Closed eminence closed 11 years ago
Hi Andrew,
Thanks for your example. There are other similar issues with Mergely's character-by-character comparison (LCS algorithm) for changes between lines. I'm working on a better? one that does word-by-word comparison. In most cases, I think this would produce better results. It might be ready in a couple of weeks.
Jamie
On 10 April 2013 04:12, Andrew Chin notifications@github.com wrote:
Mergely sometimes it a little too eager when trying to find common text. Here's a silly, made up example:
http://www.mergely.com/HA4vHsTS/
Line 5 (on the left, or line 4 on the right), the content is pretty much entirely different. However, mergely highlights it to show that "ar and "o" are in common between both lines. This in practice, is useless, since these are letters in the middle of works that happen to be common. I'd prefer it if mergely simply marked these entire lines as different (like is does for line 4left, 3right)
Is this a reasonable tweak to make?
— Reply to this email directly or view it on GitHubhttps://github.com/wickedest/Mergely/issues/5 .
Changed the algorithm to work word-by-word.
Mergely sometimes it a little too eager when trying to find common text. Here's a silly, made up example:
http://www.mergely.com/HA4vHsTS/
Line 5 (on the left, or line 4 on the right), the content is pretty much entirely different. However, mergely highlights it to show that "ar and "o" are in common between both lines. This in practice, is useless, since these are letters in the middle of works that happen to be common. I'd prefer it if mergely simply marked these entire lines as different (like is does for line 4left, 3right)
Is this a reasonable tweak to make?