Fix conversion of ip address to int

[mabn]

Properly convert bytes to unsigned integers

Fixes #198

[codecov-io]

Codecov Report

Merging #199 into master will not change coverage.
The diff coverage is 100%.

@@            Coverage Diff            @@
##             master     #199   +/-   ##
=========================================
  Coverage     81.34%   81.34%           
  Complexity      470      470           
=========================================
  Files            77       77           
  Lines          1812     1812           
  Branches        214      214           
=========================================
  Hits           1474     1474           
  Misses          249      249           
  Partials         89       89

Impacted Files	Coverage Δ	Complexity Δ
...ore/src/main/java/com/uber/jaeger/utils/Utils.java	78.94% <100%> (ø)	6 <0> (ø)	⬇️

---

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 74faa98...f3abc85. Read the comment docs.

[yurishkuro]

for my education, since octet is already of type byte, why does & 0xFF make a difference?

[mabn]

octet is promoted to integer via binary numeric promotion (https://docs.oracle.com/javase/specs/jls/se7/html/jls-5.html#jls-5.6.2), but it's a signed integer so the mask is filled with 1s on the left for negative numbers. & 0xFF converts it to unsigned (It's a byte so no need to care about int overflow. If it mattered then it should be 0xFFL). At least that's how I understand that.

In Java 8 that would instead be: Byte.toUnsignedInt(octet)

[yurishkuro]

the magic number is hard to validate, how about a sanity check

assertEquals(expectedIp, (10 << 24) | (137 << 16) | (1 << 8) | 2); // sanity check

[mabn]

Makes sense. I've also changed assertions to assertThat(x, equals(y)), I think that they are more readable

[yurishkuro]

Thanks!