Rendering bulgarian texts ends up as UTF-16 escaped

Hello,

I’m using thymeleaf 3.0.9.RELEASE and I’m rendering the following Bulgarian text: “Безплатна доставка на следващия ден” and it ends up rendered as \u0411\u0435\u0437\u043F\u043B\u0430\u0442\u043D\u0430 \u0434\u043E\u0441\u0442\u0430\u0432\u043A\u0430 \u043D\u0430 \u0441\u043B\u0435\u0434\u0432\u0430\u0449\u0438\u044F \u0434\u0435\u043D

Here’s my method:

    @Override
    public String parseContent(@Nonnull final String content, final Map<String, Object> params) {
        Assert.notNull(content, "Parameter content must be non-null");
        return templateEngine.process("<span th:inline=\"javascript\">" + content + "</span>", createSpringWebContext(params));
    }

Where content is

“Купете стоки на стойност повече от [[${currencyFormatter.print(promotion.activeSubtotalThreshold, locale)}]] за да получите [[${promotion.quantity}]] of [[${promotion.product.getName(locale)}]] БЕЗПЛАТНО!”

and promotion is a variable in the context with name in Bulgarian locale being “Безплатна доставка на следващия ден”.

I found out the problem is because when you are creating your JacksonStandardJavaScriptSerializer you are specifying this:

this.mapper.enable(JsonGenerator.Feature.ESCAPE_NON_ASCII);

making it like this:

this.mapper.disable(JsonGenerator.Feature.ESCAPE_NON_ASCII);

works perfectly fine. I will submit a PR.

[ptahchiev]

Hi @danielfernandez

thank you for the response - it works perfectly fine with no javascript inlining. I guess I was confused because this used to work with thymeleaf 2.x

Once again, thank you for your fast response.

[danielfernandez]

OK, but what I mean is that apparently you are rendering that text inside an HTML element (<span>), but you are using th:inline="javascript", which considers what is processed inside as JavaScript code and therefore applies Jackson and, in your case, JavaScript-escaping. Thus the unicode escapes, which are perfectly valid JavaScript escaped text.

So I’m confused about your use of th:inline="javascript" in a <span> element. If you are doing that only because you want your [[${...}]] expressions resolved, and seeing you are using Thymeleaf 3.0, you don’t really have to use inlining at all.

Assuming you are using the StringTemplateResolver, this:

return templateEngine.process("<span>" + content + "</span>", createSpringWebContext(params));

or even simply this:

return templateEngine.process(content, createSpringWebContext(params));

…should work for you. There will be escaping anyway for your Bulgarian characters, but it will be HTML escaping and not JavaScript, which I presume is what you are looking for.