package parser

import (
“regexp”
“strconv”
“ycProject/crawler/engine”
“ycProject/crawler/model”
)
var ageRe = regexp.MustCompile(<td><span class="label">年龄：</span>([\d]+)岁</td>)

var marriageRe = regexp.MustCompile(<td><span class="label">婚况：</span>([^<]+)</td>)

var IncomeRe = regexp.MustCompile(<td><span class="label">月收入：</span>([^<]+)元</td>)
var weightRe = regexp.MustCompile(<td><span class="label">体重：</span><span field="">(\d+)KG</span></td>)
var heightRe = regexp.MustCompile(<td><span class="label">身高：</span>([\d]+)CM</td>)
var genderRe = regexp.MustCompile(<td><span class="label">性别：</span><span field="">([^<]+)</span></td>)
var xingzuoRe = regexp.MustCompile( 星座：([^<]+)`)

var educationRe = regexp.MustCompile(<td><span class="label">学历：</span>([^<]+)</td>)
var occupationRe = regexp.MustCompile(<td><span class="label">职业：</span><span field="">([^<]+)</span></td>)
var hukouRe = regexp.MustCompile(<td><span class="label">籍贯：</span>([^<]+)</td>)
var houseRe = regexp.MustCompile(<td><span class="label">住房条件：</span><span field="">([^<]+)</span></td>)
var carRe = regexp.MustCompile(<td><span class="label">是否购车：</span><span field="">([^<]+)</span></td>)

func ParseProfile(contents []byte) engine.ParseResult {
profile := model.Profile{}
age, err := strconv.Atoi(
extractString(contents, ageRe))
if err == nil {
profile.Age = age
}
height, err := strconv.Atoi(extractString(contents, heightRe))
if err == nil {
profile.Height = height
}
weight, err := strconv.Atoi(extractString(contents, weightRe))
if err == nil {
profile.Weight = weight
}

profile.Income = extractString(
	contents, IncomeRe)
profile.Gender = extractString(
	contents, genderRe)
profile.Car = extractString(
	contents, carRe)
profile.Education = extractString(
	contents, educationRe)
profile.Hukou = extractString(
	contents, hukouRe)
profile.House = extractString(
	contents, houseRe)
profile.Marriage = extractString(
	contents, marriageRe)
profile.Occupation = extractString(
	contents, occupationRe)
profile.Xingzuo = extractString(
	contents, xingzuoRe)                           
result := engine.ParseResult{
	Items: []interface{} {profile},
}
return result

}

func extractString(contents []byte, re *regexp.Regexp) string {
match := re.FindSubmatch(contents)
if len(match) >= 2 {
return string(match[1])
} else {
return “”
}
}
这个是做测试结果