Add ability to set lib xml parameters

[bytestream]

Motivation

This library fails to parse documents with high nesting levels. Use of LIBXML_PARSEHUGE is required in such cases.

Resolves another unmerged PR #204
Closes #200

[bytestream]

@stof any thoughts on my three PRs? They should all be pretty harmless (new methods, previous functionality unchanged)

[Thorry84]

@stof Any update on this? I need this as well.

[Imran-imtiaz48]

Positive Aspects:

1. Constructor Logic: The constructor effectively checks for the existence of the CssSelectorConverter class before instantiating it. This is good practice for maintaining compatibility and preventing runtime errors when the class is not available.
2. Setter Method: The setLibXmlOptions() method follows a clean approach for setting the libXmlOptions parameter and returning the instance (self), which allows for method chaining. This is a useful design pattern that promotes cleaner code and better readability.

Suggestions for Improvement:

1. Type Hinting in Constructor: You could consider adding type hinting for the $cssConverter property by declaring it explicitly as a type or interface. It will improve code readability and IDE support.

   private ?CssSelectorConverter $cssConverter = null;

   This also initializes it as null by default, which can help avoid uninitialized property warnings.

2. LibXml Options Parameter Type: You could add a type declaration for the $options parameter in the setLibXmlOptions() method. Since PHP 7.0, scalar type declarations like int can be used:

   public function setLibXmlOptions(int $options): self

3. DocBlock Enhancements: Adding a DocBlock comment for the constructor would improve the documentation and clarify the purpose of the $cssConverter. It could describe what the CssSelectorConverter is and why it's being conditionally instantiated.

   /**
    * Constructor method.
    * Initializes CssSelectorConverter if available.
    */

4. Error Handling for Non-Existence of CssSelectorConverter: Although you check if the class exists, it might be helpful to either log a message or throw a custom exception if it doesn’t, depending on the context of the application.

Conclusion:

This code demonstrates good object-oriented practices. With the addition of type hinting and better documentation, it will be even more robust and maintainable. Consider adding error handling for potential edge cases such as the absence of the CssSelectorConverter class. Great work overall!

[Thorry84]

Did you just review a patch from 3.5 years ago using AI? Why?

[oliverklee]

@Thorry84 My best guess is that this might be related to Hacktoberfest, which started today Or because @Imran-imtiaz48 gamifies their GitHub contributions, which encourages reviews (and sounds like fun!). 🙂

Anyway, I welcome this review, even if the PR is a bit on the stale side and the PR probably needs an update. (We still should review this PR as humans, though.)

[bytestream]

I will update the PR if @stof explicitly says he will merge it