|
|
#391 | |
|
Addict
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() Posts: 292
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
|
Quote:
This is fixed in the code for the next release. A switch with an ops:default no longer draws that RSC-005. One without an ops:default will get the same message as in epubcheck: the switch is incomplete because its default element is missing. That also answers your first post: the old message was wrong about what was wrong. One case is still not covered: a switch directly inside a list or a table, where only list items or rows may stand. If one of your books has that, please tell me and I will handle it as well. |
|
|
|
|
|
|
#392 |
|
Addict
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() Posts: 292
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
|
epubveri 0.45.0 - one fix: an EPUB 2 ops:switch is no longer a wrong error.
https://github.com/veripublica/epubv...es/tag/v0.45.0 If you use one of the plugins, there is nothing to do: all three fetch the latest release on their own. epubveri still never writes to a book. What changed. 413Michele reported in #387 that a cover page with an ops:switch drew RSC-005, "element switch is not allowed here", even with a correct ops:default. That was a wrong error on a valid book: epubveri had no ops:switch for EPUB 2 at all. OPS 2.0.1 section 2.6.3 defines it. A switch holds one or more ops:case elements and then one ops:default, and the document is valid if it would still be valid with each switch replaced by what its ops:default contains. epubveri now checks it exactly that way, as epubcheck does. So a switch may stand anywhere its default's content may stand. That includes a list or a table, where its default holds list items, rows or cells. In #391 I said that case was not covered yet; it is now. A switch with no ops:default gets the same message as in epubcheck: the switch is incomplete because its default element is missing. What can now fail: text written directly inside ops:case or ops:default, outside any element. The schema that section 2.6.3 names as normative allows only elements there, even inside a paragraph, and epubcheck reports it too. Wrap the text in a span. How I checked. I built 27 small EPUB 2 books, each with a different switch shape, valid and invalid. On all 27, epubveri 0.45.0 and epubcheck 5.4.0 give the same verdict and the same number of errors. None of my own books uses a switch, so on all of them the output is the same as 0.44.0. If a book gets a finding you did not expect, the ID it names should make it quick to show me. crates.io: https://crates.io/crates/epubveri npm: https://www.npmjs.com/package/@verip.../epubveri-wasm |
|
|
|
|
|
#393 |
|
Addict
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() Posts: 292
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
|
epubveri 0.46.0 - no verdict changes; one missed CSS finding is now found, and a few advisories are more accurate.
https://github.com/veripublica/epubv...es/tag/v0.46.0 If you use one of the plugins, there is nothing to do: all three fetch the latest release on their own. epubveri still never writes to a book. What can now be reported. A CSS rule whose { is never closed was missed when the stylesheet happened to end in a } that belongs to something else, such as a broken string or an unclosed [ or url(. The } looked like the rule's own, so epubveri said nothing. Such a rule is now reported like any other unclosed rule: CSS-008, an error in EPUB 2 and usage in EPUB 3. I found no real book with this shape; it came up in a review of the CSS parser. With --advisory only. These come from styloria 0.21, the CSS parser epubveri uses: - A pseudo-class that Selectors 4 does not allow after a pseudo-element, such as p::before:first-child, is reported, because browsers drop the whole rule. Hover-style pseudo-classes after a pseudo-element are still fine. - Two wrong advisories are gone: x: art(a): art(b) is valid, and ::-webkit-scrollbar:horizontal is not dropped by the browsers that know it.- column-width: min-content and the other sizing keywords are reported as invalid values, since the CSS working group removed them. So is ui-serif (or ui-sans-serif, ui-monospace, ui-rounded) used as a font family name in @font-face. How I checked. On epubcheck's own test books and my own books, the output without --advisory is the same as 0.45.0. The two wrong advisories were found by epub-test-books, in its new CSS sets. If a book gets a finding you did not expect, the ID it names should make it quick to show me. crates.io: https://crates.io/crates/epubveri npm: https://www.npmjs.com/package/@verip.../epubveri-wasm |
|
|
|
|
|
#394 |
|
Bibliophagist
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() Posts: 55,954
Karma: 182837452
Join Date: Jul 2010
Location: Vancouver
Device: Kobo Sage, Libra Colour, Lenovo M8 FHD, Paperwhite 4, Tolino epos
|
Something which I've noted on and off depending on whether I run both epubcheck and epubveri and when I run them in my workflow.
epubcheck reports the following error: Code:
OEBPS/Text/cover.xhtml 7 219 Col: 25: ERROR(RSC-005): Error while parsing file: element "meta" missing one or more required attributes; expected attribute "about", "accesskey", "autocapitalize", "autocorrect", "autofocus", "class", "contenteditable", "datatype", "dir", "draggable", "enterkeyhint", "headingoffset", "headingreset", "hidden", "http-equiv", "id", "inert", "inlist", "inputmode", "is", "itemid", "itemprop", "itemref", "itemscope", "itemtype", "lang", "media", "name", "nonce", "ns:alphabet", "ns:ph", "onauxclick", "onbeforeinput", "onbeforematch", "onbeforetoggle", "onblur", "oncancel", "oncanplay", "oncanplaythrough", "onchange", "onclick", "onclose", "oncommand", "oncontextlost", "oncontextmenu", "oncontextrestored", "oncopy", "oncuechange", "oncut", "ondblclick", "ondrag", "ondragend", "ondragenter", "ondragleave", "ondragover", "ondragstart", "ondrop", "ondurationchange", "onemptied", "onended", "onerror", "onfocus", "onformdata", "ongotpointercapture", "oninput", "oninvalid", "onkeydown", "onkeypress", "onkeyup", "onload", "onloadeddata", "onloadedmetadata", "onloadstart", "onlostpointercapture", "onmousedown", "onmouseenter", "onmouseleave", "onmousemove", "onmouseout", "onmouseover", "onmouseup", "onpaste", "onpause", "onplay", "onplaying", "onpointercancel", "onpointerdown", "onpointerenter", "onpointerleave", "onpointermove", "onpointerout", "onpointerover", "onpointerrawupdate", "onpointerup", "onprogress", "onratechange", "onreset", "onresize", "onscroll", "onscrollend", "onsecuritypolicyviolation", "onseeked", "onseeking", "onselect", "onslotchange", "onstalled", "onsubmit", "onsuspend", "ontimeupdate", "ontoggle", "ontouchcancel", "ontouchend", "ontouchmove", "ontouchstart", "onvolumechange", "onwaiting", "onwheel", "popover", "prefix", "property", "resource", "slot", "spellcheck", "style", "tabindex", "title", "translate", "typeof", "vocab", "writingsuggestions", "xml:base", "xml:lang" or "xml:space" (with xmlns:ns="http://www.w3.org/2001/10/synthesis") Code:
<meta content="true"/> Code:
<meta name="cover" content="true"/> |
|
|
|
|
|
#395 |
|
Addict
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() Posts: 292
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
|
DNSB, about the <meta content="true"/> in cover.xhtml (post #394): thank you. epubcheck is right and epubveri was wrong to say nothing. HTML requires a meta element to have exactly one of name, http-equiv, charset or itemprop, and this one has none.
epubveri 0.48.0, out today, reports it as RSC-005, an error, the same ID epubcheck uses. Your search and replace to name="cover" is the right fix, and that form still passes. EPUB 2 books are not affected: XHTML 1.1 only requires content on meta, and both tools accept it there. https://github.com/veripublica/epubv...es/tag/v0.48.0 If you use one of the plugins, there is nothing to do: all three fetch the latest release on their own. epubveri still never writes to a book. Fewer errors. CSS-001, the error for direction or unicode-bidi in EPUB 3 CSS, is now reported only when the declaration survives CSS parsing. The EPUB working group settled this reading this week (w3c/epub-specs issue 3115). A declaration that CSS throws away, such as direction: sideways or one inside a rule with a broken selector, no longer counts. Eight books in epub-test-books were rejected for this and now pass in epubveri; epubcheck still rejects them, and I have opened an issue there (w3c/epubcheck issue 1781). What can now be reported. - The meta element case above, and HTML's other two meta rules: two of name, http-equiv, charset and itemprop on one element, and name or itemprop without content. Of my own books, only one draws any of these, a book with Adobe DRM meta elements that both tools already reject. - In an EPUB 3 book, OriginalLength="0" on a Compression element in META-INF/encryption.xml. The spec says it must be a positive integer. How I checked. On epubcheck's own test books and my own books, nothing else changes. epub-test-books has sets for each case, which you can run in either tool. crates.io: https://crates.io/crates/epubveri npm: https://www.npmjs.com/package/@verip.../epubveri-wasm |
|
|
|
![]() |
|
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| [Plugin] epubveri - Simple epubveri wrapper | Doitsu | Plugins | 61 | 11-19-2026 12:15 PM |
| kepubify v4: A fast, standalone EPUB to Kobo EPUB converter | geek1011 | Kobo Reader | 50 | 11-08-2026 08:21 AM |
| Epub:type not allowed in epub validator | wDr | Editor | 0 | 07-01-2016 10:03 PM |
| ePub Validator | odedta | ePub | 9 | 06-01-2014 04:35 AM |
| epub validator | fiona86 | Conversion | 2 | 06-24-2011 12:34 PM |