perf: lazy init cookies and ip when first time use it

[dead-horse]

No description provided.

[dead-horse]

• bechmark test result before

  1 async middleware
  12319.35

  5 async middleware
  11579.15

  10 async middleware
  12569.11

  15 async middleware
  11661.07

  20 async middleware
  9762.09

  30 async middleware
  9716.56

  50 async middleware
  8148.58

  100 async middleware
  5868.63

• benchmark test result after

  1 async middleware
  13943.32

  5 async middleware
  13606.49

  10 async middleware
  12345.93

  15 async middleware
  11159.54

  20 async middleware
  10569.92

  30 async middleware
  10238.07

  50 async middleware
  8820.32

  100 async middleware
  6254.53

about 5%~10% performance improvement

[codecov]

Codecov Report

Merging #1216 into master will not change coverage.
The diff coverage is 100%.

@@          Coverage Diff           @@
##           master   #1216   +/-   ##
======================================
  Coverage     100%    100%           
======================================
  Files           5       5           
  Lines         383     391    +8     
======================================
+ Hits          383     391    +8

Impacted Files	Coverage Δ
lib/application.js	100% <ø> (ø)	⬆️
lib/request.js	100% <100%> (ø)	⬆️
lib/context.js	100% <100%> (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 74170ca...604b501. Read the comment docs.

[fengmk2]

Please add tests to coverage setters.