Add breakSubstring function to Data.ByteString.Lazy.

[mrijkeboer]

This is a copy of the version in Data.ByteString with the following changes:

• breakByte -> break
• unsafeDrop -> drop
• unsafeHead -> head
• unsafeIndex -> index
• unsafeSplitAt -> splitAt
• unsafeTake -> take

[mrijkeboer]

It seems that after my last commit, the continuous integration tests fail on something other than the changes of this pull request.

[SeanRBurton]

It is not reasonable to just port the Data.ByteString implementation; the algorithm there requires constant-time indexing for good performance.

[SeanRBurton]

Instead of using indexing, you should keep references to the start and end of the span being considered, and iterate them in lock-step using tail (like a zipWith).

[sjakobi]

@mrijkeboer Do you intend to address the issues that @SeanRBurton pointed out? Otherwise it might be time to close this PR…

(I'm setting the status to "Changes requested" for clarity)

[mrijkeboer]

@sjakobi It looks like I won't have time to address @SeanRBurton issues any time soon, so I'll close this PR.