aquatic_ws: protocol: add info hash vector parsing; refactor

aquatic_ws/src/lib/handler.rs

@@ -194,28 +194,23 @@ pub fn handle_scrape_requests(
     requests: Drain<(ConnectionMeta, ScrapeRequest)>,
 ){
     messages_out.extend(requests.map(|(meta, request)| {
-        let num_info_hashes = request.info_hashes
-            .as_ref()
-            .map(|v| v.len())
-            .unwrap_or(0);
+        let num_info_hashes = request.info_hashes.len();
 
         let mut response = ScrapeResponse {
             files: HashMap::with_capacity(num_info_hashes),
         };
 
-        // If request.info_hashes is None, don't return scrape for all
+        // If request.info_hashes is empty, don't return scrape for all
         // torrents, even though reference server does it. It is too expensive.
-        if let Some(info_hashes) = request.info_hashes {
-            for info_hash in info_hashes {
-                if let Some(torrent_data) = torrents.get(&info_hash){
-                    let stats = ScrapeStatistics {
-                        complete: torrent_data.num_seeders,
-                        downloaded: 0, // No implementation planned
-                        incomplete: torrent_data.num_leechers,
-                    };
+        for info_hash in request.info_hashes {
+            if let Some(torrent_data) = torrents.get(&info_hash){
+                let stats = ScrapeStatistics {
+                    complete: torrent_data.num_seeders,
+                    downloaded: 0, // No implementation planned
+                    incomplete: torrent_data.num_leechers,
+                };
 
-                    response.files.insert(info_hash, stats);
-                }
+                response.files.insert(info_hash, stats);
             }
         }
 

aquatic_ws/src/lib/protocol/deserialize.rs

@@ -0,0 +1,199 @@
+use serde::{Deserializer, de::{Visitor, SeqAccess}};
+
+use super::InfoHash;
+
+
+struct TwentyByteVisitor;
+
+impl<'de> Visitor<'de> for TwentyByteVisitor {
+    type Value = [u8; 20];
+
+    fn expecting(&self, formatter: &mut std::fmt::Formatter) -> std::fmt::Result {
+        formatter.write_str("string consisting of 20 bytes")
+    }
+
+    fn visit_str<E>(self, value: &str) -> Result<Self::Value, E>
+        where E: ::serde::de::Error,
+    {
+        let mut arr = [0u8; 20];
+
+        let bytes = value.as_bytes();
+
+        if bytes.len() == 20 {
+            arr.copy_from_slice(&bytes);
+
+            Ok(arr)
+        } else {
+            Err(E::custom(format!("not 20 bytes: {}", value)))
+        }
+    }
+}
+
+
+pub fn deserialize_20_bytes<'de, D>(
+    deserializer: D
+) -> Result<[u8; 20], D::Error>
+    where D: Deserializer<'de>
+{
+    deserializer.deserialize_any(TwentyByteVisitor)
+}
+
+
+pub struct InfoHashVecVisitor;
+
+
+impl<'de> Visitor<'de> for InfoHashVecVisitor {
+    type Value = Vec<InfoHash>;
+
+    fn expecting(&self, formatter: &mut std::fmt::Formatter) -> std::fmt::Result {
+        formatter.write_str("string or array of strings consisting of 20 bytes")
+    }
+
+    fn visit_str<E>(self, value: &str) -> Result<Self::Value, E>
+        where E: ::serde::de::Error,
+    {
+        match TwentyByteVisitor::visit_str::<E>(TwentyByteVisitor, value){
+            Ok(arr) => Ok(vec![InfoHash(arr)]),
+            Err(err) => Err(E::custom(format!("got string, but {}", err)))
+        }
+    }
+
+    fn visit_seq<A>(self, mut seq: A) -> Result<Self::Value, A::Error>
+        where A: SeqAccess<'de>
+    {
+        let mut info_hashes: Self::Value = Vec::new();
+
+        while let Ok(Some(value)) = seq.next_element::<&str>(){
+            let arr = TwentyByteVisitor::visit_str(
+                TwentyByteVisitor, value
+            )?;
+
+            info_hashes.push(InfoHash(arr));
+        }
+
+        Ok(info_hashes)
+    }
+
+    fn visit_none<E>(self) -> Result<Self::Value, E>
+        where E: ::serde::de::Error
+    {
+        Ok(vec![])
+    }
+}
+
+
+/// Empty vector is returned if value is null or any invalid info hash
+/// is present
+pub fn deserialize_info_hashes<'de, D>(
+    deserializer: D
+) -> Result<Vec<InfoHash>, D::Error>
+    where D: Deserializer<'de>,
+{
+    Ok(deserializer.deserialize_any(InfoHashVecVisitor).unwrap_or_default())
+}
+
+
+#[cfg(test)]
+mod tests {
+    use serde::Deserialize;
+
+    use super::*;
+
+    fn info_hash_from_bytes(bytes: &[u8]) -> InfoHash {
+        let mut arr = [0u8; 20];
+
+        arr.copy_from_slice(bytes);
+
+        InfoHash(arr)
+    }
+
+    #[test]
+    fn test_deserialize_20_bytes(){
+        let input = r#""aaaabbbbccccddddeeee""#;
+
+        let expected = info_hash_from_bytes(b"aaaabbbbccccddddeeee");
+        let observed: InfoHash = serde_json::from_str(input).unwrap();
+
+        assert_eq!(observed, expected);
+
+        let input = r#""1aaaabbbbccccddddeeee""#;
+        let res_info_hash: Result<InfoHash, _> = serde_json::from_str(input);
+
+        assert!(res_info_hash.is_err());
+
+        let input = r#""aaaabbbbccccddddeeeö""#;
+        let res_info_hash: Result<InfoHash, _> = serde_json::from_str(input);
+
+        assert!(res_info_hash.is_err());
+    }
+
+    #[derive(Debug, PartialEq, Eq, Deserialize)]
+    struct Test {
+        #[serde(deserialize_with = "deserialize_info_hashes", default)]
+        info_hashes: Vec<InfoHash>,
+    }
+
+
+    #[test]
+    fn test_deserialize_info_hashes_vec(){
+        let input = r#"{
+            "info_hashes": ["aaaabbbbccccddddeeee", "aaaabbbbccccddddeeee"]
+        }"#;
+
+        let expected = Test {
+            info_hashes: vec![
+                info_hash_from_bytes(b"aaaabbbbccccddddeeee"),
+                info_hash_from_bytes(b"aaaabbbbccccddddeeee"),
+            ]
+        };
+
+        let observed: Test = serde_json::from_str(input).unwrap();
+
+        assert_eq!(observed, expected);
+    }
+
+    #[test]
+    fn test_deserialize_info_hashes_str(){
+        let input = r#"{
+            "info_hashes": "aaaabbbbccccddddeeee"
+        }"#;
+
+        let expected = Test {
+            info_hashes: vec![
+                info_hash_from_bytes(b"aaaabbbbccccddddeeee"),
+            ]
+        };
+
+        let observed: Test = serde_json::from_str(input).unwrap();
+
+        assert_eq!(observed, expected);
+    }
+
+    #[test]
+    fn test_deserialize_info_hashes_null(){
+        let input = r#"{
+            "info_hashes": null
+        }"#;
+
+        let expected = Test {
+            info_hashes: vec![]
+        };
+
+        let observed: Test = serde_json::from_str(input).unwrap();
+
+        assert_eq!(observed, expected);
+    }
+
+    #[test]
+    fn test_deserialize_info_hashes_missing(){
+        let input = r#"{}"#;
+
+        let expected = Test {
+            info_hashes: vec![]
+        };
+
+        let observed: Test = serde_json::from_str(input).unwrap();
+
+        assert_eq!(observed, expected);
+    }
+}

aquatic_ws/src/lib/protocol/mod.rs

@@ -1,64 +1,31 @@
 use hashbrown::HashMap;
-use serde::{Serialize, Deserialize, Deserializer, de::Visitor};
+use serde::{Serialize, Deserialize};
+
+pub mod deserialize;
+
+use deserialize::*;
 
 
-struct TwentyAsciiBytesVisitor;
-
-impl<'de> Visitor<'de> for TwentyAsciiBytesVisitor {
-    type Value = [u8; 20];
-
-    fn expecting(&self, formatter: &mut std::fmt::Formatter) -> std::fmt::Result {
-        formatter.write_str("string consisting of 20 bytes")
-    }
-
-    fn visit_str<E>(self, value: &str) -> Result<Self::Value, E>
-    where
-        E: ::serde::de::Error,
-    {
-        let mut arr = [0u8; 20];
-
-        let bytes = value.as_bytes();
-
-        if value.is_ascii() && bytes.len() == 20 {
-            arr.copy_from_slice(&bytes);
-
-            Ok(arr)
-        } else {
-            Err(E::custom(format!("not 20 ascii bytes: {}", value)))
-        }
-    }
-}
-
-
-fn deserialize_20_ascii_bytes<'de, D>(
-    deserializer: D
-) -> Result<[u8; 20], D::Error>
-    where D: Deserializer<'de>
-{
-    deserializer.deserialize_any(TwentyAsciiBytesVisitor)
-}
-
-
-#[derive(Clone, Copy, Hash, PartialEq, Eq, Serialize, Deserialize)]
+#[derive(Debug, Clone, Copy, Hash, PartialEq, Eq, Serialize, Deserialize)]
 #[serde(transparent)]
 pub struct PeerId(
-    #[serde(deserialize_with = "deserialize_20_ascii_bytes")]
+    #[serde(deserialize_with = "deserialize_20_bytes")]
     pub [u8; 20]
 );
 
 
-#[derive(Clone, Copy, Hash, PartialEq, Eq, Serialize, Deserialize)]
+#[derive(Debug, Clone, Copy, Hash, PartialEq, Eq, Serialize, Deserialize)]
 #[serde(transparent)]
 pub struct InfoHash(
-    #[serde(deserialize_with = "deserialize_20_ascii_bytes")]
+    #[serde(deserialize_with = "deserialize_20_bytes")]
     pub [u8; 20]
 );
 
 
-#[derive(Clone, Copy, PartialEq, Eq, Serialize, Deserialize)]
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize, Deserialize)]
 #[serde(transparent)]
 pub struct OfferId(
-    #[serde(deserialize_with = "deserialize_20_ascii_bytes")]
+    #[serde(deserialize_with = "deserialize_20_bytes")]
     pub [u8; 20]
 );
 
@@ -173,7 +140,8 @@ pub struct ScrapeRequest {
     // If omitted, scrape for all torrents, apparently
     // There is some kind of parsing here too which accepts a single info hash
     // and puts it into a vector
-    pub info_hashes: Option<Vec<InfoHash>>,
+    #[serde(deserialize_with = "deserialize_info_hashes", default)]
+    pub info_hashes: Vec<InfoHash>,
 }