v0.15.2

Fixes

This patch release updates some dependencies to fix the security issues introduced by log-driver package. No other changes since the previous release.

v0.15.1

Fixes

• Updated dependencies to mitigate possible security issues.

v0.15.0

Features

In this release, still beta, we have added some small compatible improvements to the v1 API and refreshed auto generated code. This version will likely be released to general availability after some time.

@google-cloud/vision v0.14.0

This should be considered a subsequent and hopefully final release candidate. It has no breaking changes; we just simply want to give the API introduced in 0.13.0 slightly more bake time before declaring GA.

Features

Beta Endpoint

• #11

This release adds the v1p1beta1 endpoint. This is a new backend endpoint which adds features which the backend API considers to be in public beta. These features are being prepared for general release and should be mostly stable; however, changes may occur if necessary.

By default, this client library will use the v1 (general availability) endpoint, rather than silently opt you into using beta features. If you wish to use v1p1beta1, you can opt into it explicitly when requiring the library:

const vision = require('@google-cloud/vision').v1p1beta1;

The v1p1beta1 endpoint is a superset of v1; it also contains all v1 features. The goal is that you can change nothing but your require statement, and when these features go GA, you only have to remove the version designation.

@google-cloud/vision v0.13.0

This is the (hopefully) final release candidate before declaring a stable release.

⚠️ Breaking Changes!

Request Format

• PR: #2555, #5

We learned from a few mistakes we made in the previous release, and have updated the API to be simpler and more consistent. We now require the entire AnnotateImageRequest to be sent to annotateImage, rather than just the image component. This allows for cases where you need to specify something else (e.g. imageContext) without having to fall back on the batch method.

To make up for this, we also reinstated the ability to send a simple string or buffer in lieu of the entire request in the single-feature methods (#5), which should make this a net reduction of code.

Simple Cases

Here is a migration guide for simple cases:

Before

var image = {
  source: {imageUri: 'gs://path/to/image.jpg'},
};
vision.labelDetection(image).then(data => {
  let apiResponse = data[0];
  // ...
});

After

vision.labelDetection('gs://path/to/image.jpg').then(data => {
  let apiResponse = data[0];
  // ...
});

Complex Cases

In complex cases (for example, where you needed another key in the request), you will need to send the entire request object, except for the parts that get added in by the method (e.g. the features key).

Carrying forward the previous example:

Before

var image = {
  source: {imageUri: 'gs://path/to/image.jpg'},
};
vision.labelDetection(image).then(data => {
  let apiResponse = data[0];
  // ...
});

After

var request = {
  image: {
    source: {imageUri: 'gs://path/to/image.jpg'},
  },
};
vision.labelDetection(request).then(data => {
  let apiResponse = data[0];
  // ...
});

Refer to the nodejs-vision API reference documentation for details.

Features

Runnable samples

There are now runnable samples in the samples/ directory. These samples demonstrate the use of this API and offer an additional resource for getting started. These also power the readme, and are automatically tested to ensure that they are up-to-date and correct.

Implementation Details

• Change to a new linter (ESLint) and code style formatter (prettify).
• Update docs to use JSDoc 3.
• Migrate to the googleapis/nodejs-vision repository.

@google-cloud/vision v0.12.0

⚠️ Breaking Changes!

The majority of our Vision API has changed (PR: #2298). Please refer to the API documentation for more information on how to use the module.

Usage

Usage has changed to more closely resemble the underlying API surface. Migration will generally look like this:

Before

vision.detectLabels('gs://path/to/image.jpg').then(data => {
  let labels = data[0];
  let apiResponse = data[1];
  // ...
});

After

var image = {
  source: {imageUri: 'gs://path/to/image.jpg'},
};
vision.labelDetection(image).then(data => {
  let apiResponse = data[0];
  // ...
});

Refer to the nodejs-vision API reference documentation for details.

@google-cloud/vision v0.11.5

Dependencies

• Update dependencies.

@google-cloud/vision v0.11.4

Implementation Details

• Update retry configuration (#2390)

@google-cloud/vision v0.11.3

Metadata

• Update documentation.

Dependencies

• Update dependencies.

@google-cloud/vision v0.11.2

Implementation Details

• Add routing headers for the backend. (#2228)