Advanced first and last sentence functions - php

I'm trying to retrieve the first and last sentences from database text entries.
The code I have works fine in this example:
$text = "He was doing ok so far, but this one had stumped him. He was a bit lost..."
The functions:
function first_sentence($content) {
$pos = strpos($content, '.');
if($pos === false) {
return $content;
}
else {
return substr($content, 0, $pos+1);
}
} // end function
// Get the last sentence
function last_sentence($content) {
$content = array_pop(array_filter(explode('.', $content), 'trim'));
return $content;
} // end function
The last sentence function takes into account any trailing...'s at the end of a sentence, but neither can cope with the following:
$text = "Dr. Know-all was a coding master, he knew everything and was reputed the world over. But the Dr. was in trouble..."
Result:
First sentence: Dr.
Last sentence: was in trouble
I need to modify the functions to take into account things like 'Dr.' and other such abbreviations if that's possible, so the last text variable would come out as:
First sentence: Dr. Know-all was a coding master, he knew everything and was reputed the world over
Last sentence: But the Dr. was in trouble
Can it be done? Any help much appreciated!

You can exclude some word by replacing them.
<?
function first_sentence($content) {
$pos = strpos($content, '.');
if($pos === false) {
return $content;
}
else {
return substr($content, 0, $pos+1);
}
} // end function
// Get the last sentence
function last_sentence($content) {
$content = array_pop(array_filter(explode('.', $content), 'trim'));
return $content;
} // end function
$text = "Dr. Know-all was a coding master, he knew everything and was reputed the world over. But the Dr. was in trouble...";
$tmp = str_replace("Dr.","Dr____",$text);
echo $tmm ."\n";
echo str_replace("Dr____","Dr.",first_sentence($tmp ))."\n";
echo str_replace("Dr____","Dr.",last_sentence($tmp ));
?>
WORKING CODE

Maybe you thought about that..
Can you make a function to encode/decode the $content before you search for the sentences;
function encode_content($content){
return $encoded_content = str_replace("Dr.", "Dr#;#", $content);
}
And after you retrieve the sentences, decode again:
function decode_content($content){
return $encoded_content = str_replace("Dr#;#", "Dr." , $content);
}

You can check your substr length, and return only if it is longer than 3 characters (point included). If it is less or equal, you can use a whitelist in order not to stumble upon words such as "No", "Me", "Us", "Oh"... Scrabble dictionaries should be able to help you there :)

Just to answer my own question after putting some new functions together given the answers so far
function encode_text($content){
$search = array("Dr.", "i.e.", "Mr.", "Mrs.", "Ms."); // put our potential problems in an array
$replace = array("Dr#;#", "i#e#", "Mr#;#", "Mrs#;#", "Ms#;#"); // make them good for first and last sentence functions
$encoded_content = str_replace($search, $replace, $content);
return $encoded_content;
} // end encode
Then we just swap the search and replace variables around to make our decode function. Now they can be used with the first and last sentence functions above, and it works a charm. Adding stuff to the arrays is simple, thinking of what's appropriate to add tho is less so :)
Cheers!

Related

PHP keyword replacement using preg_replace and a FOR loop

I'm completely lost on what's wrong here. Any help will be appreciated.
I'm making a keyword replacement tool for a wordpress site. My problem is that my preg_replace seems to only run on the last element in the array when going through the post content.
The two get_options below are just textarea fields with each keyword and replacement on a separate line using character returns.
function keyword_replace($where) {
$keywords_to_replace = get_option('keywords_to_replace');
$keyword_links = get_option('keyword_links');
$KWs = explode("\n", $keywords_to_replace);
$URLs = explode("\n", $keyword_links);
$pattern = array();
$replacement = array();
for($i=0; $i< count($KWs); $i++) {
$pattern2 = '/<a[^>]*>(.*?)'.$KWs[$i].'(.*?)</a>/';
if(preg_match($pattern2, $where)) {
continue;
} else {
$pattern[$i] = '\'(?!((<.*?)|(<a.*?)))(\b'. $KWs[$i] . '\b)(?!(([^<>]*?)>)|([^>]*?</a>))\'si';
$replacement[$i] .= ''.$KWs[$i].'';
}
}
return preg_replace($pattern, $replacement, $where, -1);
}
add_filter('the_content','keyword_replace');
A var_dump() returns all the correct information, and nothing appears to be skipped in the dump. I'm stuck trying to figure out why it doesn't loop through the whole array.
Thanks for your help,
Rafael
I had a friend fix this for me. Just in case anyone else comes across a problem similar to mine, the fix was to use the trim() command around $KWs[$i] inside $pattern[$i].
$pattern[$i] = '\'(?!((<.*?)|(<a.*?)))(\b'. trim($KWs[$i]) . '\b)(?!(([^<>]*?)>)|([^>]*?</a>))\'si';
It all came down to windows vs. mac on character returns.

Unspin text in php from spun text on sentence level

I need to neatly output spun text in a php page.
I already have the prespun text in {hi|hello|greetings} format.
I have a php code that i found elsewhere, but it does not output the spun text on sentence level, where two {{ come.
Here is the code that needs fixing.
<?php
function spinText($text){
$test = preg_match_all("#\{(.*?)\}#", $text, $out);
if (!$test) return $text;
$toFind = Array();
$toReplace = Array();
foreach($out[0] AS $id => $match){
$choices = explode("|", $out[1][$id]);
$toFind[]=$match;
$toReplace[]=trim($choices[rand(0, count($choices)-1)]);
}
return str_replace($toFind, $toReplace, $text);
}
echo spinText("{Hello|Hi|Greetings}!");;
?>
The output will be randomly chose word: Hello OR Hi OR Greetings.
However, if there is a sentence level spinning, the output is messed up.
E.g.:
{{hello|hi}.{how're|how are} you|{How's|How is} it going}
The output is
{hello.how're you|How is it going}
As you can see the text has not been spun completely.
Thank you
This is a recursive problem, so regular expressions aren't that great; but recursive patterns can help though:
function bla($s)
{
// first off, find the curly brace patterns (those that are properly balanced)
if (preg_match_all('#\{(((?>[^{}]+)|(?R))*)\}#', $s, $matches, PREG_OFFSET_CAPTURE)) {
// go through the string in reverse order and replace the sections
for ($i = count($matches[0]) - 1; $i >= 0; --$i) {
// we recurse into this function here
$s = substr_replace($s, bla($matches[1][$i][0]), $matches[0][$i][1], strlen($matches[0][$i][0]));
}
}
// once we're done, it should be safe to split on the pipe character
$choices = explode('|', $s);
return $choices[array_rand($choices)];
}
echo bla("{{hello|hi}.{how're|how are} you|{How's|How is} it going}"), "\n";
See also: Recursive patterns

Replace multiple occurrences of a string with different values

I have a script that generates content containing certain tokens, and I need to replace each occurrence of a token, with different content resulting from a separate loop.
It's simple to use str_replace to replace all occurrences of the token with the same content, but I need to replace each occurrence with the next result of the loop.
I did see this answer: Search and replace multiple values with multiple/different values in PHP5?
however it is working from pre-defined arrays, which I don't have.
Sample content:
This is an example of %%token%% that might contain multiple instances of a particular
%%token%%, that need to each be replaced with a different piece of %%token%% generated
elsewhere.
I need to replace each occurrence of %%token%% with content generated, for argument's sake, by this simple loop:
for($i=0;$i<3;$i++){
$token = rand(100,10000);
}
So replace each %%token%% with a different random number value $token.
Is this something simple that I'm just not seeing?
Thanks!
I don't think you can do this using any of the search and replace functions, so you'll have to code up the replace yourself.
It looks to me like this problem works well with explode(). So, using the example token generator you provided, the solution looks like this:
$shrapnel = explode('%%token%%', $str);
$newStr = '';
for ($i = 0; $i < count($shrapnel); ++$i) {
// The last piece of the string has no token after it, so we special-case it
if ($i == count($shrapnel) - 1)
$newStr .= $shrapnel[$i];
else
$newStr .= $shrapnel[$i] . rand(100,10000);
}
I know this is an old thread, but I stumbled across it while trying to achieve something similar. If anyone else sees this, I think this is a little nicer:
Create some sample text:
$text="This is an example of %%token%% that might contain multiple instances of a particular
%%token%%, that need to each be replaced with a different piece of %%token%% generated
elsewhere.";
Find the search string with regex:
$new_text = preg_replace_callback("|%%token%%|", "_rand_preg_call", $text);
Define a callback function to change the matches
function _rand_preg_call($matches){
return rand(100,10000);
}
Echo the results:
echo $new_text;
So as a function set:
function _preg_replace_rand($text,$pattern){
return preg_replace_callback("|$pattern|", "_rand_preg_call", $text);
}
function _rand_preg_call($matches){
return rand(100,10000);
}
I had a similar issue where I had a file that I needed to read. It had multiple occurrences of a token, and I needed to replace each occurrence with a different value from an array.
This function will replace each occurrence of the "token"/"needle" found in the "haystack" and will replace it with a value from an indexed array.
function mostr_replace($needle, $haystack, $replacementArray, $needle_position = 0, $offset = 0)
{
$counter = 0;
while (substr_count($haystack, $needle)) {
$needle_position = strpos($haystack, $needle, $offset);
if ($needle_position + strlen($needle) > strlen($haystack)) {
break;
}
$haystack = substr_replace($haystack, $replacementArray[$counter], $needle_position, strlen($needle));
$offset = $needle_position + strlen($needle);
$counter++;
}
return $haystack;
}
By the way, 'mostr_replace' is short for "Multiple Occurrence String Replace".
You can use the following code:
$content = "This is an example of %%token%% that might contain multiple instances of a particular %%token%%, that need to each be replaced with a different piece of %%token%% generated elsewhere.";
while (true)
{
$needle = "%%token%%";
$pos = strpos($content, $needle);
$token = rand(100, 10000);
if ($pos === false)
{
break;
}
else
{
$content = substr($content, 0,
$pos).$token.substr($content, $pos + strlen($token) + 1);
}
}

how to CaPiTaLiZe every other character in php?

I want to CaPiTaLiZe $string in php, don't ask why :D
I made some research and found good answers here, they really helped me.
But, in my case I want to start capitalizing every odd character (1,2,3...) in EVERY word.
For example, with my custom function i'm getting this result "TeSt eXaMpLe" and want to getting this "TeSt ExAmPlE".
See that in second example word "example" starts with capital "E"?
So, can anyone help me? : )
Well I would just make it an array and then put it back together again.
<?php
$str = "test example";
$str_implode = str_split($str);
$caps = true;
foreach($str_implode as $key=>$letter){
if($caps){
$out = strtoupper($letter);
if($out <> " ") //not a space character
$caps = false;
}
else{
$out = strtolower($letter);
$caps = true;
}
$str_implode[$key] = $out;
}
$str = implode('',$str_implode);
echo $str;
?>
Demo: http://codepad.org/j8uXM97o
I would use regex to do this, since it is concise and easy to do:
$str = 'I made some research and found good answers here, they really helped me.';
$str = preg_replace_callback('/(\w)(.?)/', 'altcase', $str);
echo $str;
function altcase($m){
return strtoupper($m[1]).$m[2];
}
Outputs: "I MaDe SoMe ReSeArCh AnD FoUnD GoOd AnSwErS HeRe, ThEy ReAlLy HeLpEd Me."
Example
Here's a one liner that should work.
preg_replace('/(\w)(.)?/e', "strtoupper('$1').strtolower('$2')", 'test example');
http://codepad.org/9LC3SzjC
Try:
function capitalize($string){
$return= "";
foreach(explode(" ",$string) as $w){
foreach(str_split($w) as $k=>$v) {
if(($k+1)%2!=0 && ctype_alpha($v)){
$return .= mb_strtoupper($v);
}else{
$return .= $v;
}
}
$return .= " ";
}
return $return;
}
echo capitalize("I want to CaPiTaLiZe string in php, don't ask why :D");
//I WaNt To CaPiTaLiZe StRiNg In PhP, DoN'T AsK WhY :D
Edited: Fixed the lack of special characters in the output.
This task can be performed without using capture groups -- just use ucfirst().
This is not built to process multibyte characters.
Grab a word character then, optionally, the next character. From the fullstring match, only change the case of the first character.
Code: (Demo) (or Demo)
$strings = [
"test string",
"lado lomidze needs a solution",
"I made some research and found 'good' answers here; they really helped me."
]; // if not already all lowercase, use strtolower()
var_export(preg_replace_callback('/\w.?/', function ($m) { return ucfirst($m[0]); }, $strings));
Output:
array (
0 => 'TeSt StRiNg',
1 => 'LaDo LoMiDzE NeEdS A SoLuTiOn',
2 => 'I MaDe SoMe ReSeArCh AnD FoUnD \'GoOd\' AnSwErS HeRe; ThEy ReAlLy HeLpEd Me.',
)
For other researchers, if you (more simply) just want to convert every other character to uppercase, you could use /..?/ in your pattern, but using regex for this case would be overkill. You could more efficiently use a for() loop and double-incrementation.
Code (Demo)
$string = "test string";
for ($i = 0, $len = strlen($string); $i < $len; $i += 2) {
$string[$i] = strtoupper($string[$i]);
}
echo $string;
// TeSt sTrInG
// ^-^-^-^-^-^-- strtoupper() was called here

Content Spinning using PHP?

Can anyone please help me? Say if I had this text or a smaller section stored in a variable, how can I randomise the words in the '{ }' ?
For example, the first one is "{important|essential|crucial|critical|vital|significant}" how can I make PHP choose one of those words randomly and then echo it? Thanks for helping me. :)
http://webarto.com/62/random-sentence-spinning-function
function get_random($matches)
{
$rand = array_rand($split = explode("|", $matches[1]));
return $split[$rand];
}
function show_randomized($str)
{
$new_str = preg_replace_callback('/\{([^{}]*)\}/im', "get_random", $str);
if ($new_str !== $str) $str = show_randomized($new_str);
return $str;
}
Applied on your text file... http://ideone.com/rkuf6
strip off initial and ending curly braces, you can use trim()
explode the resulting string on | using explode()
use array_rand() for the array you had in last step
Will not work with nested({a|x {b|c} y|z})!
function doStuff($from){
$to="";
while(($pos=strpos($from,'{'))!==false){
$to.=substr($from,0,$pos);
$from=substr($from,$pos);
$closepos=strpos($from,'}');
$arr=explode('|',substr($from,1,$closepos-1));
$to.=$arr[array_rand($arr)];
$from=substr($from,$closepos+1);
}
return $to.$from;
}

Categories