Thank you. Lingua::EN::Contraction is good enough for for my purpose. (My work consists in comparing two English sentences and check if they are the same, but first I needed to normalize them, so this module does this, which is more than OK)
Maybe the inverse ("what's" to "what is") is not possible because there are some words like person's which doesn't (always) mean person is. (I'm not a native English speaker, so I may be wrong about this)
| [reply] [d/l] [select] |
It's not just the confusability between "possessive 's vs. contracted 's", but also confusability between "'s contracted from is vs. 's contracted from has" -- for example:
- John's father is old. (possessive)
- John's old. (contracted from "John is")
- John's been sick. (contracted from "John has")
There's also some possible ambiguity with 'd, although this is relatively rare:
- He'd never done that before. (contracted from "He had")
- He'd never do that again. (contracted from "He would")
Apart from those cases, every other English contraction has a distinct full form.
If you're just trying to see whether two strings are identical except for contracted vs. uncontracted forms, it should work if you normalize both strings by applying contraction wherever possible, then seeing if they match.
I suppose there's a way to do it by expanding the contractions instead, but that's a lot more complicated.
| [reply] |
Shouldn't most of these s contractions (except the possessive case) be restricted to spoken English?
Cheers Rolf
( addicted to the Perl Programming Language)
| [reply] |
person's which doesn't (always) mean person is.
You are correct. The following 2 sentences are equivalent:
That person's code smells.
The code of that person smells.
| [reply] [d/l] |