I have a problem with a PHP Code. This loop only executes the last regular expression in the file and when I change the sequence of expressions in the file and another expression becomes last, only this new last expression is executed.
foreach(file('general.txt') as $line) {
$text = preg_replace("/" . $line . "/", "", $text);
}
The file general.txt contains lines of regular expressions, everything tested. But in this loop, it doesn't work anymore.
Do you maybe know why this is like this? I have tried a lot, but didn't figure out why...
Thank you
Simon
you need to trim your lines as follows:
foreach(file('general.txt') as $line) {
$text = preg_replace("/" . trim($line) . "/", "", $text);
}
Instead of using the file() function, you can use fopen and stream_get_line that removes the newline sequence. To do that, you must know the newline sequence used in your pattern file. Exemple with a Windows newline sequence:
$fh = fopen('patterns.txt', 'r');
if ($fh) {
$nl = "\r\n";
while ( false !== $line = stream_get_line($fh, 2048, $nl) ) {
$str = preg_replace('/' . $line . '/', '', $str);
}
}
A significant advantage over trim: you can use patterns that start or end with whitespaces.
Related
I have problems with preg_replace. The $insert_marker text will not be recognized and is caused by "$". If I remove the 2 $ characters, it works. So what is the problem?
function insert_into_file($file_path, $insert_marker, $text, $after = true) {
$contents = file_get_contents($file_path);
$new_contents = preg_replace($insert_marker, ($after) ? '$0' . $text : $text . '$0', $contents);
return file_put_contents($file_path, $new_contents);
}
$file_path = ".htaccess";
$insert_marker = "/##-- $Id: _.htaccess 10934 2017-08-31 12:11:28Z serpent_driver $/";
$text = "\n##added text";
$num_bytes = insert_into_file($file_path, $insert_marker, $text, true);
if ($num_bytes === false) {
echo "Could not insert into file $file_path.";
} else {
echo "Insert successful!";
}
$ is a special character for regex (it means end of line), you have to escape it: \$ and because you are using a variable that contains the regex and pass it as a parameter of the function, you have to triple escape:
$insert_marker = "/##-- \\\$Id: _.htaccess 10934 2017-08-31 12:11:28Z serpent_driver \\\$/";
It's a little hard to understand your question, but I figured out that you mean the $Id part of $insert_marker is causing issues. You need to escape the $ with a backslash, like so: $insert_marker = "/##-- \$Id: _.htaccess 10934 2017-08-31 12:11:28Z serpent_driver $/";
I have a text file, called logs.txt, I'm trying to create a script in PHP which is supposed to check if a string which starts with foo|| exists. If it exists, it should be replaced with a specific string, otherwise a specific string will be added at the end of the file.
This is the code I tried to make:
<?php
function replaceInFile($what, $with, $file){
$buffer = "";
$fp = file($file);
foreach($fp as $line){
$buffer .= preg_replace("|".$what."[A-Za-z_.]*|", $what.$with, $line);
}
fclose($fp);
echo $buffer;
file_put_contents($file, $buffer);
}
replaceInFile("foo||", "foo||hello", "logs.txt");
?>
but it doesn't really do what I want. Can someone help me on fixing the code? Any help is appreciated.
This should work:
$newline = preg_replace('/^' . $what . '(.*)$/', $with . '${1}', $line, 1 , $count);
$buffer .= ($count == 1) ? $newline : $line . $with;
The common Delimiters is / no | specially if you have | in your search criteria. Than you have to create a Capturing Group this is between ( and ). Now you can use this Capturing Group and in your replacement as ${ + number of the Capturing Group + }. In your case you only have one.
I'm trying to make a function in PHP that can delete code within two tags from all .js file within one folder and all its subfolders. So far everything works except preg_replace(). This is my code:
<?php
deleteRealtimeTester('test');
function deleteRealtimeTester($folder_path)
{
foreach (glob($folder_path . '/*.js') as $file)
{
$string = file_get_contents($file);
$string = preg_replace('#//RealtimeTesterStart(.*?)//RealtimeTesterEnd#', 'test2', $string);
$file_open = fopen($file, 'wb');
fwrite($file_open, $string);
fclose($file_open);
}
$subfolders = array_filter(glob($folder_path . '/*'), 'is_dir');
if (sizeof($subfolders) > 0)
{
for ($i = 0; $i < sizeof($subfolders); $i++)
{
echo $subfolders[$i];
deleteRealtimeTester($subfolders[$i]);
}
}
else
{
return;
}
}
?>
As mentioned I want to delete everything inside these tags and the tags themselve:
//RealtimeTesterStart
//RealtimeTesterEnd
It is important that the tags contains the forward slashes and also that if a file contains multiple of these tags, only code from //RealtimeTesterStart to //RealtimeTesterEnd is deleted and not from //RealtimeTesterEnd to //RealtimeTesterStart.
I hope that someone can help me.
You could also change your regex to use the [\s\S] character set which can be used to match any character, including line breaks.
So have the following
preg_replace('#\/\/RealtimeTesterStart[\s\S]+\/\/RealtimeTesterEnd#', '', $string);
This would remove the contents of //RealtimeTesterStart to //RealtimeTesterEnd and the tags themselves.
I'm assuming that //RealtimeTesterStart, //RealtimeTesterEnd and the code in between are on different lines? In PCRE . does NOT match newlines. You need to use the s modifier ( and you don't need the () unless you need the captured text for the replacement):
#//RealtimeTesterStart.*?//RealtimeTesterEnd#s
Also, look at GLOB_ONLYDIR for glob instead of array_filter. Also, also, maybe file_put_contents instead of fopen etc.
Maybe something like:
foreach (glob($folder_path . '/*.js') as $file) {
$string = file_get_contents($file);
$string = preg_replace('#//RealtimeTesterStart.*?//RealtimeTesterEnd#s', 'test2', $string);
file_put_contents($file, $string);
}
foreach(glob($folder_path . '/*', GLOB_ONLYDIR) as $subfolder) {
deleteRealtimeTester($subfolder);
}
I have a php program that looks at a log file and prints it to a page (code below). I don't want the user of said website to be able to look at any line containing a /. I know I could use trim to delete certain characters, but is there a way to delete the entire line? For example, I want to keep something like "Hello" and delete something like /xx.xx.xx.xx connected. All the lines I wish to delete have the same common key, /. Peoples names in said log file have <>s around them, so I must use htmlspecialcharacters
$file = file_get_contents('/path/to/log', true);
$file = htmlspecialchars($file);
echo nl2br($file);
Thanks for your help!
EDIT:
Thanks for all of the answers, currently tinkering with them!
EDIT2:
final code:
<?php
$file = file_get_contents('/path/to/log', true);
// Separate by line
$lines = explode(PHP_EOL, $file);
foreach ($lines as $line) {
if (strpos($line, '/') === false) {
$line = htmlspecialchars($line . "\n");
echo nl2br($line);
}
}
?>
Do you mean, like this?
$file = file_get_contents('/path/to/log', true);
// Separate by line
$lines = explode(PHP_EOL, $file);
foreach ($lines as $line) {
if (strpos($line, '/') === false) {
// If the line doesn't contain a "/", echo it
echo $line . PHP_EOL;
}
}
For anyone wondering, PHP_EOL is the PHP constant for "end of line" and promotes consistency between different systems (Windows, UNIX, etc.).
If you are iterating through the file line by line you can check with preg_match if the line contains /character and skip the echo if it does. If not, first split them at new line and iterate over that array.
If you don't want to split the file you can probably use preg_replace with a regexp such as (^|\n).*/.*(\n|$) and replace with empty string.
Use the str_replace function -
http://php.net/manual/en/function.str-replace.php. Alternate solution (before escaping the special characters) -
/* pattern /\/[0-9]+\.[0-9]+\.[0-9]+\.[0-9]+\sconnected/ = /xx.xx.xx.xx connected */
/* pattern will be replaced with "newtext" */
$file = file_get_contents("/path/to/log", true);
$lines = explode("\n", $file);
foreach ($lines as $line)
$correctline = preg_replace( '/\/[0-9]+\.[0-9]+\.[0-9]+\.[0-9]+\sconnected/', 'newtext', $line );
echo $correctline;
}
<?php
$file = file_get_contents("/path/to/log", true);
$lines = explode("\n", $file);
foreach ($lines AS $num => $line)
{
if ( strpos($line, "/") === false ) // Line doesn't contain "/"
{
echo htmlspecialchars($line) . "\n";
}
}
?>
I'm using fputcsv in PHP to output a comma-delimited file of a database query. When opening the file in gedit in Ubuntu, it looks correct - each record has a line break (no visible line break characters, but you can tell each record is separated,and opening it in OpenOffice spreadsheet allows me to view the file correctly.)
However, we're sending these files on to a client on Windows, and on their systems, the file comes in as one big, long line. Opening it in Excel, it doesn't recognize multiple lines at all.
I've read several questions on here that are pretty similar, including this one, which includes a link to the really informative Great Newline Schism explanation.
Unfortunately, we can't just tell our clients to open the files in a "smarter" editor. They need to be able to open them in Excel. Is there any programmatic way to ensure that the correct newline characters are added so the file can be opened in a spreadsheet program on any OS?
I'm already using a custom function to force quotes around all values, since fputcsv is selective about it. I've tried doing something like this:
function my_fputcsv($handle, $fieldsarray, $delimiter = "~", $enclosure ='"'){
$glue = $enclosure . $delimiter . $enclosure;
return fwrite($handle, $enclosure . implode($glue,$fieldsarray) . $enclosure."\r\n");
}
But when the file is opened in a Windows text editor, it still shows up as a single long line.
// Writes an array to an open CSV file with a custom end of line.
//
// $fp: a seekable file pointer. Most file pointers are seekable,
// but some are not. example: fopen('php://output', 'w') is not seekable.
// $eol: probably one of "\r\n", "\n", or for super old macs: "\r"
function fputcsv_eol($fp, $array, $eol) {
fputcsv($fp, $array);
if("\n" != $eol && 0 === fseek($fp, -1, SEEK_CUR)) {
fwrite($fp, $eol);
}
}
This is an improved version of #John Douthat's great answer, preserving the possibility of using custom delimiters and enclosures and returning fputcsv's original output:
function fputcsv_eol($handle, $array, $delimiter = ',', $enclosure = '"', $eol = "\n") {
$return = fputcsv($handle, $array, $delimiter, $enclosure);
if($return !== FALSE && "\n" != $eol && 0 === fseek($handle, -1, SEEK_CUR)) {
fwrite($handle, $eol);
}
return $return;
}
Using the php function fputcsv writes only \n and cannot be customized. This makes the function worthless for microsoft environment although some packages will detect the linux newline also.
Still the benefits of fputcsv kept me digging into a solution to replace the newline character just before sending to the file. This can be done by streaming the fputcsv to the build in php temp stream first. Then adapt the newline character(s) to whatever you want and then save to file. Like this:
function getcsvline($list, $seperator, $enclosure, $newline = "" ){
$fp = fopen('php://temp', 'r+');
fputcsv($fp, $list, $seperator, $enclosure );
rewind($fp);
$line = fgets($fp);
if( $newline and $newline != "\n" ) {
if( $line[strlen($line)-2] != "\r" and $line[strlen($line)-1] == "\n") {
$line = substr_replace($line,"",-1) . $newline;
} else {
// return the line as is (literal string)
//die( 'original csv line is already \r\n style' );
}
}
return $line;
}
/* to call the function with the array $row and save to file with filehandle $fp */
$line = getcsvline( $row, ",", "\"", "\r\n" );
fwrite( $fp, $line);
As webbiedave pointed out (thx!) probably the cleanest way is to use a stream filter.
It is a bit more complex than other solutions, but even works on streams that are not editable after writing to them (like a download using $handle = fopen('php://output', 'w'); )
Here is my approach:
class StreamFilterNewlines extends php_user_filter {
function filter($in, $out, &$consumed, $closing) {
while ( $bucket = stream_bucket_make_writeable($in) ) {
$bucket->data = preg_replace('/([^\r])\n/', "$1\r\n", $bucket->data);
$consumed += $bucket->datalen;
stream_bucket_append($out, $bucket);
}
return PSFS_PASS_ON;
}
}
stream_filter_register("newlines", "StreamFilterNewlines");
stream_filter_append($handle, "newlines");
fputcsv($handle, $list, $seperator, $enclosure);
...
alternatively, you can output in native unix format (\n only) then run unix2dos on the resulting file to convert to \r\n in the appropriate places. Just be careful that your data contains no \n's . Also, I see you are using a default separator of ~ . try a default separator of \t .
I've been dealing with a similiar situation. Here's a solution I've found that outputs CSV files with windows friendly line-endings.
http://www.php.net/manual/en/function.fputcsv.php#90883
I wasn't able to use the since I'm trying to stream a file to the client and can't use the fseeks.
windows needs \r\n as the linebreak/carriage return combo in order to show separate lines.
I did eventually get an answer over at experts-exchange; here's what worked:
function my_fputcsv($handle, $fieldsarray, $delimiter = "~", $enclosure ='"'){
$glue = $enclosure . $delimiter . $enclosure;
return fwrite($handle, $enclosure . implode($glue,$fieldsarray) . $enclosure.PHP_EOL);
}
to be used in place of standard fputcsv.