Can't get title and subject from metadata

[borgir]

• PHP Version: 7.3.21
• PDFParser Version: 2.2.0

Description:

When trying to get this specific PDF metadata, I only get the author. Also need the title and subject which are there as you can see on the screenshot below:
https://prnt.sc/aOy-Sl6sSIUW

PDF input

https://www.alliancehealthplan.org/document-library/67033/

Expected output & actual output

Expected output:

Array
(
    [Author] => Tasha Jennings
    [Company] => Alliance Health
    [CreationDate] => 2022-04-21T11:25:35-04:00
    [Creator] => Acrobat PDFMaker 22 for PowerPoint
    [Pages] => 3
    [Title] => Enrollment and Client Update Overview
    [Subject] => Enrollment and Client Update Overview
)

Actual output:

Array
(
    [Author] => Tasha Jennings
    [Company] => Alliance Health
    [CreationDate] => 2022-04-21T11:25:35-04:00
    [Creator] => Acrobat PDFMaker 22 for PowerPoint
    [Pages] => 3
)

Code

    try {
        $parser = new \Smalot\PdfParser\Parser();
        $pdf = $parser->parseFile($file_path);
        $meta_data = $pdf->getDetails();
    } catch (Exception $e) {
        $field['value'] = 'Error: ' . $e->getMessage();
    }

Note: It works as expected, for example, with this file:
https://www.alliancehealthplan.org/document-library/70434/

Output:

Array
(
    [Author] => Hewett-RobinsonL
    [Company] => SRAHEC
    [CreationDate] => 2022-04-19T10:36:25-04:00
    [Creator] => Acrobat PDFMaker 22 for Word
    [Keywords] => 
    [ModDate] => 2022-04-19T10:36:26-04:00
    [Producer] => Adobe PDF Library 22.1.149
    [SourceModified] => 2022-04-19T14:36:12+00:00
    [Subject] => Social Determinants of Health: Now is the Time
    [Title] => Social Determinants of Health: Now is the Time
    [Pages] => 2
)

[k00ni]

Did this work before? If not, its a feature request, isn't it?

[borgir]

Hi @k00ni ! Thank you for the reply.
It works with some PDF files. With this in particular, it doesn't work.
I've update the question with and example that works correctly.
Thks.

[borgir]

Hi!
Any clue on what might be the problem?
Thks

[k00ni]

No sorry.

[GreyWyvern]

This is happening because there is a 'name' in the metadata that includes a hexadecimal-encoded space: "Document#20Type". The fix proposed by @philippze adds a position increment that loops and eventually bypasses the offending metadata name instead of parsing it.

The correct fix is to allow the '#' character in property names, and then convert them to the correct characters.

From Element.php

            ...
            if (!$only_values) {
                if (!preg_match('/\G\s*(?P<name>\/[A-Z#0-9\._]+)(?P<value>.*)/si', $content, $match, 0, $position)) {
                    break;
                } else {
                    $name = preg_replace_callback(
                        '/#(\d\d)/',
                        function($m) {
                            return \chr(base_convert($m[1], 16, 10));
                        },
                        ltrim($match['name'], '/')
                    );
                    $value = $match['value'];
                    $position = strpos($content, $value, $position + \strlen($match['name']));
                }
            } else {
            ...

The result of getDetails() will then be:

array(9) {
  ["Author"]=>
  string(14) "Tasha Jennings"
  ["Company"]=>
  string(15) "Alliance Health"
  ["CreationDate"]=>
  string(25) "2022-04-21T11:25:35-04:00"
  ["Creator"]=>
  string(34) "Acrobat PDFMaker 22 for PowerPoint"
  ["Document Type"]=>
  string(9) "Templates"
  ["ModDate"]=>
  string(25) "2022-04-21T11:25:36-04:00"
  ["Producer"]=>
  string(26) "Adobe PDF Library 22.1.149"
  ["Subject"]=>
  string(37) "Enrollment and Client Update Overview"
  ["Title"]=>
  string(37) "Enrollment and Client Update Overview"
}