Issue #9 - Dead Letter Exchange, Dead Letter Routing and Alternative Exchange Policies

[RedMu]

No description provided.

[codecov]

Codecov Report

Merging #131 into master will increase coverage by 0.14%.
The diff coverage is 97.22%.

@@             Coverage Diff              @@
##             master     #131      +/-   ##
============================================
+ Coverage     94.20%   94.35%   +0.14%     
- Complexity      455      492      +37     
============================================
  Files            38       40       +2     
  Lines          1070     1151      +81     
  Branches         49       52       +3     
============================================
+ Hits           1008     1086      +78     
- Misses           45       46       +1     
- Partials         17       19       +2     

Impacted Files	Coverage Δ	Complexity Δ
...a/com/github/fridujo/rabbitmq/mock/MockPolicy.java	93.10% <93.10%> (ø)	15.00 <15.00> (?)
...ridujo/rabbitmq/mock/tool/ParameterMarshaller.java	96.55% <96.55%> (ø)	21.00 <21.00> (?)
...com/github/fridujo/rabbitmq/mock/AmqArguments.java	100.00% <100.00%> (ø)	11.00 <8.00> (-13.00)
...b/fridujo/rabbitmq/mock/MockConnectionFactory.java	100.00% <100.00%> (ø)	8.00 <6.00> (+5.00)
...ava/com/github/fridujo/rabbitmq/mock/MockNode.java	95.55% <100.00%> (+0.61%)	36.00 <6.00> (+6.00)
...va/com/github/fridujo/rabbitmq/mock/MockQueue.java	95.39% <100.00%> (+0.19%)	82.00 <2.00> (+1.00)
...o/rabbitmq/mock/exchange/BindableMockExchange.java	92.59% <100.00%> (+1.28%)	14.00 <3.00> (+1.00)
...jo/rabbitmq/mock/exchange/MockDefaultExchange.java	66.66% <100.00%> (+4.16%)	4.00 <1.00> (+1.00)

---

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 8930b81...bdc641f. Read the comment docs.

[ledoyen]

Codacy found an issue: JUnit tests should include assert() or fail()

[ledoyen]

Please do not.

I have nothing but bad experiences with Lombok.
I know that some people may find it useful to write less code, however code that is written behave as expected (vs generation of all possible constructors for ex.).
At best it messes up source jar.

[RedMu]

I'm happy to take it out, it shouldn't be packaged in the source jar though.

I've got to finish ttl and then I've done all the policies so I'll probably push it all back as one.

[ledoyen]

it shouldn't be packaged in the source jar though

That is not what I meant, rather that sources using Lombok are not valid raw Java.
When navigating sources of a third-party library, IDEs match / compile source code with no knowledge of Lombok (or other tools like it) generating stuff at compile time.

This ends up with broken navigation, (navigating to a constructor that does not exists for instance), line numbers not matching those in stacktraces and inaccurate breakpoints.

[ledoyen]

I'll probably push it all back as one

If you can split into feature commits (1 commit = prod+test) to ease review I would be very grateful !