avoid casting CssNumber to CssLength or other. Use the CssCheckableValue interface for that.
Also CssColor and potentially other values needs to handle calc() and attr()

Nested rules don't seem to validate correctly. Consider the following example:

@media print {
  /* hide navigation controls when printing */
  #navigation { display: none }
  @media (max-width: 12cm) {
    /* keep notes in flow when printing to narrow pages */
    .note { float: none }
  }
}

The validator reports ambiguous errors:

7		Parse Error @media (max-width: 12cm) { /* keep notes in flow when printing to narrow pages */ .note { float: none } }
8		Parse Error }

Has anyone else encountered problems with the CSS validator (https://jigsaw.w3.org/css-validator/validator) suddenly marking pages that validated perfectly a week or more ago with invalid code? The invalid code marked by the validator seems to involve the clear property and its both value.

I did an experiment this morning by entering in these styles in the CSS validator on 11/4/15:

p {
clear: left;
}
Result: validated

p {
clear: right;
}
Result: validated

p {
clear: both;
}
Result: Error!

According to last W3 recommendation, thos values are valid:
https://www.w3.org/TR/SVG/painting.html#SpecifyingPaint

If the CSS contains the strings "<style>" and "</style>", then the validator attempts to parse the whole file as HTML, even if type was manually selected to be CSS.

Reduced testcase:

.test::before {
   content: "<style>";
}
.test::after {
   content: "</style>";
}

Workaround: split up tags into multiple strings

.test::before {
   content: "<" "style>";
}
.test::after {
   content: "<" "/style>";
}

Currently it clutters up the root directly something fierce.

On validation I get the error:

Property user-select doesn't exist : none

According to this this spec, it is valid.

I'm referring to the circled field, which is currently blank for the css-validator repository:

Many sites these days unfortunately still need to accommodate IE9 (or, $DEITY forbid, IE8), and authors must frequently resort to CSS browser hacks to work around bugs/deficiencies in these browsers. However, they would still like to benefit from the CSS Validator's valuable error detection facilities when working on the CSS for their sites. I therefore propose the validator should add a couple options (disabled by default, and perhaps hidden in an "Advanced" pane) that would squelch errors about a couple of the simplest yet most common CSS hacks used to deal with these browsers. Specifically:

• An option to squelch errors about IE<=9's proprietary filter: CSS property (not to be confused with the same-named standard property from Filter Effects Module Level 1)
• An option to squelch errors about the \9 CSS property-value hack for IE6-8

These currently lead to nasty "Value Error"s or "Parse Error"s, as opposed to more easily manageable "Property foo doesn't exist", "Feature foo doesn't exist for media bar", or "Sorry, the at-rule @-foo is not implemented." errors.

There is precedent for adding (small, limited) old-IE-specific accommodations to Web validators/specs; see for example validator/validator#12 & https://www.w3.org/Bugs/Public/show_bug.cgi?id=27091

Could we please have an option to treat unknown properties (not detected as vendor extensions) as warnings instead of errors? After all they are perfectly valid, they'll just be ignored if not supported, but warnings are useful in case there's a misspelling.

The following CSS:

h3 {
visibility: none;
adobe-hyphenate: none;
hyphens: none;
-epub-hyphenate: none;
}

should only give an error for visibility: none (unvalid value for valid property), the rest are at most warnings, regardless of the CSS level.

Hello,

As can be seen at https://caniuse.com/#feat=css-variables, today CSS variables ("custom properties") is a widely supported feature in all modern standards-compliant browsers. (Of course, Internet Explorer is not in any way modern or standards-compliant.)

I am proudly coding all my sites by hand in raw HTML5/CSS3 according to the W3C standards and recently started using CSS variables heavily (a much-awaited feature!), which results in dozens of validation "errors" whenever someone clicks one of my ✔ CSS3 links in the footer of my pages, which links to the validation results for that page using your validator.

I take pride in having no or very few validation errors in my pages. Having to explain to users that these "errors" are not actual errors and are more akin to a bug in the validator is pretty awkward. So please finally add support for CSS variables, a widely supported and crucial feature.

(edit: I just noticed that the CSS variables specification is "only" at the Candidate Recommendation stage (since 3 december 2015), but still...)

When a redirect is encountered, HTTPUrl attempt to get a connection to the path referenced in the Location header. Detection of a relative URL is required, else connections are attempted to invalid urls such as http:/some.relative.path.

This can be reproduced by initiating a validation request to https://www.usaa.com/inet/pages/bank_loan_auto_main.

Results will be displayed as:
w3c CSS Validator Error
Target: https://www.usaa.com/inet/pages/bank_loan_auto_main

I/O Error: Unable to contact target server :80 after 3 tries.

The CSS Validator should support Grid display properties like:

        @supports (display: grid){
            body{
                display: grid;
                grid-template-rows: 100px 1fr 40px;
                gird-template-columns: 170px 1fr;
                grid-template-areas: 
                  "abc"
                  "def"
                  "hij";
            }
            header {
                grid-area: header;
            }
        }

I am surprised nobody else has reported this yet but the validator is showing calc(25% - 4px); and margin-left -4px; values as invalid css when I am assured it is not invalid and it is just the validator that has not been updated in ages. Almost everyone on the planet using a certified wordpress theme the latest ones will be having this problem.

hey there,

thanks for this service! I have a small problem:

with
@charset "utf-8";

.⛺e{
   font-family:Raleway,sans-serif;
   color:#d9b75b;
}

I got inside "direct-input":

but same inside uri check

please help, thx!

greetings
crazyx13th

I received following error on https://jigsaw.w3.org/css-validator/validator?uri=https://www.ixigo.com

URI : https://edge.ixigo.com/css/39d5e349508c3f42076f957d2b968946/ixigo/ui.home.v3.style.css
-1 javax.net.ssl.SSLHandshakeException: Received fatal alert: handshake_failure

Although https://edge.ixigo.com/css/39d5e349508c3f42076f957d2b968946/ixigo/ui.home.v3.style.css is accessible otherwise.

See https://lists.w3.org/Archives/Public/www-validator-css/2017Dec/0003.html

Line: 106
Parse Error @font-face { font-family: 'fontello'; font-weight: normal;
font-style: normal; src: url("../fonts/fontello.eot?21718288"); src:
url("../fonts/fontello.eot?21718288#iefix") format("embedded-opentype"),
url("../fonts/fontello.woff2?21718288") format("woff2"),
url("../fonts/fontello.woff?21718288") format("woff"),
url("../fonts/fontello.ttf?21718288") format("truetype"),
url("../fonts/fontello.svg?21718288#fontello") format("svg") }

Line: 500
Parse Error }

See http://www.w3.org/TR/filter-effects-1/#FilterProperty
It is currently not recognized; the validator reports a "Property filter doesn't exist" error; see https://jigsaw.w3.org/css-validator/validator?uri=http%3A%2F%2Fbennettfeely.com%2Ffilters%2F&profile=css3&usermedium=all&warning=no&vextwarning=true&lang=en

When a block of CSS containing

background: linear-gradient(left, #7ac0da 0, #a4d4e6 100%);

gets fed into the W3C CSS Validation Service by direct input it gets caught as

Value Error : background left is not a color value )

CSS documentation over at Mozilla, however, seems to imply that such input is in fact valid.

flex: 0 1 auto; is reported as "auto is not a flex-flow value : 0 1 auto"

flex is grow/shrink/basis: https://www.w3.org/TR/css-flexbox-1/#flex-property

Despite unset being a CSS-wide keyword it still throws a Value Error on the CSS Validation Service.

Example input:

div {
  display: unset;
}

For example

Властивість% s є невідомим розширенням

while using it on https://jigsaw.w3.org/css-validator/validator with -webkit-transform.

So I'm using the "Nu Html Checker" to check for html5 errors by just using the jar and giving it a file as an argument.

Now the documentation for css-validator states that I could use

java -jar css-validator.jar http://www.w3.org/

But this does not work, as there seem to be libraries missing?
Where can I see what libraries I would need?
Can I just download them and put them in this folder or would this not work?

Thank you

Validating http://elektronotdienst-nuernberg.de/bugs/prefix.html gives me this warning:

Property -apple-system,system-ui,segoe ui,arimo,liberation sans,helvetica,sans-serif is an unknown vendor extension

Thats of course wrong, because "-apple-system" is a font family and not a property

IF the pseudo-classes for a required field are used, they do not validated.

Examples of what I am trying to use on a form that has required fields on them.
input:required:invalid
textarea:required:invalid
input:focus:invalid
textarea:focus:invalid
input:required:valid
textarea:required:valid
input:focus:valid
textarea:focus:valid

The invalid ones will turn the field a color like RED denoting that something is missing. When data has been entered into that field the color changes to something else like GREEN.

I'm a college professor. I use the validators as part of my grading rubric. Students have alerted me that they cannot validate their CSS right now because of this error. Please advise.

This would make it easy to use this app with travis builds

Following valid ruleset yields errors:

@media (min-width: 992px) and (max-width: 1199px) {
  #foo {
    position:absolute;
    animation: foobar 5s ease-in-out infinite alternate;
  }

@keyframes foobar {
    0% {
        transform: scale(1.0);
    }
    100% {
        transform: scale(2.0);
    }
  }
}

Errors:

W3C CSS Validator results for TextArea (CSS level 3)
Sorry! We found the following errors (2)
URI : TextArea
17 Parse Error @Keyframes foobar { 0% { transform: scale(1.0); } 100% { transform: scale(2.0); } }
18 Parse Error }

There are no errors when parsing one of the following CSS inputs alone:

@media (min-width: 992px) and (max-width: 1199px) {
  #foo {
    position:absolute;
    animation: foobar 5s ease-in-out infinite alternate;
  }
}

@keyframes foobar {
    0% {
        transform: scale(1.0);
    }
    100% {
        transform: scale(2.0);
    }
  }

Hi,
I noticed the following issue with your current master:
When building a jar file using ant: ant jar, the created css-validator.jar is missing a dependency resulting in a crash:

roiex@Robins-PC:/mnt/e/GitHub/css-validator$ java -jar css-validator.jar  http://localhost:4000
{output=text, profile=css3, vextwarning=false, warning=2, medium=all, lang=en}
Exception in thread "main" java.lang.NoClassDefFoundError: org/apache/commons/lang/StringUtils
        at org.apache.velocity.runtime.resource.ResourceManagerImpl.initialize(ResourceManagerImpl.java:161)
        at org.apache.velocity.runtime.RuntimeInstance.initializeResourceManager(RuntimeInstance.java:730)
        at org.apache.velocity.runtime.RuntimeInstance.init(RuntimeInstance.java:263)
        at org.apache.velocity.runtime.RuntimeSingleton.init(RuntimeSingleton.java:112)
        at org.apache.velocity.app.Velocity.init(Velocity.java:74)
        at org.w3c.css.css.StyleSheetGenerator.<clinit>(StyleSheetGenerator.java:92)
        at org.w3c.css.css.StyleReportFactory.getStyleReport(StyleReportFactory.java:18)
        at org.w3c.css.css.CssValidator.handleRequest(CssValidator.java:194)
        at org.w3c.css.css.CssValidator.main(CssValidator.java:163)
Caused by: java.lang.ClassNotFoundException: org.apache.commons.lang.StringUtils
        at java.net.URLClassLoader.findClass(URLClassLoader.java:381)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:424)
        at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:335)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:357)
        ... 9 more

Output of java -version:

openjdk version "1.8.0_151"
OpenJDK Runtime Environment (build 1.8.0_151-8u151-b12-0ubuntu0.16.04.2-b12)
OpenJDK 64-Bit Server VM (build 25.151-b12, mixed mode)

As you may notice this is not a direct dependency of css-validator, but an indirect dependency of apache velocity.
Given that you should perhaps consider updating your dependencies (all of them), (and maybe compile using java 8 because even java 8 got old/common by now), or even switch to gradle instead of using ant.

I opened this Repo in eclipse and for gods sake, please run a formatter on this codebase, the code is un-readable in some classes, not to mention a lot of dead code, if(true)/if("" != null), raw generics statements etc. etc. etc.

I'm assuming some code has been generated by a decompiler? Definitely looks like it.

Anyways, enough ranting,
TL;DR: Priority 1 should be to fix this dependency, priority 2 should be to make your code readable.

I'd be happy to support you with any of the potential tasks I mentioned, if you want me to run a formatter add a dependency/fix the current one, create a gradle script or something I would be happy to do that, just tell me :)

I really think this Project is worth maintaining, because it is indeed pretty useful to have as some sort of test for static (or dynamic) websites being developed under version control.

Can you please add support for @ supports rule?

This CSS 2.1 test file checks that URL() functional notation escapes like U\r\4c ("green.png"); are legal:
http://test.csswg.org/suites/css2.1/20110323/html4/uri-015.htm
but the CSS validator reports a parse error on this file

Allowing URL escaping was documented in this errata
https://www.w3.org/Style/css2-updates/REC-CSS2-20110607-errata.html#s.4.1.1a

The validator is unable to process empty input files and produces the following error both on direct input and empty file input:

CSS
Validator
Error
Target: No file

I/O Error: You have sent an invalid request.

I'd like to have the automatic printing (reverted in 1eed592) as a cli option, so users can disable it if it's too verbose.
For stuff like travis builds etc. only the error log is needed, when everything works fine the log is not needed and just clutters most of the log.

Low Priority though.

I believe that this is a regression, as the CSS 2.1 attr() function used to validate. Some recent change in 2015 has stopped it from validating. Neither CSS 2.1-style attr(attribute) nor CSS 3.0-style attr(attribute <type>) validate correctly.

Reduced testcase:

.test::before {
   content: attr(title);
}

This can be done via https://github.com/w3c/css-validator/releases

You can+should also include a JAR for the release via the indicated field of GitHub release UI:

According to W3 last recommendation, stroke-width accepts a value:
https://www.w3.org/TR/SVG/types.html#DataTypeLength

Has anyone else encountered problems with the CSS validator (https://jigsaw.w3.org/css-validator/validator) suddenly marking pages that validated perfectly a week or more ago with invalid code? The invalid code marked by the validator seems to involve the clear property and its both value.

I did an experiment this morning by entering in these styles in the CSS validator on 11/4/15:

p {
clear: left;
}
Result: validated

p {
clear: right;
}
Result: validated

p {
clear: both;
}
Result: Error!

This CSS 2.1 test suite file checks that multiple text-decoration styles are valid:
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-021.htm

The CSS validator reports this as invalid:
Value Error : text-decoration overline is not a color value : underline blink overline

The same issue is reported in a number of other test suite files:

http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-022.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-027.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-028.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-033.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-034.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-042.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-044.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-045.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-046.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-048.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-050.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-051.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-052.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-054.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-056.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-057.htm
http://test.csswg.org/suites/css2.1/20110323/html4/text-decoration-058.htm

This issue is explained at http://stackoverflow.com/q/15481088/2657549

#something {
    background: url(data:image/png;base64,ABCDEF);
}

with a correct base64 encoding of an image instead of ABCDEF is valid according to http://www.w3.org/TR/CSS21/syndata.html#uri but throws an error when going through the validator:

 Value Error : background url(data:image/png;base64,ABCDEF) is an incorrect URL url(data:image/png;base64,ABCDEF) 

Is that a bug?

The value of a CssTransform property object is a CssLayerList instance. The CssLayerList.toString() method concatenates multiple values by using ", " as separator.
https://github.com/w3c/css-validator/blob/master/org/w3c/css/values/CssLayerList.java#L87

While that is appropriate for CssFont and other properties with CssLayerList values, for the "transform" property the function list must be whitespace-only separated (https://www.w3.org/TR/css-transforms-1/#typedef-transform-list) - or even not separated at all?
Browsers do not accept multiple transform-function values separated by commas, while whitespace is accepted.

Since CssLayerList is shared between properties that require and some what forbid commas, I guess either a distinct class or a new toString() method need to be introduced.

Please distribute a LICENSE file for the software in the git repository.

When parsing errors occur on a test initiated from Unicorn, Unicorn is unable to handle the error and will only show something like:

A critical error has occurred.
class org.apache.xmlbeans.XmlException
error: Unexpected character encountered (lex state 3): '&'
org.apache.xmlbeans.impl.store.Locale$SaxLoader.load(Locale.java:3486)
org.apache.xmlbeans.impl.store.Locale.parseToXmlObject(Locale.java:1276)
org.apache.xmlbeans.impl.store.Locale.parseToXmlObject(Locale.java:1263)
org.apache.xmlbeans.impl.schema.SchemaTypeLoaderBase.parse(SchemaTypeLoaderBase.java:345)
org.w3.x2009.x10.unicorn.observationresponse.ObservationresponseDocument$Factory.parse(Unknown Source)
org.w3c.unicorn.response.impl.DefaultResponseXBeans.(DefaultResponseXBeans.java:60)
sun.reflect.GeneratedConstructorAccessor1015.newInstance(Unknown Source)
sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)
java.lang.reflect.Constructor.newInstance(Constructor.java:526)
org.w3c.unicorn.response.ResponseFactory.getResponse(ResponseFactory.java:24)
org.w3c.unicorn.request.URIRequest.doRequest(URIRequest.java:141)
org.w3c.unicorn.RequestThread.run(RequestThread.java:66)

This can be fixed by properly escaping $span_value in ucn.properties

From:

#if ( $link_value )
        $error_content.get("link_before_parse_error") :  $error_content.get("link_name_parse_error") ($link_value)
#end
        $esc.xml($error_content.get("ErrorMsg"))
#if ( $span_value )
        $span_value
#end

To:

#if ( $link_value )
        $error_content.get("link_before_parse_error") :  $error_content.get("link_name_parse_error") ($link_value)
#end
        $esc.xml($error_content.get("ErrorMsg"))
#if ( $span_value )
        $esc.xml($span_value)
#end

Using minifiers that have the W3C CSS validator as checking tool give problems when you have the @support in your css. They break the css files.

In my situation we used cloudflare, but i read about more situations where it happened.

Requires #17.

When validating w3.org/TR/2016/CR-secure-contexts-20160915 with profile CSS level 3 + SVG, the validator complains: “# is not a fill/stroke value” four times.

The offending chunk seems to be:

.secure {
  fill: #88FF88;
}
.non-secure {
  fill: #FF8888;
  stroke-dasharray: 5px, 5px;
}
text.rejection {
  fill: #FF0000;
  font-weight: 700;
  font-size: 2em;
}
g path {
  stroke-width: 2;
  stroke: #666666;
  fill-opacity: 0;
  stroke-dasharray: 5px, 5px;
}

The CSS Validator results for checking https://www.hex.co.il/ with the default options show this:

org.w3c.css.properties.css3.Css3Style cannot be cast to org.w3c.css.properties.svg.SVGBasicStyle

The problem doesn’t occur if the profile value is css3svg; it occurs if it’s just css3.

When that happens, the stack trace for the underlying exception looks like this:

java.lang.ClassCastException: org.w3c.css.properties.css3.Css3Style cannot be cast to org.w3c.css.properties.svg.SVGBasicStyle
        at org.w3c.css.properties.css.CssWritingMode.addToStyle(CssWritingMode.java:84)
        at org.w3c.css.parser.CssStyle.setProperty(CssStyle.java:83)
        at org.w3c.css.parser.CssSelectors.addProperty(CssSelectors.java:314)
        at org.w3c.css.css.StyleSheet.addProperty(StyleSheet.java:94)
        at org.w3c.css.css.StyleSheetParser.handleRule(StyleSheetParser.java:96)
        at org.w3c.css.parser.CssFouffa.handleRule(CssFouffa.java:815)
        at org.w3c.css.parser.analyzer.CssParser.ruleSet(CssParser.java:2532)
        at org.w3c.css.parser.analyzer.CssParser.afterImportDeclaration(CssParser.java:669)
        at org.w3c.css.parser.analyzer.CssParser.parserUnit(CssParser.java:547)
        at org.w3c.css.parser.CssFouffa.parseStyle(CssFouffa.java:374)
        at org.w3c.css.css.StyleSheetParser.parseStyleElement(StyleSheetParser.java:316)
        at org.w3c.css.css.HTMLParserStyleSheetHandler.endElement(HTMLParserStyleSheetHandler.java:401)
        at nu.validator.saxtree.TreeParser.endElement(TreeParser.java:132)
        at nu.validator.saxtree.Element.revisit(Element.java:110)
        at nu.validator.saxtree.TreeParser.parse(TreeParser.java:95)
        at nu.validator.htmlparser.sax.HtmlParser.parse(HtmlParser.java:428)
        at org.w3c.css.css.HTMLParserStyleSheetHandler.parse(HTMLParserStyleSheetHandler.java:695)
        at org.w3c.css.css.DocumentParser.<init>(DocumentParser.java:153)
        at org.w3c.css.css.CssValidator.main(CssValidator.java:154)

The source that causes a ClassCastException exception to be thrown is here:

Apparently when profile is css3 and that code is reached, style’s an instance of Css3Style (and can’t be cast to SVGBasicStyle because that extends Css3Style, instead of the other way around).

And I notice that same SVGBasicStyle s = (SVGBasicStyle) style cast in 22 other classes in org/w3c/css/properties/css/…

There’s some odd behaviour parsing the font: shorthand using different variants of zero (0, 0%, 0pt, 0.0) for font size, and a value error on "font: 0 Ahem" which some CSS 2.1 test suite files depend on being valid

div { font: 0 Arial; }
Value Error : font 0 is not a font-weight value : 0 Arial

div { font: 0.0 Arial; }
Value Error : font Invalid RGB function : 0.0 Arial

div { font: 0pt Arial; }
Congratulations! No Error Found.

div { font: 0% Arial; }
Congratulations! No Error Found.

Chrome treats font: 0 Arial; as font: 0px Arial; and some of the CSS 2.1 test suite depends on this happening. For example, these tests check that font: 0 Ahem; is overridden by a subsequent font: inherit:

http://test.csswg.org/suites/css2.1/20110323/html4/font-039.htm
http://test.csswg.org/suites/css2.1/20110323/html4/font-040.htm
http://test.csswg.org/suites/css2.1/20110323/html4/font-041.htm
http://test.csswg.org/suites/css2.1/20110323/html4/font-042.htm
http://test.csswg.org/suites/css2.1/20110323/html4/font-043.htm

There's a potential ambiguity in the CSS grammar for the font: shorthand since font-weight allows an integer font-weight and 0 also is a legal font-size value (but no other integer is a legal font-size). In CSS 3 Fonts this is resolved because font-weight numeric values are restricted to the integers 100 | 200 | 300 | 400 | 500 | 600 | 700 | 800 | 900. In CSS 4 font-weight values are numbers greater than 0 and less than 1000, and all other values are treated as parse errors (which also resolves the ambiguity)

https://www.w3.org/TR/css-fonts-3/#propdef-font
https://www.w3.org/TR/css-fonts-4/#font-weight-prop

In CSS 3

"font: 0 Arial" is legal and equivalent to "font: 0px Arial"
"font: 12 Arial" is illegal and the 12 could either be reported as a font-size with a missing unit, or an invalid font-weight.
"font: 1001 Arial" is illegal and the 1001 could either be reported as a font-size with a missing unit, or an invalid font-weight.

In CSS 4

"font: 0 Arial" is legal and equivalent to "font: 0px Arial"
"font: 12 Arial" is legal, and equivalent to font-weight:12 font-family:Arial
"font: 1001 Arial" is illegal and the 1001 could either be reported as a font-size with a missing unit, or an invalid font-weight.