From 9d1374c34256b6e1d9ebbd565b7891510d753221 Mon Sep 17 00:00:00 2001
From: Sam Denty <sam@samdenty.com>
Date: Mon, 26 Dec 2022 16:01:35 +0000
Subject: [PATCH] 0.4.5: highest priority for alt containing site_name

---
 Cargo.lock   | 37 ++++++++++++++++++++++++++++++++++++-
 Cargo.toml   |  3 ++-
 src/icons.rs | 41 ++++++++++++++++++++++++++++++++++-------
 3 files changed, 72 insertions(+), 9 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index 2b22014..6974a7d 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -1638,7 +1638,7 @@ checksum = "7bd3e3206899af3f8b12af284fafc038cc1dc2b41d1b89dd17297221c5d225de"
 
 [[package]]
 name = "site_icons"
-version = "0.4.4"
+version = "0.4.5"
 dependencies = [
  "byteorder",
  "clap",
@@ -1658,6 +1658,7 @@ dependencies = [
  "serde",
  "serde_json",
  "serde_with",
+ "tldextract",
  "tokio",
  "url",
  "vec1",
@@ -1783,6 +1784,26 @@ version = "0.1.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "8eaa81235c7058867fa8c0e7314f33dcce9c215f535d1913822a2b3f5e289f3c"
 
+[[package]]
+name = "thiserror"
+version = "1.0.38"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6a9cd18aa97d5c45c6603caea1da6628790b37f7a34b6ca89522331c5180fed0"
+dependencies = [
+ "thiserror-impl",
+]
+
+[[package]]
+name = "thiserror-impl"
+version = "1.0.38"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1fb327af4685e4d03fa8cbcf1716380da910eeb2bb8be417e7f9fd3fb164f36f"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
 [[package]]
 name = "time"
 version = "0.3.17"
@@ -1825,6 +1846,20 @@ version = "0.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "cda74da7e1a664f795bb1f8a87ec406fb89a02522cf6e50620d016add6dbbf5c"
 
+[[package]]
+name = "tldextract"
+version = "0.6.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ec03259a0567ad58eed30812bc3e5eda8030f154abc70317ab57b14f00699ca4"
+dependencies = [
+ "idna 0.2.3",
+ "log",
+ "regex",
+ "serde_json",
+ "thiserror",
+ "url",
+]
+
 [[package]]
 name = "tokio"
 version = "1.22.0"
diff --git a/Cargo.toml b/Cargo.toml
index f4a2ed1..5dd64a5 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -1,6 +1,6 @@
 [package]
 name = "site_icons"
-version = "0.4.4"
+version = "0.4.5"
 authors = ["Sam Denty <sam@samdenty.com>"]
 edition = "2018"
 license = "GPL-3.0"
@@ -34,6 +34,7 @@ mime = { package = "mime_4", version = "0.4.0-a.0" }
 serde = { version = "1.0", features = ["derive", "rc"] }
 serde_json = "1.0"
 futures = "0.3.25"
+tldextract = "0.6.0"
 
 [target.'cfg(target_arch = "wasm32")'.dependencies]
 reqwest = { package = "reqwest-wasm", version = "0.11.15", features = [
diff --git a/src/icons.rs b/src/icons.rs
index f0ed689..d3812ba 100644
--- a/src/icons.rs
+++ b/src/icons.rs
@@ -10,8 +10,10 @@ use reqwest::{header::*, IntoUrl};
 use scraper::{ElementRef, Html};
 use serde::Deserialize;
 use std::convert::TryInto;
+use std::iter;
 use std::task::Poll;
 use std::{collections::HashMap, error::Error, pin::Pin, task::Context};
+use tldextract::TldOption;
 use url::Url;
 
 pub struct Icons {
@@ -227,14 +229,25 @@ impl Icons {
             weight += 1;
           }
 
-          let mentions_logo = |attr_name| {
-            ancestors.iter().any(|ancestor| {
-              ancestor
-                .attr(attr_name)
-                .map(|attr| regex!("logo([^s]|$)").is_match(&attr.to_lowercase()))
-                .unwrap_or(false)
-            })
+          let mentions = |attr_name, is_match: Box<dyn Fn(&str) -> bool>| {
+            ancestors
+              .iter()
+              .chain(iter::once(&elem_ref.value()))
+              .any(|ancestor| {
+                ancestor
+                  .attr(attr_name)
+                  .map(|attr| is_match(&attr.to_lowercase()))
+                  .unwrap_or(false)
+              })
           };
+
+          let mentions_logo = |attr_name| {
+            mentions(
+              attr_name,
+              Box::new(|attr| regex!("logo([^s]|$)").is_match(attr)),
+            )
+          };
+
           if mentions_logo("class") || mentions_logo("id") {
             weight += 3;
           }
@@ -245,6 +258,20 @@ impl Icons {
             weight += 1;
           }
 
+          if let Some(site_name) = url
+            .domain()
+            .and_then(|domain| TldOption::default().build().extract(domain).unwrap().domain)
+          {
+            // if the alt contains the site_name then highest priority
+            if site_name
+              .to_lowercase()
+              .split('-')
+              .any(|segment| mentions("alt", Box::new(move |attr| attr.contains(segment))))
+            {
+              weight += 10;
+            }
+          }
+
           Some((elem_ref, weight))
         })
         .collect();