makary/phpbb
1
0
Fork 0
mirror of https://github.com/phpbb/phpbb.git synced 2025-06-08 04:18:52 +00:00
Code Issues Projects Releases Packages Wiki Activity

[ticket/15403] Refactor get_context

Browse source 
PHPBB-15403
This commit is contained in:
Ruben Calvo 2024-07-13 15:34:37 +02:00
parent 89afa0cb5e
commit f41c8eef47
No known key found for this signature in database
2 changed files with 73 additions and 55 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

106
phpBB/includes/functions_content.php
View file

@ -325,83 +325,101 @@ function bump_topic_allowed($forum_id, $topic_bumped, $last_post_time, $topic_po
*
* @return string Context of the specified words separated by "..."
*/
function get_context($text, $words, $length = 400)
function get_context(string $text, array $words, int $length = 400): string
{
$text_length = utf8_strlen($text);
// Replace all spaces/invisible characters with single spaces
$text = preg_replace("/\s+/", ' ', $text);
if ($length <= 0)
{
return '...';
}
// we need to turn the entities back into their original form, to not cut the message in between them
$text = html_entity_decode($text);
// Replace all spaces/invisible characters with single spaces
$text = preg_replace("/\s+/u", ' ', $text);
$text_length = utf8_strlen($text);
// Get first occurrence of each word
$word_indizes = [];
$word_indexes = [];
foreach ($words as $word)
{
$pos = utf8_stripos($text, $word);
if ($pos !== false)
{
$word_indizes[$pos] = $word;
$word_indexes[$pos] = $word;
}
}
// If there are coincidences
if (!empty($word_indizes))
if (!empty($word_indexes))
{
ksort($word_indizes);
ksort($word_indexes);
$wordnum = count($word_indizes);
// Size of the fragment of text per word
$characters_per_word = (int) ($length / $wordnum);
$num_indexes = count($word_indexes);
$characters_per_word = (int) ($length / $num_indexes) + 2; // 2 to leave one character of margin at the sides to don't cut words
// Get text fragments
// Get text fragment indexes
$fragments = [];
$start = $end = 0;
foreach ($word_indizes as $indize => $word)
foreach ($word_indexes as $index => $word)
{
// Check if the next word can be inside the current fragment of text
if ($end + $characters_per_word + utf8_strlen($word) < $indize)
$word_length = utf8_strlen($word);
$start = max(0, min($text_length - 1 - $characters_per_word, (int) ($index + ($word_length / 2) - ($characters_per_word / 2))));
$end = $start + $characters_per_word;
// Check if we can merge this fragment into the previous fragment
$last_element = array_pop($fragments);
if ($last_element !== null)
{
$fragment = utf8_substr($text, $start, $end-$start);
[$prev_start, $prev_end] = $last_element;
if ($start != 0)
if ($prev_end + $characters_per_word >= $index + $word_length)
{
$fragment = '... ' . $fragment;
$start = $prev_start;
$end = $prev_end + $characters_per_word;
}
$fragments[] = $fragment;
$start = $indize - ($characters_per_word / 2);
// Start fragment at the beginning of a word
$end = $start = ($start > 0) ? (utf8_strpos($text, ' ', $start - 1) + 1) : 0;
}
$end += $characters_per_word;
// End fragment at the end of a word
$substring = utf8_substr($text, $start, $end - $start);
$end = $start + utf8_strrpos($substring, ' ');
}
$fragment = utf8_substr($text, $start, $end-$start);
if ($start != 0)
else
{
$fragment = '... ' . $fragment;
$fragments[] = $last_element;
}
if ($end < $text_length)
}
$fragments[] = [$start, $end];
}
}
else
{
$fragment .= ' ...';
// There is no coincidences, so we just create a fragment with the first $length characters
$fragments[] = [0, $length];
$end = $length;
}
// Get the last fragment
$fragments[] = $fragment;
$output = [];
foreach ($fragments as [$start, $end])
{
$fragment = utf8_substr($text, $start, $end - $start + 1);
return htmlentities(implode('', $fragments));
$offset = $start;
// Find the first valid alphanumeric character in the fragment to don't cut words
if ($start > 0)
{
preg_match('/[^a-zA-Z0-9][a-zA-Z0-9]/u', $fragment, $matches, PREG_OFFSET_CAPTURE);
$start = $offset + (int) $matches[0][1] + 1; // first valid alphanumeric character
}
return htmlentities($text_length >= $length + 3 ? utf8_substr($text, 0, $length) . ' ...' : $text);
// Find the last valid alphanumeric character in the fragment to don't cut words
if ($end < $text_length - 1)
{
preg_match_all('/[a-zA-Z0-9][^a-zA-Z0-9]/u', $fragment, $matches, PREG_OFFSET_CAPTURE);
$end = $offset + end($matches[0])[1]; // last valid alphanumeric character
}
$output[] = utf8_substr($text, $start, $end - $start + 1);
}
return htmlentities(implode(' ... ', $output)) . ($end < $text_length - 1 ? ' ...' : '');
}
/**

6
tests/functions_content/get_context_test.php
View file

@ -39,7 +39,7 @@ class phpbb_functions_content_get_context_test extends TestCase
'text' => 'This is a sample text containing several words, but none of them match the given words.',
'words' => ['nonexistent'],
'length' => 50,
'expected' => 'This is a sample text containing several words, bu ...',
'expected' => 'This is a sample text containing several words ...',
],
'desired length equal to text length' => [
'text' => 'Exact length text.',
@ -57,13 +57,13 @@ class phpbb_functions_content_get_context_test extends TestCase
'text' => 'This is a sample text containing &amp; and &lt; and &gt; entities.',
'words' => ['sample', 'entities'],
'length' => 50,
'expected' => 'This is a sample text containing &amp; and &lt; and ...',
'expected' => 'This is a sample text ... and &lt; and &gt; entities.',
],
'text with multiple spaces and special characters' => [
'text' => 'This is a sample text containing several words.',
'words' => ['sample', 'several'],
'length' => 50,
'expected' => 'This is a sample text containing several words ...',
'expected' => 'This is a sample text containing several words.',
],
'empty text' => [
'text' => '',