Fix up Buffer's handling of all the encodings#645

Merged

tm-rampart merged 13 commits intomasterfrom

nvw-buffer-encoding

Dec 15, 2014

Commits on Dec 15, 2014

big pass over buffer input/output encoding handling, should be mostly there but failing some tests

natevw
authored and
tcr
committed
fix up ill-fated conflation between ascii and binary (actually could have shared code in tm_str_to_X case but…meh?)

natevw
authored and
tcr
committed
share trivially shareable 7/8-bit conversion code

natevw
authored and
tcr
committed
add/fix a few more tests and update to current count

natevw
authored and
tcr
committed
fix up some halfway-unconverted utf16le code

natevw
authored and
tcr
committed
fix up overcorrection in UTF-16 conversion fix up

natevw
authored and
tcr
committed
improve base64 hygiene

natevw
authored and
tcr
committed
indent

natevw
authored and
tcr
committed
avoid abort_trap on invalid base64, but…

natevw
authored and
tcr
committed
…node.js (for whatever reason O_o) actually does not throw but returns empty buffer on malformed base64

natevw
authored and
tcr
committed
excise dead code, https://github.com/tessel/runtime/pull/645#discussion_r20481702

natevw
authored and
tcr
committed
have tm_utf16 support native [and big] endian as well as the LE the buffer methods need, this way https://github.com/tessel/runtime/pull/566 will be able to share our tm_str_to_utf16 implemenattion

natevw
authored and
tcr
committed
Restores compilation ability at -O0. Does not require HTTP to be UTF8-safe.
tcr
committed