[김보겸] 시스템 설계 면접 완벽 가이드 7, 8주차 정리

2026-system-design-interview/07주차/김보겸/weighted-trie-demo/.gitignore

@@ -0,0 +1,4 @@
+backend/.venv/
+frontend/node_modules/
+frontend/dist/
+frontend/.vite/

2026-system-design-interview/07주차/김보겸/weighted-trie-demo/README.md

@@ -0,0 +1,34 @@
+# Weighted Trie Demo
+
+FastAPI가 가중치 트라이를 만들고 React가 그 트라이를 한 번 받아 브라우저 메모리에서 자동완성을 수행하는 최소 데모입니다.
+
+## 실행
+
+터미널 1:
+
+```bash
+cd weighted-trie-demo/backend
+python3 -m venv .venv
+source .venv/bin/activate
+pip install -r requirements.txt
+uvicorn main:app --reload --host 127.0.0.1 --port 8000
+```
+
+터미널 2:
+
+```bash
+cd weighted-trie-demo/frontend
+npm install
+npm run dev
+```
+
+브라우저에서 `http://127.0.0.1:5173`을 엽니다.
+
+## 핵심 흐름
+
+1. 서버는 `WORDS` 목록의 각 단어를 한 글자씩 트라이에 삽입합니다.
+2. 각 prefix 노드의 `top` 배열에는 그 prefix에서 가장 높은 가중치의 후보 5개를 미리 저장합니다.
+3. 클라이언트는 `/api/trie`를 앱 시작 시 한 번만 가져옵니다.
+4. 사용자가 입력할 때 React는 네트워크 요청 없이 `children[char]`를 따라 내려가고, 도착 노드의 `top`을 화면에 보여줍니다.
+
+예를 들어 `ap`를 입력하면 `root -> a -> p` 노드까지 이동하고, 그 노드에 저장된 `apple`, `application`, `app` 같은 후보가 가중치 순서로 반환됩니다.

2026-system-design-interview/07주차/김보겸/weighted-trie-demo/backend/main.py

@@ -0,0 +1,129 @@
+from __future__ import annotations
+
+from typing import List
+
+from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
+from pydantic import BaseModel
+
+
+MAX_SUGGESTIONS_PER_NODE = 5
+
+
+class WeightedWord(BaseModel):
+    word: str
+    weight: int
+    description: str
+
+
+WORDS = [
+    WeightedWord(word="apple", weight=92, description="사과"),
+    WeightedWord(word="app", weight=71, description="앱"),
+    WeightedWord(word="application", weight=83, description="응용 프로그램"),
+    WeightedWord(word="apply", weight=64, description="지원하다/적용하다"),
+    WeightedWord(word="apricot", weight=41, description="살구"),
+    WeightedWord(word="banana", weight=88, description="바나나"),
+    WeightedWord(word="band", weight=49, description="밴드"),
+    WeightedWord(word="bank", weight=74, description="은행"),
+    WeightedWord(word="bar", weight=38, description="막대/바"),
+    WeightedWord(word="cat", weight=68, description="고양이"),
+    WeightedWord(word="catalog", weight=53, description="목록"),
+    WeightedWord(word="catch", weight=46, description="잡다"),
+    WeightedWord(word="dog", weight=57, description="개"),
+    WeightedWord(word="door", weight=63, description="문"),
+    WeightedWord(word="dorm", weight=31, description="기숙사"),
+]
+
+
+def word_weight(item: WeightedWord) -> int:
+    return item.weight
+
+
+app = FastAPI(title="Weighted Trie Demo")
+
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["http://localhost:5173", "http://127.0.0.1:5173"],
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+
+def build_weighted_trie(words: List[WeightedWord]) -> dict:
+    def new_node() -> dict:
+        node = {}
+        node["top"] = []
+        node["terminal"] = None
+        node["children"] = {}
+
+        return node
+
+    def remember_best_words(node: dict, item: WeightedWord) -> None:
+        top_words = node["top"]
+        top_words.append(item)
+        top_words.sort(key=word_weight, reverse=True)
+
+        while len(top_words) > MAX_SUGGESTIONS_PER_NODE:
+            top_words.pop()
+
+    def to_json(node: dict) -> dict:
+        top_json = []
+        for candidate in node["top"]:
+            top_json.append(candidate.model_dump())
+
+        terminal_json = None
+        terminal_word = node["terminal"]
+        if terminal_word is not None:
+            terminal_json = terminal_word.model_dump()
+
+        children_json = {}
+        child_chars = list(node["children"].keys())
+        child_chars.sort()
+
+        for char in child_chars:
+            child_node = node["children"][char]
+            children_json[char] = to_json(child_node)
+
+        result = {}
+        result["top"] = top_json
+        result["terminal"] = terminal_json
+        result["children"] = children_json
+
+        return result
+
+    root = new_node()
+
+    for item in words:
+        current = root
+        remember_best_words(current, item)
+
+        for char in item.word.lower():
+            children = current["children"]
+            if char not in children:
+                children[char] = new_node()
+
+            current = children[char]
+            remember_best_words(current, item)
+
+        current["terminal"] = item
+
+    return to_json(root)
+
+
+@app.get("/api/words")
+def list_words() -> list[WeightedWord]:
+    sorted_words = list(WORDS)
+    sorted_words.sort(key=word_weight, reverse=True)
+
+    return sorted_words
+
+
+@app.get("/api/trie")
+def get_trie() -> dict:
+    response = {}
+    response["version"] = "demo-1"
+    response["maxSuggestionsPerNode"] = MAX_SUGGESTIONS_PER_NODE
+    response["node"] = build_weighted_trie(WORDS)
+
+    return response

2026-system-design-interview/07주차/김보겸/weighted-trie-demo/backend/requirements.txt

@@ -0,0 +1,2 @@
+fastapi==0.115.6
+uvicorn[standard]==0.34.0

2026-system-design-interview/07주차/김보겸/weighted-trie-demo/frontend/index.html

@@ -0,0 +1,12 @@
+<!doctype html>
+<html lang="ko">
+  <head>
+    <meta charset="UTF-8" />
+    <meta name="viewport" content="width=device-width, initial-scale=1.0" />
+    <title>Weighted Trie Demo</title>
+  </head>
+  <body>
+    <div id="root"></div>
+    <script type="module" src="/src/main.tsx"></script>
+  </body>
+</html>