regex - replace in java with a regex doesn't replace from left to right

Question

I'm trying to do a markdown parser. Basically, right now I just want to transform something like: "this is a *italic* text" into "this is a italic text".

I have basically this:

html_text = html_text.replaceAll("\*(.+)\*", "$1");

Here's the problem. If I use replaceAll, in a string such as "this *is* a *test* ok." it doesn't replace it 2 times, only once. Anyway, it can be fixed using replaceFirst() several times, so it's no biggie, still I don't get why it's replaceAll() and it doesn't replace all...

Anyway, using replaceFirst() I get one replacement, with the result as follows: "this is* a *test ok." I don't know much about regex, but I want it to replace from left to right not use some arbitrary rule. This is, the first block it should find is "*is*" not "*is* a *test*" giving as a result "this is a *test* ok.". Anyway, doing it another time the output is: "this is a test ok." which is wrong. A left to right would give the right one which is: "this is a test ok."

Since I don't know much about regex and I've been looking for a while I've decided to just ask you guys.

TL.DR.: I want replaceFirst() to replace using a left to right order, not some arbitrary one.

Edit: A solution is not allowing the * symbol inside the search. replaceAll() works fine. Still allowing the search from left to right would solve the problem much more easily, so is it possible?

score 0 · Accepted Answer

これを試して：

\*([^*]+)\*(アスタリスク以外を 1 つ以上取る)

またはこれ：

\*(.+?)\*(貪欲でない + のバージョンなので、アスタリスクまでの最小の一致が必要です)

あなた(.+)は貪欲なので、最後のアスタリスクに行き、あなたが望む以上のものをキャプチャします. 私の考えでは、貪欲でない正規表現の方がエレガントですが、どちらでも機能します。

regex - replace in java with a regex doesn't replace from left to right

1 に答える 1

Related

Reference