[libcxx-commits] [libcxx] [libc++] Install modules. (PR #75741)

[Ben Boeckel via libcxx-commits]

================
@@ -0,0 +1,26 @@
+{
+  "version": 1,
+  "revision": 1,
+  "modules": [
+    {
+      "logical-name": "std",
+      "source-path": "@LIBCXX_MODULE_RELATIVE_PATH@/std.cppm",
+      "is-standard-library": true,
+      "local-arguments": {
+        "system-include-directories": [
+          "@LIBCXX_MODULE_RELATIVE_PATH@"
+        ]
+      }
+    },
+    {
+      "logical-name": "std.compat",
+      "source-path": "@LIBCXX_MODULE_RELATIVE_PATH@/std.compat.cppm",
+      "is-std-library": true,
----------------
mathstuf wrote:

FWIW, [P1689](https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2022/p1689r5.html) uses kebab-case as well. Note that 18 is in rc3; I don't know how much "Python APIs are easier" is to make a release blocker versus "the file is not consistent with itself". Something for release managers to decide.

Note that there's nothing stopping a typed model from acting on it as-is because a function can transform these into identifier names (IME, the explicitness from things like `#[serde(rename = "…")]` are better than forcing the interchange format to bend to a specific language's ergonomics).

https://github.com/llvm/llvm-project/pull/75741