Replace banned APIs with safe ones

[ffried]

New Pull Request Checklist

• I have read and understood the CONTRIBUTING guide

• I have read the Documentation

• I have searched for a similar pull request in the project and found none

• I have updated this branch with the latest master to avoid conflicts (via merge from master or rebase)

• I have added the required tests to prove the fix/feature I am adding

• I have updated the documentation (if necessary)

• I have run the tests and they pass

• I have run the lint and it passes (pod lib lint)

This merge request fixes / refers to the following issues: #1050

Pull Request Description

This replaces the banned APIs in the list mentioned in #1050 with their safe counterparts.

[lolgear]

@ffried
Could you merge #1054 first? ( If it is necessary ).

[ffried]

@lolgear done

[lolgear]

I neither like 1 as a length of desired string nor 2 as a max length of string in strnlen.
Maybe it is better to use macros here or const/static?
I don't know, it is weird :/

[ffried]

What's wrong with these values? 1 as length is simply malloc(size * 1). And since strnlen will have to check for \0 in the string, 2 as max len will make sure it doesn't look too far but gives up whenever the string is any longer than 1. Since this is an arbitrary choice, I've added the comment there.

The first is just "normal" IMHO. The second is arbitrary, but only for this specific location, so macros or static const won't do any good either. That's what comments are for.

[OSSPolice]

	1 Warning
⚠️	The library files were changed, but the tests remained unmodified. Consider updating or adding to the tests to match the library changes.

Generated by 🚫 Danger

[codecov-io]

Codecov Report

Merging #1056 into master will decrease coverage by 0.78%.
The diff coverage is 33.33%.

@@            Coverage Diff             @@
##           master    #1056      +/-   ##
==========================================
- Coverage   39.65%   38.87%   -0.79%     
==========================================
  Files          27       27              
  Lines        4017     4018       +1     
==========================================
- Hits         1593     1562      -31     
- Misses       2424     2456      +32

Impacted Files	Coverage Δ
Classes/DDLog.m	59.15% <0%> (ø)	⬆️
Classes/DDTTYLogger.m	0% <0%> (ø)	⬆️
Classes/Extensions/DDFileLogger+Buffering.m	79.16% <100%> (+0.21%)	⬆️
Classes/DDASLLogger.m	13.46% <0%> (-42.31%)	⬇️
Classes/DDOSLogger.m	57.69% <0%> (-19.24%)	⬇️

---

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 168419e...25b5649. Read the comment docs.

[lolgear]

@ffried
Fine, right.

I think about "pure" solution here.

1. Check first symbol of string to not to be null.
2. Check first symbol to be '/'.
3. Check second symbol to be null.

// If I am correct about comparison here. Not Sure
if (name != '\0' && name == '/' && *(++name) == '\0') {}

[lolgear]

Vanishing dust in C runtime - Nice blow!

[lolgear]

@ffried
Ah!
I see, you calloc 1s instead of sizeof(char) :)

[ffried]

@lolgear You're actually right. We could express these as (char *)calloc(_appLen + 1, sizeof(char))... That sounds better indeed. Let's change that.

[lolgear]

@ffried
What about C pointer arithmetics?
Do you admire it? :)

[ffried]

@lolgear What about it? 🙂

[lolgear]

@ffried

// If I am correct about comparison here. Not sure
if (name != '\0' && name == '/' && *(++name) == '\0') {}

[ffried]

@lolgear I think it would be correct. But from a readability perspective much worse than strnlen(name, 2) == 1 😉

[lolgear]

@ffried
So, let's merge it?

[ffried]

@lolgear Fine by me. Let's wait for the CI 🙂

[sushichop]

LGTM👍
Now I found another banned API(function). So I'll create a PR for it later.