[RESOLVED] search text file

chrisleon

I am trying to search for words on each line of a text file. I can not get this code to work.

$searchpieces = explode(" ", $SearchBox);
$lines = file("fileitm.txt");
foreach ($lines as $line_num => $line) {
$pieces = explode("~", $line);
foreach($searchpieces as $searchpiece){
if (isset($stuff)) { $stuff .= ") && preg_match(\"/$searchpiece/i\", \"$pieces[5]\""; }
else{ $stuff .= "\"/$searchpiece/i\", \"$pieces[5]\""; }
}//end of foreach

if(preg_match($stuff)){ print "$pieces[0] $pieces[2] $pieces[3] $pieces[4] $pieces[5] $pieces[6]<br>"; }

unset($stuff);
}//end of for each

bradgrafelman

When posting PHP code, please use the board's [PHP][/PHP] bbcode tags as they make your code much easier to read and analyze. You should go back and edit your post now to add these tags.
Where is $stuff defined? [EDIT: Oops, I see the $pieces declaration now.]
If $stuff isn't set (e.g. the else part of your if/else structure), it makes no sense to have PHP try to append data to it - it should instead create the variable. Since I have no idea what you're trying to do with the function, though, I can't guess as to how it should be written.
In your last if() statement above the unset() line, you call [man]preg_match/man with only one arguement, though the function requires two.

Since your code doesn't make sense at all, you might want to explain in words what exactly you're trying to do.

chrisleon

I have a text file, roughly 9000 records long. Each record has 10 fields, delimited by a ~. I have a front end to this, an input box that you type a search into.

In the sixth field, ie $pieces[5] after exploded, is the description for the record. For now, I just want to search the description.

So the first thing I am doing, is exploding the search for multiple words.
I am then opening up the file, running through each record and exploding each field by the ~.

The next for each, is to try to do a match on the first word, then if there is a second word, the isset then allows the second preg match to work.

If I hard code the preg_match:

if(preg_match("/green/i", $searchbox) && preg_match("/bean/i", $searchbox)){}
this works fine.

I use the append in a similar search with mysql, but this is a flat text file, so I am looking for a new way.

My problem is definitely in the preg_match with wrong parameters so that is why I posted it that way.

I have never defined my variables before, so that is not the issue.

bradgrafelman

chrisleon wrote:
I have a text file, roughly 9000 records long.

Any reason why you're not using a DB such as MySQL to make all of this a LOT easier?

chrisleon wrote:
delimited by a ~

Out of curiosity, what happens if your data has a literal '~' in it?

chrisleon wrote:
I have a front end to this, an input box that you type a search into.

I'm assuming that $SearchBox is defined to have this value above the code snippet?

chrisleon wrote:
the isset then allows the second preg match to work.

But how is it supposed to do that? As I said before, $stuff isn't defined in the code snippet you provided for us, so as far as we know the variable is guaranteed not to exist.

chrisleon wrote:
I have never defined my variables before, so that is not the issue.

Er, what? So you're saying I can come up with some arbitrary variable name, let's say $broccoli, and PHP is supposed to know what value should be in $broccoli ?

If moving this flat file to a database such as MySQL (where it'd be a simple SELECT query and you'd have the results), then I'd say you need to start from scratch. I understand how you split the search term into different 'words', and I see how you're looping through each line of the file. Beyond that, your code doesn't make any sense.

In fact, you don't even need to use preg_match() at all - you don't have any regular expression patterns (unless you're allowing these in your search box, in which case you have other problems you need to prevent). Instead, inside your inner foreach() loop, you'd simply use [man]stripos/man:

if(stripos($piece[5], $searchpiece) !== FALSE)
    echo "$pieces[0] $pieces[2] $pieces[3] $pieces[4] $pieces[5] $pieces[6]<br>";

EDIT: After looking at your code again, I believe one problem you have is that you think this:

$string = "'var1', 'var2'";
function($string);

is the same as this:

function('var1', 'var2');

when this is not the case. In the first code snippet, function is called with one argument, while in the second code snippet it is called with two. Using variables as arguments isn't like copying-and-pasting the contents of that value in the variable's location in the PHP script; the variable's value is sent to the function as a parameter.

chrisleon

That still did not help.

I got results that had the word BEAN, but not GREEN
I got results that had the word GREEN, but not BEAN

What I need is a way to type in GREEN BEAN, and get these results:

BEAN, GREEN CAN
GREEN BEANS LARGE
LARGE, BEANS GREEN
BEAN, GRN VERI-GREEN

I have done this with mysql, but unfortunately that is not an option at this point, flat file is it.

bradgrafelman

Can you show me a new copy of your code (don't forget about the bbcode tags I mentioned) please?

chrisleon

Never done the CODE wrappers before, hope this is what you meant.

$searchpieces = explode(" ", $SearchBox);

$lines = file("fileitm.txt");
foreach ($lines as $line_num => $line) {

$pieces = explode("~", $line);

foreach($searchpieces as $searchpiece){
if(strpos($pieces[5], $searchpiece) !== false){  print "$pieces[2] $pieces[3] $pieces[5]<br>"; }
}//end of foreach

}//end of for each

bradgrafelman

I tested your code, and it worked fine for me, though it did present another problem; if both words are found on a line, you'll get duplicates. If three words are searched for, and all three appear, you'll get triplets (or duplicates if only 2 of the 3 words match), and so on.

As such, you'd need to re-work the way this outputs lines; I would suggest instead storing the line number in an array as the key if the keyword is found; since array keys are unique, you wouldn't get multiples for each line. As such, you'd then output the line numbers in the array like so:

foreach($matchedArray as $line_num => $someArbitraryValue)
    echo $lines[$line_num] . "<br>";

Anyway, back to your code not working. Can you do a [man]var_dump/man on $SearchBox and give us the output?

chrisleon

Here are the some of the last results and the var dump

00994440 9.51 00994440 15.220 SALAD, BEAN FOUR 0002
00994448 8.05 00994448 40.270 BEANS, FRENCH CLEAN TRIMMED 0005
00996421 4.18 00996421 15.390 ONIONS, GREEN WASHED/TRIM 0004
00996441 4.64 00996441 17.240 ONIONS, GREEN W&T ROOTLESS 0004
00996489 .00 00996489 13.170 TOFU, SOFT (GREEN) 0012
00996595 .00 00996595 20.140 GREENS, COLLARD 0001
00997559 .00 00997559 52.930 GREENS, RAMPS WILD LB 0001
00999152 .00 00999152 9.470 BANANA, GREEN TIP 1/10# 0001
00999516 .00 00999516 59.620 GRAPES, GREEN SDLSS 0018
00999813 .00 00999813 13.330 PEPPERS, GREEN BELL MED 25/28# 0025
00999961 .00 00999961 13.910 LETTUCE, GREEN LEAF LINER MFC 0024
string(10) "GREEN BEAN"

You can see that I am getting results that have each, but not both.

chrisleon

I figured it out:

$SearchBox = "GREEN BEAN";

$searchpieces = explode(" ", $SearchBox);

$lines = file("fileitm.txt");
foreach ($lines as $line_num => $line) {

$pieces = explode("~", $line);

foreach($searchpieces as $searchpiece){
	if(isset($stuff)){  $stuff .= " && preg_match(\"/$searchpiece/i\", \"$pieces[5]\")";  }
	else{  $stuff .= "if(preg_match(\"/$searchpiece/i\", \"$pieces[5]\")";  }
}//end of foreach

$stuff .= "){print \"$pieces[0] $pieces[2] $pieces[4] $pieces[5]<br>\";}";

eval ($stuff);

unset($stuff);

}//end of for each

And here are the results:

00053252 .00 33.630 APTZR, GREEN BEAN BATTERED
00200733 .00 21.910 BEAN, GRN VERI-GREEN
00990400 .00 11.490 BEANS, GREEN REPACK
00991935 .00 25.800 BEANS, GREEN