This preg_replace how do I use it with arrays ?

This site is best viewed in a modern browser with JavaScript enabled.

This preg_replace how do I use it with arrays ?

jeddik

Hi,

I have some data to find and replace from a string.

The find and replace data is in a mysql table
so I can easily extract it and build 2 arrays,
or build an associated array

( I am not sure which, so I did both)

$sql = "SELECT word,replacement FROM theos order by word";
	$result = mysql_query($sql) or die("Could NOT select words". mysql_error() );

$words = array();
$replacements= array();
$words_reps = array();
$new_words = array();
$new_words_reps = array();

while ($row = mysql_fetch_assoc($result)) {
	extract($row);
	$words[] = $word;
	$replacements[] = $replacement.' ';
	$words_reps[$word] = $replacement ;
	}

The string to do the find and replace on is a sentence like this:

$input_data =

there is something about being the parent of a very small child a child who has not yet begun to form words that has exerted a subtle pressure on the way i think about language in general my investment in words is heavy and more or less literal in the sense that they are the means by which i make at least in theory my living since my son was born six months ago the first way in which my relationship with words has changed is this i haven't been able to get nearly enough of the words down on paper

Now I could make that array:

$input_data_arr = explode(' ', $input_data);

But I don't know if I need to.

My FIND array $words
looks like this:

2
a
about
access to
across the
advert
advice
ago
agreement

My REPLACE array $replacements
looks like this:

{2|two|only two}
{a|some|a very|a definite|a specific|a particluar}
{about|regarding|concerning|with regards to|in relation to|related to|pertaining to|on the subject of|when it comes to|in regards to|in regard to|with regard to|on the subject off|as regards to}
{access to|usage of|use of|the means to access}
{across the|around the|all over the|over the|through the|round the|all around the}
{advert|advertisement|ad|offer|promotion|special offer|promo|campaign|deal|special promotion|special deal}
{advice|guidance|suggestions|tips|assistance|information|help and advice|recommendations|help|aid|instruction|counsel}
{ago|back|in the past|earlier|previously|previous|prior}
{agreement|arrangement|contract|deal}

AND THE associated array:

( Using this to list it:

foreach($words_reps as $x=>$x_value) {
		echo $x . "-->" . $x_value;
		echo "<br>";
		}

2-->{2|two|only two}
a-->{a|some|a very|a definite|a specific|a particluar}
about-->{about|regarding|concerning|with regards to|in relation to|related to|pertaining to|on the subject of|when it comes to|in regards to|in regard to|with regard to|on the subject off|as regards to}
access to-->{access to|usage of|use of|the means to access}
across the-->{across the|around the|all over the|over the|through the|round the|all around the}
advert-->{advert|advertisement|ad|offer|promotion|special offer|promo|campaign|deal|special promotion|special deal}
advice-->{advice|guidance|suggestions|tips|assistance|information|help and advice|recommendations|help|aid|instruction|counsel}
ago-->{ago|back|in the past|earlier|previously|previous|prior}
agreement-->{agreement|arrangement|contract|deal}

In order to use preg_replace(),
I wanted to add some reg_exp so that when I search for

"a" it does NOT get found in "about" but only as "a"

( maybe I should use a \b for word boundary ? )

Anyway, I changed the FIND array to be:
$new_words

"/^2$/"
"/^a$/"
"/^about$/"
"/^access to$/"
"/^across the$/"
"/^advert$/"
"/^advice$/"
"/^ago$/"
"/^agreement$/"

I also change the associated array:
$new_words_reps:

"/^{2$/"-->{2|two|only} two}
"/^{a$/"-->{a|some|a} very|a definite|a specific|a particluar}
"/^{about$/"-->{about|regarding|concerning|with} regards to|in relation to|related to|pertaining to|on the subject of|when it comes to|in regards to|in regard to|with regard to|on the subject off|as regards to}
"/^access to$/"-->{access to|usage of|use of|the means to access}
"/^across the$/"-->{across the|around the|all over the|over the|through the|round the|all around the}
"/^{advert$/"-->{advert|advertisement|ad|offer|promotion|special} offer|promo|campaign|deal|special promotion|special deal}
"/^{advice$/"-->{advice|guidance|suggestions|tips|assistance|information|help} and advice|recommendations|help|aid|instruction|counsel}
"/^{ago$/"-->{ago|back|in} the past|earlier|previously|previous|prior}
"/^{agreement$/"-->{agreement|arrangement|contract|deal}}

OK !

So now I want to use the preg_replace()

BUT my question is - do I need to use the associated array
or do I use the two FIND and REPLACE arrays

Since these 2 arrays were generated from the same table, they have the same
natural keys - does that keep them in step when replacing ?

If I use the two arrays, I guess it should like this ...

$new_data = preg_replace($new_words,$replacements,$input_data);

If is the assocated array, I don't know how to write that !! :o

Am I correct in thinking that the input can be a string ?
Or does that need to be an array as well ?

Sorry,
A lot of questions .. but I am not sure how to do this !!

Thanks for any help !

NogDog

Take a look at my reply in your other thread to see how I used an associative array with the search regex as the key.

bionoid

Looks like you're creating an automatic Thesaurus?

jeddik

Thanks for your reply.

I will work with what to showed in the other thread and try and understand it
with the example I used here.

So - to answer my question - I should use the associative array
which has both search and replace data in it ??

If ı don't understand it, I'll ask again

bionoid: yes it is a bit like that - should be helpful in classifiying the mountains of text data I am trying to deal with

Thanks to both for your help !!

bionoid

I won't interfere with this as NogDog is already helping, but it does look seemingly similar to previous posts we've dealt with.
Just one concern I have to both of you with the array based search and replaces, is that you could potentially overlap replaced words as it goes through your replacement arrays, possibly expanding the sentence unintentionally.

NogDog

By using one array, you don't have to worry about the search terms in one array not matching up correctly with the replacement terms in a separate array.

bionoid

The requirement in this thread seems more complex than in the other. As there are now multiple replacement phrases for the same word, so now you have to keep record of what has been used, and if you reach the end do you start with the first phrase again?

jeddik

The requirement in this thread seems more complex

Yes, that is the reason for starting a new thread is that I realised that to accomplish my objective, I needed to have a more complex replacement.

BUT actually everything inside the curly brackets can be considered as one replacement phrase.

I actually just want "access to" to be replaced with "{access to|usage of|use of|the means to access}" that's all - nothing more complicated.

( But of course once replace - I don't want parts of it getting replaced again. )

I have mixed the find/replacement data so that it is now:

$words_reps = array(
'John' => 'John: 123 Main St',
'John Smith' => 'John Smith: 456 Elm St',
'John J. Smith' => 'John J. Smith: 531 Park Ave',
'a' => '{a|some|a very|a definite|a specific|a particluar}',
'about' => '{about|regarding|concerning|with regards to}',
'about to' => '{about to|nearly|on the point of|will|just about to}',
'across the' => '{across the|around the|all over the|over the|through the|round the|all around the}'
);

And the input data is:

$input_data = <<<EOD
Testing 1, 2, 3. John Smith is about to be a jerk. He is John's friend and lives across the road John J. Smith is a good person, but John is living about 1 mile away !!! The end
EOD;

And this is my code:

krsort($words_reps);
foreach($words_reps as $x=>$x_value) {
	echo $x . "-->" . $x_value;
	echo "<br>";
	} 


foreach($words_reps as $key => $val) {
	$regex = '/(?<!'.preg_quote('{marker}', '/').')'.preg_quote($key, '/').'/i';
	$regexData[$regex] = '{marker}'.$val;
	}


// The array_keys() function returns an array containing the keys.
$new_data = preg_replace(array_keys($regexData), $regexData, $input_data);
$new_data = str_replace('{marker}', '', $new_data);

echo "<br>This is the RESULT:<br><br>$new_data";

The result shows that I DO have a problem -

The "{marker}" seems to be getting replaced and
things are getting double replaced ... I think.

This is the RESULT:

Testing 1, 2, 3. John Smith: 456 Elm St is {m{a|some|a very|a definite|a specific|a particluar}rker}{{m{a|some|a very|a definite|a specific|a particluar}rker}{{a|some|a very|a definite|a specific|a particluar}bout|reg{a|some|a very|a definite|a specific|a particluar}rding|concerning|with reg{a|some|a very|a definite|a specific|a particluar}rds to} to|ne{a|some|a very|a definite|a specific|a particluar}rly|on the point of|will|just {m{a|some|a very|a definite|a specific|a particluar}rker}{{a|some|a very|a definite|a specific|a particluar}bout|reg{a|some|a very|a definite|a specific|a particluar}rding|concerning|with reg{a|some|a very|a definite|a specific|a particluar}rds to} to} be {a|some|a very|a definite|a specific|a particluar} jerk. He is John: 123 Main St's friend {a|some|a very|a definite|a specific|a particluar}nd lives {m{a|some|a very|a definite|a specific|a particluar}rker}{{a|some|a very|a definite|a specific|a particluar}cross the|{a|some|a very|a definite|a specific|a particluar}round the|{a|some|a very|a definite|a specific|a particluar}ll over the|over the|through the|round the|{a|some|a very|a definite|a specific|a particluar}ll {a|some|a very|a definite|a specific|a particluar}round the} ro{a|some|a very|a definite|a specific|a particluar}d John J. Smith: 531 Park Ave is {a|some|a very|a definite|a specific|a particluar} good person, but John: 123 Main St is living {m{a|some|a very|a definite|a specific|a particluar}rker}{{a|some|a very|a definite|a specific|a particluar}bout|reg{a|some|a very|a definite|a specific|a particluar}rding|concerning|with reg{a|some|a very|a definite|a specific|a particluar}rds to} 1 mile {a|some|a very|a definite|a specific|a particluar}w{a|some|a very|a definite|a specific|a particluar}y !!! The end

Can you see what I have got wrong ?

Thanks.

jeddik

Hi,

I seemed to have improved it a bit by using a marker that is not likely to be used
anywhere 'XxX'

So my code now is:

krsort($words_reps);
foreach($words_reps as $x=>$x_value) {
	echo $x . "-->" . $x_value;
	echo "<br>";
	} 


foreach($words_reps as $key => $val) {
	$regex = '/(?<!'.preg_quote('XxX', '/').')'.preg_quote($key, '/').'/i';
	$regexData[$regex] = 'XxX'.$val;
	}


// The array_keys() function returns an array containing the keys.
$new_data = preg_replace(array_keys($regexData), $regexData, $input_data);
$new_data = str_replace('XxX', '', $new_data);

echo "<br>This is the RESULT:<br><br>$new_data";

This is the RESULT:

Testing 1, 2, 3. John Smith: 456 Elm St is {{{a|some|a very|a definite|a specific|a particluar}bout|reg{a|some|a very|a definite|a specific|a particluar}rding|concerning|with reg{a|some|a very|a definite|a specific|a particluar}rds to} to|ne{a|some|a very|a definite|a specific|a particluar}rly|on the point of|will|just {{a|some|a very|a definite|a specific|a particluar}bout|reg{a|some|a very|a definite|a specific|a particluar}rding|concerning|with reg{a|some|a very|a definite|a specific|a particluar}rds to} to} be {a|some|a very|a definite|a specific|a particluar} jerk. He is John: 123 Main St's friend {a|some|a very|a definite|a specific|a particluar}nd lives {{a|some|a very|a definite|a specific|a particluar}cross the|{a|some|a very|a definite|a specific|a particluar}round the|{a|some|a very|a definite|a specific|a particluar}ll over the|over the|through the|round the|{a|some|a very|a definite|a specific|a particluar}ll {a|some|a very|a definite|a specific|a particluar}round the} ro{a|some|a very|a definite|a specific|a particluar}d John J. Smith: 531 Park Ave is {a|some|a very|a definite|a specific|a particluar} good person, but John: 123 Main St is living {{a|some|a very|a definite|a specific|a particluar}bout|reg{a|some|a very|a definite|a specific|a particluar}rding|concerning|with reg{a|some|a very|a definite|a specific|a particluar}rds to} 1 mile {a|some|a very|a definite|a specific|a particluar}w{a|some|a very|a definite|a specific|a particluar}y !!! The end

The last word should be "away"

But I get "{a|some|a very|a definite|a specific|a particluar}w{a|some|a very|a definite|a specific|a particluar}y"

To stop this happening do I need the \b for word boundary somewhere ?

If so ... where ?

Thanks

bionoid

Sorry, have been away for the last couple days.

As mentioned before you have overlaps on the replacements, not just with the word "away" but also the word "about" in one of the initial replacements.
I did one as well for you to look at, I have used the boundary option in the regular expression as well:

<?php

$input_data = "Testing 1, 2, 3. John Smith is about to be a jerk. He is John's friend and lives across the road John J. Smith is a good person, but John is living about 1 mile away !!! The end";
$words_reps = array(
	'John' => 'John: 123 Main St',
	'John Smith' => 'John Smith: 456 Elm St',
	'John J. Smith' => 'John J. Smith: 531 Park Ave',
	'a' => '{a|some|a very|a definite|a specific|a particluar}',
	'about' => '{about|regarding|concerning|with regards to}',
	'about to' => '{about to|nearly|on the point of|will|just about to}',
	'across the' => '{across the|around the|all over the|over the|through the|round the|all around the}'
);

echo '<p>-- Original --<br />' . $input_data . '</p>';

$sets = array();
foreach ($words_reps as $i => $j) {
	$sets[count(explode(' ', $i))][$i] = $j;
}
krsort($sets, SORT_NUMERIC);

$words = array();
foreach ($sets as $set) {foreach ($set as $word => $replace) {
	$word_index = '{WORD_' . count($words) . '}';
	$input_data = preg_replace('/\b' . preg_quote($word, '/') . '\b/', $word_index, $input_data);
	$words[$word_index] = $replace;
}}

echo '<p>-- Markers --<br />' . $input_data . '</p>';

foreach ($words as $i => $j) {
	$input_data = str_replace($i, $j, $input_data);
}

echo '<p>-- Final --<br />' . $input_data . '</p>';

?>

Output

-- Original --
Testing 1, 2, 3. John Smith is about to be a jerk. He is John's friend and lives across the road John J. Smith is a good person, but John is living about 1 mile away !!! The end

-- Markers --
Testing 1, 2, 3. {WORD_1} is {WORD_2} be {WORD_5} jerk. He is {WORD_4}'s friend and lives {WORD_3} road {WORD_0} is {WORD_5} good person, but {WORD_4} is living {WORD_6} 1 mile away !!! The end

-- Final --
Testing 1, 2, 3. John Smith: 456 Elm St is {about to|nearly|on the point of|will|just about to} be {a|some|a very|a definite|a specific|a particluar} jerk. He is John: 123 Main St's friend and lives {across the|around the|all over the|over the|through the|round the|all around the} road John J. Smith: 531 Park Ave is {a|some|a very|a definite|a specific|a particluar} good person, but John: 123 Main St is living {about|regarding|concerning|with regards to} 1 mile away !!! The end

Hope that's what you're looking for

jeddik

Hi.

Thanks a ton !

Now I was already re-testing my changes to Nogdog's suggestion

I had tried this:

krsort($words_reps);
foreach($words_reps as $x=>$x_value) {
	echo $x . "-->" . $x_value;
	echo "<br>";
	} 

foreach($words_reps as $key => $val) {
	$regex = '/(?<!'.preg_quote('XxX', '/').')'.'/\b'.preg_quote($key, '/').'/\b'.'/i';
	$regexData[$regex] = 'XxX'.$val;
       }

// The array_keys() function returns an array containing the keys.
$new_data = preg_replace(array_keys($regexData), $regexData, $input_data);
$new_data = str_replace('XxX', '', $new_data);

echo "<br>This is the RESULT:<br><br>$new_data";

I have added .'/\b'. before and after the key of $words_reps
That should be the "a" or "about" etc.

But I get an error:

Warning: preg_replace(): Unknown modifier '\' in /home/coje66/public_html/auto_pre_spin.php on line 90

Line 90 is :

$new_data = preg_replace(array_keys($regexData), $regexData, $input_data);

So ...
not sure what happened

Hi Bionoid.

Yes that works great -
I don't fully understand it

Not sure why we are counting those keys ??

I will work through it.

Would like to get Nogdog's suggestion running if possible
first as I understand it but think it just needs the boundary to
make it work !!

Thanks

NogDog

That error is because you have a "/" before your "\b", and "/" is your regexp delimiter. Not sure if you really mean to have it there, but if you do want a literal "/", you will have to escape it (or else use a different delimiter).

jeddik

Thanks - I thought I needed to escape the \ character so I put that / in

actually I saw it in bionoid's code:

$input_data = preg_replace('/\b' . preg_quote($word, '/') . '\b/', $word_index, $input_data);

Now I have taken them out it runs better, but not perfect.

My code is now:

krsort($words_reps);
foreach($words_reps as $x=>$x_value) {
	echo $x . "-->" . $x_value;
	echo "<br>";
	} 

foreach($words_reps as $key => $val) {
	$regex = '/(?<!'.preg_quote('XxX', '/').')'.'\b'.preg_quote($key, '/').'\b'.'/i';
	$regexData[$regex] = 'XxX'.$val;
	}

This is the RESULT:

Testing 1, 2, 3. John Smith: 456 Elm St is {{about|regarding|concerning|with regards to} to|nearly|on the point of|will|just {about|regarding|concerning|with regards to} to} be {a|some|a very|a definite|a specific|a particluar} jerk. He is John: 123 Main St's friend and lives {across the|around the|all over the|over the|through the|round the|all around the} road John J. Smith: 531 Park Ave is {a|some|a very|a definite|a specific|a particluar} good person, but John: 123 Main St is living {about|regarding|concerning|with regards to} 1 mile away !!! The end

The first replacement has a problem, the rest are OK

strange ???

jeddik

Its OK - got it working now.

Thanks again for helping

jeddik

Hi again.

I have been using this little script to apply names and addresses
but still have a problem ...

The replacing is occurring again with already replaced sets of data.

For example I have slightly extended the test data from above:

It is now:

$words_reps = array(
'john' => '{John: 123 Main St}',
'now' => '{now|right now|just now}',
'is' => '{is|really is|most certainly is}',
'special' => '{special|really special|very special}',
'this' => '{this|this thing|this unique}',
'this is' => '{this is|this wretched|this is hell}',
'john smith' => '{John Smith: 456 Elm St|John B Smith: 456 Elm St}',
'john j. smith' => '{John J. Smith: 531 Park Ave}',
'a' => '{a|some|a very|a definite|a specific|a particluar}',
'about' => '{about|regarding|concerning|with regards to}',
'test' => '{test|examination|in depth look|research project}',
'about to' => '{about to|nearly|on the point of|will|just about to}',
'across the' => '{across the|around the|all over the|over the|through the|round the|all around the}'
);

This is the RESULT:

{now|right now|just now} {{this|this thing|this unique} {is|really is|most certainly is}|{this|this thing|this unique} wretched|{this|this thing|this unique} {is|really is|most certainly is} hell} {a|some|a very|a definite|a specific|a particluar} {special|really special|very special} {test|examination|in depth look|research project} and {is|really is|most certainly is} all {about|regarding|concerning|with regards to} {John: 123 Main St} and his friends johns like {{John: 123 Main St} Smith: 456 Elm St|{John: 123 Main St} B Smith: 456 Elm St} {is|really is|most certainly is} {now|right now|just now} to be {a|some|a very|a definite|a specific|a particluar} jerk. and {{this|this thing|this unique} {is|really is|most certainly is}|{this|this thing|this unique} wretched|{this|this thing|this unique} {is|really is|most certainly is} hell} another {test|examination|in depth look|research project} to find out what {this|this thing|this unique} thing {is|really is|most certainly is} all {about|regarding|concerning|with regards to}. we need to understand what we can do with {this|this thing|this unique}. he {is|really is|most certainly is} {John: 123 Main St}'s friend and lives {across the|around the|all over the|over the|through the|round the|all around the} road {{John: 123 Main St} J. Smith: 531 Park Ave} {is|really is|most certainly is} {a|some|a very|a definite|a specific|a particluar} good person, but {John: 123 Main St} {is|really is|most certainly is} living {about|regarding|concerning|with regards to} 1 mile away !!! the end

When there is a replacable name within an already replaced name and address
then it gets replaced again.

Like this:

{{John: 123 Main St} Smith: 456 Elm St|{John: 123 Main St} B Smith: 456 Elm St}

That should have been:

{John Smith: 456 Elm St|John B Smith: 456 Elm St}.

ACTUALLY ...

The same problem occurs in the text as well:

{{this|this thing|this unique} {is|really is|most certainly is}|{this|this thing|this unique} wretched|{this|this thing|this unique} {is|really is|most certainly is} hell}

That should have been simply:

{this is|this wretched|this is hell}

How can I stop this nesting of replacements ?

Thanks.

bionoid

Hi, what code are you using to get those results?
If it's my code causing the error I will gladly look into it.

jeddik

This is the code ( including the data )

$input_data = <<<EOD
	NOW This is a special test and is all about John and his friends Johns like John Smith is now to be a jerk. And this is another test to find out what this thing is all about. We need to understand what we can do with this.
	He is John's friend and lives across the road John J. Smith is a good person, but John is living about 1 mile away !!! The end
	EOD;

$input_data = strtolower($input_data);


echo "This is the article:<br><br>$input_data<br><br>";

$words_reps = array(
	'john' => '{John: 123 Main St}',
	'now' =>  '{now|right now|just now}',
	'is' =>  '{is|really is|most certainly is}',
	'special' =>  '{special|really special|very special}',
	'this' =>  '{this|this thing|this unique}',
	'this is' =>  '{this is|this wretched|this is hell}',
	'john smith' => '{John Smith: 456 Elm St|John B Smith: 456 Elm St}',
	'john j. smith' => '{John J. Smith: 531 Park Ave}',
	'a' => '{a|some|a very|a definite|a specific|a particluar}',
	'about' => '{about|regarding|concerning|with regards to}',
	'test' => '{test|examination|in depth look|research project}',
	'about to' => '{about to|nearly|on the point of|will|just about to}',
	'across the' => '{across the|around the|all over the|over the|through the|round the|all around the}'
	);

echo "This is the FIND AND REPLAY ARRAY:<br>";
foreach($words_reps as $x=>$x_value) {
	echo $x . "-->" . $x_value;
	echo "<br>";
	} 

echo "<br>======================================================<br>";


echo "This is the NEW find AND replay ARRAY:<br>";


	// KRSORT() Sorts an array by key in reverse order, 
	// maintaining key to data correlations

krsort($words_reps);
foreach($words_reps as $x=>$x_value) {
	echo $x . "-->" . $x_value;
	echo "<br>";
	} 

foreach($words_reps as $key => $val) {
	$regex = '/(?<!'.preg_quote('XxX', '/').')'.'\b'.preg_quote($key, '/').'\b'.'/i';
	$regexData[$regex] = 'XxX'.$val;
	}


// The array_keys() function returns an array containing the keys.
$new_data = preg_replace(array_keys($regexData), $regexData, $input_data);
$new_data = str_replace('XxX', '', $new_data);

echo "<br>This is the RESULT:<br><br>$new_data";

I think its what you recommended - but I
may have made an error somewhere :o

Thanks.

bionoid

You're using NogDog's code there.
I have updated the one I submitted anyway for comparison:

<?php

$input_data = <<<EOD
    NOW This is a special test and is all about John and his friends Johns like John Smith is now to be a jerk. And this is another test to find out what this thing is all about. We need to understand what we can do with this.
    He is John's friend and lives across the road John J. Smith is a good person, but John is living about 1 mile away !!! The end
EOD;

$words_reps = array(
    'john' => '{John: 123 Main St}',
    'now' =>  '{now|right now|just now}',
    'is' =>  '{is|really is|most certainly is}',
    'special' =>  '{special|really special|very special}',
    'this' =>  '{this|this thing|this unique}',
    'this is' =>  '{this is|this wretched|this is hell}',
    'john smith' => '{John Smith: 456 Elm St|John B Smith: 456 Elm St}',
    'john j. smith' => '{John J. Smith: 531 Park Ave}',
    'a' => '{a|some|a very|a definite|a specific|a particluar}',
    'about' => '{about|regarding|concerning|with regards to}',
    'test' => '{test|examination|in depth look|research project}',
    'about to' => '{about to|nearly|on the point of|will|just about to}',
    'across the' => '{across the|around the|all over the|over the|through the|round the|all around the}'
);

$sets = array();
foreach ($words_reps as $i => $j) {
    $sets[count(explode(' ', $i))][$i] = $j;
}
krsort($sets, SORT_NUMERIC);

$words = array();
foreach ($sets as $set) {foreach ($set as $word => $replace) {
    $word_index = '{WORD_' . count($words) . '}';
    $input_data = preg_replace('/\b' . preg_quote($word, '/') . '\b/', $word_index, $input_data);
    $words[$word_index] = $replace;
}}

foreach ($words as $i => $j) {
    $input_data = str_replace($i, $j, $input_data);
}

echo '<p>' . $input_data . '</p>';

?>

Output

NOW This {is|really is|most certainly is} {a|some|a very|a definite|a specific|a particluar} {special|really special|very special} {test|examination|in depth look|research project} and {is|really is|most certainly is} all {about|regarding|concerning|with regards to} John and his friends Johns like John Smith {is|really is|most certainly is} {now|right now|just now} to be {a|some|a very|a definite|a specific|a particluar} jerk. And {this is|this wretched|this is hell} another {test|examination|in depth look|research project} to find out what {this|this thing|this unique} thing {is|really is|most certainly is} all {about|regarding|concerning|with regards to}. We need to understand what we can do with {this|this thing|this unique}. He {is|really is|most certainly is} John's friend and lives {across the|around the|all over the|over the|through the|round the|all around the} road John J. Smith {is|really is|most certainly is} {a|some|a very|a definite|a specific|a particluar} good person, but John {is|really is|most certainly is} living {about|regarding|concerning|with regards to} 1 mile away !!! The end

NogDog

Note that when you use the HEREDOC method of quoting (with "<<<"), the string that ends the quote ("EOD;" in this example) must not have any leading tabs/spaces before it (as corrected in bionoid's update).

jeddik

You are both very kind.
Thanks very much for your help.

Will work on this tomorrow.
Hopefully get it going