From 7d7a858f8b8580700ebafa8a3e083282b54e20d0 Mon Sep 17 00:00:00 2001
From: Teddy Wing
Date: Tue, 7 Nov 2017 18:28:31 +0100
Subject: jenkins.rs: Outline commit status update algorithm

Comment and a rough interface definition describing how commit statuses
should be updated and how Jenkins should be queried.
---
 src/jenkins.rs | 26 ++++++++++++++++++++++++++
 1 file changed, 26 insertions(+)

(limited to 'src')

diff --git a/src/jenkins.rs b/src/jenkins.rs
index e69de29..4958e78 100644
--- a/src/jenkins.rs
+++ b/src/jenkins.rs
@@ -0,0 +1,26 @@
+// maybe wait a few seconds to be sure a Jenkins job was created. This happens at the caller.
+
+// make request to [branch]-branches
+// if it comes back successfully with a `builds` hash
+// request all URLs in `builds`
+//   if its `displayName` matches [branch]-commitsha{5}
+//     check `result` ('SUCCESS', 'FAILURE', nonexistent)
+//     update GitHub commit status
+//       if pending
+//           start a thread that checks every 30 seconds for the `result` and update GitHub commit status
+//                 if time spent > 20 minutes
+//                    set GH commit status to error (timeout)
+//                 if `result` is successful or failed, update status and stop
+//   set GH status to error (no job found)
+
+// fn update_github_status(commit_ref)
+// fn get_jobs(repo_name)
+// fn af83 job name from commit_ref (separate af83 module)
+// fn update_github_commit_status(status, message) (lives in GitHub module)
+// fn request_job(url)
+// fn result_from_job(payload)
+
+#[cfg(test)]
+mod tests {
+    // fn 
+}
-- 
cgit v1.2.3