Help understanding QLexers

[Phil Thompson]

On 24/01/2021 18:03, Andrea wrote:
> On 23/01/2021 11:54, kristof.mulier at telenet.be wrote:
>> Hi Andrea,
>> Did you have a look at the QScintilla (unofficial) documentation page?
>> Here is the page about lexers:
>> https://qscintilla.com/#syntax_highlighting
>> 
>> And here about your specific question:
>> https://qscintilla.com/#syntax_highlighting/subclass_qscilexercustom/all_about_styles
>> 
>> Hope it helps.
>> 
> 
> It helps to a degree.
> I seem to understand that LexerCustom is very powerful and requires
> subclasses to
> 1) define all the styles they need
> 2) apply them in styleText() when they are needed
> 
> and there is no leak of the definition of is a number or a comment
> outside this class.
> 
> But it leaves open a few questions
> 1) why are the other lexers not inheriting form custom, what do they
> do which is different?
> 2) all other lexers seems to offer much more usable interface based on
> keywords (sets of keywords)
> 
> And in a lot of cases, this is all it is needed. Add a couple of
> options to define what is a string (" vs '), what is a comment and
> people do not need to think how to implement styleText().
> 
> So, I still think the interface is upside down.
> Custom should have been the base class.
> Then a keyword-based implementation (from which all existing lexers 
> derive from)
> And if one needs a really special lexer, then she can derive and do
> something completely different.
> 
> But I think I am very confused about how to read the documentation.
> 
> This is what I need to understand.
> https://www.scintilla.org/ScintillaDoc.html#SCI_SETKEYWORDS

The QsciLexer classes are just wrappers around the configuration of the 
corresponding Scintilla lexers. Those lexers are completely independent 
and hardcode the meaning of the different values - hence the lack of 
consistency.

Phil