Create script for merging ambr files

cspell.json

@@ -23,6 +23,7 @@
 	],
 	"words": [
 		"afterstep",
+		"ambr",
 		"Andale",
 		"APNG",
 		"appcode",

tests/merge_ambr.py

@@ -0,0 +1,39 @@
+"""Merge Amber snapshot files to include test results from both, and warn if shared results don't match.
+
+This allows applying the TDD principle to snapshot tests, in the specific case that
+tests can pass in isolation, but fail when run together.
+"""
+
+import re
+from pathlib import Path
+
+DIR = Path(__file__).parent / "__snapshots__"
+
+files = [
+    DIR / "test_snapshots.ambr",
+    DIR / "test_snapshots_ascii.ambr",
+]
+output_file = DIR / "test_snapshots_merged.ambr"
+
+re_snapshot = re.compile(r"^# name: (test_.*)\n  '''\n((?:.|\n  )*)\n  '''$", re.MULTILINE)
+
+snapshots: dict[str, str] = {}
+names: set[str] = set()
+for file in files:
+    with open(file, "r", encoding="utf-8") as f:
+        for match in re_snapshot.finditer(f.read()):
+            name = match.group(1)
+            snapshot = match.group(2)
+            if name in snapshots:
+                if snapshots[name] != snapshot:
+                    print(f"Snapshots don't match for {name}; preferring result from {files[0]}")
+            else:
+                snapshots[name] = snapshot
+                names.add(name)
+
+with open(output_file, "w", encoding="utf-8") as f:
+    f.write(f"# serializer version: 1\n")
+    for name, snapshot in snapshots.items():
+        f.write(f"# name: {name}\n  '''\n{snapshot}\n  '''\n# ---\n")
+
+print(f"Merged {len(names)} snapshots from {len(files)} files into {output_file}")