Early broken bincode implementation

Cargo.lock

@@ -32,6 +32,15 @@ version = "1.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "d468802bab17cbc0cc575e9b053f41e72aa36bfa6b7f55e3529ffa43161b97fa"
 
+[[package]]
+name = "bincode"
+version = "1.3.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b1f45e9417d87227c7a56d22e471c6206462cba514c7590c09aff4cf6d1ddcad"
+dependencies = [
+ "serde",
+]
+
 [[package]]
 name = "bitflags"
 version = "1.3.2"
@@ -136,6 +145,7 @@ dependencies = [
  "codes-agency",
  "codes-common",
  "csv",
+ "serde",
  "tera",
 ]
 
@@ -315,6 +325,7 @@ version = "0.1.0"
 name = "jichanorg-jmdict"
 version = "0.1.0"
 dependencies = [
+ "bincode",
  "codes-iso-639",
  "serde",
  "serde_json",

jmdict/.gitignore

@@ -0,0 +1 @@
+JMdict*.bin

jmdict/Cargo.toml

@@ -6,6 +6,7 @@ edition = "2021"
 # See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
 
 [dependencies]
+bincode = "1.3.3"
 serde = { version = "1.0.175", features = ["serde_derive"] }
 serde_json = "1.0.103"
 
@@ -13,5 +14,4 @@ serde_json = "1.0.103"
 git = "https://github.com/ElnuDev/rust-codes.git"
 rev = "bea9e74d0ec2aaec0b74c2fa6f8da0490a638090"
 version = "0.1.5"
-default-features = false # no serde
 features = ["part_2"]

jmdict/src/models/jmdict.rs

@@ -1,15 +1,15 @@
 use codes_iso_639::part_2::LanguageCode;
-use serde::Deserialize;
+use serde::{Serialize, Deserialize};
 
 use crate::serial::*;
 
-#[derive(Deserialize)]
+#[derive(Serialize, Deserialize)]
 #[serde(rename_all(deserialize = "PascalCase"), deny_unknown_fields)]
 pub struct Jmdict {
     pub entries: Vec<JmdictEntry>,
 }
 
-#[derive(Deserialize)]
+#[derive(Serialize, Deserialize)]
 #[serde(rename_all(deserialize = "PascalCase"), deny_unknown_fields)]
 pub struct JmdictEntry {
     pub sequence: i32,
@@ -19,14 +19,16 @@ pub struct JmdictEntry {
     pub sense: Vec<JmdictSense>,
 }
 
-#[derive(Deserialize)]
+#[derive(Serialize, Deserialize)]
 #[serde(rename_all(deserialize = "PascalCase"), deny_unknown_fields)]
 pub struct JmdictExample {
-    pub srce: JmdictExampleSource,
+    // Always empty with latest version of JMdict
+    #[serde(default, skip_serializing, deserialize_with = "err_if_not_blank")]
+    srcaswdfa: serde::de::IgnoredAny,
     pub sentences: Vec<JmdictExampleSentence>,
 }
 
-#[derive(Deserialize)]
+#[derive(Serialize, Deserialize)]
 #[serde(rename_all(deserialize = "PascalCase"), deny_unknown_fields)]
 pub struct JmdictExampleSentence {
     #[serde(deserialize_with = "lang")]
@@ -34,33 +36,26 @@ pub struct JmdictExampleSentence {
     pub text: String,
 }
 
-#[derive(Deserialize)]
-#[serde(rename_all(deserialize = "PascalCase"), deny_unknown_fields)]
-pub struct JmdictExampleSource {
-    pub id: String,
-    pub src_type: JmdictExampleSourceSrcType,
-}
-
-#[derive(Deserialize)]
+#[derive(Serialize, Deserialize)]
 #[serde(rename_all(deserialize = "PascalCase"), deny_unknown_fields)]
 pub enum JmdictExampleSourceSrcType {
     Amogus,
 }
 
-#[derive(Deserialize)]
+#[derive(Serialize, Deserialize)]
 #[serde(rename_all(deserialize = "PascalCase"), deny_unknown_fields)]
 pub struct JmdictGlossary {
     pub content: String,
     #[serde(deserialize_with = "lang_default_eng")]
     pub language: LanguageCode,
     // Always empty with latest version of JMdict
-    #[serde(default, skip_serializing, deserialize_with = "err_if_not_null")]
+    #[serde(default, skip_serializing, deserialize_with = "err_if_not_blank")]
     gender: serde::de::IgnoredAny,
     pub r#type: Option<JmdictGlossaryType>,
 }
 
 // https://github.com/FooSoft/yomichan/issues/2057
-#[derive(Deserialize)]
+#[derive(Serialize, Deserialize)]
 pub enum JmdictGlossaryType {
     #[serde(rename = "lit")]
     Literal,
@@ -72,7 +67,7 @@ pub enum JmdictGlossaryType {
     Trademark,
 }
 
-#[derive(Deserialize)]
+#[derive(Serialize, Deserialize)]
 #[serde(rename_all(deserialize = "PascalCase"), deny_unknown_fields)]
 pub struct JmdictKanji {
     pub expression: String,
@@ -82,12 +77,12 @@ pub struct JmdictKanji {
     pub priorities: Vec<String>,
 }
 
-#[derive(Deserialize)]
+#[derive(Serialize, Deserialize)]
 #[serde(rename_all(deserialize = "PascalCase"), deny_unknown_fields)]
 pub struct JmdictReading {
     pub reading: String,
     // Always empty with latest version of JMdict
-    #[serde(default, skip_serializing, deserialize_with = "err_if_not_null")]
+    #[serde(default, skip_serializing, deserialize_with = "err_if_not_blank")]
     no_kanji: serde::de::IgnoredAny,
     #[serde(deserialize_with = "nullable_list")]
     pub restrictions: Vec<String>,
@@ -97,7 +92,7 @@ pub struct JmdictReading {
     pub priorities: Vec<String>,
 }
 
-#[derive(Deserialize)]
+#[derive(Serialize, Deserialize)]
 #[serde(rename_all(deserialize = "PascalCase"), deny_unknown_fields)]
 pub struct JmdictSense {
     #[serde(deserialize_with = "nullable_list")]
@@ -126,7 +121,7 @@ pub struct JmdictSense {
     pub examples: Vec<JmdictExample>,
 }
 
-#[derive(Deserialize)]
+#[derive(Serialize, Deserialize)]
 #[serde(rename_all(deserialize = "PascalCase"), deny_unknown_fields)]
 pub struct JmdictSource {
     pub content: String,

jmdict/src/serial.rs

@@ -84,13 +84,14 @@ where
         .map_err(|err| serde::de::Error::custom(err))
 }
 
-pub fn err_if_not_null<'de, D>(deserializer: D) -> Result<serde::de::IgnoredAny, D::Error>
+pub fn err_if_not_blank<'de, D>(deserializer: D) -> Result<serde::de::IgnoredAny, D::Error>
 where
     D: Deserializer<'de>,
 {
-    match serde_json::Value::deserialize(deserializer)? {
-        Value::Null => Ok(serde::de::IgnoredAny),
-        Value::String(str) if str.is_empty() => Ok(serde::de::IgnoredAny),
-        _ => Err(serde::de::Error::custom("Expected null or empty string!")),
+    let value = serde_json::Value::deserialize(deserializer)?;
+    if value.is_null() || value == "" {
+        Ok(serde::de::IgnoredAny)
+    } else {
+        Err(serde::de::Error::custom("Expected null or empty string!"))
     }
 }

jmdict/src/tests.rs

@@ -1,10 +1,9 @@
-use std::fs;
+use std::{fs::{self, File}, io::{Write, Read}};
 
 use crate::jmdict::Jmdict;
 
-#[test]
-fn jmdict() {
-    let contents = fs::read_dir("loader")
+fn get_jmdict_json() -> String {
+    fs::read_dir("loader")
         .unwrap()
         .filter_map(|file| {
             let file = file.unwrap();
@@ -16,6 +15,23 @@ fn jmdict() {
         })
         .next()
         .unwrap()
-        .unwrap();
-    let _jmdict: Jmdict = serde_json::from_str(&contents).unwrap();
+        .unwrap()
+}
+
+const BIN_PATH: &str = "JMdict.bin";
+
+#[test]
+fn jmdict_load_and_serialize() {
+    let jmdict: Jmdict = serde_json::from_str(&get_jmdict_json()).unwrap();
+    let encoded: Vec<u8> = bincode::serialize(&jmdict).unwrap();
+    let mut file = File::create(BIN_PATH).unwrap();
+    file.write_all(&encoded).unwrap();
+}
+
+#[test]
+fn jmdict_deserialize() {
+    let mut file = File::open(BIN_PATH).unwrap();
+    let mut buffer = Vec::new();
+    file.read_to_end(&mut buffer).unwrap();
+    let _jmdict: Jmdict = bincode::deserialize(&buffer).unwrap();
 }