refactor a function for detecting HTML output

Author: afeld

scripts/diffable.py

@@ -23,6 +23,10 @@ def fix_sample(line):
     return re.sub(r"\.sample\((\d+)\)", r".sample(\1, random_state=1)", line)
 
 
+def has_html_output(output):
+    return "data" in output and "text/html" in output["data"]
+
+
 input_str = sys.stdin.read()
 notebook = json.loads(input_str)
 
@@ -51,8 +55,8 @@ def fix_sample(line):
             # ignore all the execution count numbers
             output["execution_count"] = 1
 
-        # clear HTML output, since it often has generated IDs (from displacy, plotly, etc.) that change with each execution
-        if "data" in output and "text/html" in output["data"]:
+        if has_html_output(output):
+            # clear HTML output, since it often has generated IDs (from displacy, plotly, etc.) that change with each execution
             cell["outputs"] = []
             # go to next cell
             break