Default to Subword Mode?

[ejmr]

Enabling subword-mode causes commands like M-d and other word-based commands to treat CamelCaseNames as individual words. In modern PHP programming I often see camel case, so should PHP mode enable subword-mode by default? I would appreciate feedback from users of PHP mode, no matter how long you have used it, and PHP programmers even if you do not use Emacs. Is camel case common in your experience as it is mine?

[metaturso]

That would be a lovely addition. But perhaps having it enabled as default could break the Rule of Least Astonishment. In particular to those Emacs users who have fine-tuned their muscular memory to efficiently navigate words (:

What about having it disabled and then asking the user what action to take the first time he uses M-d or M-f to navigate a word? As for example the first time one [mistakenly] invokes narrow-to-region.

[ejmr]

That's a good point. I like the idea of word-related functions asking the user if they would prefer subword behavior or not when they first try to use them. Over the next week I will (try to) write some patches to implement that and then push it out as an experimental branch for PHP mode users to try and provide feedback. Although I still welcome any other ideas and feedback about the topic.

[TatriX]

I suggest adding 'config hints' to readme with examples and maybe some default templates.
So the user can just copy and paste the desired customizations to his .emacs file.
It's easier to implement and less intrusive as well.

[ejmr]

Describing the option to use subword-mode in the README is a good idea, whether it ever becomes a default or not. Thanks for the suggestion.

[ejmr]

I have updated the README to describe Subword Mode (e273209) and defined the C-c C-w key to toggle the mode for consistency with cc-mode, which PHP Mode relies on (5a1f99d).