正则表达式匹配，但不包括

Question

I'm trying to use regular expression to only select the bolded (minus the *'s) based on the below string (carriage returns are in the string). Every string that I want to select it out of begins with ISA. I've been trying to work with this, (?<=ISA\\\\*)(\\w*) , but whenever I start adding additional characters to exclude in the (?<...) part, I no longer find any matches.

ISA*00* *00* *ZZ*SOME STRING *ZZ*99999999 *130605 * 2239*|*00501*000000001*0*P*>

REF*TJ*12345677*

REF*PQ*23432211

LX*1

Answer 1

The following should work:

^ISA(?:[^*]*[*][^*]*){8}\*(\d+)\*  

See http://regex101.com/r/lU8sC8/1

The way it works:

^ISA          — start of string has "ISA"
(?:           — non-capturing group
[^*]*[*][^*]* — zero or more non-asterisks, followed by asterisk, followed by non-asterisk
{8}           — eight of these
\*            — one more asterisk
(\d+)         — capture one or more digits
\*            — followed by another asterisk

Answer 2

If it's numbers between the 9th and 10th asterisk, then you can use:

ISA(?:[^*]*\*){9}(\d+)\*

as shown in this rubular

Answer 3

I would not use Regex to parse X12 EDI documents. A good place to start would be here

Looking at that link, you'll find that your field delimiter is ALWAYS at position 104 of the ISA line, your subdelimiter is always at position 105 of the ISA line, and your record delimiter is always at 106. (If they're not there, you don't have valid X12).

Using that, I'd do something like this (if you're looking specifically for field 7 of the ISA record):

var fieldDelimiter = line[103]; //where 'line' = your ISA Line, and remember 0 based index
var fieldSubDelim = line[104];
var recordDelimiter = line[105];

var fields = line.Split(fieldDelimiter);
var yourField = fields[7];

This also allows you to sort through your other records in a similar fashion, should you need to do so.