static site checker


known issues
bug reporting
boot notes
copyright & licence


The static site checker is an opinionated HTML nitpicker, a command–line tool to validate static HTML & XHTML websites. I built it to nitpick arts & ego, my hand–coded identity website.

It should not be used on untrusted content; its parsers are holier than Robin’s cow.

If you want to try it, here’s the current source. The build instuctions follow.

Dylan Harris
December 2023

why SSC

Why did I make the static site checker? Aren’t there a lot of other HTML validators around? When I checked a few years ago; I couldn’t find a web site validator, only web page validators. Things may have improved. Anyway, my google foo is poo.

My identity website has more than 100,000 pages. I’m too impatient to push each through a validator one by one; I want to validate my site as a whole. Furthermore, single page validators can’t catch inter–page errors, such as broken internal links, let alone hidden links (an otherwise valid link to a HIDDEN element).

Many people avoid such problems by using frameworks. I find frameworks awful. IMAO, they produce dull, boring, trite design. The visual arts world has had centuries to develop excellent form for a rectangular space. Most 21st frameworks are so crude they haven’t even absorbed 14th century visual arts’ ideas, when painters broke out of rectangular form in a rectangular frame. So much is possible, so much hasn’t happened. I want to break this dull, stultifying, archaic, mutton.

Maybe I’m making the wrong comparison, that the web isn’t about image, it’s about type. The Western visual arts never did really suss mixing writing and form (that’s not really true, but, IMAO, such arts never escaped their context). However, the Eastern visual arts most certainly did, and frameworks haven’t noticed them either.

Enough of this. Rather than criticising other people for not doing, I should do. I need to make some example sites. That’s where SSC comes in.

If I am to build a site using an experimental visual process, I can’t use frameworks. If I can’t use frameworks, I have to hand code. And there’s the key problem: HTML is such a convoluted, evolved mess, that the people who designed it, in their own design presentations, make errors. Ok, I only found this out by testing SSC on them, which conveniently illustrates that HTML is overcomplicated. I’m not going to reveal names because these people are working hard to make the web a better place. Let’s just say W3 had broken links, WhatWG referenced withdrawn ontologies, and many other authors’ sites have other internal inconsistencies. That the people who define the web make mistakes using their own design in their own documents that espouse their design, helps explain why most stick to dull, formulaic, boring, frameworks. To be fair, my HTML is far worse than any of these mild examples of technical naughtiness, which is why I had to write SSC.

I’ve yet to build a site inspired by visual art’s form and layout. My efforts have been spent building SSC, a tool to make that practical.

Since I’m here, I’ll list other issues I have with frameworks:

Dylan Harris
December 2023


Static Site Checker
(an opinionated HTML nitpicker)
version 0.1.49
(c) 2020-2023 dylan harris
see LICENCE.txt & LICENSE.txt for copyright & licence notice

ssc analyses static HTML snippets, files and sites:
— HTML living standard, Jan 2005 to Oct 2023
— HTML 1.0/+/2.0/3.0/3.2/4.00/4.01/5.0/5.1/5.2/5.3–draft
— CSS 1/2.0/2.1/2.2–draft, 2007-2023 snapshots, more
— SVG 1.0/1.1/1.2 Tiny/1.2 Full/2.0/2.x–draft Apr 2021
— MathML 1/2/3/4–draft Jul 2022
— XHTML 1.0/1.1/2.0/5.x
— finds broken links
— server side includes, mostly
— many ontologies

with opinions on:
— standard english where dialect is required
— perfectly legal but sloppy HTML
— abhorrent rudeness such as autoplay on videos

It does NOT:
— analyse or understand scripts
— analyse or understand XML or derivatives, except as noted above

It can output:
— ‘repaired’ HTML (not XHTML)
— HTML with resolved server side includes
— JSON of ontological content
— website statistical information
— deduplicated websites

ssc -h
for a usage summary.

ssc -f config_file
analyse site using preprepared configuration

ssc directory
analyse website based in directory

To build & run:
1. Follow the build instructions in build.txt
2. Gleefully run ssc. It will misbehave if you are insufficiently gleeful.

This is an alpha version of ssc. It may contain unexpected features.
If you encounter such a delight, please help improve ssc by collecting
the following information (where relevant):
— version of ssc;
— precise version of the operating system;
— hardware architecture and system information;
— detailed description of the problem;
— detailed description of the steps to recreate it;
— copy of output file showing the error;
— copy of pages/website being analysed;
— precise command used;
— configuration file(s) used, if any;
— any ndx file or other pre–existing file used during the run;
— any known workarounds or solutions;
— optionally, a dance interpretation of the ‘feature’;
and emailing everything to (if the collected files are more
than small, please use a public fileserver and email the link). Do NOT
send anything confidential. Furthermore, unless you state otherwise,
we reserve the right to publish some or all of the information sent in
future versions of ssc, usually in the test suite. If you have a fix,
you are invited to submit a pull request on github. Thank you.

SSC can be run in a CGI environment. This is intended for use with OpenBSD’s native httpd web server.
You are reminded that SSC is α software. Do NOT expose it to untrusted data
sources, such as the open web, without taking serious precautions. SSC probably has
more bugs than the Creator’s Ultimate All–Beetle Extravaganza (J.B.S.
Haldane, apocryphal : “[the Creator has] an inordinate fondness for beetles.”).

Notes on names:
— recipe: a nod to Vernor Vinge’s “A Fire Upon the Deep”;
— tea: without tea, nothing works; then there’s builders’ tea;
— sauce: makes the dull tasty; identifies incompetent pedants;
— toast: toasts code; i liked burnt toast;
— heater: i’m not stopping now;
— unii: my preferred plural of unix: to my ears, both unixes and unices
        sound like they sing castrato.
— andor: and/or sans ancienne; land of Gift (aber nicht das Gift)

build.txt        notes on building ssc
gen.txt          a model man page
usage.txt        how to use ssc
releasenotes.txt chips
LICENCE.txt      ssc licence information
LICENSE.txt      formal GPL 3 licence
more licences    licences for borrowed external content

I have a website, arts & ego, at It has
approaching 60G of original content. It contains hand coded HTMLs 2,
3, 4 & 5. It is a complete mess. Despite a long search, I could not
find any tools to properly identify its flaws. Anything I did find
was at most cursory.

Then came the cow flu*.

*corvid means crow, thus covid means cow**.

**by the rules of sympathetic spelling.

Unabashed Opportunism
If you appreciate modernist poetry or abstract photography, I’ve been
published. Click on books at arts & ego for gen.

written by dylan harris
December 2023


ssc - analyse static web site source

ssc [...] directory
ssc -f config

ssc (the Static Site Checker) is an opinionated HTML nit-picker,
intended for people, such as its author, who hand code websites. It
doesn't just check static websites for broken links, dubious syntax,
and bad semantic data, it will actively complain about things that are
perfectly legal but rather untidy, like its author.

Except when serving CGI queries, it recursively scans the directory
looking for HTML source files to analyse. It produces a list of errors,
warnings, comments, and other hints of imperfection. Once complete, it
summarises internal site inconsistencies, and can produce some simple

Scripts are ignored.


These options are only available on the command line:

-f file                 Load configuration from file, which should be
                        in .INI file format. See CONFIGURATION FILE
                        FORMAT below. This should be an absolute path.

-F                      Load the configuration file .ssc/config in the
                        current directory.

-h                      Show a summary of switches, then exit.

-H snippet              Only nitpick this snippet of html

--ontology.list         List known schema versions, then exit.

-V                      Show version details, then exit.

--validation            Show attribute extensions, then exit. Attribute
                        extensions are additional values that can be
                        associated with attributes on some X/HTML
                        elements, and is intended for use with bespoke
                        extensions of HTML.


These options are available on the command line (with dashes) and in
configuration files (without dashes). The short form alternative
switches only work on the command line.

Most binary options, e.g. those without arguments below that turn on a
feature (which may be the default), have a corresponding "no-" switch
to turn it off. The "no-" is inserted after the dot, so, for example,
the contradiction to "--general.class" is "". When
both are specified, perhaps in a configuration file and on the command
line, the "no-" switch always applies.

--corpus.article        Prefer the content of 
when gathering corpus text. --corpus.body Prefer the content of when gathering corpus text. This is the default. --corpus.main Prefer the content of
when gathering corpus text. --corpus.output file Dump XML corpus of site into file. This is intended for use by a local search engine. If none of --corpus.article, --corpus.body, or --corpus.main are specified, the content of is used. If more than one are specified, then the text collected depends on a page's content. This is incompatible with --shadow.update. --css.adjust X Use CSS Colour Adjustment level X, where X is 0 or 3. --css.anchor X Use CSS Scrollbar Anchoring level X, where X is 0 or 3. --css.animation X Use CSS Animation level X, where X is 0, 3 or 4. --css.background X Use CSS Backgrounds and Borders level X, where X is 0 or 3. X Use CSS Box Alignment level X, where X is 0 or 3. X Use CSS Box Model level X, where X is 0, 3 or 4. X Use CSS Box Sizing level X, where X is 0, 3 or 4. --css.cascade X Use CSS Cascading and Inheritance level X, where X is 0, 3, 4, 5 or 6. --css.colour X Use CSS Colour level X, where X is 0, 3, 4 or 5. --css.compositing X Use CSS Compositing and Blending level X, where X is 0 or 3. --css.cond-rule X Use CSS Conditional Rules level X, where X is 0, 3, 4, or 5. --css.contain X Use CSS Contain level X, where X is 0, 3, 4 or 5: see --css.version for gen. --css.content X Use CSS Generated Content level X, where X is 0 or 3. --css.cs X Use CSS Counter Style level X, where X is 0 or 3. --css.custom X Use CSS Custom Properties for Cascading Variables level X, where X is 0 or 3. --css.device X Use CSS Device Adaption level X, where X is 0 or 3. --css.display X Use CSS Display level X, where X is 0 or 3. --css.ease X Use CSS Easing Functions level X, where X is 0 or 3. --css.exclude X Use CSS Exclusions level X, where X is 0 or 3. --css.extension ext Presume files with extension '.ext' are CSS files. --css.fbl X Use CSS Flexible Box Layout level X, where X is 0 or 3. --css.filter X Use CSS Filter Effects level X, where X is 0 or 3. --css.float X Use CSS Page Floats level X, where X is 0 or 3. --css.font X Use CSS Fonts level X, where X is 0, 3, 4 or 5. --css.frag X Use CSS Fragmentation level X, where X is 0, 3. or 4 --css.grid X Use CSS Grid level X, where X is 0, 3 or 4: see --css.version for gen. --css.highlight X Use CSS Custom Highlights level X, where X is 0 or 3. --css.image X Use CSS Images level X, where X is 0, 3 or 4. --css.inline X Use CSS Inline Layout level X, where X is 0 or 3. --css.line-grid X Use CSS Line Grid level X, where X is 0 or 3. --css.list X Use CSS Lists and Counters level X, where X is 0 or 3. --css.logic X Use CSS Logical Properties level X, where X is 0 or 3. --css.marquee X Use CSS Marquee level X, where X is 0 or 3. --css.masking X Use CSS Masking level X, where X is 0 or 3. X Use CSS Media Queries level X, where X is 0, 3, 4 or 5. Test against the CSS Mobile Profile. --css.multi-column X Use CSS Multi-Column level X, where X is 0 or 3. --css.namespace X Use CSS Namespaces level X, where X is 0 or 3. --css.nes X Use CSS Non-Element Selectors level X, where X is 0 or 3. --css.overflow X Use CSS Overflow level X, where X is 0, 3 or 4. --css.overscroll X Use CSS Overscroll Behaviour level X, where X is 0 or 3. X Use CSS Paged Media level X, where X is 0 or 3. --css.position X Use CSS Positions level X, where X is 0 or 3. --css.present X Use CSS Presentation Levels level X, where X is 0 or 3. --css.print Test against the CSS Print Profile. --css.region X Use CSS Regions level X, where X is 0 or 3. --css.rhythm X Use CSS Rhythmic Sizing level X, where X is 0 or 3. --css.round X Use CSS Round Display level X, where X is 0 or 3. --css.ruby X Use CSS Ruby Annotations level X, where X is 0 or 3. --css.scope X Use CSS Scoping level X, where X is 0 or 3. --css.scrollbar X Use CSS Scrollbar Style level X, where X is 0 or 3. --css.sda X Use CSS Scroll-Driven Animations Style level X, where X is 0 or 3. --css.selector X Use CSS Selectors level X, where X is 0, 3 or 4. --css.shadow X Use CSS Shadow Parts level X, where X is 0 or 3. --css.shape X Use CSS Shapes level X, where X is 0, 3 or 4. --css.snap X Use CSS Scroll Snap level X, where X is 0 or 3. --css.spatial X Use CSS Spatial Navigation level X, where X is 0 or 3. --css.speech X Use CSS Speech level X, where X is 0 or 3. X Use CSS Style level X, where X is 0 or 3. --css.syntax X Use CSS Syntex level X, where X is 0 or 3 --css.table X Use CSS Tables level X, where X is 0 or 3 (this is an experimental spec, likely to change). --css.text X Use CSS Text level X, where X is 0, 3 or 4. --css.text-dec X Use CSS Text Decoration level X, where X is 0, 3 or 4. Test against the CSS TV Profile. --css.transform X Use CSS Transforms level X, where X is 0, 3 or 4: see --css.version for gen. --css.transition X Use CSS Transitions level X, where X is 0 or 3. --css.ui X Use CSS Basic User Interface level X, where X is 0, 3 or 4. --css.value X Use CSS Values and Units level X, where X is 0, 3 or 4. --css.verify Verify CSS files (replaces --general.css). --css.version X Presume version X of CSS, where X is one of: 1 CSS 1.0 2.0 CSS 2.0 2.1 CSS 2.1 2.2 CSS 2.2 (Feb 2022 draft) 3 all CSS level 3 so far 4 all CSS level 4 so far 5 all CSS level 5 so far 6 all CSS level 6 so far 2007 2010 2015 2015+ 2015++ 2017 2017+ 2017++ 2018 2018+ 2018++ 2020 2020+ 2020++ 2021 2021+ 2021++ 2022 2022+ 2022++ 2023 2023+ 2023++ The years are CSS snapshots, whether the year itself for stable modules, with + for wobbly modules, and ++ for wibbly-wobbly modules, as per the corresponding W3 CSS snapshots (the terminology in those snapshots is inconsistent, hence our use of the scientific terms wobbly and wibbly-wobbly). For levels 3, 4, 5 and 6, note that extensions that are part of neither CSS 1 nor CSS 2.x specifications are numbered three and upwards in ssc, for internal consistency. If you wish to use an extension named ... level 1, that is not part of CSS 1, specify 3. Similarly, for those named level 2 that are not part of any CSS 2 specification, etc.. --css.view X Use CSS View Transitions level X, where X is 0 or 3. --css.wc X Use CSS Will Change level X, where X is 0 or 3. --css.writing X Use CSS Writing Mode level X, where X is 0, 3 or 4. --general.class Nitpick class values --general.classic Report all classes used, not just those in CSS files. --general.cgi Check environment variables for snippets of -W HTML. SSC expects environment variables as produced by OpenBSD's native httpd, produced using <FORM METHOD=GET …>. Do NOT let ssc anywhere near untrusted data. Ignores many options such as shadowing. --general.custom EL Define a custom element <EL> for verifying the IS attribute. May be repeated. --general.datapath dir Look for any configuration, caches, and other -p dir useful files, in this directory. --general.defthrd N If --general.thread is not given, then set the -N N number of threads to N. --general.error x If nits of the specified category or worse are -E generated, then, on exit, return an error code. Values are: 'catastrophe', 'error' (the default), 'warning', 'info', or 'comment'. --general.exclude xxx Ignore all paths containing xxx. May be repeated. Case independent under Windows only. .DS_Store is always excluded in darwin (macos). ---general.file XXX File for persistent data. Requires -N. See also --general.datapath. Default extension: .ndx. --general.ignore EL ignore attributes and content of the element <EL>. May be repeated. Report launch context when starting. --general.lang LA If an X/HTML file does not have a language / dialect specified (e.g. "en" for generic English, "en-IE" for Irish English, "lb-LU" for Luxembourgish, "ma" for Marain, etc.), default to 'LA'. If not given, the default is your system default, or, if none, then "en-US" (the standard American dialect of English). --general.maxfilesize n Do not process HTML source files that exceed n bytes in size (default: 4M). Specify 0 for unlimited, although be warned that ssc is stunningly stupid in such circumstances and may even attempt to load files bigger than available memory. --general.output Output to the specified file. If this switch is -o file not used, standard output is used. --general.progress Dump progress information to standard output. -D This can interfere with formatted output. --general.rdfa Check RDFa attributes. --general.rel Only mention <LINK> REL values, found neither in the living standard nor at, in debug output. --general.rpt Report CSS files that are opened. --general.slob Ignore perfectly legal yet inefficient, indeed thoroughly slobby, HTML, such as being far too lazy to get round to bothering to close elements. --general.spec Reset the values of most switches to false. -j --general.ssi Process Server Side Includes (SSIs). Note ssc -I cannot process SSIs directives with formulae. Processing SSIs may cause incorrect line numbers to be mentioned when an issue is reported. --general.test Output data in automated test format. Used by -T ssc-test. Not generally useful. Documented so you can avoid it! --general.thread N Use N threads when running. Defaults to a value -n N appropriate for the hardware. Too high a value can cause problems. See also --general.defthrd. --general.vcs Excludes, as per --general.exclude: .bazaar .bk CVS .cvsignore _darcs .fslckout .git .gitattributes .gitignore .gitmodules .pijul RCS SCCS .svn --general.verbose x Output nits to the specified verbosity: -v 'catastrophe', 'error', 'warning', 'info', 'comment' (the default), or '0' for silence. Additional values are available when debugging. Each level includes its preceding level, so, for example, 'warning' will also output 'catastrophe' and 'error' nits. --html.force If <!DOCTYPE. …> is missing, force presumption of --html.version value, not HTML 1/tags Don't mention certain Internet Explorer features. --html.rfc1867 Ignore the RFC 1867 (INPUT=FILE) extension when processing HTML 2.0 --html.rfc1942 Ignore the RFC 1942 (tables) extension when processing HTML 2.0. --html.rfc1980 Ignore the RFC 1980 (client side image maps) extension when processing HTML 2.0. --html.rfc2070 Ignore the RFC 2070 (internationalisation) extension when processing HTML 2.0. --html.safari Don't mention certain early Safari features. --html.tags When an HTML file is loaded that contains no DOCTYPE, ssc normally presumes HTML 1. This switch tells it to presume the file conforms to an earlier HTML Tags specification (the one at CERN). This is overridden by --html.version. --html.title n If <TITLE> text is longer than n characters, -z n say so. This applies to text enclosed by a <TITLE> element under <HEAD>, not the value of TITLE attributes. --html.version X If no doctype (or xml header) is specified, presume version X of HTML. X can be: tags HTML tags (1991, informal) 1 HTML 1.0 (Jun 1993 draft) 1.0 HTML 1.0 (Jun 1993 draft) + HTML Plus (Nov 1993 draft) 2 HTML 2.0 2.0 HTML 2.0 3 HTML 3.2 3.0 HTML 3.0 (Mar 1995 draft) 3.2 HTML 3.2 4 HTML 4.01 4.0 HTML 4.0 4.1 HTML 4.01 4.2 XHTML 1.0 4.3 XHTML 1.1 core 4.4 XHTML 2.0 (Dec 2010 draft) 5 recent WhatWG HTML 5 5.0 W3 HTML 5.0 5.1 W3 HTML 5.1 5.2 W3 HTML 5.2 5.3 W3 HTML 5.3 (Oct 2018 draft) 2005/1/1 WhatWG WebApps draft (Jan 2005) ... (halfly) 2007/1/1 WhatWG WebApps draft (Jan 2007) 2007/7/1 WhatWG HTML 5 (Jul 2007) ... (halfly) 2021/1/1 WhatWG HTML 5 (Jan 2021) ... (quarterly) 2023/7/1 WhatWG HTML 5 (Jul 2023) XHTML 1.0 XHTML 1.0 XHTML 1.1 XHTML 1.1 core XHTML 2.0 (Dec 2010 draft) XHTML 5.x XHTML corresponding to equivalent W3 HTML Although you can specify exact dates for versions of the WhatWG HTML 5 living standard, currently only broad versions published in January and July are supported (quarterly from 2021). Certain versions of HTML offer variants, such as loose and strict definitions. ssc picks those up from the <!DOCTYPE …> in the HTML file, if any, and then carefully ignores them. Validation of XHTML is even less strict. Just to remind you, there are no guarantees of accuracy (or inaccuracy). Copies of the appropriate standards can be found online. A copy of the copies referenced during ssc's development can be found at --link.301 Normally, when ssc checks external links -3 (--link.external), it does not report http forwarding errors 301 and 308. Use this switch to have it do so. --link.check Check internal links, e.g. those within the -l website being analysed. --link.example Report links to faux domains, as defined by RFC 2606 (note ssc also reports links to, & --link.external Check external links, e.g. those not on the -e site being checked. Note that ssc will NOT check RFC 2606 links, such as (see --link.example). --link.forward Report HTTP forwarding errors encountered when checking external links (e.g. 301 and 308) --link.ignore DOMAIN When checking external links, ignore this domain. May be repeated. --link.local Report links to local domains, such as domains ending in .lan, .home, .corp, and others. --link.once Only report each broken external link once. If, -O for example, the site has a number of references to a page that does not exist, ssc will only report the first instance of the broken link. Note that, even if it reports every occurrence of the link, it will only check it the first time it's encountered (requires --link.external). --link.pretend FILE Pretend links containing xxx exist. May be repeated. DOMAIN Report links to domain and its descendents. May be repeated. --link.revoke Do not check whether links' https certificates -r have been revoked (requires --link.external). --link.xlink Check crosslink IDs on the site being analysed. -X For example, if a link goes to /index.html#id, then, when this switch is set, ssc will verify that the id exists and that it is not hidden. --math.core Presume Math 4 core (May 2022 draft). --math.draft D Presume this draft of MathML 4. The following drafts supported: 2020 December 2020 draft 2022 August 2022 draft --math.version N Presume version N of MathML (1, 2, 3 pr 4). The following versions are supported: 0 try to work it out from the (HTML) version of the file being analysed 1 MathML 1 2 MathML 2 3 MathML 3 4 MathML 4 (see --math.draft) --microdata.export Export microdata encountered. This data is exported in JSON format (not JSON-LD). --microdata.root DIR When exporting microdata with --microdata.export, write files into the directory DIR. ssc will create the directory tree structure as appropriate. --microdata.verify Check microdata found in WhatWG microdata -m attributes (itemprop, itemtype, etc.). Note that ssc only knows about certain ontologies (see --ontology.list) --microdata.virtual v=d When exporting microdata using --microdata.export, export the contents of virtual directory 'v' to 'd'. 'v' must match a directory identified with --site.virtual. For example: --microdata.virtual virtual=X:\virtual. --microformat.export Export microformat data encountered in JSON format. This option will write files in the same directory as the source, with the extension .json. --microformat.verify Verify Microformats data in class and rel -M attributes (see --microformat.version x Presume microformats version x. The following values are currently accepted: 1 microformats version 1 only 2 microformats version 2 only 3 both microformats versions 1 and 2 --nits.catastrophe n redefine nit n as a catastrophe; may be repeated (the value of n can be determined using --nits.nids below). --nits.comment n Redefine nit n as a comment; may be repeated (the value of n can be determined using --nits.nids). --nits.debug n Redefine nit n as a debug message; may be repeated (the value of n can be determined using --nits.nids). --nits.error n Redefine nit n as an error; may be repeated (the value of n can be determined using --nits.nids). --nits.extra Report additional nits. --nits.format F Specify the output format; F is a template file (see OUTPUT TEMPLATE below). n Redefine nit n as information; may be repeated (the value of n can be determined using --nits.nids). --nits.nids Output nit ids, which can be used to redefine nits. --nits.override F Use this output format, not the one specified by --nits.format. F is a template file (see OUTPUT TEMPLATE below). This switch is intended to aid automation. --nits.quote X Specify quote style when using nit.format. X can be 'text' or 'html'. --nits.root By default, seek nit output template files in the website root. --nits.silence n Silence nit n; may be repeated (the value of n can be determined using --nits.nids). --nits.unique Do not output repeated nits, even if they may contain additional information. --nits.warning n Redefine nit n as a warning; may be repeated (the value of n can be determined using --nits.nids). Output debug nits (intended for automation). --rdfa.version When checking RDFa files, presume this version (default: 1.1.3). Note, RDFa analyis is incomplete, and only intended for supporting HTML analysis. --ontology.ONT X.Y Presume version X.Y of ontology ONT. For example: --ontology.xsd 1.1 defaults usage of XSD to version 1.1. The versions apply to RDFa, microdata, and microformats (using class) analysis. If .Y is omitted, .0 is presumed. X must be present. Unspecified defaults are derived from the HTML version. For a list of possible values, use --ontology.list. At the time of writing, the following ontology versions can be verified. Note that single version ontologies cannot have their version changed: adms 1.0,2.0 article 12,14,18,22 as 1.0,2.0 basic 1.0-1.3,2.1,3.0 (see below) bfo 2.0,2020 (see below) bibo 1.3 biro 1.1 book 12,14,18,22 cc 1.0 cito 2.8 content 1.0 crs 1.0 (see below) csvw 1.0 ctag 1.0 daq 1.0 ddi 1.0 dbp 1.0 dbp-owl 1.0 dbr 1.0 dc11 1.0,1.1 dcam 1.0 dcat 1.0,2.0 dcmi 1.0 dcterms 1.0,1.1 ddi 1.0 doap 1.0 dqv 1.0 describedby 1.0 duv 1.0 earl 1.0 event 1.0 exif 1.0-3.0 (see below) exifex 2.21-3.0 (see below) foaf 0.1-0.99 frbr_core 1.0 gr 1.0 grddl 1.0 gs1 1.1-1.5 ical 1.0 icaltzd 1.0 jsonld 1.0,1.1 ldp 1.0 license 1.0 locn 1.0 ma 1.0 mf 1.0-2.255 music 12,14,18,22 oa 1.0 odrl 1.0 og 10,12,14,18,22 (see below) org 1.0 owl 1.0,2.0 pam 2.0 (see below) pcm 3.1 (see below) pcmm 3.0 (see below) pcv 1.0(see below) pdf 1.0 (see below) photoshop 1.0 (see below) pim 1.0-3.0 (see below) pmi 3.0 (see below) poetry 1.0,1.1 prism 1.0-3.0 (see below) prism-ad 3.0 (see below) prl 1.0-2.0 (see below) prm 3.0 (see below) prs 3.1 (see below) profile 12,14,18,22 prov 1.0 psv 1.0 (see below) ptr 1.0 pur 2.1-3.0 (see below) qb 1.0 rdf 1.0-1.3 rdfa 1.0-1.3 rdfg 1.0 rdfs 1.0 rev 1.0 rif 1.0 role 1.0 rr 1.0 schema 0.10-23.0 (see below) sd 1.0 sioc 1.0 sioc_s 1.0 sioc_t 1.0 skos 1.0 skosxl 1.0 sosa 1.0 ssn 1.0 stdim 1.0 (see below) stevt 1.0 (see below) stfnt 1.0 (see below) stjob 1.0 (see below) stref 1.0 (see below) stver 1.0 (see below) taxo 1.0 tiff 6.0 time 1.0 v 1.0 vann 1.0,1.1 vcard 1,2,3,4 (see below) video 12,14,18,22 void 1.0 wdr 1.0 wdrs 1.0 website 12,14,18,22 wwg 1.0 xhv 1.0 xml 1.0 xmp 1.0 (see below) xmpdm 1.0 (see below) xmpg 1.0 (see below) xmpgimg 1.0 (see below) xmpidq 1.0 (see below) xmpmm 1.0 (see below) xmprights 1.0 (see below) xmptpg 1.0 (see below) xsd 1.0,1.1 The various Adobe ontologies (crs, pdf, photoshop, stdim, stevt, stfnt, stjob, stref, stver, smp, xmpdm, xmpg, xmpgimg, xmpidq, xmpmm, xmprights, xmptpg) have only been partially applied. They do not seem to have been designed for microdata, hence the partial implementation: the goal is to enable hoovering to JSON. BFO (Basic Format Ontology) versions should be specified as follows: Use For 2.0 2.0 2.2 2020 BFO 2020 uses OBO's machine code style identifiers. Given the history of computing science, as a convenience for users, and with my experience of both devops and maintaining code, identifiers following the standard ontology naming convention are also accepted. Since this is unofficial, both standard English and American dialects' spellings are processed. The Exif & ExifEx ontologies have the following versions: Use For 1.0 1.0 (exif only) 1.1 1.1 (exif only) 2.0 2.0 (exif only) 2.10 2.1 (exif only) 2.20 2.2 (exif only) 2.21 2.21 2.30 2.3 2.31 2.31 2.32 2.32 3.0 3.0 Manufacturers' extensions to EXIF are omitted, with exceptions. Open Graph versions correspond to snapshots of the specs from 2010, 2012, 2014, 2018 & 2022. The various Prism ontologies (pam, pamp, pcm, pcmm, pcv, pim, pmi, prism, prism_ad, prl, prm, prs, psv, pur) have only been partially applied: some specifications are unavailable, some specifications break HTML5 syntax. Prism was not designed for microdata, hence the partial implementation: the goal is to enable hoovering to JSON. Most versions of schema ( should be specified by their version number, but this doesn't work with early versions, which should be specified a follows: Use For 0.10 June 2011 0.15 July 2011 0.20 August 2011 0.25 September 2011 0.30 October 2011 0.35 November 2011 0.40 December 2011 0.45 January 2012 0.50 February 2012 0.55 March 2012 0.60 April 2012 0.91-0.99 as version number 1.0 1.0a 1.1 1.0b 1.2 1.0c 1.3 1.0d 1.4 1.0e 1.5 1.0f 1.10 1.1 1.20 1.2 1.30 1.3 1.40 1.4 1.50 1.5 1.60 1.6 1.70 1.7 1.80 1.8 1.90 1.9 1.91 as version number ... 23.0 as version number vCard versions correspond to RDFa specs, published in 2001, 2006, 2010 & 2014. They do NOT correspond to vCard data format specifications. --shadow.changed When shadowing a site that has been previously shadowed, only copy/link files that have changed. --shadow.comment Do not delete comments when writing shadow pages. --shadow.copy X Create a shadow directory structure from source HTML files, with errors removed and some things tidied up. X can be: no copy nothing (default) pages write 'fixed' source files, ignore non source files hard set up hard links to non-source files (requires source and shadow directories to be on the same disk) (see below) soft set up soft links to non-source files (see below) all copy non HTML files too dedu copy non HTML files too, but deduplicate them, changing links in HTML source if necessary (see below) report report duplicates (no shadowing) ssc cannot convert between versions of HTML, nor between HTML and XHTML. Link options are only available on systems that support links. --shadow.enable Enable shadowing (set by other shadow options). If shadowing is enabled, but shadow.root is not set, SSC will litter the site source directories with .ndx files. --shadow.file f Write ssc's shadow cache to file f, to accelerate future shadowing of the same content updated. --shadow.ignore ext When shadowing, ignore files with this extension (may be repeated). Add a comment at or near the top of each shadowed HTML file noting its generation time. --shadow.msg text Insert a comment containing text at the top of each generated page. Note that, if any SSI include file is updated, the comment will appear whether or not the original page has changed. --shadow.root dir Where to write the shadowed site. Leave excess/repeated spaces and blank lines in the shadowed files untidily untouched. --shadow.ssi Do NOT resolve Server Side Includes when shadowing, even if --general.ssi is set. --shadow.update Only examine files that have changed since the -u last time ssc ran. This is incompatible with --corpus.file. This requires --shadow.file. Nits of files that have not changed will not be reported. --shadow.virtual v=d When shadowing virtual directories, output the shadow of virtual directory 'v' to directory 'd'. 'v' must match a directory set up using --site.virtual. --site.domain domain The domain name of the site is 'domain'. This -S domain can be repeated. This is used to identify any URL that is apparently external but is actually internal to the site. --site.extension ext Treat files with this extension as X/HTML -x ext source files. This may be repeated. Files with extension .html are always checked. --site.index file This is the name of the default file in a -i file directory. This can be repeated. This is used for checking internal links. --site.root dir This is the root of the website to analyse. ssc -g dir will recursively scan the directory analysing any HTML files it finds. The default is the current directory. --site.virtual v=d The virtual directory 'v' is located in actual -L v=d directory 'd' on the local filesystem. For example: --site.virtual virtual=D:\actual --spell.accept XXX XXX is a correct spelling of a word (or a list of words) in all languages. --spell.cased Nitpick correctly spelt but wrongly cased words. --spell.check Check text spelling. Uses external spelling checkers, so results will be inconsistent between systems. --spell.dict LANG,DICT Unix only. Associate dictionary DICT with LANG. For example, if the standard English dictionary is en_GB-large: --spell.dict en-GB,en_GB-large (Under Windows, ssc uses the OS dictionaries.) If "no", do not use the ICU libraries at all (they are rather slow). This will increase the inaccuracy and incorrectness of the spell checks. --spell.list FN,LANG The file FN contains a list of valid spellings for language LANG (which may include country info). If LANG is omitted, the valid spellings apply to all languages. For example: --spell.list villages.txt,en-IE --spell.list dorfer.txt,de --spell.list letzstied.txt --spell.path PATH Unix only. Path to spelling executable. Hunspell or a compatible program is expected. If none is specified, ssc will seek hunspell. (Under Windows, ssc uses the OS spellchecker.) --stats.abbr Output abbreviation report, so you can verify the same abbreviations have the same expansions across the site. --stats.all Output all statistics reports. --stats.annotation Output annotation report. --stats.attribute Output element attribute report, which expands the element report to output information about attributes used. --stats.category Output category report, which output the total quantity of nits reported by nit category. --stats.character-variant Output character variant report. --stats.class Output class report, which allows to you see which classes are defined in CSS but not used, which classes are used but not defined, as well as a count of both for all classes encountered. --stats.content-name Output content name report. --stats.counter-style Output counter style report. --stats.css-property Output css property report, which gives you an idea of the sophistication of the CSS used on the site. --stats.custom-media Output custom media report, which lists all named custom media definitions encountered. --stats.custom-property Output custom property report, which lists all named custom property definitions encountered. --stats.definition Output definitions report, so you can verify the same terms have the same definitions across the site. --stats.element Output element report., which totals all elements encountered across the site. --stats.error Output counts of errors, warnings, etc.. --stats.export F Export to file F. --stats.file Output file report, which reports the number of pages processed, and summerises file sizes. --stats.font Output font report, which lists all fonts used across the site. --stats.font-family Output font family report, which lists all font families named across the site. --stats.highlight Output highlight report. --stats.historical-form Output historical font form report. Output id report, allowing you to identify which ids are styled but not mentioned. --stats.itemid Output itemid report, which gives you an idea of the ontological significance and depth of the site. --stats.keyframe Output keyframe report, which lists all named keyframes. --stats.layer Output layer report, which lists all named layers. --stats.meta Produce statistics on usage in . Note that pragmas reported (http-equiv) are those found in the HTML source, not those returned by the HTTP protocol. Remember that many web servers (not all) will remove some pragmas when serving pages. Output name/value pairs report, which helps you identify inconsistencies between definitions across the site. --stats.ontology Output ontology report, which gives an insight into the ontological depth of the site being analysed. --stats.ornament Output ornament report, which reports all named CSS font ornaments encountered. Produce statistics for each source file encountered. Output page name report, which reports all named CSS page-names encountered. --stats.palette Output palette report, which reports all named CSS palettes encountered. Output ontology property count report, as an addendum to --stats.ontology. --stats.reference Output reference report, which identifies, as precisely as it can, which versions of HTML, XHTML, CSS, etc., are found. --stats.region Output region report, which reports all CSS named regions encountered. --stats.scroll-anim Output scroll animation report, which reports all CSS named scroll animations encountered. --stats.selected Output a selected set of reports; may be modified by other stats switches. --stats.statement Output CSS statement report, which summarises all CSS statements encountered. --stats.styleset Output styleset report, which reports all CSS named stylesets encountered. --stats.stylistic Output stylistic report, which reports all CSS named stylistics encountered, excluding the band themselves. --stats.summary Produce a summary of overall statistics for the website, including grand totals. --stats.swash Output swash report, which reports all CSS named swashes encountered. --stats.version Output version report, which summarises versions of HTML, SVG, MathML, etc., encountered. --stats.view Output view report, which reports all CSS named views encountered. --svg.version x Presume any SVG code encountered is this version, unless the SVG code itself specifies a version. Versions recognised: 1.0 1.1 1.2 (really 1.2/tiny) 1.2/tiny 1.2/full (May 2004 draft, incomplete, any conflict with 1.2/tiny always resolves in favour of 1.2/tiny) 2.0 2.1 (April 2021 draft) If this switch is not used, and some SVG code does not identify its version, the version is derived from the version of the host X/HTML code. --validation.attribute ATT Add the custom attribute ATT. This attribute will be ignored, not validated. --validation.charset CH Accept CH as a charset. May be repeated. --validation.class CL Add the valid class CL. May be repeated. --validation.color COL Accept COL as a colour. May be repeated. --validation.colour COL Accept COL as a colour. May be repeated. CC Accept CC as a valid two-letter country code. May be repeated. --validation.currency CUR Accept CUR as a valid currency. May be repeated. --validation.element EL Accept <EL> as a valid element. This element will be ignored, not validated. May be repeated. --validation.element-attribute EL,ATT Accept the known attribute ATT on the element <EL>. Doesn't work with namespaces (names containing ':'). May be repeated. --validation.extension EXT Accept the extension EXT as a mimetype file extension. May be repeated. --validation.ff FEATURE Accept FEATURE as a CSS font feature. These should normally be four characters long. May be repeated. --validation.ff VARIATION Accept VARIATION as a CSS font variation. These should normally be four characters long. May be repeated. --validation.httpequiv HEQ Accept HEQ as a valid macro for httpequiv on <META> elements. May be repeated. --validation.lang LANG Accept LANG as a valid language code. May be repeated. --validation.minor x When validating W3 HTML 5 source code, using -m x this minor version of W3 HTML 5. Valid values are 0, 1, 2, and 3 (draft). WhatWG versions are determined by date, corresponding roughly to the date of the (online) publication of the specific version. See the --html.version switch. --validation.metaname M Accept M as valid for the NAME attribute of the <META> element. The VALUE will be ignored. May be repeated. --validation.microdata Validate ( microdata. --validation.mimetype MT Accept MT as a valid mimetype. May be repeated. --validation.sgml SGML Accept SGML as a valid SGML schema identification (as found in <!DOCTYPE …>). May be repeated. --validation.XXX YYY Accept YYY as a valid value for attribute type XXX. For a list of possible values of XXX, use the command line switch --validation. CONFIGURATION FILE FORMAT If a configuration file is used, it should be in INI file format. All content is optional. Section and option names are derived from the long form switch name, which consists of --SECTION.OPTION, laid out in the format: [SECTION] OPTION=yes OPTION=123456 Switches that do not have a long form version cannot be used in a configuration file. Each ssc test (in the recipe/toast folder) has a configuration file; browse them for examples. ENVIRONMENT If you set --general.cgi, ssc will check these environment variables: QUERY_STRING Run under OpenBSD's httpd server. See notes below. SSC_CONFIG If no configuration file is given on the command line, use this one SSC_ARGS Preliminary command line parameters If, when SSC is run, the environment variable QUERY_STRING is set to an OpenBSD httpd server CGI value that includes the parameter html.snippet, then SSC will nitpick that snippet only. Some other parameters are processed, including general.verbose and html.version. EXIT STATUS If no significant nits are found, ssc exits with 0, otherwise it exits with a value > 0. See the --general.error switch. OUTPUT TEMPLATE The --nit.format switch allows control of output format. It takes a file name. The format of that text file is a sequence of fixed section names, enclosed in square brackets on their own lines, each optionally followed by text. In that text, certain specific identifiers, enclosed in brace pairs, are substituted. For example: [dog-section] My pet dog {{dog-name}} is a {{bad-dog}}. For examples, browse recipe/toast/output/*.nit If no file is specified, or if the file cannot be loaded, a default template is used. Note also the --nit.quote switch. EXAMPLES To verify the version of ssc: ssc -V To check the static web side source directory /home/site/wwwroot: ssc /home/site/wwwroot To check a static HTML/XHTML website for, that uses server side includes, in the current directory, with verification of external links, with rather verbose output: ssc -e -I -x html -x shtml -s -v 5 -i index.shtml To check a static web side in the current directory, with a virtual directory, verifying microformats: ssc -L vitual=/home/site/virtual -M To check a static web site using a configuration file: ssc -f config.file A simple configuration file [general] verbose=5 output=ssc.out class= [site] extension=html index=index.html root=~/www/htdocs The website for can be found in ~/www.htdocs. Standard index files are called index.html, and html files always have the html extension. The configuration outputs errors, warnings and info messages to ssc.out. It analyses class identifiers, so will scan CSS files for class names. Links and virtual directories [general] verbose=5 output=ssc.out [site] extension=html index=index.html root=~/www/htdocs virtual=/net=tests/virtual [link] check= external= xlink= Beyond the previous example, SSC will also check links, including external links and crosslinked ids. The configuration file specifies a virtual directory. When a link refers to the local directory /net, SSC will seek the corresponding file(s) in tests/virtual (relative to the current directory, not the root). Microformats [general] verbose=5 output=ssc.out class= [site] extension=html index=index.html root=~/www/htdocs [microformat] verify= The only new thing here is to turn on microformats analysis (class analysis is required). Microdata [general] verbose=5 output=ssc.out class= [site] extension=html index=index.html root=~/www/htdocs Actually, SSC will report on any microdata it finds, no matter what. Note that, if you use itemref in your pages, ssc may give false warnings in the section referenced by the itemref. This is because ssc does not always know that the referenced data is not intended to be used in its own right, unless you put it under a TEMPLATE element. You can export the microdata encountered in JSON file format by using the export setting (not shown). Statistics [general] ssi= verbose=5 output=ssc.out class= [site] extension=html extension=shtml index=index.html root=~/www/htdocs [stats] summary= Turn on summary statistics, which produces a grand total at the end of a complete run (add ""page=" below summary for individual page stats). Note that server side includes have been turned on using general.ssi, and files with the SHTML extension will be treated as web pages. MathML 2 This configuration file could be used to test MathML 2. [general] verbose=5 output=m2.out class= [link] check= [site] extension=html index=index.html root=~/www/htdocs [math] version=2 [validation] citype=function citype=list citype=logical citype=matrix citype=set citype=var-x citype=vector What's interesting here is the validation section. The MathML documentation lists a specific set of values that can be used with the TYPE attribute on the CI element. The examples in the specification give this attribute many additional, and apparently illegal, values. (This apparent inconsistency is not uncommon amongst web specifications.) The way to avoid SSC complaining about these extra values is to give CI TYPE additional values in the configuration file, as noted. Indeed, many enumerated attribute values can be extended in this way. Use the --validation switch to get a complete list. arts & ego This is the configuration file I used to analyse my site. [general] verbose=info output=~/www/live.out ignore=pre rpt= progress= class= classic= no-rel= ssi= [css] verify= extension=css version=2023++ [html] version=2023/10/01 title=80 rfc1942= rfc2070= ie= safari= [link] no-external= xlink= pretend=cgi-bin local= [nits] silence=use_double_quote_code silence=use_quote_code silence=missing_itemtype [shadow] copy=dedu root=~/www/live file=~/www/live.ndx ignore=inc info= msg=arts & ego (c) 1978-2023 dylan harris [site] extension=shtml extension=html extension=htm extension=asp index=index.shtml root=|/Sites [spell] no-check= [stats] summary= [validation] lang=ma fontname=Marain fontname=droid-sans-mono fontname=ArialMT fontname=Arial-BoldMT fontname=Times-Italic fontname=Times-Roman fontname=TimesNewRomanPSMT fontname=TrebuchetMS fontname=TrebuchetMS-Italic My site has been built by hand over the decades, and is full of errors. I wrote ssc because I was frustrated that I could find no tool to properly check it. Unfortunately, SSC is too successful: it finds a gadzillion errors in arts & ego. I am slowly making the repairs. OpenBSD website Perhaps you've grabbed a copy of the OpenBSD website from github ( [general] verbose=info output=~/www.out git= progress= sloven= [css] version=2023++ device=3 [html] version=2023/10/01 title=80 force= [link] check= no-xlink= [nits] comment=ftp_protocol [site] root=~/github/www index=index.html [stats] summary= POTENTIAL GOTCHAS For HTML, ssc defaults to a version of the living standard that's contemporary to the version of ssc. The same applies to other document types. In all cases, though, if the document itself specifies a particular version (such as HTML 3.2), then that applies. The version indicated by the DOCTYPE can be very broad. For example, the HTML 5 doctype specifies HTML 5 alone, but not which flavour (W3 or living standard), nor which revision. If you are verifying HTML, and it contains some CSS, MathML, SVG, or other content, then ssc will attempt to guess the version of the latter from the former (unless you tell it which to use). All this is why that, if you want to be precise about document verification, specify versions, either with a switch or an entry in a configuration file. If your HTML contains MathML/SVG/etc. snippets, then specify those versions too. SEE ALSO tidy linkchecker HISTORY ssc ( is written by Dylan Harris (

known issues

SSC is α software. It doesn’t do what it’s supposed to do, and what it’s supposed to do is wrong.

Note that github hosts a list of known issues.

* How can such a dangerous animal have such a cuddly name? It’s like calling a Hound of Hell ‘Fluffy’, or Death’s horse Binky.

bug reporting

SSC is α software. It may contain unexpected features. If you encounter such a delight, please help improve ssc by collecting the following information (where relevant):

and emailing everything to (if the collected files are more than small, please use a public fileserver and email the link). Do NOT send anything confidential. Furthermore, unless you request otherwise, we reserve the right to publish some or all of the information sent in future versions of ssc, usually in the test suite. If you have a fix, you are invited to submit a pull request on github. Thank you.


static site checker
(c) 2020-2023 Dylan Harris

SSC can be built on various unii with CMake and gcc or clang,
supporting C++ 17 or better, or with Visual Studios 2017 / 2019 / 2022
under Windows. I have built & tested a 64-bit version in selected OSs
on intel/amd & arm64 (m2) architecture. 

Although ssc builds and runs with older compilers on older systems, not
all features are available on them.


Common dependencies
You need boost version 1.75 or better (, a recent
version of the ICU libraries (, Microsoft's
GSL library ( (or define NO_GSL), and
a recent version of libcurl (*. Usually, an Operating
System's package system has an appropriate version ready to install.

*libcurl requires a thread-safe underlying SSL library: see

You may need to set these environment variables:
- BOOST: if you're not using your operating system's packaged flavour
  of boost, then set BOOST to your boost source root directory;
- ICU_ROOT: if you're not using your operating system's packaged ICU,
  set ICU_ROOT to your ICU source root directory;
- GSL: set it to your GSL root directory.

Building SSC under unii, including macos, requires a development
installation of hunspell (

Once you've got them, you can run cmake from the project root

The Windows build, by default, uses the native Windows spellchecker,
although, except for Windows 11, that doesn't seem to work so well in
multilingual contexts.


To build from Visual Studio, navigate to recipe/tea, open the
appropriate .sln file, then build. Only Visual Studios 2017, 2019 and
2022, have been built & tested, for amd64 (x64) and arm64 (M2), under
Windows 10 & 11.

Note the included solutions presume Microsoft's VCPKG manager. There
are alternatives.

On low memory machines, disable the /MP switch.

Unii & mock Unii
You will need CMake 3.19 or better. On Linux, you will also need
lsb-release. These can be found in most distributions' standard
packages. (For macos, I used macports, but brew is good too.)
From the home ssc directory, compile a normal build thus:
cmake .
make install

For a debug build:
cmake -DCMAKE_BUILD_TYPE=Debug .
make install

If everything works correctly, then everything will be built, a series
of tests run, with a final result at the very end saying no failures.
Having said that, given SSC is alpha, don't be too surprised to see
some warnings or some final test errors. Note in particular that
complaints about being unable to find or copy files during testing are
not of concern, these come from scripts that set up or tear down
individual tests, and the standard commands used sometimes complain if
they can't find files they're supposed to delete, which is a bit silly
given that means things are already in the desired state.

The following have, at some point, been successfully built as x64 amd:
FreeBSD:  13.2 / 13.1 / 12.4 / 12.3
Linux:    Alma Linux 9, AOSC, Centos 8 / 9, SUSE Tumbleweed,
          Ubuntu 20.04 / 22.04
MacOS:    Sonoma, Ventura, Monterey, Big Sur, Catalina, Mojave, High
          Sierra, Sierra, El Capitan, Yosemite
OpenBSD:  7.4 / 7.3 / 7.2 / 7.1 / 7.0 / 6.9 / 6.8

The following have been successfully built as arm64:
Linux:    Centos 9
MacOS:    Sonoma, Ventura

The current version of ssc requires the current version of an operating
system. Older operating systems require older versions of ssc. Not all
features work on all systems.

Centos 9
The appropriate CMake command is:
  cmake . -DFLAVOUR=CentosOSStream -DFLAVOUR_VER=9
(beware the standard English spelling of flavour.)

You may need to increase significantly the available memory setting
for your build account in login.conf.


Under Visual Studio, run ssc??-test using these arguments:

  -v -x $(ProjectDir)..\..\ssc.exe
    -f $(ProjectDir)..\toast\ssc-test\win.lst
(all one line)

Add '-d' if you want the test utility to retain intemediate files.

Under CMake, run ctest:
  ctest -V
(which runs ssc-test for you, using nix.lst).

The testing utility is rather dim; it will test features that are not
built, causing failures.

Spelling test results depend on the dictionaries installed.

Supporting libraries

If you can't find a copy of Microsoft's GSL in your system's standard
package suite, then grab a current copy from its github repository
(, then unpack, build and install it.
In Windows, remember to add its root directory to your local path.

This is in every standard unix flavour repository that I've met, but
not so easy for Windows. I suggest the following:

1  Install the Microsoft repository vcpkg (
2  Run up the visual studio 64bit native command prompt
3  Run the vcpkg bootstrap batch file, noting the comments
4  Enter "vcpkg install curl:x64-windows"
5  Add an environment variable to your account for CURL, perhaps:
6  Add the installation bin directory to your local path, perhaps:
7  If you plan to debug, also add debug\bin, perhaps:

Adapt the architecture as necessary. If you omit it, vcpkg might
carefully ignore your hardware and install x86.

Boost is to C++ as breakfast is to the working day.

Most package managers support it, including vcpkg. Many people build
their own versions using the sources found at


If everything works correctly, then everything will be built, a series of tests run, with a final result at the very end saying no failures. Having said that, given SSC is α, don’t be too surprised to see some warnings or some final test errors. Worse, some tests have dependencies that vary across systems, which can cause spurious test failures.






















































































































boot notes

Notes on folder names:

These reference documents are hoovered from various open source sites. They’re collected here for convenience; at all times the originals are correct. The subjects are: aria, activity streams, bibo, creative commons, charsets, content (RDF), content security policy, cascading style sheets, csvw, common tag, dataset quality, dbpedia, dublin core, data catalogue, did, document object model, domain, data quality, data usage, earl, ebu, fibo, foaf, good relations, grddl, HTML 1, HTML 2, HTML 3, HTML 4, HTTP, ical, its, javascript, json, lang, link relations, locn, ma-ont, marinetlo, mathML, media capture, microdata, mime, music, ns, web annotation, odrl, open graph, ontologies, openmath, org, other, owl, p3p, powder, prov, pso, qb, rddl, RDFa, RDFa, rif,, sd, sioc, skos, sm, smil, smpte, sosa, ssn, svg, time, ttml, url, vann, vcard, void, W3, webgl, webmention, whatwg, XHTML, xhv, XML, xsd, xsl, XSLT.

copyright & licence

Any dispute shall be resolved in accordance with the law of the Grand Duchy of Luxembourg.


SSC, static site checker,
copyright (c) 2020-2023 dylan harris

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
GNU General Public License for more details.

You should have received a copy of the GNU General Public
License along with this program; if not, write to the Free Software
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA


Some test files come from (some directly, in W3 documents, etc.), and are licensed as follows:


By obtaining and/or copying this work, you (the licensee) agree that you have read, understood, and will comply with the following terms and conditions.

Permission to copy, modify, and distribute this work, with or without modification, for any purpose and without fee or royalty is hereby granted, provided that you include the following on ALL copies of the work or portions thereof, including modifications:

    The full text of this NOTICE in a location viewable to users of the redistributed or derivative work.
    Any pre-existing intellectual property disclaimers, notices, or terms and conditions. If none exist, the W3C Software and Document Short Notice should be included.
    Notice of any changes or modifications, through a copyright statement on the new code or document such as
    "This software or document includes material copied from or derived from [title and URI of the W3C document]. Copyright © [YEAR] W3CÆ (MIT, ERCIM, Keio, Beihang)."




The name and trademarks of copyright holders may NOT be used in advertising or publicity pertaining to the work without specific, written prior permission. Title to copyright in this work will at all times remain with copyright holders.

This version:

Previous version:

This version makes clear that the license is applicable to both software and text, by changing the name and substituting "work" for instances of "software and its documentation." It moves "notice of changes or modifications to the files" to the copyright notice, to make clear that the license is compatible with other liberal licenses.


Some test files come from (some directly, in WhatWG documents, etc.), and are licensed under a Creative Commons Attribution 4.0 International License. See for details.

Some test files are derived from pages at They are licensed under a Creative Commons Attribution 4.0 International License. Browse for details.

Some test files are derived from pages at They are licensed under a Creative Commons Attribution 4.0 International License. Browse for details.