AutoHotkey Community

[Joy2DWorld]

doc today:

[Joy2DWorld]

also, in case it has not already been updated:

current doc:

[Chris]

I don't agree that the current wordings are wrong. I wrote the docs colloquially and concisely to help users understand and use RegEx -- not to explain how regular expressions work internally.

On the other hand, I welcome second opinions: if PhiLho or anyone else thinks the current wordings are wrong or misleading, please let me know.

[corrupt]

I think it's misleading. If they are not the same then they are not the same.

[Titan]

I think Chris' definition is correct. For a more precise explanation see http://www.pcre.org/pcre.txt section titled 'PCRE MATCHING ALGORITHMS'.
_________________

RegExReplace("irc.freenode.net/autohotkey", "^(?=(.(?=[\0-r\[]*((?<=\.).))))(?:[c-\x73]{2,8}(\S))+((2)|\b[^\2-]){2}\D++$", "$u3$1$3$4$2")

[PhiLho]

You are arguing on the wording of an example. Which is correct.
Chris is right, his doc is just a quick reference, and cannot be exhaustive, and isn't a tutorial.
_________________
vPhiLho := RegExReplace("Philippe Lhoste", "^(\w{3})\w*\s+\b(\w{3})\w*$", "$1$2")

[corrupt]

The doc states that \ and ' escaped characters are the same. Do they always have the same effect in all cases? If they don't then the docs are not correct. If the docs are not meant to be specific to AutoHotkey's implementation of RegEx then they should be deleted and a link to an alternate site should be added instead. This is where people will look for specific differences between RegEx usage in other languages vs usage in AutoHotkey. For this reason, as much detail as possible should be provided for syntax/behaviour that is specific to AutoHotkey. I'm not sure I understand why the RegEx sections should be an exception. If anything, these are the areas where the documentation should go into more detail, not less detail. I wouldn't worry about the current documentation on RegEx being unintentionally interpreted as being any form of tutorial .

[PhiLho]

I was commenting on the first part. Indeed, Joy2DWorld pointed out `n and \n can be different in some cases.
Now, even if the RegEx quick reference might be handy, somehow I agree with corrupt, perhaps no reference, or one concentrating on specific gotchas, might be better. The Net is full of tutorials / references anyway.
On the other hand, it can be frustrating for newcomers to regexes, not to have at least the basis of the syntax of a function...
As always, a delicate balance.
_________________
vPhiLho := RegExReplace("Philippe Lhoste", "^(\w{3})\w*\s+\b(\w{3})\w*$", "$1$2")

[corrupt]

[PhiLho]

Well, your idea of "advanced documentation", listing all the obscure gotchas that Chris don't want to expand, but explains in the Bug Reports section (or elsewhere) is good.
The Wiki looks like a good place to do this.
That would be a iFAQ (infrequently asked questions)...
_________________
vPhiLho := RegExReplace("Philippe Lhoste", "^(\w{3})\w*\s+\b(\w{3})\w*$", "$1$2")

[Joy2DWorld]

Maybe this is helpful, and can suggest the proper doc wording:

`n is NOT a REGEX escape sequence.

`n is the equivalent of inserting a chr(13) into the relevant string.


\n is NOT equivalent to inserting a chr(13) into the relevant string.



the x) option affects the regex response to chr(13) in the needle string. Thus the distinction is important.


`n is an ESCAPE SEQUENCE FOR AHK, *NOT* for REGEX !



there is no shame in that!!! it's even kind of cool.... and the more info given.. allows users more power... more understanding and options.....


hope this helps.

oh and...

[Grumpy]

[Lexikos]