Case sensitivig

[tomron]

Add an option to run it case insensitive

[intgr]

Unfortunately it's not as easy as it sounds. The rules from Wikipedia generally can deal with capitalized initial letters, but the replacements are case-sensitive, they cannot deal with allcaps words or mixed capitalization. For example this rule:

<Typo word="Activity" find="\b([Aa])c(?:itivit|tivt|tvit)(y|ies)\b" replace="$1ctivit$2"/>

In case-insensitive mode the word "ACTIVTY" would turn into "ActivitY"

What's the use case for this? How do you imagine this working?

[tomron]

My use case is that users' texts which don't really care about capitalization, I believe it is very common. Specially when you read data from social networks.

Another example - German native speakers which write in English and are not used to non capitalizing nouns.

[tomron]

BTW a possible way to implement it is to take it as an argument (say -i) and when loading the rules do -
r = regex.compile(find, regex.IGNORECASE)
It will replace it with case sensitive string but it will also identify strings with different capitalization.

[intgr]

I'm not convinced yet. Implementing it like that, it would be easily misunderstood by users and I suspect I would get bug reports of it doing the Wrong Thing with allcaps words, turning them to "ActivitIES" for example.

[tomron]

Ok, I'll fork it then.