vault backup: 2025-12-26 02:09:22

stroma/quiz/utils/question_parser.py

@@ -0,0 +1,89 @@
+import dataclasses
+import pathlib
+
+import frontmatter
+import mistune
+
+from quiz.utils.obsidian_embed_plugin import obsidian_embed
+
+markdown = mistune.create_markdown(renderer="ast", plugins=[obsidian_embed])
+
+
+class Node:
+    def __init__(self, token):
+        self.type = token["type"]
+        self.raw = token.get("raw", "")
+        self.attrs = token.get("attrs", {})
+        self.children = [Node(token=child) for child in token.get("children", [])]
+
+    def __repr__(self) -> str:
+        attrs = []
+        if self.raw:
+            attrs.append(f"raw={self.raw!r}")
+        if self.attrs:
+            attrs.append(f"attrs={self.attrs!r}")
+        if self.children:
+            attrs.append(f"children={self.children!r}")
+        # block_text -> BlockText
+        pretty = self.type.replace("_", " ").title().replace(" ", "")
+        return f"{pretty}(" + ", ".join(attrs) + ")"
+
+    @property
+    def text(self) -> str:
+        if self.type == "text":
+            return self.raw
+        texts = []
+        for child in self.children:
+            texts.append(child.text)
+        return "".join(texts)
+
+
+@dataclasses.dataclass
+class ParsedQuestion:
+    metadata: dict = dataclasses.field(default_factory=dict)
+    raw_content: str = ""
+    nodes: list[Node] = dataclasses.field(default_factory=list)
+
+
+def parse_question(path: pathlib.Path):
+    raw = path.read_text(encoding="utf-8")
+    return parse_question_from_content(raw)
+
+
+def parse_question_from_content(content_str: str):
+    """Parse question from a content string instead of a file."""
+    metadata, content = frontmatter.parse(content_str)
+    tokens = markdown(content)
+    question = ParsedQuestion(
+        metadata=metadata,
+        raw_content=content,
+        nodes=[Node(token=token) for token in tokens],
+    )
+    return question
+
+
+def main():
+    root = pathlib.Path(__file__).parent.parent.parent.parent
+    print(root)
+    exams = root / "content" / "Anatomi & Histologi 2" / "Gamla tentor"
+    for file in sorted(exams.glob("*/*.md")):
+        if len(file.stem) > 2:
+            continue
+        question = parse_question(file)
+        print(question.metadata, repr(question.raw_content))
+        continue
+        for node in question.nodes:
+            match node.type:
+                case "heading":
+                    print("Heading:", repr(node.text))
+                case "paragraph":
+                    print("Paragraph:", repr(node.text))
+                case "list":
+                    print("List:")
+                    for child in node.children:
+                        print(" - List item:", repr(child.text))
+                case "block_code" if node.attrs["info"] == "spoiler-block:":
+                    print("Spoiler:", repr(node.raw.rstrip()))
+
+if __name__ == "__main__":
+    main()