XPath (Hanyar XML) harshe ne na bayyana W3C don adireshin da zaɓar sassa na takaddar XML, kuma har ma yana iya kewaya HTML. Abin da ya ke bayyana shi ne hanyar bayyana, kamar //a/@href, wanda ke tafiya cikin itacen takaddar don isa ga abubuwa, halaye ko rubutu. Kayan aikin scraping kamar Scrapy suna goyon bayan XPath don fitar da bayanai daga shafukan da aka sauke.

Sabon sigar, XPath 3.1, ta zama Shawarar W3C a ranar 21 ga Maris 2017. Yana aiki akan XML kuma, a wannan sigar, bayanan JSON.

Asalin Syntax na XPath

  • Hanya: //title yana zaɓar kowanne title abu a ko'ina cikin takaddar.
  • Rubutu: //title/text() yana dawo da rubutun da ke cikin waɗannan abubuwan.
  • Halaye: @ yana zaɓar halaye, don haka //a/@href yana dawo da wuraren haɗi.
  • Ka'idoji: Kwance kwance suna tace sakamakon, don haka //div[@class="price"] yana kiyaye kawai div abubuwan tare da wannan ajin.

XPath a cikin Web Scraping

Masu zaɓin Scrapy suna karɓar XPath da CSS, kuma Scrapy yana fassara tambayoyin CSS zuwa XPath a cikin gida ta amfani da ɗakin karatu na cssselect. Masu zaɓin sa suna amfani da ɗakin karatu na lxml a ƙarƙashin hoods. A cikin masu bincike, XPath na iya kewaya DOM na HTML da takardun SVG ta hanyar XPathExpression interface, a matsayin madadin hanyoyi kamar querySelectorAll().

Ayyukan XPath da XSLT

XPath an tsara shi don a haɗa shi cikin harshe mai masauki kamar XSLT. W3C yana bayyana ɗakin karatu guda ɗaya, a cikin fn namespace, don XPath, XQuery da XSLT. Ayyukan da suka dace a cikin scraping sun haɗa da:

  • contains() da starts-with(): Daidaiton substring.
  • normalize-space() da string-length(): Aiki akan ƙimar rubutu.
  • count(): Yana ƙidaya abubuwa.

Misali, //a[contains(@href, "product")] yana kiyaye kawai hanyoyin haɗi waɗanda URL ɗin su ke dauke da "samfur".

Sharuɗɗan da suka shafi