Merge pull request #1 from LangChain-OpenTutorial/soo-dev

feat: create new feature

Author: sooyoung-wind

.gitignore

@@ -1,2 +1,3 @@
 __pycache__/
-.DS_Store
+.DS_Store
+.python-version

.pre-commit-config.yaml

@@ -0,0 +1,14 @@
+# See https://pre-commit.com for more information
+# See https://pre-commit.com/hooks.html for more hooks
+repos:
+-   repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v5.0.0
+    hooks:
+    -   id: trailing-whitespace
+    -   id: end-of-file-fixer
+    -   id: check-yaml
+    -   id: check-added-large-files
+-   repo: https://github.com/psf/black
+    rev: 24.10.0
+    hooks:
+    -   id: black

README.md

@@ -12,3 +12,14 @@ python mdconverter.py --filename <notebook-file-path>
 python mdconverter.py --filename sample/05-GoogleGenerativeAI.ipynb
 ```
 
+개발 중...
+```python
+
+from mdconverter.mdconverter_class import Mdconverter
+app = Mdconverter()
+app.filename = "sample/05-GoogleGenerativeAI.ipynb"
+app.output_filename = 'sample/New-05-GoogleGenerativeAI.md'
+app.css_filename = "css/styles.css"
+
+app.run()
+```

css/styles.css

@@ -0,0 +1,15 @@
+.custom {
+    background-color: #008d8d;
+    color: white;
+    padding: 0.25em 0.5em 0.25em 0.5em;
+    white-space: pre-wrap;       /* css-3 */
+    white-space: -moz-pre-wrap;  /* Mozilla, since 1999 */
+    white-space: -pre-wrap;      /* Opera 4-6 */
+    white-space: -o-pre-wrap;    /* Opera 7 */
+    word-wrap: break-word;
+}
+
+pre {
+    background-color: #027c7c;
+    padding-left: 0.5em;
+}

mdconverter/__init__.py

@@ -4,25 +4,15 @@
 from typing import List, Dict, Any, Optional
 
 
-def get_default_css() -> str:
+def get_default_css(css_filename: str) -> str:
     """기본 CSS 스타일을 반환합니다."""
-    return """<style>
-    .custom {
-        background-color: #008d8d;
-        color: white;
-        padding: 0.25em 0.5em 0.25em 0.5em;
-        white-space: pre-wrap;       /* css-3 */
-        white-space: -moz-pre-wrap;  /* Mozilla, since 1999 */
-        white-space: -pre-wrap;      /* Opera 4-6 */
-        white-space: -o-pre-wrap;    /* Opera 7 */
-        word-wrap: break-word;    
-    }
-    
-    pre {
-        background-color: #027c7c;
-        padding-left: 0.5em;
-    }
-</style>"""
+    try:
+        with open(css_filename, "r", encoding="utf-8") as f:
+            css_content = f.read()
+        return f"<style>\n{css_content}\n</style>"
+    except FileNotFoundError:
+        print("Could not find styles.css file.")
+        return ""
 
 
 def _read_notebook_file(filename: str) -> Dict[str, Any]:
@@ -92,7 +82,7 @@ def _process_code_cell(cell: Dict[str, Any], cells: List[str]) -> None:
         cells.append("```\n\n")
 
 
-def _convert_markdown_from_notebook(filename: str, output_filename: str) -> str:
+def _process_converter(filename: str, output_filename: str, css_filename: str) -> str:
     """노트북을 마크다운으로 변환합니다."""
     notebook = _read_notebook_file(filename)
     cells: List[str] = []
@@ -103,8 +93,10 @@ def _convert_markdown_from_notebook(filename: str, output_filename: str) -> str:
         elif cell["cell_type"] == "markdown":
             cells.extend(cell["source"])
             cells.append("\n")
+        else:
+            print(f"Unknown cell type. Verification Needed. file : {filename}")
 
-    final_output = f"{get_default_css()}\n\n{''.join(cells)}"
+    final_output = f"{get_default_css(css_filename)}\n\n{''.join(cells)}"
 
     with open(output_filename, "w") as f:
         f.write(final_output)
@@ -113,13 +105,15 @@ def _convert_markdown_from_notebook(filename: str, output_filename: str) -> str:
 
 
 def convert_markdown_from_notebook(
-    filename: str, post_fix: str = "-(NEED-REVIEW)"
+    filename: str,
+    css_filename: str = "css/styles.css",
+    post_fix: str = "-(NEED-REVIEW)",
 ) -> str:
     """노트북 파일을 마크다운으로 변환하는 메인 함수입니다."""
 
     output_filename = filename.replace(".ipynb", f"{post_fix}.md")
 
-    return _convert_markdown_from_notebook(filename, output_filename)
+    return _process_converter(filename, output_filename, css_filename)
 
 
 if __name__ == "__main__":
@@ -129,6 +123,7 @@ def convert_markdown_from_notebook(
     parser.add_argument(
         "--filename", required=True, help="변환할 주피터 노트북 파일 경로"
     )
+    parser.add_argument("--css", default="css/styles.css", help="css file path")
     args = parser.parse_args()
 
     convert_markdown_from_notebook(args.filename)

mdconverter.py renamed to mdconverter/mdconverter.py

@@ -0,0 +1,50 @@
+from typing import Any
+from mdconverter.mdconverter import get_default_css
+from nbconvert import MarkdownExporter
+import nbformat
+
+
+class Mdconverter:
+    def __init__(self) -> None:
+        self.css_filename = ""
+        self.filename = ""
+        self.output_filename = ""
+
+    def run(self):
+        # TODO making mdconverter using mdconvert.py
+        # self.run_mdconverter()
+        self.run_ndconverter(
+            filename=self.filename,
+            output_filename=self.output_filename,
+            css_filename=self.css_filename,
+        )
+
+    def run_ndconverter(
+        self, filename: str, output_filename: str, css_filename: str
+    ) -> None:
+        notebook_content = self.load_ipynb(filename)
+        script = self.markdown_exporter(notebook_content)
+        self.ndconverter_script = self.add_prefix_css(script, css_filename)
+        self.save_script(self.ndconverter_script, output_filename)
+
+    def load_ipynb(self, filename: str) -> Any:
+        with open(filename, "r", encoding="utf-8") as f:
+            notebook_content = nbformat.read(f, as_version=4)
+        return notebook_content
+
+    def markdown_exporter(self, notebook_content: Any) -> str:
+        exporter = MarkdownExporter()
+        # MarkdownExporter를 사용하여 md로 변환
+        (script, resources) = exporter.from_notebook_node(notebook_content)
+        return script
+
+    def add_prefix_css(self, script: str, css_filename: str):
+        return f"{get_default_css(css_filename)}\n\n{''.join(script)}"
+
+    def save_script(self, script: str, output_filename: str) -> None:
+        # 변환된 스크립트 저장
+        with open(output_filename, "w", encoding="utf-8") as f:
+            f.write(script)
+
+
+# TODO Creating Multi loader