xml parsing script dying with "Premature end of script headers" error

[GazK]

I have been using an xml parsing script to parse a number of rss feeds
and return relevant results to a database. The script has worked well
for a couple of years, despite having very crude error-trapping (if it
finds an error in one of the xml files, the script stops). Recently, the
script has stopped working because one of the xml files is badly formed.

So I decided to rewrite the script with better error trapping; the
script should continue with the well-formed xml files and send me an
email telling me what happened.

The prototype script is failing with a "Premature end of script headers"
error. I am trying to work out if:

- this is a problem with my script, or
- a problem with the web server configuration

I have been over the code with as close as I have to a fine toothcomb,
and I can't see anything which would cause a problem.

Here is my code:

<?php

# code to parse multiple RSS .xml files, identify stories with keywords
in them, and enter those stories into DB

##### the first section of code is unchanged from the previous (working)
version

# SELECT, INSERT user privs for this page
$privs = "insert";

# create list of RSS feeds to parse
$feedsource = array(
"http://news.bbc.co.uk/rss/newsonline_uk_edition/uk/rss091.xml",
"http://newsrss.bbc.co.uk/rss/newsonline_uk_edition/england/rss.xml",
"http://newsrss.bbc.co.uk/rss/newsonline_uk_edition/scotland/rss.xml",
"http://newsrss.bbc.co.uk/rss/newsonline_uk_edition/wales/rss.xml",
"http://newsrss.bbc.co.uk/rss/newsonline_uk_edition/northern_ireland/rss.xml"
);

# these provide db connection and various query functions used below
include("../includes/config.inc");
include("../includes/sql.inc");

$insideitem = FALSE;
$tag = "";
$title = "";
$description = "";
$textdump = "";
$link = "";
$itemcount = FALSE;
$body1 = "";

function startElement($parser, $name, $attrs) {
global $insideitem, $tag, $title, $description, $link;
if ($insideitem) {
$tag = $name;
} elseif ($name == "ITEM") {
$insideitem = TRUE;
}
}

function endElement($parser, $name) {
global $insideitem, $tag, $title, $description, $link, $keywords, $feed;
$numkeywords = count($keywords);
$duplicate = FALSE;

if ($name == "ITEM") {
for($counter=0; $counter < $numkeywords; $counter++) {

# create regex which matches a whole word anywhere in a string
$regex = "/\b(" . $keywords[$counter] . ")\b/";

if(preg_match($regex, $title) || preg_match($regex, $description)) {
# if title or description string of parsed story matches the word
# get all news stories from db
$result = getTotalNews();
while ($row = mysql_fetch_array($result)) {
# loop through each existing news story
if($row[txtLink] == trim($link)) {
# if new link matches existing link, flag as duplicate
$duplicate = TRUE;
}
}
if($duplicate == FALSE) {
$itemcount = TRUE;
$datetime = date("Y-m-d H:i:s");
$title = trim(str_replace("'", "\'", $title));
$description = trim(str_replace("'", "\'", $description));
$link = trim($link);
$result = insertNews($title, $description, $link, $feed, $datetime);
$body1 .= "Item added: ";
$body1 .= $title;
$body1 .= " (link: ";
$body1 .= $link;
$body1 .= ") - ";
$body1 .= $description;
$body1 .= "\n\n";
mail("in*****@invalid.co.uk", "News Item Added: " . $title, $body1,
"FROM: ne******@railwaysarchive.co.uk");
} else {
$duplicate = FALSE;
}
}
}
$title = "";
$description = "";
$link = "";
$insideitem = FALSE;
}
}

function characterData($parser, $data) {
global $insideitem, $tag, $title, $description, $link;
if ($insideitem) {
switch ($tag) {
case "TITLE":
$title .= $data;
break;
case "DESCRIPTION":
$description .= $data;

break;
case "LINK":
$link .= $data;
break;
}
}
}

##### from here onwards the script has been rewritten

# initialise feed counter
$count = 0;
$passed = TRUE;
$body = "RSS parse results:\n";

foreach ($feedsource as $feed) {
# loop through each RSS file in turn
$xml_parser = xml_parser_create();
xml_set_element_handler($xml_parser, "startElement", "endElement");
xml_set_character_data_handler($xml_parser, "characterData");

if(fopen("$feed", "r")) {
# if file can be opened

$fp = fopen("$feed", "r");
$body .= "Success opening " . $feed . "\n";

while ($data = fread($fp, 4096)) {
# loop through feed contents

if(xml_parse($xml_parser, $data, feof($fp))) {
# success

$body .= "Success parsing " . $feed . "\n";

} else {
# fail

$body .= "Failed to parse " . $feed . ": XML error " .
xml_error_string(xml_get_error_code($xml_parser)) . " at line " .
xml_get_current_line_number($xml_parser) . "\n";
$passed = FALSE;

}

}

} else {
# failed to open file

$body .= "Failed to open " . $feed . "\n";
$passed = FALSE;

}

# close file
fclose($fp);
# free up xml parser
xml_parser_free($xml_parser);

}

if($passed) {
# if no errors
$passText = "no errors";
} else {
$passText = "ERRORS";
}

$subject = "Newsfeed report: " . $passText . " at " . date("d-m-Y G:i");

$to = "in*****@invalid.com";

mail($to, $subject, $body);

?>

[Curtis]

GazK wrote:

I have been using an xml parsing script to parse a number of rss feeds
and return relevant results to a database. The script has worked well
for a couple of years, despite having very crude error-trapping (if it
finds an error in one of the xml files, the script stops). Recently, the
script has stopped working because one of the xml files is badly formed.

So I decided to rewrite the script with better error trapping; the
script should continue with the well-formed xml files and send me an
email telling me what happened.

The prototype script is failing with a "Premature end of script headers"
error. I am trying to work out if:

- this is a problem with my script, or
- a problem with the web server configuration

I have been over the code with as close as I have to a fine toothcomb,
and I can't see anything which would cause a problem.

Here is my code:

<?php

[snipped some function and variable declarations]

##### from here onwards the script has been rewritten

# initialise feed counter
$count = 0;
$passed = TRUE;
$body = "RSS parse results:\n";

foreach ($feedsource as $feed) {
# loop through each RSS file in turn
$xml_parser = xml_parser_create();
xml_set_element_handler($xml_parser, "startElement", "endElement");
xml_set_character_data_handler($xml_parser, "characterData");

if(fopen("$feed", "r")) {
# if file can be opened

$fp = fopen("$feed", "r");

First of all, this is not a good way to test if fopen() succeeded.

if ($fp = fopen($feed, 'r')) {
...
}

Also, doing "$var" is a bad habit, because you may run into some
unexpected typing troubles:

var_dump("$int"); // string
var_dump($int); // integer

$body .= "Success opening " . $feed . "\n";

while ($data = fread($fp, 4096)) {
# loop through feed contents

Here's where your problem probably lies. You should not parse your RSS
data until you're finished collecting all the data. What happens when
your RSS data exceeds the buffer? The answer is that the
while-statement will start another iteration to get more data,
continuing in this manner until EOF is reached. This will cause
xml_parse() and other xml functions to attempt to operate on the
incomplete RSS feed.

Instead, use file_get_contents(), and eliminate the loop entirely.

Here's a small scale example of what *might* be happening to you, with
your current approach:

<?php
$rss = <<<EORSS
<?xml version="1.0"?>
<rss version="2.0">
<foo>
<bar>baz</bar>
</foo>
</rss>
EORSS;
$file = 'fread_test.rss';
$bufferTooSmall = ceil(strlen($rss) / 2);

// write the data - error checking removed for brevity
file_put_contents($file, $rss);

if ($fp = fopen($file, 'r')) {
$i = 1;
while ($data = fread($fp, $bufferTooSmall)) {
echo "Iteration $i:\n$data\n\n";
$i++;
}
fclose($fp);
}
?>

if(xml_parse($xml_parser, $data, feof($fp))) {
# success

$body .= "Success parsing " . $feed . "\n";

} else {
# fail

$body .= "Failed to parse " . $feed . ": XML error " .
xml_error_string(xml_get_error_code($xml_parser)) . " at line " .
xml_get_current_line_number($xml_parser) . "\n";
$passed = FALSE;

}

}

} else {
# failed to open file

$body .= "Failed to open " . $feed . "\n";
$passed = FALSE;

}

# close file
fclose($fp);
# free up xml parser
xml_parser_free($xml_parser);

}

if($passed) {
# if no errors
$passText = "no errors";
} else {
$passText = "ERRORS";
}

$subject = "Newsfeed report: " . $passText . " at " . date("d-m-Y G:i");

$to = "in*****@invalid.com";

mail($to, $subject, $body);

?>

--
Curtis
$eMail = str_replace('sig.invalid', 'gmail.com', $from);

[GazK]

Curtis wrote:

GazK wrote:

>I have been using an xml parsing script to parse a number of rss feeds
and return relevant results to a database. The script has worked well
for a couple of years, despite having very crude error-trapping (if it
finds an error in one of the xml files, the script stops). Recently,
the script has stopped working because one of the xml files is badly
formed.

So I decided to rewrite the script with better error trapping; the
script should continue with the well-formed xml files and send me an
email telling me what happened.

The prototype script is failing with a "Premature end of script
headers" error. I am trying to work out if:

- this is a problem with my script, or
- a problem with the web server configuration

I have been over the code with as close as I have to a fine toothcomb,
and I can't see anything which would cause a problem.

Here is my code:

<?php

[snipped some function and variable declarations]

>##### from here onwards the script has been rewritten

# initialise feed counter
$count = 0;
$passed = TRUE;
$body = "RSS parse results:\n";

foreach ($feedsource as $feed) {
# loop through each RSS file in turn
$xml_parser = xml_parser_create();
xml_set_element_handler($xml_parser, "startElement", "endElement");
xml_set_character_data_handler($xml_parser, "characterData");
if(fopen("$feed", "r")) {
# if file can be opened
$fp = fopen("$feed", "r");

First of all, this is not a good way to test if fopen() succeeded.

if ($fp = fopen($feed, 'r')) {
...
}

Also, doing "$var" is a bad habit, because you may run into some
unexpected typing troubles:

var_dump("$int"); // string
var_dump($int); // integer

> $body .= "Success opening " . $feed . "\n";
while ($data = fread($fp, 4096)) {
# loop through feed contents

Here's where your problem probably lies. You should not parse your RSS
data until you're finished collecting all the data. What happens when
your RSS data exceeds the buffer? The answer is that the while-statement
will start another iteration to get more data, continuing in this manner
until EOF is reached. This will cause xml_parse() and other xml
functions to attempt to operate on the incomplete RSS feed.

Instead, use file_get_contents(), and eliminate the loop entirely.

Here's a small scale example of what *might* be happening to you, with
your current approach:

<?php
$rss = <<<EORSS
<?xml version="1.0"?>
<rss version="2.0">
<foo>
<bar>baz</bar>
</foo>
</rss>
EORSS;
$file = 'fread_test.rss';
$bufferTooSmall = ceil(strlen($rss) / 2);

// write the data - error checking removed for brevity
file_put_contents($file, $rss);

if ($fp = fopen($file, 'r')) {
$i = 1;
while ($data = fread($fp, $bufferTooSmall)) {
echo "Iteration $i:\n$data\n\n";
$i++;
}
fclose($fp);
}
?>

> if(xml_parse($xml_parser, $data, feof($fp))) {
# success
$body .= "Success parsing " . $feed . "\n";
} else {
# fail
$body .= "Failed to parse " . $feed . ":
XML error " . xml_error_string(xml_get_error_code($xml_parser)) . " at
line " . xml_get_current_line_number($xml_parser) . "\n";
$passed = FALSE;
}
}

} else {
# failed to open file
$body .= "Failed to open " . $feed . "\n";
$passed = FALSE;
}

# close file
fclose($fp);
# free up xml parser
xml_parser_free($xml_parser);
}

if($passed) {
# if no errors
$passText = "no errors";
} else {
$passText = "ERRORS";
}

$subject = "Newsfeed report: " . $passText . " at " . date("d-m-Y G:i");

$to = "in*****@invalid.com";

mail($to, $subject, $body);

?>

Curtis, thanks for the assistance. I will give the file_get_contents()
approach a go - it looks much simpler in any case.

Garry

[GazK]

GazK wrote:

Curtis wrote:

>GazK wrote:

>>I have been using an xml parsing script to parse a number of rss
feeds and return relevant results to a database. The script has
worked well for a couple of years, despite having very crude
error-trapping (if it finds an error in one of the xml files, the
script stops). Recently, the script has stopped working because one
of the xml files is badly formed.

So I decided to rewrite the script with better error trapping; the
script should continue with the well-formed xml files and send me an
email telling me what happened.

The prototype script is failing with a "Premature end of script
headers" error. I am trying to work out if:

- this is a problem with my script, or
- a problem with the web server configuration

I have been over the code with as close as I have to a fine
toothcomb, and I can't see anything which would cause a problem.

Here is my code:

<?php

[snipped some function and variable declarations]

>>##### from here onwards the script has been rewritten

# initialise feed counter
$count = 0;
$passed = TRUE;
$body = "RSS parse results:\n";

foreach ($feedsource as $feed) {
# loop through each RSS file in turn
$xml_parser = xml_parser_create();
xml_set_element_handler($xml_parser, "startElement", "endElement");
xml_set_character_data_handler($xml_parser, "characterData");
if(fopen("$feed", "r")) {
# if file can be opened
$fp = fopen("$feed", "r");

First of all, this is not a good way to test if fopen() succeeded.

if ($fp = fopen($feed, 'r')) {
...
}

Also, doing "$var" is a bad habit, because you may run into some
unexpected typing troubles:

var_dump("$int"); // string
var_dump($int); // integer

>> $body .= "Success opening " . $feed . "\n";
while ($data = fread($fp, 4096)) {
# loop through feed contents

Here's where your problem probably lies. You should not parse your RSS
data until you're finished collecting all the data. What happens when
your RSS data exceeds the buffer? The answer is that the
while-statement will start another iteration to get more data,
continuing in this manner until EOF is reached. This will cause
xml_parse() and other xml functions to attempt to operate on the
incomplete RSS feed.

Instead, use file_get_contents(), and eliminate the loop entirely.

Here's a small scale example of what *might* be happening to you, with
your current approach:

<?php
$rss = <<<EORSS
<?xml version="1.0"?>
<rss version="2.0">
<foo>
<bar>baz</bar>
</foo>
</rss>
EORSS;
$file = 'fread_test.rss';
$bufferTooSmall = ceil(strlen($rss) / 2);

// write the data - error checking removed for brevity
file_put_contents($file, $rss);

if ($fp = fopen($file, 'r')) {
$i = 1;
while ($data = fread($fp, $bufferTooSmall)) {
echo "Iteration $i:\n$data\n\n";
$i++;
}
fclose($fp);
}
?>

>> if(xml_parse($xml_parser, $data, feof($fp))) {
# success
$body .= "Success parsing " . $feed . "\n";
} else {
# fail
$body .= "Failed to parse " . $feed . ":
XML error " . xml_error_string(xml_get_error_code($xml_parser)) . "
at line " . xml_get_current_line_number($xml_parser) . "\n";
$passed = FALSE;
}
}

} else {
# failed to open file
$body .= "Failed to open " . $feed . "\n";
$passed = FALSE;
}

# close file
fclose($fp);
# free up xml parser
xml_parser_free($xml_parser);
}

if($passed) {
# if no errors
$passText = "no errors";
} else {
$passText = "ERRORS";
}

$subject = "Newsfeed report: " . $passText . " at " . date("d-m-Y G:i");

$to = "in*****@invalid.com";

mail($to, $subject, $body);

?>

Curtis, thanks for the assistance. I will give the file_get_contents()
approach a go - it looks much simpler in any case.

Garry

Update - script is now morking much more reliably. Old script has been
binned. Thanks!