#Copyright (c) 2025 Huawei Technologies Co., Ltd.
#This program is free software, you can redistribute it and/or modify it under the terms and conditions of
#CANN Open Software License Agreement Version 2.0 (the "License").
#Please refer to the License for details. You may not use this file except in compliance with the License.
#THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED,
#INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.
#See LICENSE in the root of the software repository for the full text of the License.

import sys, os
sys.stdout.reconfigure(encoding='utf-8')

cpp = r'D:\AI\Vector_error_testing_angent\output\RmsNormGrad\aclnnRmsNormGrad.cpp'

with open(cpp, 'rb') as f:
    data = f.read()

# 1. Normalize line endings: CRLF -> LF (for the file itself)
data = data.replace(b'\r\n', b'\n')

# 2. Check for escaped \n in printf/LOG_PRINT - should be the 2-byte sequence 0x5C 0x6E
#    After text-mode write, Python's \n (which is LF=0x0A in the join) becomes CRLF=0x0D0x0A in file
#    After step 1, file has LF line endings. Now check the printf arguments.

#    The C++ escape \n is: backslash + n = bytes 0x5C 0x6E in the .cpp file
#    After binary write (which we now use), the Python source \\\\n becomes 0x5C 0x5C 0x6E
#    which is WRONG for C++ (literal \\n, not newline)
#    And \\n in Python source becomes 0x5C 0x6E which is CORRECT for C++

# Count patterns
has_cxx_newline = 0  # 0x5C 0x6E = correct C++ \n
has_double = 0       # 0x5C 0x5C 0x6E = wrong, literal \\n
has_cr = 0           # 0x0D 0x0A = CRLF line ending (should be gone after step 1)

idx = 0
while True:
    idx = data.find(b'\x5c\x6e', idx)
    if idx < 0:
        break
    # Check if preceded by another backslash
    if idx > 0 and data[idx-1:idx] == b'\x5c':
        has_double += 1
    else:
        has_cxx_newline += 1
    idx += 2

has_cr = data.count(b'\r\n')

print(f'C++ newline (0x5C 0x6E): {has_cxx_newline}')
print(f'Double backslash (0x5C 0x5C 0x6E): {has_double}')
print(f'CRLF remaining: {has_cr}')

# 3. If we have 0x5C 0x5C 0x6E, replace with 0x5C 0x6E
if has_double > 0:
    data = data.replace(b'\x5c\x5c\x6e', b'\x5c\x6e')
    print(f'Fixed {has_double} double-backslash -> single')

# 4. Write back as binary (LF line endings)
with open(cpp, 'wb') as f:
    f.write(data)

# Verify
with open(cpp, 'rb') as f:
    newdata = f.read()
new_double = 0
idx = 0
while True:
    idx = newdata.find(b'\x5c\x5c\x6e', idx)
    if idx < 0:
        break
    new_double += 1
    idx += 3

new_crlf = newdata.count(b'\r\n')
print(f'After fix: double-backslash={new_double}, CRLF={new_crlf}')
print('Done')