RegEx用于在MATLAB中捕获科学数字

Question

I have the impression I would have a more elegant and compact way of parsing numbers using regexp. I know how to parse numbers in several formats,

str_expr = '[\-\+\d\.E]+';
fruit    = regexp(str_text, str_expr, 'match')

And I know how to get a look behind,

(?<=test)expr

But, what if I want to match several numbers after certain look-behind "test"? I usually do it in two regexp, one for the look behind, and the second for parsing the numbers. For example:

str_text = 'bla, ble,... bli, 1     -0.1243E-02    23.123E+03'

% pre-digest: look behind
str_reg1 = '(?<=bli,).*'
fruit    = regexp(str_text, str_reg1, 'match')

% parse numbers
str_reg2 = '[\-\+\d\.E]+'
fruit    = regexp(fruit{1}, str_reg2, 'match')

If I try to do it in a single step, I only catch the first number

% 1st try..
str_reg_try1 = '(?<=bli,)[\-\+\d\.E]+'
fruit        = regexp(fruit{1}, str_reg_try1, 'match')

% 2nd try..
str_reg_try2 = '(?<=bli,)([\-\+\d\.E]+)+'
fruit        = regexp(fruit{1}, str_reg_try1, 'tokens')

How do I solve this problem?

Answer 1

If my guess might be close, then here, we are looking to collect the digits. I'm positive that we might want to break the string to two or more parts before we pass it to RegEx engine:

bla, ble,... bli,
1     -0.1243E-02    23.123E+03

Then, in this case, we may not even need regular expression.

---

Or we might be wanting to just integrate some simple expressions such as:

(.+?bli,)|([0-9-E.\-+]+)

Demo

Also, I think the advice from bobble bubble in the comment is a much better choice ( link ):

(?:\G(?!^)|bli,)\s+\K[-+\d.E]+

RegEx

If this expression wasn't desired, it can be modified or changed in regex101.com .

RegEx Circuit

jex.im visualizes regular expressions: