<?php
/*
* Reading a BZIP2 file can be tricky, and I never seen a complete example of
* code that account for any possible failure that may happen accessing a file
* in general, and decoding compressed data in this specific case.
* The example that follows is my attempt to address this gap.
* Some things that worth noting are:
* - Encoding/decoding errors must be detected with bzerrno().
* - bzopen() may fail returning FALSE if the file cannot be created or read,
* but succeeds also if the file is not properly encoded.
* - bzread() may fail returning FALSE if it fails reading from the source, but
* it returns the empty string on end of file and on encoding error.
* - bzread() may still return corrupted data with no error whatsoever until the
* BZIP2 algo encounters the first hash code, so data retrieved cannot be
* trusted until the very end of the file has been reached.
*/
// Safety first:
error_reporting(-1);
// On error, set $php_errormsg:
ini_set("track_errors", "1");
/**
* Reads and displays on stdout the content of a BZIP2 compressed file with
* full error detection.
* @param string $fn Filename.
* @return void
*/
function displaysBZIP2File($fn)
{
echo "Reading $fn:\n";
$bz = @bzopen($fn, "r");
if( $bz === FALSE ){
echo "ERROR: bzopen() failed: $php_errormsg\n";
return;
}
$errno = bzerrno($bz);
if( $errno != 0 ){
// May detect "DATA_ERROR_MAGIC" (not a BZIP2 file), or "DATA_ERROR"
// (BZIP2 decoding error) and maybe others BZIP2 errors too.
echo "ERROR: bzopen(): BZIP2 decoding failed: ", bzerrstr($bz), "\n";
@bzclose($bz);
return;
}
while(! feof($bz) ) {
$s = bzread($bz, 100);
if( $s === FALSE ){
echo "ERROR: bzread() failed: $php_errormsg\n";
@bzclose($bz);
return;
}
$errno = bzerrno($bz);
if( $errno != 0 ){
// May detect "DATA_ERROR" (BZIP2 decoding error) and maybe others
// BZIP2 errors too.
echo "ERROR: bzread(): BZIP2 decoding failed: ", bzerrstr($bz), "\n";
@bzclose($bz);
return;
}
echo "read: ", var_export($s, true), "\n";
}
if( ! bzclose($bz) ){
echo "ERROR: bzclose() failed: $php_errormsg\n";
}
}
// Target file:
$fn = "test.bz2";
// Test 1: writes and read a good BZIP2 file:
file_put_contents($fn, bzcompress("Content of the file."));
displaysBZIP2File($fn); // works ok.
// Test 2: invalid content, not a BZIP2 file:
file_put_contents($fn, "This ia plain text file, no compression at all!");
displaysBZIP2File($fn); // ERROR: bzread(): BZIP2 decoding failed: DATA_ERROR_MAGIC
// Test 3: creates a corrupted BZIP2 file:
$plain = str_repeat("Quite random string. ", 1000);
$compressed = bzcompress($plain);
$compressed_corrupted = $compressed;
$compressed_corrupted[(int)(strlen($compressed)/2)] = 'X'; // put random char in middle
file_put_contents($fn, $compressed_corrupted);
displaysBZIP2File($fn);
// Only after some Kbytes of garbage, it tells:
// ERROR: bzread(): BZIP2 decoding failed: DATA_ERROR
// Safe coding against headache, ever.
?>