you can take care
of it in your load callback
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#5412 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAlaSrPVCx8C8zxCtf-F5EfrrAzxWNUPks5u286pgaJpZM4Y4N3N>
.
--
—
Oliver Nassar
|
|
yes error should not happen. |
|
Okay thank you
On Sat, Dec 8, 2018 at 11:13 AM Andrea Bogazzi ***@***.***> wrote:
yes error should not happen.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#5412 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAlaSmSsZc5A2PufiFo3L7uEc-S-rDBxks5u2-UugaJpZM4Y4N3N>
.
--
—
Oliver Nassar
|
|
the main problem is that here we are parsing that attribute as a number, 8. |
|
i think is safe to assume that handling xlink:href and href as a special case that does not need to pass in |
|
I assumed the issue was referencing an image that doesn’t exist?
On Sun, Dec 9, 2018 at 6:41 AM Andrea Bogazzi ***@***.***> wrote:
i think is safe to assume that handling xlink:href and href as a special
case that does not need to pass in parseUnit is safe enough. Could look
patchy, but at least is explicit and coherent with what we do for other non
number properties.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#5412 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAlaSmlWWdzwbTlnQc2BRff77t_c1LNFks5u3PbcgaJpZM4Y4N3N>
.
--
—
Oliver Nassar
|
|
well that was handled and failed later. it was failing before because we were using parseFloat on that url that was returning '8' as a number and throwing on .indexOf |
|
Woah very strange
On Sun, Dec 9, 2018 at 10:49 AM Andrea Bogazzi ***@***.***> wrote:
well that was handled and failed later. it was failing before because we
were using parseFloat on that url that was returning '8' as a number and
throwing on .indexOf
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#5412 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAlaSurSWPPq6V_0UVBusOoTgo6o-JYWks5u3TEJgaJpZM4Y4N3N>
.
--
—
Oliver Nassar
|
Version
Kitchen Sink
Test Case
On Kitchen Sink demo, use the following vector code:
Information about environment
Chrome Browser
Steps to reproduce
Copy and paste the above vector code into the Kitchen Sink demo (loading a vector).
Expected Behavior
It works.
Actual Behavior
It doesn't, and produces this error in the console:
The text was updated successfully, but these errors were encountered: