Compress cached JSON files

mypy/metastore.py

@@ -14,6 +14,7 @@
 import os
 import time
 from abc import abstractmethod
+from gzip import compress, decompress
 from typing import TYPE_CHECKING, Any, Iterable
 
 if TYPE_CHECKING:
@@ -93,8 +94,8 @@ def read(self, name: str) -> str:
         if not self.cache_dir_prefix:
             raise FileNotFoundError()
 
-        with open(os.path.join(self.cache_dir_prefix, name)) as f:
-            return f.read()
+        with open(os.path.join(self.cache_dir_prefix, name), "rb") as f:
+            return decompress(f.read()).decode()
 
     def write(self, name: str, data: str, mtime: float | None = None) -> bool:
         assert os.path.normpath(name) != os.path.abspath(name), "Don't use absolute paths!"
@@ -106,8 +107,8 @@ def write(self, name: str, data: str, mtime: float | None = None) -> bool:
         tmp_filename = path + "." + random_string()
         try:
             os.makedirs(os.path.dirname(path), exist_ok=True)
-            with open(tmp_filename, "w") as f:
-                f.write(data)
+            with open(tmp_filename, "wb") as f:
+                f.write(compress(data.encode(), 1, mtime=0))
             os.replace(tmp_filename, path)
             if mtime is not None:
                 os.utime(path, times=(mtime, mtime))

mypy/test/test_ref_info.py

@@ -5,6 +5,7 @@
 import json
 import os
 import sys
+from gzip import decompress
 
 from mypy import build
 from mypy.modulefinder import BuildSource
@@ -33,8 +34,8 @@ def run_case(self, testcase: DataDrivenTestCase) -> None:
         major, minor = sys.version_info[:2]
         ref_path = os.path.join(options.cache_dir, f"{major}.{minor}", "__main__.refs.json")
 
-        with open(ref_path) as refs_file:
-            data = json.load(refs_file)
+        with open(ref_path, "rb") as refs_file:
+            data = json.loads(decompress(refs_file.read()))
 
         a = []
         for item in data:

test-data/unit/daemon.test

@@ -309,7 +309,7 @@ Found 1 error in 1 file (checked 2 source files)
 $ {python} -c "import sys; sys.stdout.write(open('log').read())"
 -- make sure the meta file didn't get updated. we use this as an imperfect proxy for
 -- whether the source file got rehashed, which we don't want it to have been.
-$ {python} -c "x = open('.mypy_cache/3.11/bar.meta.json').read(); y = open('asdf.json').read(); assert x == y"
+$ {python} -c "x = open('.mypy_cache/3.11/bar.meta.json', 'rb').read(); y = open('asdf.json', 'rb').read(); assert x == y"
 
 [case testDaemonSuggest]
 $ dmypy start --log-file log.txt -- --follow-imports=error --no-error-summary