Question

Assuming a tab-separated values (TSV) file with a header line, how would one create a PHP array with the header fields as the key and the data fields as the data?

Assuming $txtArray contains all the lines in the file,

$hdrArray = explode( "\t", $txtArray[0]);
$i = 0;
foreach ($hdrArray as $hdr) {
  $heads[$hdr] = '';
  $headerNames[$i++] = $hdr;
}
for ($i = 1; $i < (count($txtArray) - 1); $i++ ) {
  $datArray = explode( "\t", $txtArray[$i]);
  if (count($datArray) > 1) {
    for($j = 0; $j < count($datArray); $j++) {
      $heads[$headerNames[$j]] = $datArray[$j];
    }
  }
  # process the line
}

I’ve got $heads containing field_name => field_data for all the fields in each line of the file. Is there a better way to code this?

Answer 1

What qualifies as ‘better’?

You could use a regex split to make it a little more robust but if you have control over the source CSV you shouldn’t have to worry about dirty data.

One obvious optimization I see is to cache the count() result.

Use:

for ($i = 1, $c=count($txtArray); $i < c - 1); $i++)

Instead of:

for ($i = 1; $i < (count($txtArray) - 1); $i++ )

Every time you call count() it’s re-calculating the result. Doing the calculation once should be sufficient so you just save the result.

I don’t see why you need:

if (count($datArray) > 1)

If you’re working with ‘clean’ data, it should have a fixed number of values-per-row so counting them and checking for none is unnecessary. To speed things up you could cache the row length by counting the number of rows in the header.

After:

$hdrArray = explode( "\t", $txtArray[0]);

Do:

$c2 = count($hdrArray);

Then use it in the second for loop:

for($j = 0; $j < $c2; $j++)

If you do have to worry about empty rows it would probably be faster to search for an empty line and skip it in the loop.

Like this:

// skip the row if the $datArray contains an empty array
if($datArray == array()) {
    continue;
}
$heads[$headerNames[$j]] = $datArray[$j];       

Altogether you get:

$hdrArray = explode( "\t", $txtArray[0]);
$c2 = count($hdrArray);

// it has an iterator variable...
// I don't understand why you wouldn't use a for loop here
$i = 0;
foreach ($hdrArray as $hdr) {
    $heads[$hdr] = '';
    $headerNames[$i++] = $hdr;
}

for ($i = 1, $c = count($txtArray); $i < $c - 1; $i++) {
    $datArray = explode( "\t", $txtArray[$i]);
    for($j = 0; $j < $c2; $j++)
        // skip the row if the $datArray contains an empty array
        if($datArray == array()) {
            continue;
        }
        $heads[$headerNames[$j]] = $datArray[$j];
    }      
}

I’m assuming your first implementation worked, and the source data is actually CSV (ie fixed number of rows/columns.

All I did was apply some simple (and common) optimizations to cut down on the number of unnecessary calculations. Pretty basic stuff that you get used to seeing after a while.