Convert HTML to JSON with this .NET package.
This version has been archived and is available to download from the release page.
This version was abandoned. See branch jsonize-2.0.0 if you are interested.
Version 3.0.0 introduces breaking changes to the Jsonize project.
The project has been completely rewritten to decouple, simplify, and keep up with new standards.
Version 3.*.* will drop the JackWFinlay portion of the namespace, e.g. JackWFinlay.Jsonize is now Jsonize.
The project now splits the parsing and serialization into separate areas of concern,
as noted by the introduction of the IJsonizeParser and IJsonizeSerializer interfaces,
found in the Jsonize.Abstractions package.
These can be implemented by anyone,
but a brand new parser has been written using AngleSharp as its HTML engine.
This is supplied as the Jsonize.Parser package.
There is available a serialization helper that conforms to the same standard as previous versions,
using the System.Text.Json library.
There is also the Jsonize.Serializer.Json.Net package,
which implements a basic serializer wrapping the Newtonsoft.Json package with some useful functions.
Feel free to implement your own serializer.
The Jsonize package simply wraps the parser and serializer functions into one.
Jsonize no longer will grab any content from the internet for you;
you must supply the HTML as a string (or alternatively a Stream from 3.1.0) to Jsonizer class methods.
From version 3.1.0, you can now pass a CancellationToken
to allow cancellation of the parsing methods.
This is an optional parameter, so will not break any existing code.
From version 3.1.0 you can now directly pass in a Stream object for your HTML to be parsed.
The methods accepting a Stream are overloads on the same method names as before.
There's no real performance increase, but you'll no longer have to process the Stream to a string yourself first!
Version 3.0.0 introduced a major performance regression that could make the run time many hundred times worse
(compared to version 1.0.9)!.
Please upgrade to version 3.1.0 as 3.0.0 will be deprecated once the new package is pushed to NuGet.
You will also get some nice extra methods for working with Stream objects.
Get the NuGet packages:
| Package | Build Status | NuGet Version |
|---|---|---|
| Jsonize |  |
 |
| Jsonize.Abstractions | |
 |
| Jsonize.Parser | |
 |
| Jsonize.Serializer | |
 |
| Jsonize.Serializer.Json.Net | |
 |
An example to get the site "https://jackfinlay.com" as a JSON string:
private static async Task<string> Testy(string q = "")
{
using (var client = new HttpClient())
{
string url = @"https://jackfinlay.com";
HttpResponseMessage response = await client.GetAsync(url);
string html = await response.Content.ReadAsStringAsync();
// The use of the parameterless constructors will use default settings.
JsonizeParser parser = new JsonizeParser();
JsonizeSerializer serializer = new JsonizeSerializer();
Jsonizer jsonizer = new Jsonizer(parser, serializer);
return jsonizer.ParseToStringAsync();
}
}Alternatively, get the response as a JsonizeNode:
return jsonizer.ParseToJsonizeNodeAsync();You can control the output with a JsonizeParserConfiguration object, which is passed as a parameter to the constructor of the IJsonizeParser of choice:
JsonizeParserConfiguration parserConfiguration = new JsonizeParserConfiguration()
{
NullValueHandling = NullValueHandling.Ignore,
EmptyTextNodeHandling = EmptyTextNodeHandling.Ignore,
TextTrimHandling = TextTrimHandling.Trim,
ClassAttributeHandling = ClassAttributeHandling.Array
}
JsonizeConfiguration jsonizeConfiguration = new JsonizeConfiguration
{
Parser = new JsonizeParser(parserConfiguration),
Serializer = new JsonizeSerializer()
};
Jsonizer jsonizer = new Jsonizer(jsonizeConfiguration);Results are in the form:
{
"nodeType":"Node type e.g. Document, Element, or Comment",
"tag":"If node is Element this will display the tag e.g p, h1 ,div etc.",
"text":"If node is of type Text, this will display the text in that node.",
"attr":{
"name":"value",
"class": []
},
"children":[
{
"nodeType":"Node type e.g. Document, Element, or Comment",
"tag":"If node is Element this will display the tag e.g p, h1 ,div etc.",
"text":"If node is of type Text, this will display the text in that node.",
"child": []
}
]
}The following HTML:
<!DOCTYPE html>
<html>
<head>
<title>Jsonize</title>
</head>
<body>
<div id="parent" class="parent-div">
<div id="child1" class="child-div child1">Some Text</div>
<div id="child2" class="child-div child2">Some Text</div>
<div id="child3" class="child-div child3">Some Text</div>
</div>
</body>
</html>Becomes:
{
"nodeType": "Document",
"tag": null,
"text": null,
"attr": {},
"children": [
{
"nodeType": "DocumentType",
"tag": "html",
"text": null,
"attr": {},
"children": []
},
{
"nodeType": "Element",
"tag": "html",
"text": null,
"attr": {},
"children": [
{
"nodeType": "Element",
"tag": "head",
"text": null,
"attr": {},
"children": [
{
"nodeType": "Element",
"tag": "title",
"text": "Jsonize",
"attr": {},
"children": []
}
]
},
{
"nodeType": "Element",
"tag": "body",
"text": null,
"attr": {},
"children": [
{
"nodeType": "Element",
"tag": "div",
"text": null,
"attr": {
"id": "parent",
"class": [
"parent-div"
]
},
"children": [
{
"nodeType": "Element",
"tag": "div",
"text": "Some Text",
"attr": {
"id": "child1",
"class": [
"child-div",
"child1"
]
},
"children": []
},
{
"nodeType": "Element",
"tag": "div",
"text": "Some Text",
"attr": {
"id": "child2",
"class": [
"child-div",
"child2"
]
},
"children": []
},
{
"nodeType": "Element",
"tag": "div",
"text": "Some Text",
"attr": {
"id": "child3",
"class": [
"child-div",
"child3"
]
},
"children": []
}
]
}
]
}
]
}
]
}- Add Documentation.
MIT
See license.md for details.