From 7008f06d45693e035c74be5e4516cd40ebc6f265 Mon Sep 17 00:00:00 2001
From: Anson Biggs <anson@ansonbiggs.com>
Date: Wed, 24 Apr 2024 05:07:40 +0000
Subject: [PATCH] more algorithm

---
 src/main.rs | 20 ++++++++++++++++++--
 1 file changed, 18 insertions(+), 2 deletions(-)

diff --git a/src/main.rs b/src/main.rs
index 5e5209c..b625e3f 100644
--- a/src/main.rs
+++ b/src/main.rs
@@ -11,6 +11,7 @@ mod site_generator;
 mod utilities;
 mod web_fetchers;
 use rayon::prelude::*;
+use std::collections::HashMap;
 
 fn main() -> Result<(), Box<dyn Error>> {
     simple_logger::init_with_level(log::Level::Info).unwrap();
@@ -33,7 +34,7 @@ fn main() -> Result<(), Box<dyn Error>> {
 
     entries.par_iter_mut().for_each(|entry| {
         if entry.image_url.is_some() {
-            entry.score += 300;
+            entry.score += 1440;
         } else {
             match web_fetchers::fetch_social_image(entry.link.clone()) {
                 Ok(social_image_url) => {
@@ -41,15 +42,30 @@ fn main() -> Result<(), Box<dyn Error>> {
                 }
                 Err(error) => {
                     log::info!("{}: {}", error, entry.link.clone());
-                    entry.score += -600;
+                    entry.score += -1440;
                 }
             }
         }
     });
 
     entries.retain(|entry| entry.score.is_positive());
+
+    // Cunt occurences of main urls
+    let url_counts = entries.iter().fold(HashMap::new(), |mut acc, post| {
+        *acc.entry(post.main_url.clone()).or_insert(0) += 1;
+        acc
+    });
+
+    // Punish blogs that post really often
+    entries.iter_mut().for_each(|entry| {
+        entry.score = (entry.score / url_counts.get(&entry.main_url).unwrap()) as i64;
+    });
+
     entries.sort();
 
+    // Remove bottom 10% from list
+    entries.truncate(entries.len() - (entries.len() as f64 * 0.1).ceil() as usize);
+
     let index = site_generator::generate_index(entries.clone());
     let index_path = Path::new("output/index.html");
     DirBuilder::new()