LabVIEW
Error in Skipping text in Scan from String
Solved!
Hi RTSLVU,
Thanks for replying.
I actually want to get the output in two separate output indicator. I am attaching the VI again for previous LV versions.
Could you give us some more examples, or maybe elaborate on what you're trying to do? If you need to extract a name, how are you determining what the name is? Just anything after "name is "? And I assume for age, you want the word "Age: " then a number, is that right?
Also, in my experience Scan from String is great for well-formatted data, but not so much for "flexible" data. It works great for values formatted into strings using Format Into String, but if you're looking for something more flexible it's not great.
For that, I'd recommend using Match Regular Expression. For example, finding the phrase "name is" then grabbing whatever comes afterwards is clunky (if not impossible) for Scan from String, but is trivial with Match Regular Expression.
If you remove the asterisk, wire the top indicator, and just use everything else you already had, it gets pretty close:
The asterisk was what was making it not work, since %* isn't a part of a valid format string.
If you do it without that, you get close ("Anna. " is the output, i.e. "Anna" plus a period and a space) but not exact.
However, the way you're doing it is still a bit weird. I do agree with the other poster that either Match Pattern or Match Regular Expression is more likely to get a good result.
Have you tried using the "Edit scan string" right-click option? It can make it a lot easier to see what you're doing:
For instance, right now, you are scanning for:
1. All characters until right before a capital A
2. All characters until right before a capital H
3. The exact string "Her Age: " is skipped
4. All characters in a row that look like a decimal number
If you change to something like Match Pattern, you can get better results. For instance, you could search for:
1. "name is", search to find this, discard it and anything before it.
2. "[A-Za-z]+", getting you the next set of letters, stopping at any space or punctuation
3. "Age:", searching for the exact string "Age:" and discarding it and anything else along the way
4. "[0-9]+", looking for any number of digits in a row, then stopping. Or, skip that and just convert the remaining text after the "Age:" search to a decimal number.
@Hk637 wrote:
If you change to something like Match Pattern, you can get better results. For instance, you could search for:
1. "name is", search to find this, discard it and anything before it.
2. "[A-Za-z]+", getting you the next set of letters, stopping at any space or punctuation
3. "Age:", searching for the exact string "Age:" and discarding it and anything else along the way
4. "[0-9]+", looking for any number of digits in a row, then stopping. Or, skip that and just convert the remaining text after the "Age:" search to a decimal number.
Would you please write just the formating here ? I still did not get it how these section works.
Thanks in advance.
Look here:
https://www.ni.com/docs/en-US/bundle/labview-api-ref/page/functions/match-pattern.html
There is a section there that lists the formatting. It is similar to regular expressions, but leaves out some options in order to let it run faster.
For example, I listed that the search for the name would be "[A-Za-z]+". The brackets mean "match any character in the brackets". The hyphens between A and Z and a and z mean "any character between these or including these". The plus at the end means "keep matching characters in a row until you reach one that doesn't match". So the complete meaning is essentially to search for one word, not caring about capitalization, but not starting until letters start and ending on the last character that is a letter, whether the character after that is a number, space, or punctuation.
"Match Pattern" basically splits the string you pass in into three sections:
- Before the match
- The part that matches
- After the match
By giving it patterns, you can find certain areas of text that meet broad patterns, then break them down as needed. This is unlike Scan from String, which has much less flexibility when it comes to certain things.
You can see above the implementation I described.
