正则表达式向后提取字符串

Question

So I have this ugly string which I'm picking up off the wire:

{"feedtype": "playlist", "base_url": " http://feeds.xhis.com/rteavgen/player/ ", "feed_title": "Single Item Playlist", "feedid": "playlist", "alt_url": " http://www.xhis.com/player/#v=10322367 ", "platform": "iptv", "current_date": "2014-11-14T12:24:39.84167", "full_url": " http://feeds.xhis.com/rteavgen/player/playlist?type=iptv&showId=10343367 ", "shows": [{"itemid": 10332367, "showid": 11544367, "valid_start": "2014-11-13T21:37:39", "ispodcast": 0, "programmeid": 1, "BRINumber": "ih011305791", "duration": 2053247, "id": 10323367, "media:group": [{"rte:server": " http://vod.hds.xhis.com/hds-vod ", "medium": "video", "url": "/2014/1113/20141113-dumbydoozle_cl10344367_10344406_260_/manifest.f4m", "type": "video/mp4", "i

It's sorta JSONy - the string I get isn't always guaranteed to be complete, so I can't parse it. Also, the protocol could change.

Anyway, I'm trying to do this:

• find "manifest.f4m"
• extract the string: "/2014/1113/20141113-dumbydoozle_cl10344367_10344406_260_/manifest.f4m"

Once I have the location of manifest.f4m, I'm done.

---

So I'm trying to formulate a regex to do this reliably, but I'm having terrible trouble...

Here's my regex sofar:

/(?<=\/)manifest.f4m(?=("|\s))/

It matches "manifest.f4m" (with either a " or a space after it).

I'm a bit stuck with the lookbehind - I want to look back to the first "/" and extract the entire string that is pointed to by "url".

Though maybe there's a much better way of doing all this?

Answer 1

So I came up with this regex:

[-A-Za-z0-9+&@#\/%?=~_|!:,.;]+[-A-Za-z0-9+&@#\/%=~_|]manifest\.f4m(?=("|\s))

It seems to work rather well.

http://regex101.com/r/iT7vG2/2

Answer 2

您能否仅从url：部分开始并使用非捕获组，我想至少它会出现。我对照您的示例对其进行了测试，并且似乎可以正常工作

\b(?:url.+)(/.+manifest\.f4m)