From d341c66390e38401a2f17522fc89977e2f72226c Mon Sep 17 00:00:00 2001
From: husky <nikocs@voremicrocomputers.com>
Date: Fri, 14 Mar 2025 14:29:46 -0700
Subject: [PATCH] prioritize english

---
 vorebot/src/webparse/mod.rs | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/vorebot/src/webparse/mod.rs b/vorebot/src/webparse/mod.rs
index ba84f74..0d34b0c 100644
--- a/vorebot/src/webparse/mod.rs
+++ b/vorebot/src/webparse/mod.rs
@@ -220,6 +220,17 @@ pub async fn web_parse(
     // i guess we're good
     driver.goto(url).await.map_err(|_| ())?;
 
+    let html_element = driver.find(By::Tag("html")).await.map_err(|_| ())?;
+
+    if let Some(lang) = html_element.attr("lang").await.ok().flatten() {
+        if !lang.starts_with("en") && !lang.starts_with("unknown") {
+            // i.e. non-english language
+            // fixme: remove this once we start expanding to non-english-speaking markets?
+            warn!("skipping {} due to {} language (currently prioritizing english", url, lang);
+            return Err(());
+        }
+    }
+
     let meta_elements = driver.find_all(By::Tag("meta")).await.map_err(|_| ())?;
 
     let title = driver.title().await.map_err(|_| ())?;