Release

.github/workflows/black_lint.yaml

@@ -0,0 +1,24 @@
+name: Black Lint Check
+
+on: [pull_request]
+
+jobs:
+  lint:
+    runs-on: ubuntu-latest
+
+    steps:
+    - name: Checkout repository
+      uses: actions/checkout@v2
+
+    - name: Set up Python
+      uses: actions/setup-python@v2
+      with:
+        python-version: 3.11
+
+    - name: Install Black
+      run: |
+        python -m pip install --upgrade pip
+        pip install black
+
+    - name: Run Black Lint Check
+      run: black --check .

.github/workflows/output.yaml

@@ -0,0 +1,39 @@
+name: ファイル出力
+
+on: workflow_dispatch
+
+jobs:
+  output:
+    runs-on: ubuntu-latest
+
+    steps:
+    - name: Checkout repository
+      uses: actions/checkout@v2
+
+    - name: Set up Python
+      uses: actions/setup-python@v2
+      with:
+        python-version: 3.11
+
+    - name: Install dependencies
+      run: |
+        python -m pip install --upgrade pip
+        pip install -r requirements.txt
+
+    - name: Output
+      run: |
+        mkdir -p assets/outputs
+        find assets/outputs -name *.xlsx -type f -exec rm {} \;
+        cd src
+        find ../assets/inputs/ -name *.xlsx -type f -exec sh -c 'python japanese_school_parser.py "$0" "../assets/outputs/$(basename "$0" .xlsx).json"' {} \;
+
+    - name: Commit and push
+      env:
+        GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+      run: |
+        git remote set-url origin https://github-actions:${GITHUB_TOKEN}@github.com/${GITHUB_REPOSITORY}
+        git config --global user.name "GITHUB_ACTOR"
+        git config --global user.email "${GITHUB_ACTOR}@users.noreply.github.com"
+        git add .
+        git commit -m "[自動生成]パース出力"
+        git push origin HEAD:${{ github.ref_name }}

.github/workflows/test.yaml

@@ -0,0 +1,28 @@
+name: テスト
+
+on: [pull_request]
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+
+    steps:
+    - name: Checkout repository
+      uses: actions/checkout@v2
+
+    - name: Set up Python
+      uses: actions/setup-python@v2
+      with:
+        python-version: 3.11
+
+    - name: Install dependencies
+      run: |
+        python -m pip install --upgrade pip
+        pip install -r requirements.txt
+
+    - name: Run pytest
+      env:
+        PYTHONPATH: ../
+      run: |
+        cd src/tests
+        pytest

.gitignore

@@ -158,3 +158,5 @@ cython_debug/
 #  and can be added to the global gitignore or merged into this file.  For a more nuclear
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
 #.idea/
+
+~$*

.vscode/launch.json

@@ -0,0 +1,21 @@
+{
+    // Use IntelliSense to learn about possible attributes.
+    // Hover to view descriptions of existing attributes.
+    // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
+    "version": "0.2.0",
+    "configurations": [
+        {
+            "name": "Python: Debug",
+            "type": "python",
+            "request": "launch",
+            "program": "japanese_school_parser.py",
+            "cwd": "${workspaceFolder}/src",
+            "args": [
+                "tests/files/multi_sheets1.xlsx",
+                "output.json",
+            ],
+            "console": "integratedTerminal",
+            "justMyCode": false
+        }
+    ]
+}

.vscode/settings.json

@@ -0,0 +1,14 @@
+{
+  "editor.indentSize": 4,
+  "editor.tabSize": 4,
+  "python.testing.pytestEnabled": true,
+  "python.testing.autoTestDiscoverOnSaveEnabled": true,
+  "python.testing.pytestArgs": [
+    "tests"
+  ],
+  "python.testing.cwd": "src",
+  "[python]": {
+    "editor.defaultFormatter": "ms-python.black-formatter"
+  },
+  "python.formatting.provider": "none",
+}

README.md

@@ -1 +1,102 @@
-# JapaneseSchoolPaser
+# JapaneseSchoolPaser
+文部科学省が公開している[大学一覧](https://www.mext.go.jp/a_menu/koutou/ichiran/mext_01853.html)のエクセルファイルから、大学情報をパースしデータとして使いやすい形で出力するプログラムです。
+
+### 使い方
+使用には[Pythonの実行環境](https://www.python.org/downloads/)が必要です。
+
+1. [文部科学省のサイト](https://www.mext.go.jp/a_menu/koutou/ichiran/mext_00006.html)よりエクセルファイルをダウンロードします。
+1. Json出力の場合、以下のpythonコマンドを実行し、出力を行います。
+    ```bash
+    python japanese_school_parser.py <入力Excelファイルパス> <出力ファイルパス>
+    ```
+
+    ```bash
+    # 例
+    python japanese_school_parser.py 20220415_mxt_daigakuc01_000021808_03-7.xlsx 20220415_mxt_daigakuc01_000021808_03-7.json
+    # => コマンドを実行したディレクトリに20220415_mxt_daigakuc01_000021808_03-7.jsonが生成されます。
+    ```
+    Pythonのdict型で出力を行いたい場合は以下の関数を呼び出します。
+    ```python
+    schools = parse_schools_to_dict(source_path)
+    ```
+
+
+## 実装状況
+### 出力形式
+- [x] Json
+- [x] Python dict型
+- [ ] yaml
+
+### 出力内容
+#### 学校基本情報
+- [x] 学校コード
+- [x] 学長
+- [ ] 学長任期
+- [ ] 所在地
+- [ ] 電話番号
+
+#### 設置者情報
+- [ ] 設置者
+- [ ] 役職
+- [ ] 役職者名
+
+#### 学部・研究科所在地(キャンパス名など)
+- [ ] 名称
+- [ ] 所在地
+- [ ] 電話番号
+
+#### 国立大学附置研究所
+- [ ] 附置研究所名
+- [ ] 所在地住所
+- [ ] 電話番号
+- [ ] 設置年月日
+- [ ] 備考
+
+#### 国立大学附置研究所附属施設
+- [ ] 附置研究所名
+- [ ] 附属施設
+- [ ] 所在地住所
+- [ ] 電話番号
+
+#### 学部
+- [x] 学部
+- [x] 学科
+- [ ] 都道府県
+- [ ] 市区町村
+- [ ] 修業年限
+- [ ] 入学定員
+- [ ] 編入定員
+- [ ] 編入(夜間)
+
+#### 研究科
+- [x] 研究科
+- [x] 専攻
+- [ ] 都道府県
+- [ ] 市区町村
+- [ ] 夜間昼夜
+- [ ] 修士/博士 前期課程
+- [ ] 博士後期課程
+- [ ] 博士課程(一貫制)
+- [ ] 専門職学位課程
+- [ ] 編入定員
+
+#### 学部沿革
+- [ ] 年月
+- [ ] 沿革
+
+#### 大学院沿革
+- [ ] 年月
+- [ ] 沿革
+
+
+## 貢献
+現状作成者が個人的に使用する部分のみ実装しています。
+
+Issueでご要望があれば追加実装を行います。
+
+Pullrequestもお待ちしています。
+
+### 備考
+出典：[文部科学省ホームページ](https://www.mext.go.jp/)
+
+本プログラムは上記出典のデータを加工して出力しています。

assets/outputs/multi_sheets1.json

@@ -0,0 +1 @@
+{"schools": [{"classification": "\u79c1\u7acb", "faculties": [{"departments": [{"name": "\u9020\u5f62\u5b66\u79d1"}], "name": "\u82b8\u8853\u5b66\u90e8"}, {"departments": [{"name": "\u30d3\u30b8\u30e5\u30a2\u30eb\u30c7\u30b6\u30a4\u30f3\u5b66\u79d1"}, {"name": "\u30a4\u30e9\u30b9\u30c8\u5b66\u79d1"}, {"name": "\u30d7\u30ed\u30c0\u30af\u30c8\u30c7\u30b6\u30a4\u30f3\u5b66\u79d1"}, {"name": "\u5efa\u7bc9\u5b66\u79d1"}], "name": "\u30c7\u30b6\u30a4\u30f3\u5b66\u90e8"}, {"departments": [{"name": "\u30de\u30f3\u30ac\u5b66\u79d1"}, {"name": "\u30a2\u30cb\u30e1\u30fc\u30b7\u30e7\u30f3\u5b66\u79d1"}], "name": "\u30de\u30f3\u30ac\u5b66\u90e8"}, {"departments": [{"name": "\u7dcf\u5408\u4eba\u6587\u5b66\u79d1"}], "name": "\u4eba\u6587\u5b66\u90e8"}, {"departments": [{"name": "\u30dd\u30d4\u30e5\u30e9\u30fc\u30ab\u30eb\u30c1\u30e3\u30fc\u5b66\u79d1"}], "name": "\u30dd\u30d4\u30e5\u30e9\u30fc\u30ab\u30eb\u30c1\u30e3\u30fc\u5b66\u90e8"}, {"departments": [{"name": "\u30e1\u30c7\u30a3\u30a2\u8868\u73fe\u5b66\u79d1"}], "name": "\u30e1\u30c7\u30a3\u30a2\u8868\u73fe\u5b66\u90e8"}, {"departments": [{"name": "\u4eba\u6587\u5b66\u79d1"}, {"name": "\u30b0\u30ed\u30fc\u30d0\u30eb\u30b9\u30bf\u30c7\u30a3\u30fc\u30ba\u5b66\u79d1"}], "name": "\u56fd\u969b\u6587\u5316\u5b66\u90e8"}], "graduate_schools": [{"majors": [{"name": "\u82b8\u8853\u5c02\u653b"}], "name": "\u82b8\u8853\u7814\u7a76\u79d1"}, {"majors": [{"name": "\u30c7\u30b6\u30a4\u30f3\u5c02\u653b"}, {"name": "\u5efa\u7bc9\u5c02\u653b"}], "name": "\u30c7\u30b6\u30a4\u30f3\u7814\u7a76\u79d1"}, {"majors": [{"name": "\u30de\u30f3\u30ac\u5c02\u653b"}], "name": "\u30de\u30f3\u30ac\u7814\u7a76\u79d1"}, {"majors": [{"name": "\u4eba\u6587\u5b66\u5c02\u653b"}], "name": "\u4eba\u6587\u5b66\u7814\u7a76\u79d1"}], "name": "\u4eac\u90fd\u7cbe\u83ef\u5927\u5b66", "president": "\u30a6\u30b9\u30d3\u30fb\u30b5\u30b3", "school_code": "F126310107644"}, {"classification": "\u79c1\u7acb", "faculties": [{"departments": [{"name": "\u937c\u7078\u5b66\u79d1"}], "name": "\u937c\u7078\u5b66\u90e8"}, {"departments": [{"name": "\u67d4\u9053\u6574\u5fa9\u5b66\u79d1"}, {"name": "\u6551\u6025\u6551\u547d\u5b66\u79d1"}], "name": "\u4fdd\u5065\u533b\u7642\u5b66\u90e8"}, {"departments": [{"name": "\u770b\u8b77\u5b66\u79d1"}], "name": "\u770b\u8b77\u5b66\u90e8"}], "graduate_schools": [{"majors": [{"name": "\u937c\u7078\u5b66\u5c02\u653b"}, {"name": "\u81e8\u5e8a\u937c\u7078\u5b66\u5c02\u653b"}], "name": "\u937c\u7078\u5b66\u7814\u7a76\u79d1"}, {"majors": [{"name": "\u67d4\u9053\u6574\u5fa9\u5b66\u5c02\u653b"}], "name": "\u4fdd\u5065\u533b\u7642\u5b66\u7814\u7a76\u79d1"}], "name": "\u660e\u6cbb\u56fd\u969b\u533b\u7642\u5927\u5b66", "president": "\u77e2\u91ce\u3000\u5fe0", "school_code": "F126310107653"}]}

requirements.txt

@@ -0,0 +1,7 @@
+et-xmlfile==1.1.0
+exceptiongroup==1.1.3
+iniconfig==2.0.0
+openpyxl==3.1.2
+pluggy==1.3.0
+pytest==7.4.2
+tomli==2.0.1

src/japanese_school_parser.py

@@ -0,0 +1,42 @@
+import json
+from typing import Any
+
+from openpyxl import load_workbook
+from openpyxl.workbook import Workbook
+from openpyxl.worksheet.worksheet import Worksheet
+
+from models.school import School
+from parser.school_parser import SchoolParser
+
+
+def parse_schools_to_dict(path: str) -> dict[str, Any]:
+    schools = parse_schools_to_model(path)
+
+    return {"schools": [school.to_dict() for school in schools]}
+
+
+def parse_schools_to_model(path: str) -> list[School]:
+    book: Workbook = load_workbook(path)
+    schools: list[School] = []
+
+    sheet_names = book.sheetnames
+    for sheet_name in sheet_names:
+        sheet: Worksheet = book[sheet_name]
+        schools.append(SchoolParser(sheet).parse())
+
+    book.close()
+
+    return schools
+
+
+def output_json(source_path: str, out_file_path: str) -> None:
+    schools = parse_schools_to_dict(source_path)
+    with open(out_file_path, "w") as file:
+        json.dump(schools, file)
+
+
+if __name__ == "__main__":
+    import sys
+
+    args = sys.argv
+    output_json(args[1], args[2])

src/models/base_info.py

@@ -0,0 +1,47 @@
+from enum import Enum
+from typing import Optional
+from models.model import Model
+
+
+class SchoolClassification(Enum):
+    NATIONAL = "国立"
+    PUBLIC = "公立"
+    PRIVATE = "私立"
+
+    @classmethod
+    def from_str(cls, class_str: str):
+        for value in list(cls):
+            if value.value == class_str:
+                return value
+
+
+class BaseInfo(Model):
+    def __init__(
+        self, name="", school_code="", president="", classification=None
+    ) -> None:
+        self.__name = name
+        self.__school_code: str = school_code
+        self.__president: str = president
+        self.__classification: Optional[SchoolClassification] = classification
+
+    def _register_base_info(self, base_info):
+        self.__name = base_info.name
+        self.__school_code: str = base_info.school_code
+        self.__president: str = base_info.president
+        self.__classification: Optional[SchoolClassification] = base_info.classification
+
+    @property
+    def name(self) -> str:
+        return self.__name
+
+    @property
+    def classification(self) -> SchoolClassification:
+        return self.__classification
+
+    @property
+    def school_code(self) -> str:
+        return self.__school_code
+
+    @property
+    def president(self) -> str:
+        return self.__president

src/models/faculty.py

@@ -0,0 +1,25 @@
+from models.model import Model
+
+
+class Department(Model):
+    def __init__(self, name: str) -> None:
+        self.__name: str = name
+        pass
+
+    @property
+    def name(self) -> str:
+        return self.__name
+
+
+class Faculty(Model):
+    def __init__(self, name: str, departments: list[Department]) -> None:
+        self.__name: str = name
+        self.__departments: list[Department] = departments
+
+    @property
+    def name(self) -> str:
+        return self.__name
+
+    @property
+    def departments(self) -> list[Department]:
+        return self.__departments

src/models/graduate_school.py

@@ -0,0 +1,25 @@
+from models.model import Model
+
+
+class Major(Model):
+    def __init__(self, name: str) -> None:
+        self.__name: str = name
+        pass
+
+    @property
+    def name(self) -> str:
+        return self.__name
+
+
+class GraduateSchool(Model):
+    def __init__(self, name: str, majors: list[Major]) -> None:
+        self.__name: str = name
+        self.__majors: list[Major] = majors
+
+    @property
+    def name(self) -> str:
+        return self.__name
+
+    @property
+    def majors(self) -> list[Major]:
+        return self.__majors