From efd4b95fab1be1a6c58d4a8379dde24bc516c58d Mon Sep 17 00:00:00 2001
From: Aaryan Shukla <aaryanshukla@google.com>
Date: Mon, 1 Jul 2024 22:55:42 +0000
Subject: [PATCH] [libc] created integration tests for hdrgen

- added sorting to script so C headers are outputted in alpha order
---
 libc/newhdrgen/tests/expected_output/string.h | 102 ++++++
 libc/newhdrgen/tests/input/string.h.def       |  18 ++
 libc/newhdrgen/tests/input/test_string.yaml   | 303 ++++++++++++++++++
 libc/newhdrgen/tests/test_integration.py      |  42 +++
 libc/newhdrgen/yaml_to_classes.py             |   6 +-
 5 files changed, 469 insertions(+), 2 deletions(-)
 create mode 100644 libc/newhdrgen/tests/expected_output/string.h
 create mode 100644 libc/newhdrgen/tests/input/string.h.def
 create mode 100644 libc/newhdrgen/tests/input/test_string.yaml
 create mode 100644 libc/newhdrgen/tests/test_integration.py

diff --git a/libc/newhdrgen/tests/expected_output/string.h b/libc/newhdrgen/tests/expected_output/string.h
new file mode 100644
index 00000000000000..4dcb693c353e46
--- /dev/null
+++ b/libc/newhdrgen/tests/expected_output/string.h
@@ -0,0 +1,102 @@
+//===-- C standard library header string.h --------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIBC_STRING_H
+#define LLVM_LIBC_STRING_H
+
+#include "__llvm-libc-common.h"
+
+#include "llvm-libc-macros/null-macro.h"
+
+#include <llvm-libc-types/size_t.h>
+
+__BEGIN_C_DECLS
+
+void * memccpy(void *__restrict, const void *__restrict, int, size_t);
+
+void * memchr(const void *, int, size_t);
+
+int memcmp(const void *, const void *, size_t);
+
+void * memcpy(void *__restrict, const void *__restrict, size_t);
+
+void * memmem(const void *, size_t, const void *, size_t);
+
+void * memmove(void *, const void *, size_t);
+
+void * mempcpy(void *__restrict, const void *__restrict, size_t);
+
+void * memrchr(const void *, int, size_t);
+
+void * memset(void *, int, size_t);
+
+void * memset_explicit(void *, int, size_t);
+
+char * stpcpy(char *__restrict, const char *__restrict);
+
+char * stpncpy(char *__restrict, const char *__restrict, size_t);
+
+char * strcasestr(const char *, const char *);
+
+char * strcat(char *__restrict, const char *__restrict);
+
+char * strchr(const char *, int);
+
+char * strchrnul(const char *, int);
+
+int strcmp(const char *, const char *);
+
+int strcoll(const char *, const char *);
+
+char * strcpy(char *__restrict, const char *__restrict);
+
+size_t strcspn(const char *, const char *);
+
+char * strdup(const char *);
+
+char * strerror(int);
+
+char * strerror_r(int, char *, size_t);
+
+size_t strlcat(const char *__restrict, const char *__restrict, size_t);
+
+size_t strlcpy(const char *__restrict, const char *__restrict, size_t);
+
+size_t strlen(const char *);
+
+char * strncat(char *, const char *, size_t);
+
+int strncmp(const char *, const char *, size_t);
+
+char * strncpy(char *__restrict, const char *__restrict, size_t);
+
+char * strndup(const char *, size_t);
+
+size_t strnlen(const char *, size_t);
+
+char * strpbrk(const char *, const char *);
+
+char * strrchr(const char *, int);
+
+char * strsep(char * *__restrict, const char *__restrict);
+
+char * strsignal(int);
+
+size_t strspn(const char *, const char *);
+
+char * strstr(const char *, const char *);
+
+char * strtok(char *__restrict, const char *__restrict);
+
+char * strtok_r(char *__restrict, const char *__restrict, char * *__restrict);
+
+size_t strxfrm(char *__restrict, const char *__restrict, size_t);
+
+__END_C_DECLS
+
+#endif // LLVM_LIBC_STRING_H
diff --git a/libc/newhdrgen/tests/input/string.h.def b/libc/newhdrgen/tests/input/string.h.def
new file mode 100644
index 00000000000000..1bd2687db2beac
--- /dev/null
+++ b/libc/newhdrgen/tests/input/string.h.def
@@ -0,0 +1,18 @@
+//===-- C standard library header string.h --------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIBC_STRING_H
+#define LLVM_LIBC_STRING_H
+
+#include "__llvm-libc-common.h"
+
+#include "llvm-libc-macros/null-macro.h"
+
+%%public_api()
+
+#endif // LLVM_LIBC_STRING_H
diff --git a/libc/newhdrgen/tests/input/test_string.yaml b/libc/newhdrgen/tests/input/test_string.yaml
new file mode 100644
index 00000000000000..2754f856bc8077
--- /dev/null
+++ b/libc/newhdrgen/tests/input/test_string.yaml
@@ -0,0 +1,303 @@
+header: string.h
+macros: []
+types:
+  - type_name: size_t
+enums: []
+objects: []
+functions:
+  - name: memcpy
+    standards: 
+      - stdc
+    return_type: void *
+    arguments:
+      - type: void *__restrict
+      - type: const void *__restrict
+      - type: size_t
+  - name: memmove
+    standards: 
+      - stdc
+    return_type: void *
+    arguments:
+      - type: void *
+      - type: const void *
+      - type: size_t
+  - name: memcmp
+    standards: 
+      - stdc
+    return_type: int
+    arguments:
+      - type: const void *
+      - type: const void *
+      - type: size_t
+  - name: memchr
+    standards: 
+      - stdc
+    return_type: void *
+    arguments:
+      - type: const void *
+      - type: int
+      - type: size_t
+  - name: memset
+    standards: 
+      - stdc
+    return_type: void *
+    arguments:
+      - type: void *
+      - type: int
+      - type: size_t
+  - name: memset_explicit
+    standards: 
+      - stdc
+    return_type: void *
+    arguments:
+      - type: void *
+      - type: int
+      - type: size_t
+  - name: strcpy
+    standards: 
+      - stdc
+    return_type: char
+    arguments:
+      - type: char *__restrict
+      - type: const char *__restrict
+  - name: strncpy
+    standards: 
+      - stdc
+    return_type: char *
+    arguments:
+      - type: char *__restrict
+      - type: const char *__restrict
+      - type: size_t
+  - name: strcat
+    standards: 
+      - stdc
+    return_type: char *
+    arguments:
+      - type: char *__restrict
+      - type: const char *__restrict
+  - name: strncat
+    standards: 
+      - stdc
+    return_type: char *
+    arguments:
+      - type: char *
+      - type: const char *
+      - type: size_t
+  - name: strcmp
+    standards: 
+      - stdc
+    return_type: int
+    arguments:
+      - type: const char *
+      - type: const char *
+  - name: strcoll
+    standards: 
+      - stdc
+    return_type: int
+    arguments:
+      - type: const char *
+      - type: const char *
+  - name: strncmp
+    standards: 
+      - stdc
+    return_type: int
+    arguments:
+      - type: const char *
+      - type: const char *
+      - type: size_t
+  - name: strxfrm
+    standards: 
+      - stdc
+    return_type: size_t
+    arguments:
+      - type: char *__restrict
+      - type: const char *__restrict
+      - type: size_t
+  - name: strchr
+    standards: 
+      - stdc
+    return_type: char *
+    arguments:
+      - type: const char *
+      - type: int
+  - name: strcspn
+    standards: 
+      - stdc
+    return_type: size_t
+    arguments:
+      - type: const char *
+      - type: const char *
+  - name: strdup
+    return_type: char *
+    arguments:
+      - type: const char *
+  - name: strndup
+    standards: 
+      - stdc
+    return_type: char *
+    arguments:
+      - type: const char *
+      - type: size_t
+  - name: strpbrk
+    standards: 
+      - stdc
+    return_type: char *
+    arguments:
+      - type: const char *
+      - type: const char *
+  - name: strrchr
+    standards: 
+      - stdc
+    return_type: char *
+    arguments:
+      - type: const char *
+      - type: int
+  - name: strspn
+    standards: 
+      - stdc
+    return_type: size_t
+    arguments:
+      - type: const char *
+      - type: const char *
+  - name: strstr
+    standards: 
+      - stdc
+    return_type: char *
+    arguments:
+      - type: const char *
+      - type: const char *
+  - name: strtok
+    standards: 
+      - stdc
+    return_type: char *
+    arguments:
+      - type: char *__restrict
+      - type: const char *__restrict
+  - name: strerror
+    standards: 
+      - stdc
+    return_type: char *
+    arguments:
+      - type: int
+  - name: strlen
+    standards: 
+      - stdc
+    return_type: size_t
+    arguments:
+      - type: const char *
+  - name: memccpy
+    standards: 
+      - POSIX
+    return_type: void *
+    arguments:
+      - type: void *__restrict
+      - type: const void *__restrict
+      - type: int
+      - type: size_t
+  - name: mempcpy
+    standards: 
+      - POSIX
+    return_type: void *
+    arguments:
+      - type: void *__restrict
+      - type: const void *__restrict
+      - type: size_t
+  - name: stpcpy
+    standards: 
+      - POSIX
+    return_type: char *
+    arguments:
+      - type: char *__restrict
+      - type: const char *__restrict
+  - name: stpncpy
+    standards: 
+      - POSIX
+    return_type: char *
+    arguments:
+      - type: char *__restrict
+      - type: const char *__restrict
+      - type: size_t
+  - name: strnlen
+    standards: 
+      - POSIX
+    return_type: size_t
+    arguments:
+      - type: const char *
+      - type: size_t
+  - name: strtok_r
+    standards: 
+      - POSIX
+    return_type: char *
+    arguments:
+      - type: char *__restrict
+      - type: const char *__restrict
+      - type: char ** __restrict
+  - name: strsignal
+    standards: 
+      - POSIX
+    return_type: char *
+    arguments:
+      - type: int
+  - name: memmem
+    standards: 
+      - GNUExtensions
+    return_type: void *
+    arguments:
+      - type: const void *
+      - type: size_t
+      - type: const void *
+      - type: size_t
+  - name: memrchr
+    standards: 
+      - GNUExtensions
+    return_type: void *
+    arguments:
+      - type: const void *
+      - type: int
+      - type: size_t
+  - name: strerror_r
+    standards: 
+      - GNUExtensions
+    return_type: char *
+    arguments:
+      - type: int
+      - type: char *
+      - type: size_t
+  - name: strcasestr
+    standards: 
+      - GNUExtensions
+    return_type: char *
+    arguments:
+      - type: const char *
+      - type: const char *
+  - name: strchrnul
+    standards: 
+      - GNUExtensions
+    return_type: char *
+    arguments:
+      - type: const char *
+      - type: int
+  - name: strlcat
+    standards: 
+      - BSDExtensions
+    return_type: size_t
+    arguments:
+      - type: char *__restrict
+      - type: char *__restrict
+      - type: size_t
+  - name: strlcpy
+    standards: 
+      - BSDExtensions
+    return_type: size_t
+    arguments:
+      - type: char *__restrict
+      - type: char *__restrict
+      - type: size_t
+  - name: strsep
+    standards: 
+      - BSDExtensions
+    return_type: char *
+    arguments:
+      - type: char **__restrict
+      - type: char *__restrict
+
diff --git a/libc/newhdrgen/tests/test_integration.py b/libc/newhdrgen/tests/test_integration.py
new file mode 100644
index 00000000000000..ca485dc0377c1d
--- /dev/null
+++ b/libc/newhdrgen/tests/test_integration.py
@@ -0,0 +1,42 @@
+import subprocess
+import unittest
+from pathlib import Path
+
+class TestHeaderGenIntegration(unittest.TestCase):
+    def setUp(self):
+        self.output_dir = Path('tests/output')
+        self.maxDiff = None  
+
+    def run_script(self, yaml_file, h_def_file, output_dir):
+        result = subprocess.run([
+            'python3', 'yaml_to_classes.py', yaml_file, h_def_file, '--output_dir', str(output_dir)
+        ], capture_output=True, text=True)
+
+        print("STDOUT:", result.stdout)
+        print("STDERR:", result.stderr)
+        result.check_returncode() 
+
+    def compare_files(self, generated_file, expected_file):
+        with generated_file.open('r') as gen_file:
+            gen_content = gen_file.read()
+        with expected_file.open('r') as exp_file:
+            exp_content = exp_file.read()
+        
+        self.assertEqual(gen_content, exp_content)
+
+    def test_generate_header(self):
+        # this is for example, will find a way to test everything at once
+        yaml_file = Path('tests/input/test_string.yaml')
+        h_def_file = Path('tests/input/string.h.def')
+        expected_output_file = Path('tests/expected_output/string.h')
+        output_file = self.output_dir / 'string.h'
+
+        if not self.output_dir.exists():
+            self.output_dir.mkdir(parents=True)
+
+        self.run_script(yaml_file, h_def_file, self.output_dir)
+
+        self.compare_files(output_file, expected_output_file)
+
+if __name__ == '__main__':
+    unittest.main()
diff --git a/libc/newhdrgen/yaml_to_classes.py b/libc/newhdrgen/yaml_to_classes.py
index 9b52c9cf9bb7c7..65a990c8a87678 100644
--- a/libc/newhdrgen/yaml_to_classes.py
+++ b/libc/newhdrgen/yaml_to_classes.py
@@ -46,11 +46,13 @@ def yaml_to_classes(yaml_data):
             Enumeration(enum_data["name"], enum_data.get("value", None))
         )
 
-    for function_data in yaml_data.get("functions", []):
+    functions = yaml_data.get("functions", [])
+    sorted_functions = sorted(functions, key=lambda x: x['name'])
+    for function_data in sorted_functions:
         arguments = [arg["type"] for arg in function_data["arguments"]]
         guard = function_data.get("guard", None)
         attributes = function_data.get("attributes", None)
-        standards = (function_data.get("standards", None),)
+        standards = function_data.get("standards", None)
         header.add_function(
             Function(
                 function_data["return_type"],