Gracefully handle errors in a lockfile

[matklad]

Closes #2715

Question: why internal errors are hidden by default? I think if the unexpected has happened you most likely want to know some details. See also #2756.

[rust-highfive]

r? @alexcrichton

(rust_highfive has picked a reviewer for you, use r? to override)

[alexcrichton]

@bors: r+ afae9f3

Thanks! Yeah Cargo's story with errors is a bit up in the air, but the idea is that "internal" errors are somewhat hidden by default as they're not useful, but we've come around to thinking we should actually never hide anything sometimes, so it's a tough balance...

Basically our thinking is that if the top of an error chain is internal, then that's massively bad and we throw out tons of contextual information. If the top is a "human error", though, we can print all those and ignore the internal ones by default (as the human ones describe it)

[bors]

⌛ Testing commit afae9f3 with merge c205132...

[bors]

☀️ Test successful - cargo-cross-linux, cargo-linux-32, cargo-linux-64, cargo-mac-32, cargo-mac-64, cargo-win-gnu-32, cargo-win-gnu-64, cargo-win-msvc-32, cargo-win-msvc-64
Approved by: alexcrichton
Pushing c205132 to master...