Giter VIP home page Giter VIP logo

pagser's Introduction

Pagser

go-doc-img travis-img go-report-card-img Coverage Status

Pagser inspired by page parser

Pagser is a simple, extensible, configurable parse and deserialize html page to struct based on goquery and struct tags for golang crawler.

Contents

Install

go get -u github.com/foolin/pagser

Or get the specified version:

go get github.com/foolin/pagser@{version}

The {version} release list: https://github.com/foolin/pagser/releases

Features

  • Simple - Use golang struct tag syntax.
  • Easy - Easy use for your spider/crawler/colly application.
  • Extensible - Support for extension functions.
  • Struct tag grammar - Grammar is simple, like `pagser:"a->attr(href)"`.
  • Nested Structure - Support Nested Structure for node.
  • Configurable - Support configuration.
  • Implicit type conversion - Automatic implicit type conversion, Output result string convert to int, int64, float64...
  • GoQuery/Colly - Support all goquery project, such as go-colly.

Docs

See Pagser

Usage

package main

import (
	"encoding/json"
	"github.com/foolin/pagser"
	"log"
)

const rawPageHtml = `
<!doctype html>
<html>
<head>
    <meta charset="utf-8">
    <title>Pagser Title</title>
	<meta name="keywords" content="golang,pagser,goquery,html,page,parser,colly">
</head>

<body>
	<h1>H1 Pagser Example</h1>
	<div class="navlink">
		<div class="container">
			<ul class="clearfix">
				<li id=''><a href="/">Index</a></li>
				<li id='2'><a href="/list/web" title="web site">Web page</a></li>
				<li id='3'><a href="/list/pc" title="pc page">Pc Page</a></li>
				<li id='4'><a href="/list/mobile" title="mobile page">Mobile Page</a></li>
			</ul>
		</div>
	</div>
</body>
</html>
`

type PageData struct {
	Title    string   `pagser:"title"`
	Keywords []string `pagser:"meta[name='keywords']->attrSplit(content)"`
	H1       string   `pagser:"h1"`
	Navs     []struct {
		ID   int    `pagser:"->attrEmpty(id, -1)"`
		Name string `pagser:"a->text()"`
		Url  string `pagser:"a->attr(href)"`
	} `pagser:".navlink li"`
}

func main() {
	//New default config
	p := pagser.New()

	//data parser model
	var data PageData
	//parse html data
	err := p.Parse(&data, rawPageHtml)
	//check error
	if err != nil {
		log.Fatal(err)
	}

	//print data
	log.Printf("Page data json: \n-------------\n%v\n-------------\n", toJson(data))
}

func toJson(v interface{}) string {
	data, _ := json.MarshalIndent(v, "", "\t")
	return string(data)
}

Run output:


Page data json: 
-------------
{
	"Title": "Pagser Title",
	"Keywords": [
		"golang",
		"pagser",
		"goquery",
		"html",
		"page",
		"parser",
		"colly"
	],
	"H1": "H1 Pagser Example",
	"Navs": [
		{
			"ID": -1,
			"Name": "Index",
			"Url": "/"
		},
		{
			"ID": 2,
			"Name": "Web page",
			"Url": "/list/web"
		},
		{
			"ID": 3,
			"Name": "Pc Page",
			"Url": "/list/pc"
		},
		{
			"ID": 4,
			"Name": "Mobile Page",
			"Url": "/list/mobile"
		}
	]
}
-------------

Configuration

type Config struct {
	TagName    string //struct tag name, default is `pagser`
	FuncSymbol   string //Function symbol, default is `->`
	Debug        bool   //Debug mode, debug will print some log, default is `false`
}

Struct Tag Grammar

[goquery selector]->[function]

Example:

type ExamData struct {
	Herf string `pagser:".navLink li a->attr(href)"`
}

1.Struct tag name: pagser
2.goquery selector: .navLink li a
3.Function symbol: ->
4.Function name: attr
5.Function arguments: href

grammar

Functions

Builtin functions

  • text() get element text, return string, this is default function, if not define function in struct tag.
  • eachText() get each element text, return []string.
  • html() get element inner html, return string.
  • eachHtml() get each element inner html, return []string.
  • outerHtml() get element outer html, return string.
  • eachOutHtml() get each element outer html, return []string.
  • attr(name) get element attribute value, return string.
  • eachAttr() get each element attribute value, return []string.
  • attrSplit(name, sep) get attribute value and split by separator to array string.
  • attr('value') get element attribute value by name is value, return string, eg: will return "xxx".
  • textSplit(sep) get element text and split by separator to array string, return []string.
  • eachTextJoin(sep) get each element text and join to string, return string.
  • eq(index) reduces the set of matched elements to the one at the specified index, return Selection for nested struct.
  • ...

More builtin functions see docs: https://pkg.go.dev/github.com/foolin/pagser?tab=doc#BuiltinFunctions

Extension functions

  • Markdown() //convert html to markdown format.
  • UgcHtml() //sanitize html

Extensions function need register, like:

import "github.com/foolin/pagser/extensions/markdown"

p := pagser.New()

//Register Markdown
markdown.Register(p)

Custom function

Function interface

type CallFunc func(node *goquery.Selection, args ...string) (out interface{}, err error)

Define global function

//global function need call pagser.RegisterFunc("MyGlob", MyGlobalFunc) before use it.
// this global method must call pagser.RegisterFunc("MyGlob", MyGlobalFunc).
func MyGlobalFunc(node *goquery.Selection, args ...string) (out interface{}, err error) {
	return "Global-" + node.Text(), nil
}

type PageData struct{
  MyGlobalValue string    `pagser:"->MyGlob()"`
}

func main(){

    p := pagser.New()

    //Register global function `MyGlob`
    p.RegisterFunc("MyGlob", MyGlobalFunc)

    //Todo

    //data parser model
    var data PageData
    //parse html data
    err := p.Parse(&data, rawPageHtml)

    //...
}

Define struct function

type PageData struct{
  MyFuncValue int    `pagser:"->MyFunc()"`
}

// this method will auto call, not need register.
func (d PageData) MyFunc(node *goquery.Selection, args ...string) (out interface{}, err error) {
	return "Struct-" + node.Text(), nil
}


func main(){

    p := pagser.New()

    //Todo

    //data parser model
    var data PageData
    //parse html data
    err := p.Parse(&data, rawPageHtml)

    //...
}

Call Syntax

Note: all function arguments are string, single quotes are optional.

  1. Function call with no arguments

->fn()

  1. Function calls with one argument, and single quotes are optional

->fn(one)

->fn('one')

  1. Function calls with many arguments

->fn(one, two, three, ...)

->fn('one', 'two', 'three', ...)

  1. Function calls with single quotes and escape character

->fn('it\'s ok', 'two,xxx', 'three', ...)

Priority Order

Lookup function priority order:

struct method -> parent method -> ... -> global

More Examples

See advance example: https://github.com/foolin/pagser/tree/master/_examples/advance

Implicit type conversion

Automatic implicit type conversion, Output result string convert to int, int64, float64...

Support type:

  • bool
  • float32
  • float64
  • int
  • int32
  • int64
  • string
  • []bool
  • []float32
  • []float64
  • []int
  • []int32
  • []int64
  • []string

Examples

Crawl page example

package main

import (
	"encoding/json"
	"github.com/foolin/pagser"
	"log"
	"net/http"
)

type PageData struct {
	Title    string `pagser:"title"`
	RepoList []struct {
		Names       []string `pagser:"h1->textSplit('/', true)"`
		Description string   `pagser:"h1 + p"`
		Stars       string   `pagser:"a.muted-link->eqAndText(0)"`
		Repo        string   `pagser:"h1 a->attrConcat('href', 'https://github.com', $value, '?from=pagser')"`
	} `pagser:"article.Box-row"`
}

func main() {
	resp, err := http.Get("https://github.com/trending")
	if err != nil {
		log.Fatal(err)
	}
	defer resp.Body.Close()

	//New default config
	p := pagser.New()

	//data parser model
	var data PageData
	//parse html data
	err = p.ParseReader(&data, resp.Body)
	//check error
	if err != nil {
		log.Fatal(err)
	}

	//print data
	log.Printf("Page data json: \n-------------\n%v\n-------------\n", toJson(data))
}

func toJson(v interface{}) string {
	data, _ := json.MarshalIndent(v, "", "\t")
	return string(data)
}

Run output:


2020/04/25 12:26:04 Page data json: 
-------------
{
	"Title": "Trending  repositories on GitHub today · GitHub",
	"RepoList": [
		{
			"Names": [
				"pcottle",
				"learnGitBranching"
			],
			"Description": "An interactive git visualization to challenge and educate!",
			"Stars": "16,010",
			"Repo": "https://github.com/pcottle/learnGitBranching?from=pagser"
		},
		{
			"Names": [
				"jackfrued",
				"Python-100-Days"
			],
			"Description": "Python - 100天从新手到大师",
			"Stars": "83,484",
			"Repo": "https://github.com/jackfrued/Python-100-Days?from=pagser"
		},
		{
			"Names": [
				"brave",
				"brave-browser"
			],
			"Description": "Next generation Brave browser for macOS, Windows, Linux, Android.",
			"Stars": "5,963",
			"Repo": "https://github.com/brave/brave-browser?from=pagser"
		},
		{
			"Names": [
				"MicrosoftDocs",
				"azure-docs"
			],
			"Description": "Open source documentation of Microsoft Azure",
			"Stars": "3,798",
			"Repo": "https://github.com/MicrosoftDocs/azure-docs?from=pagser"
		},
		{
			"Names": [
				"ahmetb",
				"kubectx"
			],
			"Description": "Faster way to switch between clusters and namespaces in kubectl",
			"Stars": "6,979",
			"Repo": "https://github.com/ahmetb/kubectx?from=pagser"
		},

        //...        

		{
			"Names": [
				"serverless",
				"serverless"
			],
			"Description": "Serverless Framework – Build web, mobile and IoT applications with serverless architectures using AWS Lambda, Azure Functions, Google CloudFunctions \u0026 more! –",
			"Stars": "35,502",
			"Repo": "https://github.com/serverless/serverless?from=pagser"
		},
		{
			"Names": [
				"vuejs",
				"vite"
			],
			"Description": "Experimental no-bundle dev server for Vue SFCs",
			"Stars": "1,573",
			"Repo": "https://github.com/vuejs/vite?from=pagser"
		}
	]
}
-------------

Colly Example

Work with colly:

p := pagser.New()


// On every a element which has href attribute call callback
collector.OnHTML("body", func(e *colly.HTMLElement) {
	//data parser model
	var data PageData
	//parse html data
	err := p.ParseSelection(&data, e.Dom)

})

Dependencies

  • github.com/PuerkitoBio/goquery

  • github.com/spf13/cast

Extensions:

  • github.com/mattn/godown

  • github.com/microcosm-cc/bluemonday

pagser's People

Contributors

dependabot[bot] avatar foolin avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar

pagser's Issues

How to select an html element based on his text content?

What should be the selector for this kind of html code:

<h2>Firstname</h2>
<p>John</p>
<!--- some random html code, with random h2 tags -->
<h2>Lastname</h2>
<p>Doe</p>

I would like to fill this struct:

type Person struct {
    Firstname string `pagser:"h2[text=Firstname]+p"`
    Lastname  string `pagser:"h2[text=Lastname]+p"`
}

Support case insensitive meta attributes?

Some meta attributes can be capitalised, causing to have to duplicate them as below:

        Author1   []string `pagser:"meta[name='author']->attrSplit(content)"`
	Author2   []string `pagser:"meta[name='Author']->attrSplit(content)"`
	Desc1     []string `pagser:"meta[name='description']->attrSplit(content)"`
	Desc2     []string `pagser:"meta[name='Description']->attrSplit(content)"`
	Keywords1 []string `pagser:"meta[name='keywords']->attrSplit(content)"`
	Keywords2 []string `pagser:"meta[name='Keywords']->attrSplit(content)"`

Any suggestion for this?

go 1.21.0 / github.com/foolin/pagser v0.1.5

Thanks!

Why there are "each" versions of functions?

I think it could be simplified so that one could do:

type PageData struct {
	Link    string   `pagser:"a->attr(href)"`
	Links   []string `pagser:"a->attr(href)"`
}

The package could inspect the type associated with the tag and if it is slice, automatically assume "each" behavior. In general I think all functions should be returning slices, which then optionally are converted to an individual value (by default the first one, but you could set using ->Eq some other index).

Parse pairwise rows

I have HTML like:

<div class=".header">...</div>
<div class=".data">...</div>
<div class=".header">...</div>
<div class=".data">...</div>
<div class=".header">...</div>
<div class=".data">...</div>

It seems I am unable to parse this into a struct like:

struct {
    Rows []struct {
        Header string
        Data string
    }
}

`->eq()` should return a Selection if it's followed by something else?

type Item struct {
       Title       string `pagser:"td->eq(0)"`
       Image       string `pagser:"td a img->attr(src)"`
       Quote       string `pagser:"td->eq(3)"`
       Description string `pagser:"td->eq(4)"`
}

If I put td->eq(2) in the tag for Image, I get the text() for the full tr. But without the ->eq(2), it's possible I may end up with a different image. Ideally don't want to make a sub-struct just for this one field if possible.

(Happy to have a go at implementing this if it's actually possible.)

Support method-chaining for builtins to reduce need for custom-functions

Currently it appears for me that for each non-trivial case, you might need a custom-function (or I miss something) I would have plenty of use-cases for method-chaining on builtin-functions e.g.

type PageData struct {
	Releases []struct {
		Version string `pagser:"div.c-page-details > div->textSplit(' ', true)->eq(3)"`
	} `pagser:"section.c-panel div.o-grid div.u-mb-16"`
}

What do you think? Would this be technically doable?

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.