I was using this library to generate descriptions based on ages and found that 18 and 110 were incorrectly indefinitized:

"18-year-old woman".indefinitize   #=> "a 18-year-old woman"   (incorrect)
"180-year-old statue".indefinitize #=> "a 180-year-old statue" (correct)
"11-year-old house".indefinitize   #=> "an 11-year-old house"  (correct)
"110-year-old house".indefinitize  #=> "an 110-year-old house" (incorrect)

I edited the regex pattern for11to include \b word boundaries so that 110 could be excluded, and included a similar pattern for 18 to correct the cases above.

With this commit, the \b boundary would include punctuation for formatted numbers such as 18,000 but not 18000:

"18,000-year-old artifact".indefinitize  #=> "an 18,000-year-old artifact"
"18.000-year-old artifact".indefinitize  #=> "an 18.000-year-old artifact"
"18000-year-old artifact".indefinitize   #=> "a 18000-year-old artifact"

I saw pull request #16 and the regex pattern given in the July 27 comment looks like it could work, as well, but I'm unsure of the status of that PR. That pattern could be a better solution, but I don't know what cases the last block with digit boundaries would match for.

I'd value your feedback if this isn't an appropriate fix, and would be interested in a discussion of a better approach. Thanks!