time.h: time() and ctime()

[yulvil]

This change is 

[codecov]

Codecov Report

Merging #253 into master will increase coverage by 0.82%.
The diff coverage is 93.75%.

@@            Coverage Diff             @@
##           master     #253      +/-   ##
==========================================
+ Coverage   80.79%   81.61%   +0.82%     
==========================================
  Files         134      135       +1     
  Lines        5336     5525     +189     
==========================================
+ Hits         4311     4509     +198     
+ Misses        835      825      -10     
- Partials      190      191       +1

Impacted Files	Coverage Δ
types/resolve.go	83.87% <ø> (+3.22%)	⬆️
program/function_definition.go	100% <ø> (ø)	⬆️
noarch/time.go	93.75% <93.75%> (ø)
transpiler/binary.go	68.78% <0%> (-2.36%)	⬇️
transpiler/enum.go	95.83% <0%> (-1.76%)	⬇️
program/program.go	70.21% <0%> (+0.64%)	⬆️
ast/position.go	82.33% <0%> (+0.75%)	⬆️
transpiler/operators.go	85.11% <0%> (+1.19%)	⬆️
ast/implicit_cast_expr.go	91.3% <0%> (+1.83%)	⬆️
... and 5 more

---

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 5b418fd...9d427af. Read the comment docs.

[elliotchance]

Reviewed 4 of 4 files at r1, 1 of 1 files at r2.
Review status: all files reviewed at latest revision, 4 unresolved discussions.

---

tests/time.c, line 18 at r2 (raw file):

    now = time(&tloc);
    is_true(now == tloc);

Please use is_eq(now, tloc) instead. This will show the value of both if it fails.

---

tests/time.c, line 20 at r2 (raw file):

    is_true(now == tloc);

    pass("%s", "time");

This is unesseary because there is an assertions above. However, it would be a good idea to also check: is_not_eq(now, 0)

---

tests/time.c, line 30 at r2 (raw file):

    time_t now = 946684798;
    s = ctime(&now);
    printf("%s", s);

Avoid printf as it will break the TAP output. You can use is_streq: https://github.com/elliotchance/c2go/blob/master/tests/tests.h#L198

---

tests/time.c, line 32 at r2 (raw file):

    printf("%s", s);

    pass("%s", "ctime");

Like the previous pass this doesn't actually check anything and can be removed.

---

Comments from Reviewable

[elliotchance]

Thanks for putting in the PR. I started to review it as you were still making changes :)

Please respond to the discussions on Reviewable (purple button at the top).

[yulvil]

Review status: 2 of 4 files reviewed at latest revision, 4 unresolved discussions.

---

tests/time.c, line 18 at r2 (raw file):

Previously, elliotchance (Elliot Chance) wrote…

Please use is_eq(now, tloc) instead. This will show the value of both if it fails.

Done.

---

tests/time.c, line 20 at r2 (raw file):

Previously, elliotchance (Elliot Chance) wrote…

This is unesseary because there is an assertions above. However, it would be a good idea to also check: is_not_eq(now, 0)

Done.

---

tests/time.c, line 30 at r2 (raw file):

Previously, elliotchance (Elliot Chance) wrote…

Avoid printf as it will break the TAP output. You can use is_streq: https://github.com/elliotchance/c2go/blob/master/tests/tests.h#L198

Done.

---

tests/time.c, line 32 at r2 (raw file):

Previously, elliotchance (Elliot Chance) wrote…

Like the previous pass this doesn't actually check anything and can be removed.

Done.

---

Comments from Reviewable

[elliotchance]

Reviewed 2 of 2 files at r3.
Review status: all files reviewed at latest revision, all discussions resolved, some commit checks failed.

---

Comments from Reviewable

[yulvil]

@elliotchance Thank you for reviewing. I am having a hard time troubleshooting the OS X error; I do not have a machine at hand.

I tried to compile similar code in Travis. The build succeeded.

• https://github.com/yulvil/clang/
• https://travis-ci.org/yulvil/clang/builds/290531110

Is it possible to add debug for the Travis/OS X build? Any help appreciated.

[elliotchance]

When I run it on my Mac:

Elliots-MacBook-Pro-3:c2go elliot$ clang tests/time.c
Elliots-MacBook-Pro-3:c2go elliot$ ./a.out
1..6
# time
1 ok - now != 0
2 ok - now != 0
3 ok - now == tloc
# ctime
Segmentation fault: 11

It looks like ctime expects that the parameter not be NULL.

When I removed https://github.com/elliotchance/c2go/pull/253/files#diff-c3bb064a4d264cc6f2d46ab300725b61R25 I did not get the segfault.

Even the same compiler on different operating systems can give different results. You can remove line 25 and 26 and your build should pass.

[elliotchance]

Reviewed 1 of 1 files at r4.
Review status: all files reviewed at latest revision, all discussions resolved.

---

Comments from Reviewable

[elliotchance]

Awesome - thanks for the PR! 👍

[yulvil]

@elliotchance Thank you for merging!