In debugging upgrading to Python 3, it would be useful to be able to override the u'' string prefix to call my own function or replace with a non-u string.

I've tried things like unichr = chr which is useful for my debugging but doesn't accomplish the above.

module.uprefix = str is the type of solution I'm looking for.

CodePudding user response：

You basically can't; as others have noted in the comments, the u-prefix is handled very early, well before anything where an in-code assignment would take effect.

About the best you could do is use ast.parse to read a module on disk (without importing it) and find all the u'' strings; it distinguishes the prefixes. That would help you find them in a Python-aware way, more reliably than just searching for u' and u", but the difference probably wouldn't be large, especially if you search with word boundaries (regex \bu['"]). Unless you somehow have a lot of u' and u" in your program that aren't the prefixes?

>>> ast.dump(ast.parse('"abc"', mode='eval'))
"Expression(body=Constant(value='abc', kind=None))"
>>> ast.dump(ast.parse('u"abc"', mode='eval'))
"Expression(body=Constant(value='abc', kind='u'))"

Per the comments, what are you trying to do? I've migrated a lot of code from Python 2 to Python 3 and never needed this... There may be a different way to achieve the same goal?