Output separate YAML files for cisagov conversion

Instead of printing one large YAML file to stdout we instead group the
read software entires by the uppercased first character of the vendor
name and output a file for each of these groups. All non-alphabet
vendor entries are grouped in the "0-9" group manually. When testing
the only entries that were non-alphabet were numeric. This will generate
files that are small enough to be edited using the GitHub web UI.

src/mdyml/convert_cisagov.py

@@ -22,6 +22,7 @@
 # Standard Python Libraries
 from datetime import datetime, timezone
 import html
+from itertools import groupby
 import logging
 import sys
 from typing import Any
@@ -36,6 +37,7 @@
 from . import DEFAULT_CVE_ID, MD_LINK_RE, ORDERED_CVE_IDS, __version__
 
 RAW_URL = "https://raw.githubusercontent.com/cisagov/log4j-affected-db/develop/SOFTWARE-LIST.md"
+SOFTWARE_LIST_FILE_FORMAT = "cisagov_{}.yml"
 
 EXPECTED_COLUMN_NAMES = [
     "vendor",
@@ -160,15 +162,31 @@ def convert() -> None:
 
         out_dict_list.append(out_dict)
 
-    doc = {"version": "1.0", "software": out_dict_list}
-
-    yaml = ruamel.yaml.YAML()
-    yaml.indent(mapping=2, offset=2, sequence=4)
-    yaml.explicit_start = True
-    yaml.explicit_end = True
-    yaml.sort_base_mapping_type_on_output = False
-    yaml.allow_unicode = True
-    yaml.dump(doc, sys.stdout)
+    out_dict_groups = {
+        k: list(g)
+        for k, g in groupby(out_dict_list, key=lambda s: s["vendor"][0].upper())
+    }
+
+    non_letter_groups = list()
+    for key in list(out_dict_groups.keys()):
+        if not key.isalpha():
+            non_letter_groups.extend(out_dict_groups[key])
+            del out_dict_groups[key]
+    out_dict_groups["0-9"] = non_letter_groups
+
+    for key, data in out_dict_groups.items():
+        filename = SOFTWARE_LIST_FILE_FORMAT.format(key)
+        logging.debug("Writing data for '%s' to '%s'", key, filename)
+        with open(filename, "w") as out_file:
+            doc = {"version": "1.0", "software": data}
+
+            yaml = ruamel.yaml.YAML()
+            yaml.indent(mapping=2, offset=2, sequence=4)
+            yaml.explicit_start = True
+            yaml.explicit_end = True
+            yaml.sort_base_mapping_type_on_output = False
+            yaml.allow_unicode = True
+            yaml.dump(doc, out_file)
 
 
 def main() -> None: