Fix the canary by using edit distance (#3406)

## Testing
- deployed to my personal account

----

_By submitting this pull request, I confirm that you can use, modify,
copy, and redistribute this contribution, under the terms of your
choice._

tools/ci-cdk/README.md

@@ -21,7 +21,7 @@ From there, you can just point the `canary-runner` to the `cdk-outputs.json` to
 
 ```bash
 cd canary-runner
-cargo run -- run --sdk-release-tag <version> --musl --cdk-outputs ../cdk-outputs.json
+cargo run -- run --sdk-release-tag <version> --musl --cdk-output ../cdk-outputs.json
 ```
 
 __NOTE:__ You may want to add a `--profile` to the `deploy` command to select a specific credential

tools/ci-cdk/canary-lambda/src/canary.rs

@@ -84,7 +84,7 @@ impl CanaryEnv {
         // Amazon Transcribe starts returning different output for the same audio.
         let expected_transcribe_result = env::var("CANARY_EXPECTED_TRANSCRIBE_RESULT")
             .unwrap_or_else(|_| {
-                "Good day to you transcribe. This is Polly talking to you from the Rust S. D. K."
+                "Good day to you transcribe. This is Polly talking to you from the Rust SDK."
                     .to_string()
             });
 

tools/ci-cdk/canary-lambda/src/latest/transcribe_canary.rs

@@ -3,12 +3,13 @@
  * SPDX-License-Identifier: Apache-2.0
  */
 
-use crate::canary::CanaryError;
 use crate::mk_canary;
+use anyhow::bail;
 use async_stream::stream;
 use aws_config::SdkConfig;
 use aws_sdk_transcribestreaming as transcribe;
 use bytes::BufMut;
+use edit_distance::edit_distance;
 use transcribe::primitives::Blob;
 use transcribe::types::{
     AudioEvent, AudioStream, LanguageCode, MediaEncoding, TranscriptResultStream,
@@ -64,18 +65,20 @@ pub async fn transcribe_canary(
         }
     }
 
-    if expected_transcribe_result != full_message.trim() {
-        Err(CanaryError(format!(
+    let dist = edit_distance(&expected_transcribe_result, full_message.trim());
+    let max_edit_distance = 10;
+    if dist > max_edit_distance {
+        bail!(
             "Transcription from Transcribe doesn't look right:\n\
             Expected: `{}`\n\
-            Actual:   `{}`\n",
+            Actual:   `{}`\n. The maximum allowed edit distance is {}. This had an edit distance of {}",
             expected_transcribe_result,
-            full_message.trim()
-        ))
-        .into())
-    } else {
-        Ok(())
+            full_message.trim(),
+            max_edit_distance,
+            dist
+        )
     }
+    Ok(())
 }
 
 fn pcm_data() -> Vec<u8> {

tools/ci-cdk/canary-runner/src/build_bundle.rs

@@ -54,6 +54,7 @@ uuid = { version = "0.8", features = ["v4"] }
 tokio-stream = "0"
 tracing-texray = "0.1.1"
 reqwest = { version = "0.11.14", features = ["rustls-tls"], default-features = false }
+edit-distance = "2"
 "#;
 
 const REQUIRED_SDK_CRATES: &[&str] = &[
@@ -451,6 +452,7 @@ uuid = { version = "0.8", features = ["v4"] }
 tokio-stream = "0"
 tracing-texray = "0.1.1"
 reqwest = { version = "0.11.14", features = ["rustls-tls"], default-features = false }
+edit-distance = "2"
 aws-config = { path = "some/sdk/path/aws-config", features = ["behavior-version-latest"] }
 aws-sdk-s3 = { path = "some/sdk/path/s3" }
 aws-sdk-ec2 = { path = "some/sdk/path/ec2" }
@@ -515,6 +517,7 @@ uuid = { version = "0.8", features = ["v4"] }
 tokio-stream = "0"
 tracing-texray = "0.1.1"
 reqwest = { version = "0.11.14", features = ["rustls-tls"], default-features = false }
+edit-distance = "2"
 aws-config = { version = "0.46.0", features = ["behavior-version-latest"] }
 aws-sdk-s3 = "0.20.0"
 aws-sdk-ec2 = "0.19.0"