Struct VideoHandler 

pub struct VideoHandler {
    pub config: Config,
    pub gcs: GcsClient,
    pub http: Client,
    pub auth: AuthProvider,
}

Video generation handler.

Handles video generation requests using the Vertex AI Veo API.

Fields

config: Config

Application configuration.

gcs: GcsClient

GCS client for storage operations.

http: Client

HTTP client for API requests.

auth: AuthProvider

Authentication provider.

Implementations

impl VideoHandler

pub async fn new(config: Config) -> Result<Self, Error>

Create a new VideoHandler with the given configuration.

Errors

Returns an error if GCS client or auth provider initialization fails.

pub fn get_generate_endpoint(&self, model: &str) -> String

Get the Vertex AI Veo API endpoint for generating videos.

pub fn get_fetch_operation_endpoint(&self, model: &str) -> String

Get the endpoint for fetching LRO status.

pub async fn generate_video_t2v( &self, params: VideoT2vParams, ) -> Result<VideoGenerateResult, Error>

Generate video from a text prompt.

Arguments

• params - Video generation parameters

Returns

• Ok(VideoGenerateResult) - Generated video with GCS URI and optional local path
• Err(Error) - If validation fails, API call fails, or output handling fails

pub async fn generate_video_i2v( &self, params: VideoI2vParams, ) -> Result<VideoGenerateResult, Error>

Generate video from an image.

Arguments

• params - Image-to-video generation parameters

Returns

• Ok(VideoGenerateResult) - Generated video with GCS URI and optional local path
• Err(Error) - If validation fails, API call fails, or output handling fails

pub async fn extend_video( &self, params: VideoExtendParams, ) -> Result<VideoGenerateResult, Error>

Extend an existing video.

Arguments

• params - Video extension parameters

Returns

• Ok(VideoGenerateResult) - Extended video with GCS URI and optional local path
• Err(Error) - If validation fails, API call fails, or output handling fails

pub async fn poll_lro( &self, operation_name: &str, model: &str, ) -> Result<LroResult, Error>

Poll a long-running operation until completion.

Uses exponential backoff with configurable parameters. Uses the fetchPredictOperation endpoint which requires the operation name in the request body.