What does a regex that begins with "*" actually do?-CodePudding

I'm trying to explain to someone why a regex that begins with the character "*" doesn't produce any results (and as far as I know never will). A simple example of the behavior I'm trying to explain:

mburr@mint-19:~/temp/grep-test$ touch foobar.jpg

#
# match using glob wildcard works as expected
#
mburr@mint-19:~/temp/grep-test$ ls *.jpg
foobar.jpg

#
# using regex doesn't match
#
mburr@mint-19:~/temp/grep-test$ ls | grep "*.jpg"
mburr@mint-19:~/temp/grep-test$

In one sense I understand and can explain what's going on: the "*" character in a regex is a quantifier that quantifies the pattern expression just before it. In this case there's no pattern before the "*" so there's nothing for it to quantify, therefore no match.

But what I can't explain is how it works in this regex. Why doesn't it throw an error if it's a senseless pattern? If it's not a senseless pattern, what could it match (if anything)?

All the searches I've done are answered with something that boils down to one or more of the following:

regexes aren't globs
a "*" isn't a wildcard like it is to the shell
just let the shell do the match

But none that explain why it isn't an error or what it could match, if anything.

CodePudding user response：

Well, some regex engines throw an error. E.g. the Javascript developer console in my browser complains if I enter /*/ but accepts /a*/ (JS has regex literals). The grep regex engine also recognizes that the * cannot possibly be the Kleene star and just treats it as an inert character (i.e. as if it were escaped).

$ echo "a" | grep "*"
# nothing
$ echo "*" | grep "*"
*

This is as specified in POSIX

9.3.3 BRE Special Characters

A BRE [basic regular expression, as used by flagless grep] special character has special properties in certain contexts. Outside those contexts, or when preceded by a <backslash>, such a character is a BRE that matches the special character itself. The BRE special characters and the contexts in which they have their special meaning are as follows:

...

*

The <asterisk> shall be special except when used:

...

As the first character of an entire BRE...

So in grep "*", since the asterisk appears as the first character, it is not considered special and simply matches one occurrence of itself.