我试图抓取node.js web应用程序中网站的描述。它似乎是完美的工作,但是node.js似乎与NCR字符有关(http://en.wikipedia.org/wiki/Numeric_character_reference)。我对链路抓取器的代码如下所示使用tagsoup时出现撇号错误
getInfo:(url) ->
errorMessage = 'Invalid Link'
request(url, (error, response, body)->
if (!error && response.statusCode == 200)
handler = new htmlparser.DefaultHandler((err, dom) ->
if (err)
res(error: errorMessage)
else
imgs = select(dom, 'img')
titletags = select(dom,'title')
descripTags = select(dom,'meta')
filteredTags = _.filter(descripTags,(tag) -> tag.attribs.name? && tag.attribs.name == 'description')
uri = response.request.uri.href
mapFunc =(imgSrc) ->
pattern = /^((http|https|ftp):\/\/)/
img = imgSrc.attribs.src
if (!pattern.test(img)) then uri.substring(0,uri.length-1) + img else img
res(
images: _.filter(_.map(imgs,mapFunc),(img)-> (img != '')) || []
title: titletags[0].children[0].raw || ''
description: if filteredTags.length != 0 then filteredTags[0].attribs.content || '' else ''
)
)
parser = new htmlparser.Parser(handler)
parser.parseComplete(body)
else
res(error: errorMessage)
)
作为一个例子,如果我把在以下网址抓住信息形式(http://www.zdnet.com),描述将ZDNet's breaking news, analysis, and research keeps business technology professionals in touch with the latest IT trends, issues and events.
。撇号是问题(被表示为'
)
我的问题是,为什么没有任何库正确解析有效的HTML NCR并将它们转换为字符串等效,并且如果没有办法解决这个问题,是否可以安全地使用其他库来替换所有出现的NCR?
我使用的库下面
request = require 'request'
htmlparser = require 'htmlparser'
select = require('soupselect').select
_ = require 'underscore'