[sgf-parsing] fix escaped characters test

Running the tests for sgf-parsing throws a DeprecationWarning related to an invalid escape sequence. This is due to the \] inclusions in the test string, which really should be \\] to work correctly.

Now from the canonical data we are getting '(;A[\\]b\\nc\\nd\\t\\te \\n\\]])', but by the time it’s rendered into the _test.py it’s become "(;A[\]b\nc\nd\t\te \n\]])", leading to the aforementioned warning.

To fix this we’d seemingly need to flag the literal as raw before parsing the json file, or we’d need to manually correct for improperly escaped non-control characters.

Unfortunately we cannot just render the strings with “r” in front of them in the Jinja template, as that ends up turning \n into two characters, \\ and n.

Open to ideas.

[yawpitch]

No worries, I just wanted to have both our thoughts down in case I’m wrong and we need to revisit in the future.

[cmccandless]

The above suggestion works on my machine. If that’s an acceptable solution, I’ll PR it.