Function generate_arrow_from_folder 

pub fn generate_arrow_from_folder(
    folder: &Path,
    output: &Path,
    detect_sequences: bool,
) -> Result<usize, Error>

Generate an Arrow file from a folder of images.

Scans the folder for image files and creates an Arrow annotation file with null annotations (for unannotated datasets). This is useful for importing existing image collections into EdgeFirst.

Arguments

• folder - Path to the folder containing images
• output - Path where the Arrow file should be written
• detect_sequences - If true, attempt to detect sequences from naming patterns

Returns

The number of samples (images) included in the Arrow file.

Sequence Detection

When detect_sequences is true, the function looks for patterns like:

• {name}_{number}.{ext} → sequence with frame number
• {sequence}/{name}_{number}.{ext} → sequence in subdirectory

Example

use edgefirst_client::format::generate_arrow_from_folder;
use std::path::Path;

// Generate Arrow file from images
let count = generate_arrow_from_folder(
    Path::new("my_images"),
    Path::new("my_dataset/my_dataset.arrow"),
    true, // detect sequences
)?;
println!("Created Arrow file with {} samples", count);