Octave - 使用正则表达式在字符串中查找单词

Question

In Octave, I am finding words ending with only whitespaces, or either a comma or a period followed by whitespace(s).

The following is my code:

str = 'Hello, I am kjd#(@*#@m, aa.aa.aa.aa. It was nice meeting you.';
regexp(str, "\[a-zA-Z]+\[,.]?\s+", 'match')

This should return the words Hello , I , am , It , was , nice , meeting , you . However, it only returns was . I'm having a hard time figuring this out.

I've also tried tried this answer: https://stackoverflow.com/a/29174222/6213337 , but it returns ans = {}(1x0) .

Any ideas? Thanks.

Answer 1

Matlab uses PCRE regex flavor, thus, the regex pattern you need can be short and compact and quite comprehensive:

str = 'Hello, I am kjd#(@*#@m, aa.aa.aa.aa. It was nice meeting you.';
regexp(str, "(?<!\\S)\\p{L}++(?!\\p{P}\\S)", 'match')
print match

See the regex and IDEONE demos.

The regex matches:

• (?<!\\S) - check if there is no non-whitespace character before the current location in string, and if there is not, go on matching....
• \\p{L}++ - any 1+ letters (possessively, not allowing backtracking, thus, the next check will only be performed once after the last letter matched) that are NOT followed with...
• (?!\\p{P}\\S) - any punctuation and then a non-whitespace ( (?!...) is a negative lookahead that fails a match if its subpattern matches to the right of the current location in the string).

Answer 2

Try this

str = 'Hello, I am kjd#(@*#@m, aa.aa.aa.aa. It was nice meeting you.';
regexp(str, "(?:^|\\s+)([a-zA-Z]+)(?=[,.]?(?:$|\\s))", 'matches')