Support External Images

[ccpandhare]

Now browsers can load in external images onto the sequencer (or local ones) using just the path of the file. Earlier, it was required to provide a DataURL to Image Sequencer.

This can be done as long as CORS Restrictions aren't violated.

Should this be done via XMLHttpRequest? I don't see the benefit of doing that actually.

Discussion started in #78

[jywarren]

Can you add a README docs and inline code example demonstrating this?

…

On Wed, Jul 26, 2017 at 9:26 PM, Chinmay Pandhare ***@***.***> wrote: Now browsers can load in external images onto the sequencer (or local ones) using just the path of the file. Earlier, it was required to provide a DataURL to Image Sequencer. This can be done as long as CORS Restrictions aren't violated. Should this be done via XMLHttpRequest? I don't see the benefit of doing that actually. ------------------------------ You can view, comment on, or merge this pull request online at: #81 Commit Summary - Support External Images via Canvas File Changes - *M* dist/image-sequencer.js <https://github.com/publiclab/image-sequencer/pull/81/files#diff-0> (47) - *A* examples/red.png <https://github.com/publiclab/image-sequencer/pull/81/files#diff-1> (0) - *M* src/ImageSequencer.js <https://github.com/publiclab/image-sequencer/pull/81/files#diff-2> (5) - *M* src/LoadImage.js <https://github.com/publiclab/image-sequencer/pull/81/files#diff-3> (42) - *M* src/modules/_nomodule/PixelManipulation.js <https://github.com/publiclab/image-sequencer/pull/81/files#diff-4> (2) - *D* test.js <https://github.com/publiclab/image-sequencer/pull/81/files#diff-5> (5) - *M* test/image-manip.js <https://github.com/publiclab/image-sequencer/pull/81/files#diff-6> (9) Patch Links: - https://github.com/publiclab/image-sequencer/pull/81.patch - https://github.com/publiclab/image-sequencer/pull/81.diff — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#81>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AABfJ-XfpWWfir-rbh2pnowj08DLueZCks5sR5LJgaJpZM4OkYEA> .

[ccpandhare]

Suppose this is my file directory:

.
├── index.html
├── images
│   └── image.jpg
└── js
    └── image-sequencer.js

And now in index.html we can directly do the following, as far as CORS Restrictions don't dome into picture:

sequencer.loadImage('images/image.jpg');
sequencer.loadImage('http://domain/path/file.jpg');

I shall add a short explanation to README.md too

[jywarren]

this is great documentation, thanks!

…

On Wed, Jul 26, 2017 at 11:05 PM, Chinmay Pandhare ***@***.*** > wrote: Suppose this is my file directory: . ├── index.html ├── images │ └── image.jpg └── js └── image-sequencer.js And now in index.html we can directly do the following, as far as CORS Restrictions don't dome into picture: sequencer.loadImage('images/image.jpg'); sequencer.loadImage('http://domain/path/file.jpg'); I shall add a short explanation to README.md too — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#81 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AABfJwmIzcu_6AVbqV9weXmG8e_czFiGks5sR6osgaJpZM4OkYEA> .

[ccpandhare]

I have updated the README.

While we are at this, Should Image Sequencer allow Users on Node.js to directly load images from a URL?

[jywarren]

yes!

…

On Thu, Jul 27, 2017 at 8:15 PM, Chinmay Pandhare ***@***.***> wrote: I have updated the README. While we are at this, Should Image Sequencer allow Users on Node.js to directly load images from a URL? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#81 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AABfJ5zbBnECVAUqpKYGJsigeCXv-TN9ks5sSNOygaJpZM4OkYEA> .

[ccpandhare]

Done that. Some notes here:

• Since images from URLs or large sized images take time to load, I have made it compulsary for loadImages(), loadImage() to get a callback. It can not be chained like before (all other method chainings are intact).

  This no longer works:

  sequencer.loadImage(URL).addSteps('step-name')

  This is how it is now:

  sequencer.loadImage(URL,function(){
      this.addSteps('step-name').run();

• Images can be loaded directly from a URL now, like this:

  sequencer.loadImage('https://domain/path/to/image.png');

  The URL need not end in .jpg or .png. Any URL which returns a header with content-type: image/FORMAT will work.

• I have documented all this in README.md

• I have written a Unit Test to confirm this new feature works as expected.

[jywarren]

Looks great! merge if you like, bump minor version number in keeping with SemVer - (new features).

Do data URLs still work synchronously? If not, and the callback is now required, that's a breaking API change, so we should release a v1.0.0 -- a major version bump. See semver.org for explanation! :-) this is great though!

[ccpandhare]

Thanks for all the kind words :-)

No, DataURLs don't work synchronously now because large DataURLs will cause a problem and can trigger addSteps before the image loads leading to a ReferenceError.

So do we call this a major update? We're still working on basic stuff, and specifications might change relatively rapidly, so maybe let's just wait for publishing a new version... What do you think about this? If we don't publish it it, we don't have to bump the version right?

Sorry if I am being naïve here, I haven't handled versions before. These are just my thoughts...

[jywarren]

It's ok to bump the version number, in semver it's just a means of communication to users and we don't have any yet!

…

On Jul 28, 2017 3:38 PM, "Chinmay Pandhare" ***@***.***> wrote: Thanks for all the kind words :-) No, DataURLs don't work synchronously now because large DataURLs will cause a problem and can trigger addSteps before the image loads leading to a ReferenceError. So do we call this a major update? We're still working on basic stuff, and specifications might change relatively rapidly, so maybe let's just wait for publishing a new version... What do you think about this? If we don't publish it it, we don't have to bump the version right? Sorry if I am being naïve here, I haven't handled versions before. These are just my thoughts... — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#81 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AABfJ9gK1PdFcDEH2kS-hgjM4wqEBWO5ks5sSjiUgaJpZM4OkYEA> .

[ccpandhare]

Yep! Okay!
Merging!