String escapes

[laurentlb]

The specification doesn't describe string literals.
I tried this code. Each implementation of Starlark has a different result.

print("\a")
print("\b")
print("\c")
...
print("\z")

TODO:

• Agree on the escapes we handle.
• What shall we do with unknown escapes (e.g. \z). All implementations and Python return two characters \ and z. This behavior will prevent us to support more escapes in the future (as it would break users).

The fact that the \ character is sometimes used as an escape, and sometimes as a literal character is confusing.

Example of bug found in user code:

s = s.replace("'", "\'").replace('"', '\"')

[laurentlb]

From the Python manual:

Changed in version 3.6: Unrecognized escape sequences produce a DeprecationWarning. In some future version of Python they will be a SyntaxError.

I'd propose to make this change in Starlark too.

[laurentlb]

cc @Quarz0 @alandonovan
We'd like to make this change in Bazel soon.

[alandonovan]

I agree that rejecting unrecognized escape sequences (like C/C++/Go/Java) is the right approach.
Migration should not be hard because the problem can be reliably identified during lexing.
I don't think we should add deprecation warnings: 99.9% of people forced to read them are powerless to do anything to address the problem.

[laurentlb]

Agree. My comment was unclear, I wanted to suggest only the syntax error (Bazel has a way to roll out incompatible changes without using warnings).