Some users of Jiffy have experienced issues when decoding large JSON
documents. Normally Jiffy expects smallish documents and returns any
strings as sub-binaries. When dealing with large documents these
sub-binary references can keep a large amount of RAM around unless the
user goes through and applies `binary:copy/1` on every string returned
from Jiffy. This however causes a large amount of CPU usage to do
something that Jiffy could do as it builds the JSON structure.

The `copy_strings` decoder option does exactly this. Instead of
returning sub-binaries Jiffy now copies every string into a newly
allocated binary. Users report that this fixes the memory issues while
also not negatively affecting performance significantly.

You can no optionally request that keys are deduplicate inside of Jiffy
instead of having to perform that operation in Erlang.

For some reason I left a 377KiB test case checked in. Replacing this
with a generated test case is obvious in hindsight.

Allow Jiffy to be used easily in projects using either rebar 2 or 3.

PropEr broke my support for R14. Turns out that EQC Mini is quite usable
so I've just switched to that. If EQC Mini exists it will be used, if
not the test is skipped gracefully.

Previously if a key was malformed UTF-8 and the user specified the
`force_utf8` option we would fail to try and encode a fixed up version
of the object. This was due to missing a clause to catch the
`invalid_object_member_key` exception. This adds the clause and a couple
tests to ensure it works.

This test has proved its point and now just exists to randomly fail some
builds depending on race conditions with the garbage collector.

Previously Jiffy would throw an error about trailing data if there is
any non-whitespace character encounter after the first term had been
decoded.

This patch adds a decoder option `return_trailer` that will instead
return a sub-binary starting at the first non-whitespace character. This
allows users to be able to decode multiple terms from a single iodata()
term.

Thanks to @vlm for the original patch.

This brings back escaping forward slashes as an option during encoding.
Default is still not to escape.

This is pretty much a generalisation of the use_nil option to support an
arbitrary atom.

This fixes a leak when encoding a bare bignum. Technically it would be
possible to hit this memory leak randomly with bignums in objects but
the chances are highly unlikely.

Thanks to @miriampena for the issue.

Fixes #69

The `val` variable is a register value that we need to be able to return
at any time from `decode_iter`. If it happened that a yield was
triggered while processing trailing whitespace the lack of persistance
caused decode to return a term intialized from a random integer value.
Obviously the Erlang VM did not enjoy this.

Thanks to @michalpalka for the report.

Fixes #66

This updates encoder errors to report the actual Erlang value that
caused the error. This should make it easier to debug errors when
generating JSON.

This implements the `use_nil` option as discussed on issue #64. Passing
the atom `use_nil` as an option to both encode and decode will replace
the atom `null` with `nil` when decoding and encode `nil` as `null` when
encoding values.

Fixes #64
Fixes #68

Timeouts apparently don't actually work unless you use a test generator.
Which makes sense in hindsight.

Finally got off my butt and started using EUnit.

Refactoring the test suite I found that the old etap was only as strict
as the `==` operator where as EUnit uses the `=:=` operator.

This patch adds initial support for decoding/encoding to/from the new
maps data type.

I'd like to thank Jihyun Yu (yjh0502) for the initial versions of this
work.

Properly decode them and accept to encode them.

This was clarified by Unicode Technical Committee:
http://www.unicode.org/versions/corrigendum9.html

A single quote input was causing segfaults due to sneaking past the
string termination logic. This patch corrects that lapse in conditional
by only parsing strings where a closing quote was found. All other
strings are rejected as invalid.

Big thanks to Jean-Charles Campagne (@jccampagne) for reporting the
issue.

I removed some columns to try and minimize the amount of test data I'm
requiring and forgot to update the test code.

I pulled in a lot of tests from the double-conversion test suite to make
sure we don't accidentally break behavior in the future.

Floating point numbers are no longer encoded as a one to one mapping
of their binary representation, but as short as possible (while still
being acurate). The double-conversion library [1] is used to do the
hard work.

The ECMAScript compatible conversion is used.

[1] https://code.google.com/p/double-conversion/

After a few requests and some reflection I've decided to stop escaping
forward slashes in strings while still accepting the escaped version
through the decoder. This appears to mimic the behavior of other popular
JSON libraries I've checked (Ruby and Python).

This prevents applications that depend on Jiffy from requiring PropEr as
a dependency just to run Jiffy's full test suite. If you want to run the
full test suite you'll have to run Jiffy's Makefile directly which
creates a `.jiffy.dev` marker that enables the full test suite.

By default Jiffy is quite strict in what it encodes. By default it will
not allow invalid UTF-8 to be produced. This can cause issues when
attempting to encode JSON that was decoded by other libraries as UTF-8
semantics are not uniformly enforced.

This patch adds an option 'force_utf8' to the encoder. If encoding hits
an error for an invalid string it will forcefully mutate the object to
contain only valid UTF-8 and return the resulting encoded JSON.

For the most part this means it will strip any garbage data from
binaries replacing it replacement codepoint U+FFFD. Although, it will
also try and the common error of encoding surrogate pairs as three-byte
sequences and reencode them into UTF-8 properly.

There was a bug that was resetting the extra amount of space needed when
encoding unicode characters with the `uescape` option enabled.

I had a silly direction mistake in a bit shift that was causing the high
portion of all combining characters to be printed as \uD800 which is
obviously wrong. This bug only affects people using the non-default
uescape option during encoding.

Numbers like 1.0 were being encoded as <<"1">> which can lead to a bit
of confusion. This merely checks if a decimial point exists and if not
it appends ".0" to the value.

jiffy:encode/2 now accepts a 'pretty' option that will format the JSON
output using two-space indents.

I need to fix this so that object sizes don't explode when generating
larger values. Basically, as the type generator recurses is should be
adjusting the size value.

Looks like there's a slight difference in the number of significant
digits supported across platforms in strtod. This just adds a couple to
force it into the bignum code.

Not only was I not actually sending atoms to Jiffy, the test will end up
exhausting the atom table and crashing the Erlang VM.

Big thanks to Dmitry Groshev [1] for help getting PropEr tests into
Jiffy. These generate valid EJSON to check parsing as well as check
that random inputs don't cause segfaults or other nasty effects.

Future improvements would be to write a JSON generator as well as a
version that introduces known errors into the binary for checking known
parsing errors as well.

[1] https://github.com/si14

Fixes: #10

Rebar's eunit command wants to try and compile anything named *.erl in
the tests/ directory which causes errors with the test expectation
files. Rather than write a patch for rebar it was easier to just rename
everything.

It was possible to pass some types of invalid UTF-8 through Jiffy's
encoder. Specifically, if uescaping isn't used, values that would decode
from 0xD800 to 0xDFFFF, 0xFFFE, 0xFFFF, and values greater than 0x10FFFF
would not be flagged as invalid. Now they are.

The encoding length for unicode values was applied to the wrong index
variable.

Thanks to Paul Guyot for the detailed bug report and fix.

Fixes #7

I was accidentally truncating numbers when casting to long on
machines that don't have longs that are 8 bytes.

Fixes #5

The encoder can now return \u escaped unicode data instead of leaving
it as UTF-8 byte sequences. This done like so:

    Eshell V5.8.3  (abort with ^G)
    1> jiffy:encode(<<240, 144, 129, 128>>, [uescape]).
    <<"\"\\uD800\\uDC40\"">>

The encode and decode functions now return the value directly without
being wrapped in a tuple on success. If there is an error, it is
thrown. This is to more closely match the semantics of term_to_binary
and binary_to_term.