python正規怎麼提取域名

Question

<script type="application/ld+json">{
    "@context": "http://schema.org",
    "@type": "SaleEvent",
    "name": "10% Off First Orders",
    "url": "https://www.myvouchercodes.co.uk/coggles",
    "image": "https://mvp.tribesgds.com/dyn/oh/Ow/ohOwXIWglMg/_/mQR5xLX5go8/m0Ys/coggles-logo.png",
    "startDate": "2017-02-17",
    "endDate": "2017-12-31",
    "location": {
        "@type": "Place",
        "name": "Coggles",
        "url": "coggles.co.uk",
        "address": "Coggles"
    },
    "description": "Get the top branded fashion items from Coggles at discounted prices. Apply this code and enjoy savings on your purchase.",
    "eventStatus": "EventScheduled"
}</script>

怎麼用python正則從這段腳本中提取coggles.co.uk域名呢,望各路高手指點顯示下身手...

Answer 1

正規實現的話只要保證你的標定/特徵是唯一的就好。但是"url"這個標誌又不是唯一的。這時候@prolifes的方法是很好的。

如果一定要正規實作呢，要用到零寬斷言（zero-width assertions），當然這個字的翻譯比較直，帶來很多誤解。它其實意思是指定位置的匹配，位置的寬度就是0嘛。

這裡我們可以看到我們所需的這個"url"在"location"裡面，可以以此為位置資訊。

程式碼如下：

re.search('(?<=location).+?"url": "([^"]+)"', string, re.DOTALL).group(1)

稍微解釋一下，
(?<=location)這個地方就是指前面得有location。後面有的話這樣寫：(?=location)
re.DOTALL這個是必須的，因為這些字串已經跨行了。他的作用是將.的字串匹配範圍擴大，包含換行符。
"([^"]+)"這個地方是我的習慣，[^"]意指所有非"的字符，這就匹配了雙引號中所有的字符串。

Answer 2

這是一段挺標準的json，粗一點，直接轉換成json

import json

str = '''
<script type="application/ld+json">{
    "@context": "http://schema.org",
    "@type": "SaleEvent",
    "name": "10% Off First Orders",
    "url": "https://www.myvouchercodes.co.uk/coggles",
    "image": "https://mvp.tribesgds.com/dyn/oh/Ow/ohOwXIWglMg/_/mQR5xLX5go8/m0Ys/coggles-logo.png",
    "startDate": "2017-02-17",
    "endDate": "2017-12-31",
    "location": {
        "@type": "Place",
        "name": "Coggles",
        "url": "coggles.co.uk",
        "address": "Coggles"
    },
    "description": "Get the top branded fashion items from Coggles at discounted prices. Apply this code and enjoy savings on your purchase.",
    "eventStatus": "EventScheduled"
}</script>
'''

d = json.loads(re.search('({[\s\S]*})', str).group(1))
print d['location']['url']