Add readme for the summariser

summariser/README.md

@@ -0,0 +1,50 @@
+# Wind Tunnel: Summariser
+
+This is a tool to summarise the results of Wind Tunnel scenario runs.
+
+Start by running one or more scenarios with `--reporter influx-file` to create metrics on InfluxDB. You should find a
+`run_summary.jsonl` file created in the root of the project. The summariser reads this to find metrics.
+
+You can then run the summariser to generate a report:
+
+```shell
+RUST_LOG=info cargo run summariser 
+```
+
+This will create a new JSON file that summarises the results from the scenarios.
+
+### Testing the summariser
+
+The summariser comes with some tooling for testing it. With a `run_summary.jsonl` that contains a run which you want to 
+use as a test case, you can run the following command:
+
+```shell
+RUST_LOG=debug cargo run --features test_data summariser
+```
+
+This will add:
+- The run summary to `summariser/test_data/1_run_summaries/<scenario-name>-<scenario run fingerprint>.json`
+- The raw that fetched from influx, as JSON to `summariser/test_data/2_query_results/<query fingerprint>.json`
+- The generated report to `summariser/test_data/3_summary_outputs/<scenario-name>-<scenario run fingerprint>.json`
+
+All of these should be added to Git, then you can write a test that loads the test data. This allows you to iterate on 
+the summariser without needing to run the scenario again or even have a running InfluxDB.
+
+Tests just look like:
+
+```rust
+#[tokio::test]
+async fn scenario_name() -> anyhow::Result<()> {
+    run_snapshot_test!("<scenario fingerprint>");
+    Ok(())
+}
+```
+
+If you make changes to the summariser, you should review the diff that this test prints and ensure it looks correct. If
+so, then rather than update test data by hand, you will be prompted to run 
+
+```shell
+UPDATE_SNAPSHOTS=1 cargo test --test snapshot
+```
+
+This will overwrite the reports with their latest version. You can review and commit the diff to match the updated code.