Debug

Debugging

assert

• 主要用在開發與測試階段
• 可用在程式的條件測試
• assert <condition>, <message> : 如果 condition 為 True，沒有作用；如果為 False，會產生錯誤，並顯示訊息
• 跳過所有 assert 語句，可以使用 python -O sample.py。

檢查變數的值

x = 5
assert x == 5, "x should be 5"

assert type(username) == str, "username must be a string"

def calculate_square_root(x):
    assert x >= 0, "The input must be non-negative."
    return x ** 0.5

print(calculate_square_root(4))  # 輸出: 2.0
print(calculate_square_root(-1)) # 引發 AssertionError 並顯示 "The input must be non-negative."

檢查函示回傳值有無偶數

def get_even_number(numbers):
    for num in numbers:
        if num % 2 == 0:
            return num
    assert False, "No even number found in the list."

numbers = [1, 3, 5, 7, 10]
print(get_even_number(numbers))  # 輸出: 10
numbers = [1, 3, 5, 7]
print(get_even_number(numbers))  # 引發 AssertionError 並顯示 "No even number found in the list."

prinf debugging

print("Processing {}".format(basename))

strace

• Linux strace Command Tutorial for Beginners (8 Examples)

# Installation on RHEL if it's not installed
yum install strace

# Tracing system calls made by a program
strace ./my-program.py
strace -o my-program.strace ./my-program

Crash

pdb

功能：

• 設定程式中斷點
• 逐行檢查程式碼
• 檢查變數
• 以互動方式評估表達式

pdb3 myprog.py

pdb-subcommands

• continue : 繼續執行直到異常的程式碼
• print() : 輸出變數的內容

(Pdb) continue
...
(Pdb) print(row)

Step 1: Set a breakpoint

import pdb


def add_numbers(a, b):
    pdb.set_trace()  # This will set a breakpoint in the code
    result = a + b
    return result


print(add_numbers(3, 4))

Setp 2: Enter the interactive debugger

• a (args): Show the arguments of the current function.

• b: Manually set a persistent breakpoint while in debugger.

• n (next): Execute the next line within the current function.

• s (step): Execute the current line and stop at the first possible occasion (e.g., in a function that is called).

• c (continue): Resume normal execution until the next breakpoint.

• p (print): Evaluate and print the expression, e.g., p variable_name will print the value of variable_name.

• Pp (pretty-print): Pretty-print the value of the expression.

• q (quit): Exit the debugger and terminate the program.

• r (return): Continue execution until the current function returns.

• tbreak: Manually set a temporary breakpoint that goes away once hit the first time.

• !: Prefix to execute an arbitrary Python command in the current environment, e.g., !variable_name = "new_value" will set variable_name to "new_value".

Step 3: Inspect variables

To inspect the variables, simply type the single character, p, then the variable name to see its current value. For instance, if you have a variable in your code named sentiment_score, just type p sentiment_score at the pdb prompt to inspect its value.

Step 4: Modify variables

A big advantage of pdb is that you can change the value of a variable directly in the debugger. For example, to change sentiment_score to 0.9, you'd type !sentiment_score = 0.9.

To confirm these changes, use a or directly probe the value with p <value name>.

 Step 5: Exit the debugger

When you’re done, simply enter q (quit) to exit the debugger and terminate the program.

Post-mortem debugging

python -m pdb your_script.py

Memory Leaks

當不再需要的記憶體未釋放時，就會發生記憶體洩漏。即使重新啟動，仍需要大量記憶體的應用程式，很可能指向記憶體洩漏

memory_profiler

第一欄顯示每一行執行時所需的記憶體數量。第二欄顯示每一行所增加的記憶體

python3 -m memory_profiler myprog.py

In Code

• 在 main() 上方加上 @profile 標籤
• @ 標籤稱為 Decorator: 在 Python 中使用它來為函數增加額外的行為，而不需要修改程式碼
• memory-profiler

from memory_profiler import profile

...
...

@profile
def main():
  ...
  ...