fix: parse and clean archive badges and markdown links to URL

pyproject.toml

@@ -25,6 +25,7 @@ classifiers = [
 ]
 dependencies = [
     "pydantic>=2.0",
+    "python-doi",
     "python-dotenv",
     "requests",
     "ruamel-yaml>=0.17.21",

src/pyosmeta/models/base.py

@@ -19,7 +19,7 @@
 )
 
 from pyosmeta.models.github import Labels
-from pyosmeta.utils_clean import clean_date, clean_markdown
+from pyosmeta.utils_clean import clean_archive, clean_date, clean_markdown
 
 
 class Partnerships(str, Enum):
@@ -403,3 +403,21 @@
             label.name if isinstance(label, Labels) else label
             for label in labels
         ]
+
+    @field_validator(
+        "archive",
+        mode="before",
+    )
+    @classmethod
+    def clean_archive(cls, archive: str) -> str:
+        """Clean the archive value to ensure it's a valid archive URL."""
+        return clean_archive(archive)
+
+    @field_validator(
+        "joss",
+        mode="before",
+    )
+    @classmethod
+    def clean_joss(cls, joss: str) -> str:
+        """Clean the joss value to ensure it's a valid URL."""
+        return clean_archive(joss)

src/pyosmeta/utils_clean.py

@@ -7,6 +7,8 @@
 from datetime import datetime
 from typing import Any
 
+import doi
+
 
 def get_clean_user(username: str) -> str:
     """Cleans a GitHub username provided in a review issue by removing any
@@ -125,3 +127,30 @@
             review_dict["date_accepted"] = value
             break
     return review_dict
+
+
+def clean_archive(archive):
+    """Clean an archive link to ensure it is a valid URL."""
+
+    def is_doi(archive):
+        try:
+            return doi.validate_doi(archive)
+        except ValueError:
+            return False
+
+    if archive.startswith("[") and archive.endswith(")"):
+        # Extract the outermost link
+        link = archive[archive.rfind("](") + 2 : -1]
+        if not link.startswith("http"):
+            return clean_archive(link)
+        return link
+    elif archive.startswith("http"):
+        return archive
+    elif link := is_doi(archive):
+        return link
+    elif archive.lower() == "n/a":
+        return None
+    elif archive.lower() == "tbd":
+        return None
+    else:
+        raise ValueError(f"Invalid archive URL: {archive}")