Download files

crates/_types/tokens/src/macros.rs

@@ -56,6 +56,13 @@ macro_rules! impl_crockford_generator {
         $t(token)
       }
 
+      #[inline]
+      pub fn entropy_suffix(&self) -> &str {
+        &self.0
+          .strip_prefix($variant.prefix())
+          .unwrap_or(&self.0)
+      }
+
       #[inline]
       pub fn entropic_character_len() -> usize {
         let token_prefix = Self::token_prefix();

crates/_types/tokens/tests/media_upload_token_tests.rs

@@ -163,4 +163,16 @@ mod crockford_traits {
     let token_string = MediaUploadToken::generate().to_string();
     assert!(token_string.starts_with(prefix));
   }
+
+  #[test]
+  fn entropy_suffix() {
+    let token = MediaUploadToken::new_from_str("mu_foo");
+    assert_eq!(token.entropy_suffix(), "foo");
+
+    let token = MediaUploadToken::new_from_str("bar");
+    assert_eq!(token.entropy_suffix(), "bar");
+
+    let token = MediaUploadToken::generate();
+    assert_eq!(token.entropy_suffix().len(), 29);
+  }
 }

crates/_types/tokens/tests/user_token_tests.rs

@@ -163,4 +163,16 @@ mod crockford_traits {
     let token_string = UserToken::generate().to_string();
     assert!(token_string.starts_with(prefix));
   }
+
+  #[test]
+  fn entropy_suffix() {
+    let token = UserToken::new_from_str("U:foo");
+    assert_eq!(token.entropy_suffix(), "foo");
+
+    let token = UserToken::new_from_str("bar");
+    assert_eq!(token.entropy_suffix(), "bar");
+
+    let token = UserToken::generate();
+    assert_eq!(token.entropy_suffix().len(), 13);
+  }
 }

crates/programs/aichatbot-sidecar/src/persistence/save_directory.rs

@@ -3,6 +3,7 @@ use log::info;
 use path_absolutize::Absolutize;
 use errors::{anyhow, AnyhowResult};
 use hashing::sha256::sha256_hash_string::sha256_hash_string;
+use tokens::tokens::news_stories::NewsStoryToken;
 use crate::persistence::hashed_directory_path::hashed_directory_path;
 
 #[derive(Clone)]
@@ -18,6 +19,7 @@ impl SaveDirectory {
     }
   }
 
+  // TODO: Move web pages for news under "news_stories/{hashed_url_path}/"
   pub fn directory_for_webpage_url(&self, url: &str) -> AnyhowResult<PathBuf> {
     let url_hash = sha256_hash_string(url)?;
     let directory_path= hashed_directory_path(&url_hash);
@@ -30,6 +32,20 @@ impl SaveDirectory {
     Ok(directory)
   }
 
+  // TODO: Move web pages for news under "news_stories/{hashed_url_path}/audio/"
+  //  This requires associating the job with the URL.
+  pub fn directory_for_audio(&self, news_story_token: &NewsStoryToken) -> AnyhowResult<PathBuf> {
+    let entropy_part = news_story_token.entropy_suffix();
+    let directory_path= hashed_directory_path(&entropy_part);
+
+    let directory = self.directory.clone()
+        .join("audio/")
+        .join(directory_path)
+        .join(news_story_token.as_str());
+
+    Ok(directory)
+  }
+
   pub fn html_file_for_webpage_url(&self, url: &str) -> AnyhowResult<PathBuf> {
     Ok(self.directory_for_webpage_url(url)?.join("webpage.html"))
   }
@@ -46,6 +62,11 @@ impl SaveDirectory {
     Ok(self.directory_for_webpage_url(url)?.join("speakable_monologue.yaml"))
   }
 
+  pub fn audio_wav_file_for_news_story(&self, news_story_token: &NewsStoryToken, sequence_order: i64) -> AnyhowResult<PathBuf> {
+    let filename = format!("{}.wav", sequence_order);
+    Ok(self.directory_for_audio(news_story_token)?.join(filename))
+  }
+
   /// This is just the first directory structure, which is sequential audio files.
   /// We'll be using a database and well-formed filesystem layout later.
   pub fn get_audio_files_dir_v1(&self) -> PathBuf {

crates/programs/aichatbot-sidecar/src/workers/audio/fakeyou_audio_download/process_single_record.rs

@@ -1,3 +1,4 @@
+use std::io::Cursor;
 use crate::shared_state::job_state::JobState;
 use errors::{anyhow, AnyhowResult};
 use fakeyou_client::api::tts_inference::{CreateTtsInferenceRequest, TtsInferenceJobStatusStatePayload};
@@ -7,6 +8,7 @@ use sqlite_queries::queries::by_table::tts_render_targets::list::tts_render_targ
 use sqlite_queries::queries::by_table::tts_render_targets::update::update_tts_render_target_successfully_downloaded::Args;
 use sqlite_queries::queries::by_table::tts_render_targets::update::update_tts_render_target_successfully_downloaded::update_tts_render_target_successfully_downloaded;
 use std::sync::Arc;
+use tokens::tokens::news_stories::NewsStoryToken;
 use tokens::tokens::tts_models::TtsModelToken;
 use tokens::tokens::tts_render_tasks::TtsRenderTaskToken;
 
@@ -46,7 +48,7 @@ pub async fn process_single_record(target: &TtsRenderTarget, job_state: &Arc<Job
           //}).await?;
         }
         Some(payload) => {
-          process_download(&target.token, &payload, job_state).await?;
+          process_download(&target, &payload, job_state).await?;
         }
       }
     },
@@ -55,7 +57,7 @@ pub async fn process_single_record(target: &TtsRenderTarget, job_state: &Arc<Job
   Ok(())
 }
 
-async fn process_download(task_token: &TtsRenderTaskToken, payload: &TtsInferenceJobStatusStatePayload, job_state: &Arc<JobState>) -> AnyhowResult<()> {
+async fn process_download(target: &TtsRenderTarget, payload: &TtsInferenceJobStatusStatePayload, job_state: &Arc<JobState>) -> AnyhowResult<()> {
   let tts_result_token = match &payload.maybe_result_token {
     None => return Ok(()), // TODO: Wait
     Some(result_token) => result_token.to_string(),
@@ -67,9 +69,25 @@ async fn process_download(task_token: &TtsRenderTaskToken, payload: &TtsInferenc
   };
 
   let audio_url = get_audio_url(&bucket_path, true);
+  let news_story_token = NewsStoryToken::new_from_str(&target.story_token);
+
+  {
+    let directory = job_state.save_directory.directory_for_audio(&news_story_token)?;
+    std::fs::create_dir_all(&directory)?;
+  }
+
+  {
+    let download_filename = job_state.save_directory
+        .audio_wav_file_for_news_story(&news_story_token, target.sequence_order)?;
+
+    let response = reqwest::get(&audio_url).await?;
+    let mut file = std::fs::File::create(download_filename)?;
+    let mut content =  Cursor::new(response.bytes().await?);
+    std::io::copy(&mut content, &mut file)?;
+  }
 
   update_tts_render_target_successfully_downloaded(Args {
-    tts_render_task_token: task_token,
+    tts_render_task_token: &target.token,
     tts_result_token: &tts_result_token,
     result_url: &audio_url,
     sqlite_pool: &job_state.sqlite_pool,