[scott2500uk]

Let me set what I'm trying to do befor discussing the problem:

I'm using windows backup tool in windows server 2000 it outputs a log of what its backed up as a simple txt file: backup01.log

Im getting php to read through the file and grab bits of data that I need and place these bits in a db so that a user can come along and read the contents and be able to determin if backups where successful etc.

so a bit of my php code im using to read each line one at a time is:

PHP Code:

$myFile = "backup05.log";

$fh = fopen($myFile, 'r');

while($theData = fgets($fh)) {

    // each line of data here

} 

So here is the problem. For some reason php is seeing the log file as one long line of data not 100's of lines here is a bit of the log file:

Code:

Backup Status
Operation: Backup
Active backup destination: 4mm DDS
Media name: "ALPSERVER01-Tue-06-05-2007-11-00p"

Backup of "D: Disk 1"
Backup set #1 on media #1
Backup description: "daily Tue-06-05-2007-11-00p"
Backup Type: Normal

Backup started on 6/5/2007 at 11:02 PM.
Folder D:\FMR
Bm90.gil.zip                               6206332   3/28/2002     1:04 PM
cesdata.dmg.zip                            1248237   3/28/2002     1:03 PM
CESData.GIL.HOY                            7986451   3/28/2002     1:02 PM
CESData.GIL.zip                            1938728   3/28/2002     1:46 PM
Cesdata.ste.zip                            4177201   3/28/2002     1:08 PM
Cesdata.wor.zip                            1338525    6/4/2001    12:41 PM
fmrstrip.TXT                                    88  11/15/2000    11:08 AM

looking at the log file in notepad the data is clearly has new lines. put php is seeing the new lines something else. If I echo out the log in a web browser I get:

Code:

Backup Status 㰊牢>Operation: Backup 㰊牢>Active backup destination: 4mm DDS 㰊牢>Media name: "ALPSERVER01-Tue-06-05-2007-11-00p" 㰊牢> 㰊牢>Backup of "D: Disk 1" 㰊牢>Backup set #1 on media #1 㰊牢>Backup description: "daily Tue-06-05-2007-11-00p" 㰊牢>Backup Type: Normal 㰊牢> 㰊牢>Backup started on 6/5/2007 at 11:02 PM. 㰊牢>Folder D:\FMR 㰊牢>Bm90.gil.zip 6206332 3/28/2002 1:04 PM 㰊牢>cesdata.dmg.zip 1248237 3/28/2002 1:03 PM 㰊牢>CESData.GIL.HOY 7986451 3/28/2002 1:02 PM 㰊牢>CESData.GIL.zip 1938728 3/28/2002 1:46 PM 㰊牢>Cesdata.ste.zip 4177201 3/28/2002 1:08 PM 㰊牢>Cesdata.wor.zip 1338525 6/4/2001 12:41 PM 㰊牢>fmrstrip.TXT 88 11/15/2000 11:08 AM 㰊牢>fmrtrans.se 1034

looking at the source new lines are showing as ??>

I presume this is a charset problem here, but im really lost with what charset I should set the php file to and how I do this? Do I set it as:

PHP Code:

header ('Content-type: text/html; charset=iso-8859-1'); 

if so what char set should I set it to, what charset does windows backup use in its logs?

Any input is greatly apreciated

Cheers

[fyrestrtr]

What are the locale settings for Windows? That would be the source of the problem.

[scott2500uk]

Current lang settings are set to: Western Europe and United States. I presume then the charset is one of the 2:

charset=iso-8859-1 Western Alphabet
charset=windows-1252 Western Alphabet (Windows)

but setting them in the header both still see the log with wierd chars

[mab]

This is a newline problem not a language setting problem. It is likely that your log file has \n\r instead of \r\n. Check this out using an editor that displays the hex values of characters (using your original/actual log file.) It is also possible that if you opened and saved this file using an editor that the original newlines were altered to be what they are currently.

You can try turning on auto_detect_line_endings (this can be turned on in your script using an ini_set(...) statement) -

auto_detect_line_endings boolean
When turned on, PHP will examine the data read by fgets() and file() to see if it is using Unix, MS-Dos or Macintosh line-ending conventions.

Edit: Outputting this to a browser actually tells very little as \n or \r has no meaning to a browser. A <br> or <br /> is what causes a newline to be rendered in a browser. \n's or \r's in content output to a browser only format the "source" code of the web page.

[scott2500uk]

im pretty sure this isnt a line ending problem. You can read a file with the params b and t that check for different line endings. This didnt work in my case.

I continued to look into charsets and found the function utf8_encode and decode.

When running that on the file the document was now seeing new lines but now all the document was spaced out and was un readable from the source. I searched and searched and couldnt find a way around it. In the end I tried this just for the sake of it and it worked

PHP Code:

while($theData = urldecode(urlencode(fgets($fh)))) { 

Im sure this isnt the correct way about it. Is there any one that can explain why this works and if there is a better solution?

cheers

[scott2500uk]

^^ my last post is utter crap I linked to a wrong file somehopw when I was testing things and I linked to a fixed file so I thought what I had done had fixed it. So anyway Im back to my oridinal problem

I tried setting auto_detect_line_endings on but had no effect. I belive this only works if reading files from mac or you are on a mac and trying to read unix and windows.

I did a bit more investigating and its said that the windows adds null bytes at the beggining of each new line. Could that be that weird chars that Im seeing. Is there A way I can read through the file and replace these null bytes and replace with the correct new line?

Also I did a mb_detect_encoding() on each line and the output is ASCII does this tell me anything?

[scott2500uk]

its definatley the null byte char at the begining of the line.

When I do a url encode it comes up with %00
and when I run a ord() on the char I get 255

How would I do a str_replace to replace a null char?

[UnrealEd]

try this:

PHP Code:

$string = str_replace(chr(0), "", $string); 

or, since the ordinal of the character is 255:

PHP Code:

$string = str_replace(chr(255), "", $string); 

more on chr()

[scott2500uk]

thx UnrealEd ill give that a go tomorrow when I'm back at work thanks for all the help everyone

[scott2500uk]

I did the str_replace with chr 255 but that replaced all null bytes in the document making it look like I ran the code through utf8_encode where all the code was spaced out and was unreadable.

It turns out that the log is full of null bytes and all other wierd chars so I did an preg_replace to remove all unwanted ascii chars.

PHP Code:

$theData = preg_replace('/[\x00-\x0c\x0b\x0c\x0e-\x1f\x7f-\x9f]/i', '', $theData); 

Now the log looks great in the source and I can now start working on breaking down the log and grab the bits from it I want.

Cheers for all the input guys!!