The Big Programming Thread - Page 889

def foo(s):
   for i in xrange(0,len(s)):
   if s[i] not in s[0:i] and s[i] not in s[i+1:len(s)]:
      return s[i]
   return "?"

• Tries to be too clever
  
• Code is horribly nested, or otherwise hard to follow
  
• Unit testing? Ha no, we just got it to compile.
  
• System/integration testing? Might not even know what this is...
  
• Code has essentially no sense of API design (vs something thought out)
  

• Real world code generally serves some ongoing purpose or solution, and evolves gradually, and thus needs to be maintained in a very different fashion. Uni code is more often one-shot. Or, if you're very unlucky, you get to inherit some ball of legacy mess from previous grad students and try to Frankenstein stuff on top for your paper, at which point you hand it off to the next poor sap. Of course, there's a good bit of Frankensteining crap code on top in industry too.
  
• Industry code is more often than not going to make prodigious use of existing libraries and institutional knowledge, rather than trying to reinvent the wheel. In Uni, often the whole point of the assignment is to reinvent the wheel.
  
• If a mistake is made in pushing industry code to production, you may start hemorrhaging thousands of dollars per minute, if it's a billion dollar product. There isn't "production" for Uni code, but if a mistake is made there, you may get knocked down half a letter grade or something. So stakes are very different.
  
• Industry code / programming is going to end up involving a lot more pieces than Uni code - actually the code bit of industry is only a small part. Docs, configs, productionization, testing, and rampant political bullshit will take up more of your time.
  
• Real world code is typically read an order of magnitude more times than it is written, and by people who might not even know the person who wrote it, and spaced out by a period of N years (5, 10, or more). (By "read", I mean humans looking at the code trying to figure out / verify what it does). Uni code is typically written and re-written frantically and then read exactly 0 or 1 times (either as part of a course, or in the PhD paper mill crunch happy fun time factory)
  
• On that note, actually I've seen a lot of PhD candidates who can't code their way out of a cardboard box, because PhD research quality code is a whole different ballgame
  

def unique(word):
    for char in word:
        if len(re.findall(char,word)) == 1:
            return char
    return '?'

function first_unique_letter($param) {
    $a = explode('', $param);
    $b = array_unique($a);
    return reset($b) ?: '?';
}

function first_unique_letter($param) {
    $a = explode('', $param);
    $b = array_unique($a);
    return reset($b) ?: '?';
}

echo first_unique_letter("tesseract");

E_WARNING : type 2 -- explode(): Empty delimiter -- at line 4

E_WARNING : type 2 -- array_unique() expects parameter 1 to be array, boolean given -- at line 5

E_WARNING : type 2 -- reset() expects parameter 1 to be array, null given -- at line 6

?

// fixed code that actually runs
function first_unique_letter($param) {
    $a = str_split($param);
    $b = array_unique($a);
    return reset($b) ?: '?';
}

echo first_unique_letter("tesseract");

t