PhpOffice\PhpSpreadsheet\Reader\Html::load not working for html files smaller than 2048 bytes.

[victortodoran]

This is:

• a bug report
• a feature request
• not a usage question (ask them on https://stackoverflow.com/questions/tagged/phpspreadsheet or https://gitter.im/PHPOffice/PhpSpreadsheet)

What is the expected behavior?

The bug refers to the PhpOffice\PhpSpreadsheet\Reader\Html::canRead($filename).
To return true if a valid html file(path) is served.

What is the current behavior?

If the filesize is smaller than 2048 the value of the self::TEST_SAMPLE_SIZE the function
will always return false since inside the Html::readEnding() nothing is read by the
fread($this->fileHandle, $blockSize).

I suspect this is caused by the fact that Html::readBegining() is reading after EOF and
fseek($this->fileHandle, $size - $blockSize); does not have the expected behaviour.

Because of this issue I'm unable to load any html file that has a size smaller than 2048.

What are the steps to reproduce?

Please provide a Minimal, Complete, and Verifiable example of code that exhibits the issue without relying on an external Excel file or a web server:

<?php

require __DIR__ . '/vendor/autoload.php';

// add code that show the issue here...

$reader = new Html();
$spreadsheet = $reader->load("test.html"); // where test.html filesize < 2048


### Which versions of PhpSpreadsheet and PHP are affected?
Happend in:
PhpSpreadsheet latest (installed with composer require)
PHP 7.0.20

[PowerKiKi]

Do you have HTML files with actual valid data that are smaller than 2048 bytes ? If so what sizes are they ?

Would you have any suggestions on how to solve this ?

[victortodoran]

It's been a while so bare with me.

"Do you have HTML files with actual valid data that are smaller than 2048 bytes ?"

The error occurred in the following context:
I was parsing large html files table by table.
I was feeding the service one table at a time.
So we are not talking about files but input.

" If so what sizes are they ?"
The size is something dynamic. In my opinion it should not be a factor.

"Would you have any suggestions on how to solve this ?"

99% sure that all validation did was check if the file starts with "<" and ends with ">".
We can both agree that it does not really qualify for a validation of html format.
I can not figure out what value or problem that validation solves.
In my opinion you can drop the whole validation function entirely.

How I remember fixing the problem:

I ended up extending the provided class.
Droped the provided validation entirely.
Did a validation based on try catch.
After playing around with the input I've noticed some cases in which exceptions were thrown by the provided service for certain inputs.
I can not guarantee the provided fix suggestion as I no longer have access to the code it self, unfortunately I can not afford the hours to investigate this now.
Next time I'll raise an issue I'll leave a possible fix as well.

Thank you.

[kifni41]

Hi,
I also experiencing similar problem.

I'm using it for 'convert' html files containing table to Xlsx file.

I got "Invalid HTML file." exception. I check the html file using some validator, all says valid.

At first i didn't suspect it's related to the size, but after debugging it i notice the are static variable TEST_SAMPLE_SIZE on Reader/Html.php

`
private function readEnding()
{
$meta = stream_get_meta_data($this->fileHandle);
$filename = $meta['uri'];

    $size = filesize($filename);
    $blockSize = self::TEST_SAMPLE_SIZE;

    fseek($this->fileHandle, $size - $blockSize);

    return fread($this->fileHandle, $blockSize);

}
`

not sure why we need to set TEST_SAMPLE_SIZE to 2048, maybe there is certain reason for that.

But in my case, since i use read html that is generated by the data, for my latest case it can be as small as 1KB.
So i'm thingking maybe 512 would be a better value.

But maybe this will affect reader performance, i haven't think about performance effect.

Regards,

[alpha-and-omega]

I have fixed this long time ago in my local fork.
Patch:
https://gist.github.com/alpha-and-omega/c5e92fcc9f551bd6d312ce7dfbeae4a2#file-0001-fixes-fseek-bug-patch

[PowerKiKi]

Thank you @alpha-and-omega for your patch. I was able to complete it with unit tests and merge it.