feat: cache kanji request, add kklc tools

.gitignore

@@ -32,5 +32,7 @@ src/**/*.webp
 .obsidian/
 .stfolder/
 
-# generated kanji
-public/static/kanji/*
+# Kanji generator
+tools/kklc/kklc.csv
+tools/kklc/target/
+public/static/kanji/

src/components/cards/kanji/data.svelte.ts

@@ -1,19 +1,18 @@
+type Ruby = Array<[string, string]>;
+
 export interface KKLCEntry {
-  entry: number;
+  id: number;
   char: string;
-  meanings: string[];
-  on: string[];
-  kun: string[];
-  examples: Array<
-    [Array<[string, string]>, string]
-  >;
+  keys: string[];
   senses: string[];
+  onyomi: string[];
+  kunyomi: string[];
+  examples: Array<[string, Ruby]>;
 }
 
 
-async function chooseKanji(): Promise<number> {
+async function chooseId(): Promise<number> {
   const date = new Date().toLocaleDateString('en');
-  console.log(date);
   const data = new TextEncoder().encode(date);
 
   const hash = await crypto.subtle.digest('SHA-256', data);
@@ -23,10 +22,30 @@ async function chooseKanji(): Promise<number> {
 
   const min = 1;
   const max = 2300;
-  const id = min + (hashValue % (max - min + 1));
-  return id;
+  return min + (hashValue % (max - min + 1));
 }
 
-export async function fetchKanji(): Promise<KKLCEntry> {
-  return await fetch(`/static/kanji/${await chooseKanji()}.json`).then(res => res.json());
+function tryGetCache(id: number) {
+  const item = localStorage.getItem('kanji');
+  if (!item) return;
+
+  const cache = JSON.parse(item);
+  if (cache.id === id) {
+    return cache.data;
+  }
+}
+
+function insertCache(id: number, data: KKLCEntry) {
+  localStorage.setItem('kanji', JSON.stringify({ id, data }));
+}
+
+export async function getKanji(): Promise<KKLCEntry> {
+  const id = await chooseId();
+
+  const cache = tryGetCache(id);
+  if (cache) return cache;
+
+  const data = await fetch(`/static/kanji/${id}.json`).then(res => res.json());
+  insertCache(id, data);
+  return data;
 }

src/components/cards/kanji/kanji.svelte

@@ -1,9 +1,9 @@
 <script lang="ts">
-  import { fetchKanji, type KKLCEntry } from './data.svelte.ts';
+  import { getKanji, type KKLCEntry } from './data.svelte.ts';
 
   let state = $state<Promise<KKLCEntry>>(new Promise(() => {}));
 
-  $effect(() => void (state = fetchKanji()));
+  $effect(() => void (state = getKanji()));
 </script>
 
 
@@ -19,26 +19,26 @@
     </div>
     <div class="info-meta">
       <div class="info-key">
-        {state.meanings.join(', ')}
+        {state.keys.join(', ')}
       </div>
       <div class="info-on">
-        {state.on.join(', ')}
+        {state.onyomi.join(', ')}
       </div>
       <div class="info-kun">
-        {state.kun.join(', ')}
+        {state.kunyomi.join(', ')}
       </div>
     </div>
   </div>
-  <table>
+  <table class="examples">
     <tbody>
-    {#each state.examples as [example, meaning]}
+    {#each state.examples as [meaning, example]}
       <tr>
-        <td>
+        <td class="examples-ja">
           <ruby>
-            {#each example as [kanji, furigana]}{kanji}<rt>{furigana || ''}</rt>{/each}
+            {#each example as [expr, ruby]}{expr}<rt>{ruby||''}</rt>{/each}
           </ruby>
         </td>
-        <td>
+        <td class="examples-en">
           {meaning}
         </td>
       </tr>

src/styles/components/_kanji.scss

@@ -15,6 +15,7 @@
   }
 
   .info-char {
+    align-self: flex-start;
     position: relative;
     padding: 0.1em;
     border: 1px dashed black;
@@ -42,8 +43,10 @@
       display: block;
     }
   }
-  .ruby-cell {
-    min-width: 5em;
+
+  .examples-ja {
+    padding-right: 0.5em;
+    max-width: 3em;
   }
 
   $side: 5em;

tools/kklc/Cargo.lock

@@ -0,0 +1,124 @@
+# This file is automatically @generated by Cargo.
+# It is not intended for manual editing.
+version = 3
+
+[[package]]
+name = "csv"
+version = "1.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ac574ff4d437a7b5ad237ef331c17ccca63c46479e5b5453eb8e10bb99a759fe"
+dependencies = [
+ "csv-core",
+ "itoa",
+ "ryu",
+ "serde",
+]
+
+[[package]]
+name = "csv-core"
+version = "0.1.11"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5efa2b3d7902f4b634a20cae3c9c4e6209dc4779feb6863329607560143efa70"
+dependencies = [
+ "memchr",
+]
+
+[[package]]
+name = "furigana"
+version = "0.1.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9f66a30b0d76741b8ddf54f5cd6bd80c635810ce9dcef147c14ee545e1b07bcd"
+
+[[package]]
+name = "itoa"
+version = "1.0.10"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b1a46d1a171d865aa5f83f92695765caa047a9b4cbae2cbf37dbd613a793fd4c"
+
+[[package]]
+name = "kklc"
+version = "0.1.0"
+dependencies = [
+ "csv",
+ "furigana",
+ "serde",
+ "serde_json",
+]
+
+[[package]]
+name = "memchr"
+version = "2.7.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "523dc4f511e55ab87b694dc30d0f820d60906ef06413f93d4d7a1385599cc149"
+
+[[package]]
+name = "proc-macro2"
+version = "1.0.78"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e2422ad645d89c99f8f3e6b88a9fdeca7fabeac836b1002371c4367c8f984aae"
+dependencies = [
+ "unicode-ident",
+]
+
+[[package]]
+name = "quote"
+version = "1.0.35"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "291ec9ab5efd934aaf503a6466c5d5251535d108ee747472c3977cc5acc868ef"
+dependencies = [
+ "proc-macro2",
+]
+
+[[package]]
+name = "ryu"
+version = "1.0.16"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f98d2aa92eebf49b69786be48e4477826b256916e84a57ff2a4f21923b48eb4c"
+
+[[package]]
+name = "serde"
+version = "1.0.196"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "870026e60fa08c69f064aa766c10f10b1d62db9ccd4d0abb206472bee0ce3b32"
+dependencies = [
+ "serde_derive",
+]
+
+[[package]]
+name = "serde_derive"
+version = "1.0.196"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "33c85360c95e7d137454dc81d9a4ed2b8efd8fbe19cee57357b32b9771fccb67"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "serde_json"
+version = "1.0.113"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "69801b70b1c3dac963ecb03a364ba0ceda9cf60c71cfe475e99864759c8b8a79"
+dependencies = [
+ "itoa",
+ "ryu",
+ "serde",
+]
+
+[[package]]
+name = "syn"
+version = "2.0.48"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0f3531638e407dfc0814761abb7c00a5b54992b849452a0646b7f65c9f770f3f"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "unicode-ident",
+]
+
+[[package]]
+name = "unicode-ident"
+version = "1.0.12"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3354b9ac3fae1ff6755cb6db53683adb661634f67557942dea4facebec0fee4b"

tools/kklc/Cargo.toml

@@ -0,0 +1,12 @@
+[package]
+name = "kklc"
+version = "0.1.0"
+edition = "2021"
+
+# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+
+[dependencies]
+serde = { version = "1.0.196", features = ["derive"] }
+serde_json = "1.0.113"
+csv = "1.3"
+furigana = "0.1.1"

tools/kklc/src/main.rs

@@ -0,0 +1,163 @@
+use std::{collections::HashMap, error::Error};
+
+use std::fs::File;
+use serde::{Serialize, Deserialize};
+use csv::{ReaderBuilder, StringRecord};
+
+
+const PATH_CSV: &str = "kklc.csv";
+
+
+type Readings = HashMap<String, Vec<String>>;
+type Ruby = Vec<(String, Option<String>)>;
+
+#[derive(Debug, Serialize, Deserialize)]
+struct KKLCEntry {
+    id: i32,
+    char: String,
+    keys: Vec<String>,
+    senses: Vec<String>,
+    onyomi: Vec<String>,
+    kunyomi: Vec<String>,
+    examples: Vec<(String, Ruby)>,
+}
+
+
+fn get_readings() -> HashMap<String, Vec<String>> {
+    ReaderBuilder::new()
+        .from_path(PATH_CSV)
+        .expect("Failed opening CSV")
+        .records()
+        .map(|record| {
+            let record = record.expect("Error reading CSV record");
+
+            let kanji = record.get(1).unwrap();
+            let onyomi = record.get(3)
+                .unwrap()
+                .split("\n\n")
+                .map(|s| s.split('(').next().unwrap().to_owned());
+            let kunyomi = record.get(4)
+                .unwrap()
+                .split("\n\n")
+                .map(|s| s.split('(').next().unwrap().to_owned());
+
+            (kanji.to_owned(), kunyomi.chain(onyomi).collect())
+        })
+        .collect()
+}
+
+fn map_vocab(str: &str) -> (String, String, String) {
+    let mut iter = str.split(" ・ ").map(String::from);
+    (
+        iter.next().unwrap()
+            .replace("¹", "")
+            .replace("²", "")
+            .replace("…", "")
+            .replace("*", "")
+            .trim()
+            .into(),
+        iter.next().unwrap()
+            .split('[').next().unwrap()
+            .split('(').next().unwrap()
+            .replace(" ", "")
+            .replace("¹", "")
+            .replace("²", "")
+            .replace("…", "")
+            .trim()
+            .into(),
+        iter.next().unwrap(),
+    )
+}
+
+fn map_ruby(fst: &str, snd: &str, readings: &Readings) -> Ruby {
+    if !fst.contains("○") {
+        let mapping = furigana::map(fst, snd, &readings)
+            .into_iter()
+            .max_by_key(|f| f.accuracy);
+
+        if let Some(mapping) = mapping {
+            if mapping.accuracy > 0 {
+                return mapping.furigana
+                    .into_iter()
+                    .map(|x| (x.segment.to_owned(), x.furigana.map(String::from)))
+                    .collect()
+            }
+        }
+    }
+
+    vec![(fst.to_owned(), Some(snd.to_owned()))]
+}
+
+fn map_example(vocab: &[(String, String, String)], readings: &Readings) -> Vec<(String, Ruby)> {
+    vocab.iter()
+        .map(|(fst, snd, word)| (word.to_owned(), map_ruby(fst, snd, readings)))
+        .collect()
+}
+
+fn map_record(record: StringRecord, readings: &Readings) -> KKLCEntry {
+    let id = record.get(0)
+        .unwrap()
+        .parse()
+        .unwrap();
+    let char: String = record.get(1)
+        .unwrap()
+        .into();
+    let keys = record.get(2)
+        .unwrap_or("")
+        .split("\n\n")
+        .map(String::from)
+        .collect();
+    let onyomi = record.get(3)
+        .unwrap_or("")
+        .split("\n\n")
+        .map(String::from)
+        .collect::<Vec<_>>();
+    let kunyomi = record.get(4)
+        .unwrap_or("")
+        .split("\n\n")
+        .map(String::from)
+        .collect::<Vec<_>>();
+    let vocab = record.get(5)
+        .unwrap_or("")
+        .split("\n\n")
+        .map(map_vocab)
+        .collect::<Vec<_>>();
+    let senses = record.get(6)
+        .unwrap_or("")
+        .split("\n\n")
+        .filter(|s| !s.is_empty())
+        .map(String::from)
+        .collect();
+
+    let examples = map_example(&vocab, readings);
+
+    KKLCEntry { id, char, keys, senses, onyomi, kunyomi, examples }
+}
+
+fn get_entries(readings: &HashMap<String, Vec<String>>) -> Vec<KKLCEntry> {
+    let mut reader = ReaderBuilder::new()
+        .from_path(PATH_CSV)
+        .expect("Error opening file");
+
+    reader.records()
+        .map(|record| {
+            let record = record.expect("Error reading CSV record");
+            map_record(record, readings)
+        })
+        .collect()
+}
+
+
+fn main() -> Result<(), Box<dyn Error>> {
+    let readings = get_readings();
+    let entries = get_entries(&readings);
+
+    for (i, entry) in entries.iter().enumerate() {
+        let path = format!("../../public/static/kanji/{}.json", i + 1);
+        let file = File::create(path).expect("Error creating file");
+
+        serde_json::to_writer(&file, entry).expect("Error writing JSON to file");
+    }
+
+    Ok(())
+}