If multi-site searching, sort questions by scorev0.2.2

4 files changed, 26 insertions, 11 deletions

diff --git a/README.md b/README.md
index 52e57c1..a383197 100644
--- a/README.md
+++ b/README.md
@@ -13,6 +13,12 @@ Going with cursive because it is way more flexible than tui-rs.
 benefit of incorporating termimad features will not be felt. But, this is
 changing [soon](https://meta.stackexchange.com/q/348746).
 
+### to stress test
+Produces a long answer with noticeable pause on markdown view:
+```
+cargo run -- --site stackoverflow --site serverfault how do I exit Vim
+```
+
 ### to troubleshoot
 ```
 export RUST_BACKTRACE=full
@@ -29,5 +35,14 @@ run into throttling issues, get a key
 so --set-api-key <KEY>
 ```
 
-Recall my api key is: `8o9g7WcfwnwbB*Qp4VsGsw((`
+### multi-site searching
+As stated in the docs linked above,
+
+> If a single IP is making more than 30 requests a second, new requests will be dropped.
+
+So, don't go crazy with the multi-site search, since it is all done in parallel.
+In particular, if you specify more than 30 sites, SE will likely ban you for a short time.
+
+
 
+**Remove this** Recall my api key is: `8o9g7WcfwnwbB*Qp4VsGsw((`
diff --git a/TODO.md b/TODO.md
index 672e993..ba1946b 100644
--- a/TODO.md
+++ b/TODO.md
@@ -1,10 +1,5 @@
 # TODO
 
-### v0.2.2
-1. Site can be multiple
-2. do tokio async on SE api
-3. add warning to README about throttling on excessive requests
-
 ### v0.3.0
 1. Duckduck go search ftw, e.g.
 ```
diff --git a/roadmap.md b/roadmap.md
index 47afdf9..efe6dfe 100644
--- a/roadmap.md
+++ b/roadmap.md
@@ -18,7 +18,7 @@
 [x] For --lucky, async parsing first q/a, then the rest
 
 ### v0.2.2
-[ ] Support multiple --site args & searches
+[x] Support multiple --site args & searches
 
 ### v0.3.0
 [ ] Add duckduckgo scraper
diff --git a/src/stackexchange.rs b/src/stackexchange.rs
index b6c6a16..ddc3c48 100644
--- a/src/stackexchange.rs
+++ b/src/stackexchange.rs
@@ -116,7 +116,7 @@ impl StackExchange {
 
     /// Parallel searches against the search/advanced endpoint across all configured sites
     async fn search_advanced(&self, limit: u16) -> Result<Vec<Question>> {
-        let results = futures::stream::iter(self.config.sites.clone())
+        futures::stream::iter(self.config.sites.clone())
             .map(|site| {
                 let clone = self.clone();
                 tokio::spawn(async move {
@@ -126,12 +126,17 @@ impl StackExchange {
             })
             .buffer_unordered(CONCURRENT_REQUESTS_LIMIT)
             .collect::<Vec<_>>()
-            .await;
-        results
+            .await
             .into_iter()
             .map(|r| r.map_err(Error::from).and_then(|x| x))
             .collect::<Result<Vec<Vec<_>>>>()
-            .map(|v| v.into_iter().flatten().collect())
+            .map(|v| {
+                let mut all_qs: Vec<Question> = v.into_iter().flatten().collect();
+                if self.config.sites.len() > 1 {
+                    all_qs.sort_unstable_by_key(|q| -q.score);
+                }
+                all_qs
+            })
     }
 
     /// Search against the site's search/advanced endpoint with a given query.