[ir] Fix ir_printer now that print() supports end

[k-ye]

Now IR looks normal:

kernel {
  $0 = offloaded  
  body {
    <f32 x1> $1 = const [1.0]
    print "[", $1, ", ", $1, "]"
  }
}

Off-topic change: I switched Metal's print from std::cout to py_cout.

Related issue = #1329 (comment)

[Click here for the format server]

---

[codecov]

Codecov Report

Merging #1350 into master will increase coverage by 0.03%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master    #1350      +/-   ##
==========================================
+ Coverage   66.51%   66.55%   +0.03%     
==========================================
  Files          37       37              
  Lines        5188     5188              
  Branches      932      932              
==========================================
+ Hits         3451     3453       +2     
+ Misses       1575     1574       -1     
+ Partials      162      161       -1     

Impacted Files	Coverage Δ
python/taichi/lang/expr.py	90.00% <100.00%> (-0.06%)	⬇️
python/taichi/lang/impl.py	90.12% <100.00%> (-0.16%)	⬇️
python/taichi/lang/snode.py	87.03% <100.00%> (ø)
python/taichi/misc/test.py	69.56% <100.00%> (+10.74%)	⬆️
python/taichi/lang/shell.py	44.06% <0.00%> (+3.38%)	⬆️

---

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 4e37b9a...c2674ab. Read the comment docs.

[k-ye]

Now the IR looks like this:

[I 07/01/20 21:56:33.645] [compile_to_offloads.cpp:operator()@23] Simplified III:
kernel {
  $0 = offloaded  
  body {
    <i32 x1> $1 = const [42]
    print "hello ", $1, " world!\n"
    print "hello ", $1, " Tai"
    <i32 x1> $4 = const [233]
    print "chi ", $4, " world!\n"
    print "hello", $1, "world!\n"
    print "  "
    print "  aswd", $1, "qwer"
  }
}

[archibate]

LGTM now.