Python strings in RAM

Question

How are string variables saved in the RAM?

for example: foo = 'abcǶ' and s = u'abc\ℙ' .

The strings contains UNICODE characters, so should it be and how is it encoded before stored in memory?

Answer 1

I'm assuming you're asking about CPython, the standard Python implementation.

The Unicode string representation was changed beginning with Python 3.3 as described in PEP 0393 . Since then, strings use the same number of bytes for all characters, either 1, 2 or 4, choosing the smallest possible for each string depending on its contents. The specific encodings used are:

• 1 byte per char: Latin-1
• 2 bytes per char: UCS-2
• 4 bytes per char: UCS-4

Before version 3.3, the Unicode string representation depended on the system, and was usually either UTF-16, UCS-4 or UCS-2, to the best of my understanding. See the above-mentioned PEP 0393 and its references for more details.