Make int and float parsing more strict

[pfarrel]

Fix for #184 reported by @anthmatic.
#178 unintentionally made int and float parsing a lot more lax. This change restores some of the original strictness to match behaviour with jinja , while keeping support for our locale specific changes. It does so by:

• Checking that all of the input string was consumed by the parse operation, to ensure that trailing symbols or characters cause a parse to fail.
• Parsing integers first as a float, then converting to an integer, so that the above change doesn't prevent decimal expressions from parsing as integers.

[codecov-io]

Codecov Report

Merging #191 into master will increase coverage by 0.18%.
The diff coverage is 100%.

@@             Coverage Diff              @@
##             master     #191      +/-   ##
============================================
+ Coverage     71.18%   71.37%   +0.18%     
- Complexity     1344     1354      +10     
============================================
  Files           215      215              
  Lines          4269     4297      +28     
  Branches        680      685       +5     
============================================
+ Hits           3039     3067      +28     
+ Misses          988      987       -1     
- Partials        242      243       +1

Impacted Files	Coverage Δ	Complexity Δ
...java/com/hubspot/jinjava/lib/filter/IntFilter.java	100% <100%> (ø)	8 <0> (+2)	⬆️
...va/com/hubspot/jinjava/lib/filter/FloatFilter.java	90.9% <100%> (+2.02%)	8 <0> (+2)	⬆️
...ava/com/hubspot/jinjava/el/ExpressionResolver.java	92.3% <0%> (+3.67%)	14% <0%> (+6%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 0b5c62e...1319105. Read the comment docs.

[mattcoley]

Can we add a test to make sure parsing values such as 1.100000 still works as expected? Should work now but just want to cover all bases for future changes.