Read xml documents by using xpath to get text

src/data.py

@@ -2,10 +2,14 @@ import ast
 from pathlib import Path
 
 import pandas as pd
+from lxml import etree
 
 
 def read_sem_eval_file(file: str | Path) -> pd.DataFrame:
-    df = pd.read_xml(file)[["text"]]
+    root = etree.parse(file)
+    documents = root.xpath("//text/text()")
+    assert isinstance(documents, list), f"cannot parse text from {file}"
+    df = pd.DataFrame({"text": documents})
     return df