don't trim whitespaces when parsing markdown contents in julia demos

[johnnychen94]

The following content is a valid markdown content, but the indentation information is lost here because whitespaces are trimmed

# !!! note
#    Although `Gray.(img)` could infer the storage type as well, `ConvertEltype(Gray)` on the other
#    hand, does not do so. To get better performance, it is recommended to specify the storage type
#    when you use this operation, e.g., `ConvertEltype(Gray{Float32})`.

[johnnychen94]

Literate requires a # prefix, so we have to type five spaces here to get a correctly rendered note block.

I mistakenly typed only four spaces here. 😢