Refactor into functions

data_fetch.py

@@ -2,43 +2,52 @@ import requests
 import json
 import pandas as pd
 
-# Repository details
-owner = "EbookFoundation"
-repo = "free-programming-books"
-
-# GitHub API endpoint
-url = f"https://api.github.com/repos/{owner}/{repo}/issues"
-
-headers = {
-    "Accept": "application/vnd.github+json",
-    # "Authorization": "Bearer YOUR_GITHUB_TOKEN"  # optional
-}
-
-# Fetch issues
-response = requests.get(url, headers=headers)
-
-if response.status_code == 200:
-    issues = response.json()
-
-    # Extract key fields from each issue
+def fetch_issues(owner, repo):
+    '''Return a list of issues from OWNER/REPO.
+    '''
+    # GitHub API endpoint
+    url = f"https://api.github.com/repos/{owner}/{repo}/issues"
+    headers = {
+        "Accept": "application/vnd.github+json",
+        # "Authorization": "Bearer YOUR_GITHUB_TOKEN"  # optional
+    }
+    # Fetch issues
+    response = requests.get(url, headers=headers)
+    if response.status_code == 200:
+        return response.json()
+    else:
+        raise f"Failed to fetch issues from {owner}/{repo}: status {response.status_code}"
+
+def serialize_issue(issue):
+    '''Serializes an issue object from the Github API to a dictionary
+    of key fields.
+    '''
+    return {
+        "id": issue.get("id"),
+        "number": issue.get("number"),
+        "title": issue.get("title"),
+        "state": issue.get("state"),
+        "created_at": issue.get("created_at"),
+        "updated_at": issue.get("updated_at"),
+        "user_login": issue["user"]["login"] if "user" in issue and issue["user"] else None,
+        "comments": issue.get("comments"),
+        "html_url": issue.get("html_url")
+    }
+
+def main():
+    # Repository details
+    owner = "EbookFoundation"
+    repo = "free-programming-books"
+
+    issues = fetch_issues(owner, repo)
     data = []
     for issue in issues:
-        data.append({
-            "id": issue.get("id"),
-            "number": issue.get("number"),
-            "title": issue.get("title"),
-            "state": issue.get("state"),
-            "created_at": issue.get("created_at"),
-            "updated_at": issue.get("updated_at"),
-            "user_login": issue["user"]["login"] if "user" in issue and issue["user"] else None,
-            "comments": issue.get("comments"),
-            "html_url": issue.get("html_url")
-        })
+        data.append(serialize_issue(issue))
 
     # Convert to DataFrame
     df = pd.DataFrame(data)
     print(df.head())
     df.to_json("github_data.json")
 
-else:
-    print(f"Failed to fetch issues: {response.status_code}")
+if __name__=="__main__":
+    main()