Windows工具来解码文件中的HTML实体

Question

Is there a command line/batch script tool for Windows that can be used to decode HTML entitles like   , ℘ , and ‰ to readable UTF-8 text?

I found this web tool ( https://mothereff.in/html-entities ) that uses javascript that can do just this but I need this done from a Windows batch file. I know of the amazing JREPL.bat utility which incorporates javascript into windows command shell to make regex replacements in files. I just can't find a similar tool for HTML entities conversion.

Edit: To the bright coders out there, I hope you can write a batch tool that can perform HTML entities decoding/encoding to help me and the future readers looking for the same solution. Here are Github pages I think can be of use: https://github.com/mathiasbynens/he https://github.com/mathiasbynens/mothereff.in/tree/master/html-entities

Answer 1

You don't need extensive applications (like JREPL.bat or my own FindRepl.bat ) or complicated programs in order to perform a replacement as simple as this one. The small Batch file below is an example that performs a replacement of 3 HTML entities:

@set @a=0 // & cscript //nologo //E:JScript "%~F0" < input.txt & goto :EOF

var rep = new Array();
rep["&#xA9;"]   = "\u00A9";
rep["&#xD306;"] = "\uD306";
rep["&#x2603;"] = "\u2603";

var f = new ActiveXObject("Scripting.FileSystemObject").CreateTextFile("output.txt", true, true);
f.Write(WScript.Stdin.ReadAll().replace(/&#xA9;|&#xD306;|&#x2603;/g,function (A) {return rep[A]}));
f.Close();

input.txt:

Foo © bar 팆 baz ☃ qux

output.txt:

Foo © bar 팆 baz ☃ qux

You only need to add as many character equivalences as you want to convert...

Answer 2

It is trivial to incorporate JScript into a batch file, so you could easily write your own custom hybrid JScript/batch script that incorporates the he.js found at https://github.com/mathiasbynens/he .

But it is even simpler to use the JREPL.BAT tool that you already mentioned. You can use the /JLIB option to load the he.js code, thus making all of the he (html-entities) functionality accessible to JREPL.

Here is a trivial example that decodes test.txt, overwriting the original file.

jrepl "^.*" "he.decode($0)" /jlib "he.js" /f test.txt /o -

This isn't the most efficient way to do it, but it is probably plenty fast enough, and it sure is convenient.

Here is another example that encodes every character in test.txt (including newlines), writing the result to out.txt

jrepl "^[\s\S]*" "he.encode($0,{encodeEverything:true})" /m /j /jlib he\he.js /f test.txt /o out.txt

You should study all the documentation for both he and JREPL to discover all the possibilities.

The regex portion in the examples might seem to be more of a hindrance then help. But it is easy to envision how it might be useful to selectively encode only portions of your input text. Or you could use the JREPL /T option to use different encoding options for different sections of text.