Refactor

.gitignore

@@ -7,4 +7,6 @@ build
 .venv
 .mypy_cache
 
-target
+target
+
+.DS_Store

Cargo.toml

@@ -1,5 +1,5 @@
 [package]
-name = "transcriber"
+name = "capski"
 version = "0.1.0"
 edition = "2024"
 

README.md

@@ -30,15 +30,15 @@ The Engineering Requirements Document (**ERD**) is available here :
 brew install ffmpeg
 ```
 
-### 2. Build Transcriber
-You can now build the transcriber using Rust Cargo.
+### 2. Install Capski
+You can now install Capski using Rust Cargo.
 ```bash
-cargo build
+cargo install --path .
 ```
 
-### 3. Run Transcriber
+### 3. Run Capski
 ```bash
-cargo run -- --input "example/input_audio.wav"
+capski --input "example/input_audio.wav"
 ```
 
 This runs the pipeline end-to-end:

src/audio/extractor.rs

@@ -0,0 +1,39 @@
+use super::Extractor;
+use anyhow::{Context, Result};
+use log::info;
+use std::process::Command;
+
+pub struct FfmpegExtractor;
+
+impl Extractor for FfmpegExtractor {
+    // Function to extract audio from a video file
+    fn extract(input: &str, output: &str) -> Result<()> {
+        info!("Extracting audio from {} to {}", input, output);
+
+        let status = Command::new("ffmpeg")
+            .args(&[
+                "-y",
+                "-i",
+                input,
+                "-vn",
+                "-acodec",
+                "pcm_s16le",
+                "-ar",
+                "16000",
+                "-ac",
+                "1",
+                output,
+            ])
+            .status()
+            .context("Failed to extract audio via FFmpeg")?;
+
+        if !status.success() {
+            anyhow::bail!(
+                "ffmpeg failed to extract audio: exited with code {}",
+                status
+            );
+        }
+
+        Ok(())
+    }
+}

src/audio/mod.rs

@@ -0,0 +1,16 @@
+mod extractor;
+mod whisper;
+
+pub use extractor::FfmpegExtractor;
+pub use whisper::WhisperCapski;
+
+use crate::types::Segment;
+use anyhow::Result;
+
+pub trait Capski {
+    fn transcribe(model_path: &str, audio_path: &str) -> Result<Vec<Segment>>;
+}
+
+pub trait Extractor {
+    fn extract(input: &str, output: &str) -> Result<()>;
+}

src/audio.rs → src/audio/whisper.rs

@@ -1,18 +1,19 @@
 use crate::types::Segment;
 
+use super::Capski;
 use anyhow::{Context, Result};
 use log::info;
-use std::process::Command;
 use whisper_rs::{FullParams, SamplingStrategy, WhisperContext, WhisperContextParameters};
 
-pub struct Audio;
+pub struct WhisperCapski;
 
-impl Audio {
+impl Capski for WhisperCapski {
     // Function to transcribe audio using the Whisper model
-    pub fn transcribe(model_path: &str, audio_path: &str) -> Result<Vec<Segment>> {
+    fn transcribe(model_path: &str, audio_path: &str) -> Result<Vec<Segment>> {
         info!("Transcribing with Whisper...");
 
-        let reader = hound::WavReader::open(audio_path).expect("failed to open file");
+        let reader = hound::WavReader::open(audio_path)
+            .with_context(|| format!("failed to open audio file: {}", audio_path))?;
 
         // Read WAV file and collect samples
         let samples: Vec<i16> = reader.into_samples::<i16>().map(|x| x.unwrap()).collect();
@@ -91,32 +92,4 @@ impl Audio {
 
         Ok(segments)
     }
-
-    // Function to extract audio from a video file
-    pub fn extract(video_path: &str, audio_path: &str) -> Result<()> {
-        info!("Extracting audio from {} to {}", video_path, audio_path);
-
-        let status = Command::new("ffmpeg")
-            .args(&[
-                "-y",
-                "-i",
-                video_path,
-                "-vn",
-                "-acodec",
-                "pcm_s16le",
-                "-ar",
-                "16000",
-                "-ac",
-                "1",
-                audio_path,
-            ])
-            .status()
-            .context("Failed to extract audio")?;
-
-        if !status.success() {
-            return Err(anyhow::anyhow!("ffmpeg failed to extract audio"));
-        }
-
-        Ok(())
-    }
 }