Grig Software

Would it be feasible for "line compare" to optionally compare down to word-level instead of the current character-level?

That is, instead of highlighting the "k" and "v" in "strike" versus "strive", could it highlight the entire words? Sometimes words are accidentally similar but their meaning is different, so the users might like to see the entire words as different.

Hope that helps,
Dan Dascalescu

dandv wrote:That is, instead of highlighting the "k" and "v" in "strike" versus "strive", could it highlight the entire words? Sometimes words are accidentally similar but their meaning is different, so the users might like to see the entire words as different.

As I see it, this would make it much more difficult to spot the actual difference in the words! Back in the old WinDiff days, I had a hard time to find small differences like single letter typos, such as these.

HTH
TonHu

TonHu wrote:As I see it, this would make it much more difficult to spot the actual difference in the words! Back in the old WinDiff days, I had a hard time to find small differences like single letter typos, such as these.

HTH
TonHu

I entirely agree: this feature should be optional. ExamDiff Pro implements it as a choice of "Compare down to line/word/character level".

But when the user is interested in words, such as when comparing word lists, it could come in handy.

Yes, I know character level highlighting can be confusing sometimes, but I also find word-level not smart enough for daily usage. To find better alternative, I have added Options/"Lines Comparison"/"Min chars to match" - set it to 3-5 and you will not see single same characters highlighted. Let me know whether you find this feature useful. Adding "By word" differencing is not a problem (in fact it is done internally as a part of "by char" comparison),
but I'd like to remove extra options.

grigsoft wrote:To find better alternative, I have added Options/"Lines Comparison"/"Min chars to match" - set it to 3-5 and you will not see single same characters highlighted. Let me know whether you find this feature useful. Adding "By word" differencing is not a problem (in fact it is done internally as a part of "by char" comparison),
but I'd like to remove extra options.

Consider these three files:

<I have a processor.>

<I have a proposal.>

<I have a plum.>

Now I don't really care that "processor" and "proposal" happen to start the same way. What is interesting is that I have a DIFFERENT something. That's why I think word (not N-character) comparisons would be useful in this case.

Yes, but consider this:

<I have a processor.> vs <I have a proccessor.>

You will spend quite a lot of time trying to figure out what is the problem. Anyway, maybe I will add this feature as a shortcut for quick toggles on/off.