How do I properly escape quotes inside HTML attributes?

Question

I have a drop down on a web page which is breaking when the value string contains a quote.

The value is "asd, but in the DOM it always appears as an empty string.

I have tried every way I know to escape the string properly, but to no avail.

<option value=""asd">test</option>
<option value="\"asd">test</option>
<option value="&quot;asd">test</option>
<option value="&#34;asd">test</option>

How do I render this on the page so the postback message contains the correct value?

Rennes

How are you generating the page? – Tutankhamen

What if you use single quotes? <option value='"asd'>test</option> – Wingate

I have to point out none of these answers say how to properly escape strings for use inside html attributes – Nightfall

@Nightfall That would depend on how the HTML was being generated. The question was about quotes, so technically the accepted answer answers the question asked. As to how to properly escape strings, I don't have a link handy for the general case, but in PHP you'd use htmlentities. – Lemuelah

possible duplicate of how to have quotation marks in html input values – Blanchblancha

possible duplicate of How to escape double quotes in title attribute – Civies

@Rennes Can you remember what browser you were testing with when you found that those bottom two examples resulted in ""? – Belgian

Answer 1

" is the correct way, the third of your tests:

<option value="&quot;asd">test</option>

You can see this working below, or on jsFiddle.

alert($("option")[0].value);

<script src="https://ajax.googleapis.com/ajax/libs/jquery/2.1.1/jquery.min.js"></script>
<select>
  <option value="&quot;asd">Test</option>
</select>

Alternatively, you can delimit the attribute value with single quotes:

<option value='"asd'>test</option>

Cerebritis

OP's fourth option, &#34;, is also a valid way to escape quotes. There's a benefit to using numeric html entities over named entities, in that named entities do not cover all characters, while numeric entities do. The full HTML4 list is at w3.org/TR/html4/sgml/entities.html . – Disembody

@atk: yes, " maps to the same character as ", but there's no benefit of using the numeric option here because " is a defined named entity. " is also easier to remember. – Cerebritis

I agree. In this particular case, it's easier to use &quot;. I intended only to point out the general case. – Disembody

I agree with atk's answer – Sorb

I tested the use of the single quotes for the html attribute, specifically in native IE8 (as this was the main html validation problem). This worked. I would suggest doing a replace in the attribute value instead though, as suggested by the first option with the named entity. Thanks! – Prefecture

My Question is reverse. Howto make value="&quot;a" displayed as &quot;a instead of "a? – Characteristically

@SIDU: change it to "a (replace the & with &) – Cerebritis

^ infinite loop – Luzern

Answer 2

Per HTML syntax, and even HTML5, the following are all valid options:

<option value="&quot;asd">test</option>
<option value="&#34;asd">test</option>
<option value='"asd'>test</option>
<option value='&quot;asd'>test</option>
<option value='&#34;asd'>test</option>
<option value=&quot;asd>test</option>
<option value=&#34;asd>test</option>

Note that if you are using XML syntax the quotes (single or double) are required.

Here's a jsfiddle showing all of the above working.

Viehmann

Answer 3

If you are using PHP, try calling htmlentities or htmlspecialchars function.

Plage

just using them may not be enough, try <option value='<?php echo htmlentities("' onmouseover='alert(123);' foo='"); ?>' /> - make sure you use it with ENT_QUOTES, this is safe: <option value='<?php echo htmlentities("' onmouseover='alert(123);' foo='", ENT_QUOTES); ?>' /> , but in addition to ENT_QUOTES you should also add ENT_SUBSTITUTE and ENT_DISALLOWED, personally i've used this wrapper for years: function hhb_tohtml(string $str):string { return htmlentities($str, ENT_QUOTES | ENT_HTML401 | ENT_SUBSTITUTE | ENT_DISALLOWED, 'UTF-8', true); } – Durwin

Don't understand this comment. – Americano

Answer 4

Another option is replacing double quotes with single quotes if you don't mind whatever it is. But I don't mention this one:

<option value='"asd'>test</option>

I mention this one:

<option value="'asd">test</option>

In my case I used this solution.

Runstadler

But if the value contains single & double quotes, this will fail – Watson

@Watson I said if the value contains double quotes, convert them to single quotes. If the value contains single quotes, then it will be no problem. – Runstadler

Answer 5

If you are using JavaScript and Lodash, then you can use _.escape(), which escapes ", ', <, >, and &.

Sidewheel

Answer 6

You really should only allow untrusted data into a whitelist of good attributes like: align, alink, alt, bgcolor, border, cellpadding, cellspacing, class, color, cols, colspan, coords, dir, face, height, hspace, ismap, lang, marginheight, marginwidth, multiple, nohref, noresize, noshade, nowrap, ref, rel, rev, rows, rowspan, scrolling, shape, span, summary, tabindex, title, usemap, valign, value, vlink, vspace, width

You really want to keep untrusted data out of javascript handlers as well as id or name attributes (they can clobber other elements in the DOM).

Also, if you are putting untrusted data into a SRC or HREF attribute, then its really a untrusted URL so you should validate the URL, make sure its NOT a javascript: URL, and then HTML entity encode.

More details on all of there here: https://www.owasp.org/index.php/Abridged_XSS_Prevention_Cheat_Sheet

Hypophosphite

I know this is late, but almost all of those attributes are deprecated in HTML4.01 and removed in 5. It may not matter now anyway, as there are better ways to protect yourself, just pointing it out. – Overtime

The question is asking about data with quote characters in it, not about untrusted data. – Proleg