ambiguity in proposed coding scheme?

Hi,

I meant to ask if anyone sees ambiguity in the proposed coding scheme of the HackerLanguage exercise.

For e.g. per the suggested coding scheme, the text "These are 1000000 boxes" will be coded as "10101001101000110010111100111100101100000011000011110010110010110000001000000100000011000101101111111100011001011110011". I am guessing this is not the only case with ambiguity here, but in general any case the original text has numbers that match binary equivalents of ASCII reps. of letters or whitespace, how does the scheme distinguish between the two for the decoder to correctly reconstruct the original message?

What am I missing? I solved this exercise assuming I could flag all instances of a letter or a whitespace in the decoded message with a flag character, that itself is escaped by doubling - https://gist.github.com/leonpanokarren/c7f494d579eaae04109fc9d097b60acf