PHP Text Interpreter

Question

I’m creating a text interpreter based on the Github do Duckduckgo in PHP.

This is one of the codes I created:

if (strpos(strtolower($qt), "rand") !== FALSE){
  if (preg_match("/^rand$/", strtolower(removeAccents($qt)), $match)){
    $result = rand(1,9999);
    $sndline = "Random number";
  }
  elseif (preg_match("/^random$/", strtolower(removeAccents($qt)), $match)){
    $result = rand(1,9999);
    $sndline = "Random number";
  }
  elseif (preg_match("/^rand *\((?<min>[0-9]+),(?<max>[0-9]+)\)$/", strtolower(removeAccents($qt)), $match)){
    $result = rand($match['min'],$match['max']);
    $sndline = "Random number";
  }
  elseif (preg_match("/^random *\((?<min>[0-9]+),(?<max>[0-9]+)\)$/", strtolower(removeAccents($qt)), $match)){
    $result = rand($match['min'],$match['max']);
    $sndline = "Random number";
  }
}

As you can see, the script performs the action if the user types rand, random, rand (1,99) and random (1,99).

The whole interpreter was written with functions preg_match (I mean, everything with Regex) but found that they overload the system.

How can I create something that is fast (without overloading the system) and also understands different "orders" given by the user without using Regex (remembering that the user can type anything in the search)?

Answer 1

I see at least three points that can be optimized with ease:

1. The if more from the outside seems unnecessary.
2. The strtolower(removeAccents($qt)) may end up running more than once (4 times if you type random (1,99)).
3. The first two cases are a comparison of equality, and you do not need to use regex for this.

Translating it into code:

$cleanQt = strtolower(removeAccents($qt));
if ($cleanQt == "rand"){
  $result = rand(1,9999);
  $sndline = "Random number";
}
elseif ($cleanQt == "random"){
  $result = rand(1,9999);
  $sndline = "Random number";
}
elseif (preg_match("/^rand *\((?<min>[0-9]+),(?<max>[0-9]+)\)$/", $cleanQt), $match)){
  $result = rand($match['min'],$match['max']);
  $sndline = "Random number";
}
elseif (preg_match("/^random *\((?<min>[0-9]+),(?<max>[0-9]+)\)$/", $cleanQt, $match)){
  $result = rand($match['min'],$match['max']);
  $sndline = "Random number";
}

And I think we can still condense the last two tests into one:

$cleanQt = strtolower(removeAccents($qt));
if ($cleanQt == "rand"){
  $result = rand(1,9999);
  $sndline = "Random number";
}
elseif ($cleanQt == "random"){
  $result = rand(1,9999);
  $sndline = "Random number";
}
elseif (preg_match("/^rand|random *\((?<min>[0-9]+),(?<max>[0-9]+)\)$/", $cleanQt), $match)){
  $result = rand($match['min'],$match['max']);
  $sndline = "Random number";
}

And yet, as qmechanik shows, you can merge the first two conditions:

$cleanQt = strtolower(removeAccents($qt));
if ($cleanQt == "rand" || $cleanQt == "random"){
  $result = rand(1,9999);
  $sndline = "Random number";
}
elseif (preg_match("/^rand|random *\((?<min>[0-9]+),(?<max>[0-9]+)\)$/", $cleanQt), $match)){
  $result = rand($match['min'],$match['max']);
  $sndline = "Random number";
}

Answer 2

In accordance with mentioned by bfavaretto, the first if is unnecessary, because right below there is another if what makes practically same thing, except that the latter uses the function removeAccents.

1. You could save some lines by declaring once only the variable sndline, outside the code block of the if.

2. The code blocks of the first two if are equal, just as the last two are also, use the logical operator or to check on the same line.

   _{Note: You can also use the operator ||, more information on Qual a diferença entre “&&” e “||” e “and” e “or” em PHP? Qual usar?}

3. The use of the function preg_match in the first two if can be replaced by equality operator == (=== for something more rigid) or strcmp, strcasecmp.

4. Another point that can be improved is to declare a variable that contains the value of strtolower(removeAccents($qt)), this way you avoid running those function several times, no need.

5. You use the function preg_match in the last two if to check whether the variable qt contains the word Rand or Random followed by numbers, you can replace it by the function filter_var, used to filter the contents of a string, as you only want numbers, use as filter FILTER_SANITIZE_NUMBER_INT.

   • Combine the function filter_var with strstr to find numbers prefixed to the word Rand or Random.

Your code with those suggestions would look like this:

function filtrarNumeros($texto){
    return filter_var($texto, FILTER_SANITIZE_NUMBER_INT);
}
$qt = strtolower(removeAccents($qt));
$sndline = "Random number";

if (($qt == "rand") or ($qt == "random")){
    $result = rand(1, 9999);
}
elseif (filtrarNumeros((strstr($qt, 'rand')) !== false) or (filtrarNumeros(strstr($qt, 'random')) !== false)){
    $tok = strtok($qt, ",");
    $min = filtrarNumeros($tok);
    $max = filtrarNumeros(strtok(','));
    $result = rand($min, $max);
}
else{
    // ....
}

Exemplo