References : Learning Python

Why use Python?

• Software quality: Readability, coherence, software quality.
• Developer productivity: Less code, fast run
• Portability: runs in all major platforms
• Many useful standard library + third party library
• Integration with other languages

Downside?

• Execution speed might not be as fast as C / C++

Technical Strength

• OOP + Functional

• Python interpreter : Python is a language, but also a software package called interpreter.

• Interpreter executes other programs.

• When python package is installed, minimal package that is going to be generated is an interpreter and a support library.

• .py for “imported” files, but for general purpose too.

• Step by Step

  1. Commanded to run
  2. Byte code compilation: Python decomposes source statements into group of byte code instructions.
     • Before version 3.2, the byte code will be stored with .pyc extension. (if python has permission to write)
     • After 3.2, it saves files in a subdirectory named __pycache__
     • Source Check: those byte code files will be run without recompiling
     • Version Check: check if Python version has changed.
     • If not writable, it will write in memory & discard it
     • Byte code is saved in files only for files that are imported.
  3. Shipped to PVM (Python Virtual Machine): Runtime engine for Python. Big code loop that iterates through your byte codes. Actually runs your code.

• Byte code is a Python-specific representation. There is no build. Not binary machine code.

• No differences in dev / execution environment. Real-time running. No need for precompiling.

• All we have in Python is runtime.

Personal thoughts: Python does not compile the whole body of code. Not even for function / class creations. This is unlike more static languages like C, C++, Java, etc. Reversely speaking, the latter languages compile & run. This is probably why running a Java compilation will end up pre-alert of exception whilst script languages execute until the error is found.