Do the property name lookup via the cache

src/lib.rs

@@ -142,7 +142,7 @@ pub mod wikipedia_infobox_analyzer {
 }
 
 #[cfg(test)]
-mod tests {
+mod tests_lib {
     use super::wikipedia_infobox_analyzer::*;
     use std::fs;
 

src/main.rs

@@ -17,13 +17,13 @@ struct Cli {
     infobox_template_file: std::path::PathBuf,
 }
 
-#[derive(Debug)]
+#[derive(Clone, Debug)]
 struct StorageDisk {
     base_path: PathBuf,
 }
 
 impl StorageDisk {
-    fn new(app_name: &str) -> Self {
+    fn new(app_name: &str) -> std::io::Result<Self> {
         // Get OS-specific cache directory, otherwise use a fallback location
         let mut path = match cache_dir() {
             Some(path) => path,
@@ -31,15 +31,10 @@ impl StorageDisk {
         };
 
         path.push(app_name);
-        fs::create_dir_all(&path)
+        match fs::create_dir_all(&path) {
-            .unwrap_or_else(|_| panic!("Failed to create storage directory: {path:?}"));
+            Ok(_) => Ok(StorageDisk { base_path: path }),
-
+            Err(_) => panic!("Failed to create storage directory: {path:?}"),
-        StorageDisk { base_path: path }
         }
-
-    #[allow(dead_code)]
-    fn delete(&self) -> std::io::Result<()> {
-        fs::remove_dir_all(&self.base_path)
     }
 
     fn store_data(&self, filename: String, data: String) -> std::io::Result<()> {
@@ -61,41 +56,68 @@ struct Cache {
 
 impl Cache {
     fn new(file_name: String, storage: StorageDisk) -> std::io::Result<Self> {
-        let mut property_labels: HashMap<String, String> = HashMap::new();
+        let property_labels: HashMap<String, String> = match storage.read_data(file_name.clone()) {
-
-        match storage.read_data(file_name.clone()) {
             Ok(raw_data) => {
-                property_labels = serde_json::from_str(&raw_data)?;
+                serde_json::from_str(&raw_data)?
             }
             Err(err) if err.kind() == io::ErrorKind::NotFound => {
-                println!("No cache found, populating the program cache with wikidata...");
+                println!("No cache found, starting an empty cache");
 
-                property_labels = HashMap::new();
+                HashMap::new()
-                property_labels.insert("P31".to_string(), "instance of".to_string());
-
-                let data = serde_json::to_string(&property_labels)?;
-                storage.store_data(file_name.clone(), data)?;
-
-                println!("Cache populated from wikidata")
             }
             Err(err) => {
                 println!("Unknown error occurred, exiting gracefully");
-                return Err(err);
+                return Err(err)
-            }
             }
+        };
 
-        Ok(Cache {
+        let cache = Cache {
             file_name,
             storage,
-            property_labels,
+            property_labels
-        })
+        };
+        cache.sync_to_storage()?;
+
+        Ok(cache)
+    }
+
+    fn sync_to_storage(&self) -> std::io::Result<()> {
+        let data = serde_json::to_string(&self.property_labels)?;
+        self.storage.store_data(self.file_name.clone(), data)?;
+        Ok(())
+    }
+
+    fn delete(&mut self) -> std::io::Result<()> {
+        fs::remove_file(self.storage.base_path.join(self.file_name.clone()))?;
+        self.property_labels = HashMap::new();
+        Ok(())
+    }
+
+    async fn lookup_property_name(&mut self, pid: u64) -> std::io::Result<String> {
+        let key = format!("P{pid}");
+        if self.property_labels.contains_key(&key) {
+            let name = self.property_labels.get(&key).unwrap().to_string();
+            return Ok(name);
+        }
+
+        // If the property name cannot be found, fetch the name
+        let name = fetch_name_for_wiki_property(pid).await;
+
+        // Update cache with new value, so it can be used the next time
+        self.property_labels
+            .insert(key, name.clone());
+        self.sync_to_storage()?;
+
+        Ok(name)
     }
 }
 
-fn main() -> Result<(), Box<dyn std::error::Error>> {
+
+#[tokio::main]
+async fn main() -> Result<(), Box<dyn std::error::Error>> {
     // Initialize cache and storage
-    let storage = StorageDisk::new("wikipedia-infobox-analyzer");
+    let storage = StorageDisk::new("wikipedia-infobox-analyzer")?;
-    let cache = Cache::new("property_labels.cache".to_string(), storage);
+    let mut _cache = Cache::new("property_labels.cache".to_string(), storage)?;
 
     // Input to the program
     let args = Cli::from_args();
@@ -119,3 +141,31 @@ fn main() -> Result<(), Box<dyn std::error::Error>> {
 
     Ok(())
 }
+
+#[cfg(test)]
+mod tests_main {
+    use super::*;
+
+    #[tokio::test]
+    async fn test_cache() {
+        let storage = StorageDisk::new("wikipedia-infobox-analyzer").unwrap();
+        let mut cache = Cache::new("property_labels.test.cache".to_string(), storage.clone()).unwrap();
+        let _ = cache.delete();
+
+        // Cache should be clear after deletion
+        assert_eq!(cache.property_labels, HashMap::new());
+
+        let name = cache.lookup_property_name(31).await.unwrap();
+        assert_eq!(name, "instance of");
+
+        // Cache should store values that are looked up
+        assert_eq!(cache.property_labels.get("P31").unwrap(), "instance of");
+
+        // Cache should be persistent if the same file is passed
+        let mut cache_new = Cache::new("property_labels.test.cache".to_string(), storage.clone()).unwrap();
+        assert_eq!(cache_new.property_labels.get("P31").unwrap(), "instance of");
+
+        // Cleanup
+        let _ = cache_new.delete();
+    }
+}