13298 Avoiding piecemeal numeral matching

[chrisjj]

How best may I get the Text Compare to avoid this piecemeal numeral matching
http://img31.imageshack.us/img31/1543/regionm.png
so to get Right 3 to align with Left 51 rather than 50?

I guessed this
http://img811.imageshack.us/img811/3073/grammaritem.png
but to no avail.

Temporary workaround: use Data Compare where possible.

[Aaron]

Hello,

Create a grammar that matches on just the "ID" part of your text. In this case, it looks like it is 14 characters long (regular expression basic grammar):
^\d{14}
Then set Everything Else as Unimportant (so only the ID is important).
Then mark the Alignment to Never align differences. This will then only align if the Important text (grammar for ID) is equal.

Leave Ignore Unimportant Differences disabled. This will still treat them as differences, though they will appear as blue.

[chrisjj]

Then set Everything Else as Unimportant (so only the ID is important).

Thanks, but some of everything else is important.

Leave Ignore Unimportant Differences disabled.

I would like that to stay available for its intended purpose.

Is there an alternative?

[Aaron]

The alignment can be altered by defining the ID as Important text and marking everything else as Unimportant (basically, creating a Key and aligning only if a perfect match, just like the Data Compare). If you leave Ignore Unimportant Text disabled, then the blue text is still marked as a difference, and can be browsed by using the Next Difference commands.

You can try marking other sections of text as Unimportant to see if it helps the alignment, but the "other important text" further down the line may throw things off. The only other re-course is to use manual alignments or the Data Compare.

[chrisjj]

OK, thanks Aaron.

[chrisjj]

The only other re-course is to use manual alignments

Please tell me how to use manual alignments to make the 00743216153824-1-2 intact in this case:



Thanks.

[Aaron]

Would you be able to email us a full example file (with just one line would work, right)?

We might be able to help create a grammar element which can influence the character alignment, but it might be an example of text alignment that could be improved and there isn't anything to help in the current version.

Please include a link back to this forum thread.

[chrisjj]

Would you be able to email us a full example file (with just one line would work, right)?

See my response at http://www.scootersoftware.com/vbull...4069#post34069 .

We might be able to help create a grammar element which can influence the character alignment, but it might be an example of text alignment that could be improved and there isn't anything to help in the current version.

I'd like the manual alignments you referred to:

The only other re-course is to use manual alignments

[Aaron]

That mentioning of "manual alignments" refers to when I thought you were trying to define an ID column and attempting to match on it, as per your "Line 3 with Line 51 rather than 50" in your original post.

Manual Alignments only work on entire lines, as also the Alignment Session Settings.

The Character by character alignment within each line is an entirely different process and does not have the same tools available.

[chrisjj]

Manual Alignments only work on entire lines...

The Character by character alignment ... does not have the same tools available.

OK, I see. Thanks.