Register Guidelines E-Books Search Today's Posts Mark Forums Read

Go Back   MobileRead Forums > E-Book Formats > ePub

Notices

Reply
 
Thread Tools Search this Thread
Old 10-13-2026, 08:55 PM   #376
Kayadelenium
Addict
Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.
 
Kayadelenium's Avatar
 
Posts: 277
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
epubveri 0.26.0 - two kinds of valid book that epubveri wrongly failed now pass, and a batch of checks now count and word their findings the way epubcheck does.

https://github.com/veripublica/epubv...es/tag/v0.26.0

If you use one of the plugins, there is nothing to do: all three fetch the latest release on their own.

Books that wrongly failed. None of these were reported by anyone; I found them while comparing epubveri with epubcheck. In an EPUB 3 book, an img with a srcset and no src was an error here; epubcheck accepts it. And an epub:switch inside a paragraph was "not allowed here"; epubcheck allows it there. One more wrong finding is gone, though it does not change a verdict: a link with only a query string, like href="?page=2", was also reported as a missing file. epubcheck reports only the query string (RSC-033), and now epubveri does the same.

What can now fail. This is a minor release because some books that passed here can now fail. In each case epubcheck already reports the same error, so a book that epubcheck passes still passes here.
- A link or image address that epubcheck's grammar rejects as a URL: a square bracket in a file name (image[1].jpg), a second # in the same address, or something before a colon that cannot be a scheme.
- An ARIA value with a stray space in it, such as aria-hidden=" true". epubcheck compares these values exactly.
- epub:type on the head or template element.
- In an EPUB 2 book, an SVG element inside an SVG title, desc or metadata, such as a tspan inside a title. SVG 1.1 allows only text there.
- In an EPUB 3 book, XHTML inside an SVG title that does not fit there, such as text followed by a body element.
- A Media Overlay clip with clipEnd="0" and no clipBegin, which makes the two equal.
None of the 544 books on my shelf changed.

Fewer repeated findings. A file that exists but is missing from the manifest (RSC-008) is now reported once, not once for every page that uses it. An element that comes before something required, such as a body with no head before it, is now one finding, "not allowed yet; missing required element", instead of two. Some foreign markup inside an SVG title, and an image address that appears in both src and srcset, were also counted twice; they are now counted once, as in epubcheck.

Only with usage messages turned on. In EPUB 3 SVG, ARIA attributes are now checked against the element's role: aria-checked needs a role that takes it, and role="checkbox" needs aria-checked. epubcheck reports these at the usage level, as informative notes, and so does epubveri. They never change a verdict.

How I know these match. For each change I built small test books, one shape per book or one shape per line, and ran them through epubcheck 5.4.0 and epubveri. That was more than 700 shapes in all, the two largest being 343 ARIA values and 283 ARIA role shapes. They now agree, finding for finding. On epubcheck's own 1086 test books, the number of books where the two tools report the same IDs but in different numbers went from 44 to 31. Most of the 31 that remain are the same finding counted differently: 17 are the font-face usage note, which epubcheck gives once per declaration and epubveri once per rule, a choice I made on purpose. Three I have not looked into yet.

If one of the new errors shows up on a book where you did not expect it, the element and attribute it names should make it quick to show me.

Last edited by Kayadelenium; 10-13-2026 at 09:02 PM.
Kayadelenium is offline   Reply With Quote
Old 10-14-2026, 12:04 PM   #377
Kayadelenium
Addict
Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.
 
Kayadelenium's Avatar
 
Posts: 277
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
epubveri 0.27.0 - text encoding is now checked in every XML file of a book, not only in the OPF, and several wrong findings about encodings are gone.

https://github.com/veripublica/epubv...es/tag/v0.27.0

If you use one of the plugins, there is nothing to do: all three fetch the latest release on their own.

Wrong findings that are gone. Nobody reported these; I found them while comparing epubveri with epubcheck. An OPF that declares Shift_JIS, windows-1254, Big5, KOI8-R or most other encodings that epubcheck's XML parser can read got a fatal "unrecognized encoding" here, and that fatal stopped every other check of the package. epubcheck reports only that the encoding is not UTF-8 (RSC-028), and now epubveri does the same. A UTF-16 SVG, a UTF-16 NCX, or a UTF-16 chapter without a byte order mark was read as if it were UTF-8 and came out as a fatal "not well-formed" that hid everything after it; a UTF-16 container.xml drew an RSC-005 as well. Each of these now gets only the encoding finding, as in epubcheck. And one book wrongly failed: an OPF that wrote encoding = "ISO-8859-1" with spaces around the = sign got RSC-028, but epubcheck does not see a declaration there and reports nothing. Now neither does epubveri.

What can now fail. This is a minor release because some books that passed here can now fail. In each case epubcheck already reports the same error, so a book that epubcheck passes still passes here.
- A chapter, SVG, NCX, Media Overlay or META-INF file that declares an encoding other than UTF-8, for example a Latin-1 chapter (RSC-028). Before, only the OPF was checked.
- A UTF-16 chapter is HTM_058 in EPUB 2 books too, and also without a byte order mark.
- A declared name that epubcheck's XML parser does not accept at all, such as SJIS, cp1252, utf8 or UTF-32, is a fatal RSC-016, as in epubcheck. Note that utf8 without the hyphen is one of them.
None of the 562 books on my shelf changed, because none of them uses an encoding other than UTF-8.

How I know these match. epubcheck's XML parser knows 341 encoding names, and no rule predicts which ones it rejects, so I built a test book for each name, once with the name in the OPF and once in a chapter, and ran all of them through epubcheck 5.4.0. With 88 more books for the other cases, that makes 770, and epubveri reports the same message IDs as epubcheck on all of them, with two exceptions. epubcheck reads the OPF and container.xml twice, so it reports their encoding twice, and epubveri reports it once. And five EBCDIC names make epubcheck say "Unable to read file", because the Java runtime has no such character set; that is Java's answer, not the parser's, so I did not copy it.

The shelf and the benchmark. I added 18 books to my test shelf, so it now has 562, and I measured both tools again on all of them for the README and the benchmark page. epubveri is about 62 times faster over the whole shelf and faster on every book. It reaches the same verdict as epubcheck on every book except 49 Project Gutenberg books, which epubcheck 5.4.0 rejects because of a confirmed bug of its own (w3c/epubcheck#1726).

If an encoding finding shows up on a book where you did not expect it, the file 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
Kayadelenium is offline   Reply With Quote
Old 10-14-2026, 06:33 PM   #378
Kayadelenium
Addict
Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.
 
Kayadelenium's Avatar
 
Posts: 277
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
epubveri 0.28.0 - MathML attributes are now checked, so a book with MathML that epubcheck rejects is now rejected here too.

https://github.com/veripublica/epubv...es/tag/v0.28.0

If you use one of the plugins, there is nothing to do: all three fetch the latest release on their own.

What can now fail. This is a minor release because some books that passed here can now fail. In each case epubcheck already reports the same error, so a book that epubcheck passes still passes here.
- MathML attributes. Until now every MathML element accepted any attribute at all. Now each element takes only the attributes epubcheck allows on it, so <mi title="...">, <math xml:lang="en"> or a definitionurl on <semantics> written in lowercase is an error (RSC-005), and so is an maction without its actiontype. The lowercase case comes from tools that treat XHTML as HTML: an HTML parser lowercases names, and MathML and SVG need some of them in camelCase. Attribute values are still not checked.
- An xml:base that points to a web address. epubveri only noticed one on the root element or in a <base> tag. epubcheck notices it on any element, and from that element to the end of the file every relative link points to the web address instead. An image after it is then a remote resource (RSC-006, and OPF-014 for the missing remote-resources property).
- In EPUB 2 books, XHTML inside the title of an SVG drawn in a chapter. The misplaced element itself was already reported; what is inside it is now checked as XHTML too.

Smaller changes. MathML whose only alternative text is a plain annotation, often TeX, now gets the accessibility note ACC-009, as in epubcheck: only alttext or annotation-xml counts. An SVG element name written in lowercase, such as lineargradient, now comes with the element and its parent named in the finding, so a repair tool can fix it; textpath and fegaussianblur, which were reported twice, are reported once.

None of the 562 books on my shelf changed, because none of them uses these shapes.

How I know these match. I built 151 small test books for these cases and ran each one through epubcheck 5.4.0 and epubveri. On 139 of them both tools report the same message IDs the same number of times. On 7 more the verdict is the same and only the count differs: in EPUB 2, epubcheck reports a misplaced SVG element a second time at its first child, and I did not copy that. The other 5 are two things I did not build, because no book on my shelf uses either: Content MathML inside annotation-xml is still not checked, and an xml:base that is a relative folder, such as sub/, is still ignored. epubcheck reports both, so on those 5 books it finds an error that epubveri misses.

If a MathML book gets a finding you did not expect, the attribute 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
Kayadelenium is offline   Reply With Quote
Old 10-15-2026, 05:56 AM   #379
JSWolf
Resident Curmudgeon
JSWolf ought to be getting tired of karma fortunes by now.JSWolf ought to be getting tired of karma fortunes by now.JSWolf ought to be getting tired of karma fortunes by now.JSWolf ought to be getting tired of karma fortunes by now.JSWolf ought to be getting tired of karma fortunes by now.JSWolf ought to be getting tired of karma fortunes by now.JSWolf ought to be getting tired of karma fortunes by now.JSWolf ought to be getting tired of karma fortunes by now.JSWolf ought to be getting tired of karma fortunes by now.JSWolf ought to be getting tired of karma fortunes by now.JSWolf ought to be getting tired of karma fortunes by now.
 
JSWolf's Avatar
 
Posts: 85,684
Karma: 154365285
Join Date: Nov 2006
Location: Roslindale, Massachusetts
Device: Kobo Libra 2, Kobo Aura H2O, PRS-650, PRS-T1, nook STR, PW3
I think this is a mistake in epubcheck. But if not, epuveeri should be doing the same. epubcheck is displaying a message for every line in the @font-face declaration except the @font-face line. epubvari is only doing so for the @font-face line. The @font-face code posted starts at line 1 in the CSS. Here is the output for that @font-face from epubcheck.

Code:
OEBPS/css/9780593983768_style.css	2	3	USAGE CSS-028: Use of Font-face declaration.
OEBPS/css/9780593983768_style.css	3	3	USAGE CSS-028: Use of Font-face declaration.
OEBPS/css/9780593983768_style.css	4	3	USAGE CSS-028: Use of Font-face declaration.
OEBPS/css/9780593983768_style.css	5	3	USAGE CSS-028: Use of Font-face declaration.
Code:
@font-face {
  font-family: Courier;
  font-style: normal;
  font-weight: normal;
  src: url(../fonts/Courier-Regular.ttf);
}
Does epubvari check <title> to see if what's in it is possibly correct? If I have an eBook that has something like <title>This is the title</title>, that would be incorrect.

Last edited by JSWolf; 10-15-2026 at 06:00 AM.
JSWolf is offline   Reply With Quote
Old 10-15-2026, 08:09 AM   #380
Kayadelenium
Addict
Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.
 
Kayadelenium's Avatar
 
Posts: 277
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
Quote:
Originally Posted by JSWolf
I think this is a mistake in epubcheck. But if not, epuveeri should be doing the same.
Both counts are true. The difference is in what each tool counts. epubcheck reports CSS-028 once for every declaration inside an @font-face rule, at that declaration's line. That is why your four descriptors give four notes on lines 2 to 5 and none on line 1. epubveri reports it once per @font-face rule, at the @font-face line. It is a usage note in both tools. It says a font is declared, not that anything is wrong, and it never changes the verdict.

Neither the EPUB specification nor the CSS one says how a validator should report this. CSS-028 is epubcheck's own informational note, so this is a choice, not a rule. I chose once per rule because you said back in #206 that this note already looks like an error in the plugin panel, and repeating it four times for one font would make that worse.

While checking epubcheck's code for this, I found something that looks like a real bug there. epubcheck attaches the font file to each note, and on a second @font-face rule the declarations that come before its own src carry the first rule's font file. That suggests the original intent was one note per font, but only epubcheck's maintainers can say. So I opened an issue at W3C asking which counting they intend, with the code lines and a small test book:

https://github.com/w3c/epubcheck/issues/1741

When they answer, epubveri will follow their intent, whichever way it goes.

If you want this to change, that issue is the right place for it. epubcheck's behaviour is decided there, by W3C, not here, and epubveri follows it. A comment from someone who reads these reports on real books every day will carry more weight there than mine. Any GitHub account can comment, and it is free. You don't need to agree with me; whether you prefer one note per rule or one per declaration, saying so there is what gets it settled for everyone.

Quote:
Originally Posted by JSWolf
Does epubvari check <title> to see if what's in it is possibly correct?
No, and epubcheck doesn't either. No specification says what a title has to say, so there is nothing to check it against. A content document's title is often a chapter name, not the book's title, and a book can really be called "This is the title". Any guess would fire on valid books, and I don't add a note unless it is true every time it fires.

What both tools do check: an empty title in an EPUB 3 content document is an error (RSC-005), and an empty dc:title in the package document is a warning (OPF-055).
Kayadelenium is offline   Reply With Quote
Old 10-22-2026, 09:24 AM   #381
Kayadelenium
Addict
Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.
 
Kayadelenium's Avatar
 
Posts: 277
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
epubveri 0.31.0 - four releases since my last post here (0.28.0). A file inside the book that cannot be read is now reported, and a few errors that epubcheck does not report are gone.

https://github.com/veripublica/epubv...es/tag/v0.31.0

If you use one of the plugins, there is nothing to do: all three fetch the latest release on their own.

What can now fail. In each case epubcheck already reports the same error, so a book that epubcheck passes still passes here.
- A file in the book whose compressed data is damaged. Before 0.29.0 such a file read as if it were absent: a damaged chapter was skipped without a word, and a damaged container.xml meant nothing was checked at all, so the book came out VALID. It is now PKG-008, FATAL, as in epubcheck; a damaged image is PKG-021 (corrupt image).
- Four selectors that epubcheck reports as CSS-008 and epubveri let through: an attribute selector carrying both the i and the s flag, :not(. a), :has(. a) and an empty :not().

Errors that are gone. Each of these could fail a book that epubcheck passes.
- A compressed mimetype file. EPUB 3.3 forbids it, but epubcheck only checks what the file says, so it is no longer an error here. It is now NEXT-016, shown only with --advisory.
- A wrong CRC on its own. epubcheck does not check it; here it could make a valid book FATAL.
- A url() inside another CSS function, such as image-set(url(cover.png) 1x). epubcheck never reads these urls, and epubveri reported a missing or undeclared target as an error. It now follows epubcheck, and with --advisory the same findings come back as NEXT-018. I reported the difference upstream: w3c/epubcheck#1743

Smaller changes.
- url( "img.png" ), with a space before the quote, is now read as a url, as epubcheck reads it. A missing target now draws RSC-007, and an image used only that way is no longer called unreferenced.
- CSS-029 (the media overlay class) now also sees rules inside @media, as in epubcheck.
- A stylesheet that declares @charset "iso-8859-1" is now read as Latin-1, as CSS says. A url in it holding an accented letter drew a false RSC-020; that finding is gone. The book still gets CSS-004 in both tools, because EPUB wants stylesheets in UTF-8 or UTF-16. How epubcheck reads a stylesheet's charset differs from the CSS spec in two places, which I reported as w3c/epubcheck#1742
- With --advisory: ADV-017, a selector whose shape CSS rejects, such as #1a, p::before span, or an element name written after an attribute selector. Browsers drop the whole rule; epubcheck accepts it, so it never touches the verdict.
- For anyone using the Rust crate: 0.29.0 depended on zip 8.6.0, which fails cargo audit, and its release notes missed that. 0.29.1 moved to zip 9.0.4. The affected code writes archives to disk, which epubveri never does, so no binary, plugin or npm package was exposed.

None of the 562 books on my shelf changed in any of these four releases; none of them has these shapes. Each case was checked on small test books run through epubcheck 5.4.0 and epubveri.

I also re-measured speed with 0.31.0 on the same 562 books: about 60 times faster than epubcheck 5.4.0 over the whole shelf. Method and the other numbers: BENCHMARK.md

If a damaged book or a stylesheet 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
Kayadelenium is offline   Reply With Quote
Old 11-01-2026, 07:22 PM   #382
Kayadelenium
Addict
Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.
 
Kayadelenium's Avatar
 
Posts: 277
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
epubveri 0.36.0 - five releases since my last post here (0.31.0). More malformed books now get the same verdict as in epubcheck, and a few errors that epubcheck does not report are gone.

https://github.com/veripublica/epubv...es/tag/v0.36.0

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 fail. In each case epubcheck 5.4.0 already reports an error, so a book that epubcheck passes still passes here.
- A URL with a scheme but without its two slashes, such as http:/example.com or https:example.com, in the NCX or a package link. epubveri did not check those places, so such a book could come out VALID. It is now RSC-020, also for ftp:, ws: and wss:.
- A query string on an image, a stylesheet link, a url in an inline style or an NCX entry: RSC-033. A path that leaves the container in the NCX or a package link: RSC-026.
- A damaged ZIP. An encrypted entry, an entry comment that is not UTF-8, a broken ZIP64 field, junk bytes before the first header or a broken end record appended to the file could all come out VALID. They are now FATAL, under PKG-027, PKG-008 or PKG-004 as in epubcheck. epubcheck's answer here comes from Java's ZIP reader, so I followed the OpenJDK source and measured on OpenJDK 27. epubcheck on another JDK may refuse less.
- An @charset, @import or @namespace inside @media or @supports: CSS-008. Browsers ignore them there.

An error that is gone. CSS-008 on a selector whose only fault is a character the CSS Syntax draft recently took out of names, such as U+2023 or a no-break space. epubcheck accepts these and browsers still read them the old way, and the draft's change is under review. It could fail a book that epubcheck passes.

Same verdict, different ID.
- A remote url() outside @font-face, such as a remote background image, is now RSC-006, not RSC-008, unless the manifest declares it as audio, video or a font.
- https:x and https:/x are invalid, and epubcheck also reads them as paths inside the book. So they now add RSC-007, and RSC-026 for https:/x, where epubveri reported them as remote or reported RSC-020 alone. Under a reading system with another root URL they would be remote, so the specification itself does not ask for those two.
- A manifest href with a fragment, such as ch1.xhtml#a, names no file: RSC-001.
- A missing target no longer adds RSC-009 or RSC-013 beside its RSC-007.

Smaller changes.
- With --advisory: ADV-019 to ADV-023, for CSS a browser drops and epubcheck accepts: a pseudo-class like a:hovr, an at-rule like @medai, an @import after the rules it must precede, a media feature like min-widht, an !important inside a keyframe. They never touch the verdict.
- For anyone using the Rust crate: 0.34.0 moved to zip 9.1.3 for three advisories, one of which reached epubveri because every book is opened through the affected call. Details are in the 0.34.0 release notes. The crate also gains check_style_text, which checks one stylesheet without a book.
- Release archives and the npm package now carry LICENSE-THIRD-PARTY.txt, with the licences of the crates compiled into them.
- The README is now short. The options, exit codes and measurements moved to AGENTS.md.

On the 562 books on my shelf, no finding was added or removed in any of these five releases; none of them has these shapes. One message, CSS-007, changed its wording on 28 of them. The cases above were checked on small test books run through epubcheck 5.4.0 and epubveri.

I also re-measured speed with 0.36.0 on the same 562 books: about 60 times faster than epubcheck 5.4.0 over the whole shelf. Method and the other numbers: BENCHMARK.md

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
Kayadelenium is offline   Reply With Quote
Old 11-08-2026, 06:28 PM   #383
Kayadelenium
Addict
Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.
 
Kayadelenium's Avatar
 
Posts: 277
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
epubveri 0.38.0 - three releases since my last post here (0.36.0). The main change is a rule, not a check: where the EPUB specification and epubcheck disagree, epubveri now follows the specification.

https://github.com/veripublica/epubv...es/tag/v0.38.0

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.

Why this matters for your books. Until now a book that passed epubcheck always passed epubveri, and anything stricter went behind --advisory. From 0.38.0, a rule the specification states with MUST or MUST NOT gives an error even where epubcheck says nothing. It also works the other way: what the specification allows passes, even where epubcheck rejects it. So a book can now pass epubcheck and fail epubveri. That happens in only a few cases, and each one is listed with its clause and the epubcheck issue under "Known differences from epubcheck" in AGENTS.md.

What can now fail (epubcheck 5.4.0 passes all of these):
- A compressed mimetype file. EPUB 3.3 says it must be stored uncompressed, and epubcheck only compares its contents. It is PKG-007. It used to be an advisory, so plugin users saw it labelled as one.
- A url() inside another CSS function, such as image-set(url(a.png) 1x), whose target is missing, undeclared, outside the container or a file: URL. epubcheck does not read urls there. It is RSC-007, RSC-008, RSC-026 or RSC-030, as for any other url(). This used to be an advisory too.
- An encryption.xml that declares the mimetype file, a META-INF file or the package document as encrypted. The specification forbids encrypting those files, and epubcheck has no check for this. epubcheck has no message ID for it either, so it is a new ID, SPEC-001. SPEC-* findings are ordinary errors, not advisories.

What now passes. Elements and attributes from another namespace in META-INF/container.xml. The specification says to remove them before checking the file. epubcheck reports them as RSC-005, and that is now reported upstream.

New checks that agree with epubcheck.
- container.xml is now checked against its schema. Before, nothing checked its structure: a version other than 1.0 in an EPUB 3 book, a missing media-type, links before rootfiles. They are now RSC-005, as in epubcheck.
- A link, image or stylesheet pointing at mimetype or into META-INF is RSC-008 or RSC-007, and a plain link to the package document is RSC-010. These references were exempt before.
- An encryption.xml with the wrong root element still marks its fonts as encrypted (RSC-004). An obfuscation entry for a missing font is RSC-007 alone, without an extra OBS-001.

Smaller changes.
- 0.37.0: with --advisory, ADV-024 for a CSS value its property cannot take, such as font-style: bold or width: 150. The plugins always pass --advisory, so plugin users see these as advisories. They never touch the verdict.
- zip 9.4.2 and flate2 1.1.11. I compared the old and new builds on 801 books, and the output was the same.
- --advisory now holds only ADV-* findings. The NEXT-* family is gone, because its two checks now report as errors under the IDs above.

Without --advisory, these three releases add or remove no finding on the 562 books on my shelf: none of them has these shapes. I checked the new cases on small test books run through epubcheck 5.4.0 and epubveri. Many of them come from epub-test-books, a set of small books, each built to show one rule of the specification.

If a book gets a finding you did not expect, the ID it names should make it quick to show me. That matters most for the cases that now fail.

crates.io: https://crates.io/crates/epubveri
npm: https://www.npmjs.com/package/@verip.../epubveri-wasm
Kayadelenium is offline   Reply With Quote
Old 11-11-2026, 01:47 PM   #384
Kayadelenium
Addict
Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.
 
Kayadelenium's Avatar
 
Posts: 277
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
epubveri 0.40.1 - three releases since my last post here (0.38.0): 0.39.0, 0.40.0 and 0.40.1. They continue the rule from 0.38.0: where the EPUB specification and epubcheck disagree, epubveri follows the specification. Most of the work went into rules that were not checked before, and into a few false errors on fixed-layout books, EPUB 2 books and stylesheets.

https://github.com/veripublica/epubv...es/tag/v0.40.1

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.

The one most likely to reach your books: an EPUB 3 img with no alt is now an error (RSC-005). EPUB 3.3 holds a content document to HTML, and HTML requires the attribute, except in the cases it lists itself: a non-empty title, a figure holding only the image with a captioned figcaption, or the generator attribute. epubcheck 5.4.0 says nothing here, so a book that passes epubcheck can fail epubveri on this alone. An empty alt="" is fine for a decorative image. In EPUB 2 nothing changes: the XHTML 1.1 DTD always required alt there, and both tools already reported it.

Other things that can now fail (all of them used to pass):
- The package document is checked against its full element tables: a link without rel or href, an attribute or child element the table does not list, sections out of order. A version of 3.0.1, 3.1 or 2.0.1 is OPF-001, because only 3.0 and 2.0 exist.
- The navigation document: a second ol in a nav, a heading after the ol, a second label in a li, or stray text, as the specification's content model says.
- EPUB 2 content documents are held to the rules OPS 2.0.1 adds on top of XHTML 1.1: a noscript must come after its script, a style type must be text/css or text/x-oeb1-css, no align="char", no SVG animation or scripting, and a param must come before the rest of an object's content. epubcheck accepts all of these except the param order.
- EPUB 2 package rules epubcheck has no check for: a malformed dc:language (OPF-092), an NCX file whose ncx element has no version attribute, an unused non-core item with no fallback (SPEC-002) and fallback attributes on the NCX item (SPEC-003).
- An EPUB 3 style type other than text/css, or a script charset other than utf-8, is RSC-005. This one agrees with epubcheck. epubveri used to accept any value.
- An XML declaration with version="1.1" is HTM-001 in every file, not only in content documents.
- A fixed-layout viewport width or height of 0 is HTM-057, because the specification asks for a positive number. epubcheck accepts 0.

False errors that are gone. Each of these books now passes:
- Fixed-layout viewports. EPUB 3.3 appendix F lets a viewport meta separate its properties with a semicolon or a space, not only a comma, and lets a property have no value. Both were HTM-047 or HTM-057 here. A name written as " viewport " now counts as a viewport, and +1200 and .5 are numbers, as CSS 2 defines them.
- EPUB 2 books in UTF-16. OCF 2.0.1 allows UTF-8 or UTF-16 for every XML file, and OPS 2.0.1 allows a UTF-16 stylesheet with a byte order mark. epubveri reported both, as epubcheck still does.
- An @charset label at the top of a stylesheet is now read the way CSS reads it. utf8, unicode-1-1-utf-8, unicode, ucs-2 and a label nobody knows all leave the sheet in UTF-8, so they are fine. latin1, us-ascii, windows-1252 and the like are still CSS-004, and the message now says which encoding the label names.
- A nav of another type with no heading, and dc-metadata in an EPUB 2 book, no longer get a finding the specification does not ask for.

Where this goes against epubcheck, each case is listed with its clause and the epubcheck issue under "Known differences from epubcheck" in AGENTS.md.

How I checked. epub-test-books has 316 published sets, small books that each show one rule of the specification together with what the specification expects. On all 316, epubveri 0.40.1 gives the verdict the specification gives; 0.38.0 did on 301, and epubcheck 5.4.0 does on 303. Many of these sets were built from cases I reported, so read that as "nothing published is open", not as a score. The sets still to come include some that every tool misses, epubveri too: ZIP files with damaged local headers, which both validators open without complaint.

If a book gets a finding you did not expect, the ID it names should make it quick to show me. That matters most for the img alt case, which epubcheck users will not have seen before.

crates.io: https://crates.io/crates/epubveri
npm: https://www.npmjs.com/package/@verip.../epubveri-wasm
Kayadelenium is offline   Reply With Quote
Old 11-13-2026, 04:49 PM   #385
Kayadelenium
Addict
Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.
 
Kayadelenium's Avatar
 
Posts: 277
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
epubveri 0.43.0 - three releases since my last post here (0.40.1): 0.41.0, 0.42.0 and 0.43.0. The rule stays the one from 0.38.0: where the EPUB specification and epubcheck disagree, epubveri follows the specification. This round removes a few errors the specification does not ask for, and adds checks for rules nobody checked before.

https://github.com/veripublica/epubv...es/tag/v0.43.0

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.

The one most likely to reach your books: in EPUB 3, a CSS parse error no longer makes a book invalid. EPUB 3.3 asks three things of a stylesheet: no direction, no unicode-bidi, and UTF-8 or UTF-16. Being free of parse errors is not one of them, and CSS itself says how a reader recovers from one. So CSS-008 is still reported, but at usage level, and it no longer changes the verdict. A nested rule such as p { span { ... } } is CSS Nesting now, not an error, and the checks look inside it. EPUB 2 is unchanged. epubcheck 5.4.0 still reports all of these as errors.

Other errors that are gone
- A script type="module" is scripted content, as HTML says, so the scripted property on its document is correct, not OPF-015. Without the property it is OPF-014. epubcheck has this the other way round.
- dir="RTL" is rtl and passes: HTML matches the value without regard to case.
- In a content document, a prefix mapped to one of the package document's vocabularies is no longer OPF-007b. Those are the package's defaults, not epub:type's.
- In EPUB 2, OCF 2.0.1's rules on file names apply instead of EPUB 3.3's: a | or a private-use character is allowed there, and a second package rootfile is a warning, not an error.
- A url() with a space or a stray quote in it, such as url(a b.png), is a broken token that CSS ignores. It no longer draws a missing-resource or invalid-URL error.

What can now fail (all of these used to pass)
- EPUB 3 content documents: an empty epub:type, a term that is only a prefix and a colon, an RDFa prefix with no colon, dir=" rtl" with a space, bdo dir="auto", and microdata itemtype, itemref or itemid without the attribute HTML says they need.
- A standalone SVG whose root is not an svg element in the SVG namespace.
- The manifest listing the mimetype file (SPEC-006). EPUB 3.3 says MUST NOT, and epubcheck gives only a usage note.
- Media overlays: elements the schema does not allow, a version other than 3.0, a second audio in a par, a repeated id, a second media:duration for the same thing, and a media:active-class written as a selector such as .active.
- The ZIP container: a "version needed to extract" other than 10, 20 or 45 (SPEC-004), a file name over 255 bytes (SPEC-005), an entry that says a data descriptor follows its data when none does (SPEC-007), and a central directory entry that points at no local header (SPEC-008). In my last post I wrote that every tool misses ZIP files with damaged local headers. Two of those cases are now reported. The other three, where a local header only disagrees with its central entry about the CRC, the size or the method, still pass: the ZIP specification has no rule that the two must match, so epubcheck's verdict stands.
- EPUB 2: a container version other than 1.0, a META-INF/manifest.xml without ODF's root, a metadata.xml element in no namespace, and a rights.xml that is not well-formed.
- Mapping a prefix to the Structural Semantics Vocabulary, also when it is written as http://idpf.org/epub/vocab/structure/# without www. That is the spelling EPUB 3.4 gives.

Where this goes against epubcheck, each case is listed with its clause and the epubcheck issue under "Known differences from epubcheck" in AGENTS.md.

How I checked. epub-test-books now has 608 published sets, small books that each show one rule of the specification together with what the specification expects. On all 608, epubveri 0.43.0 gives the verdict the specification gives. epubcheck 5.4.0 does on 544. On the 516 sets that already existed for 0.40.1, that release gave the specification's verdict on 495, and the 21 books that changed all moved to it. Many of these sets were built from cases I reported, so read that as "nothing published is open", not as a score.

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
Kayadelenium is offline   Reply With Quote
Old Today, 04:48 PM   #386
Kayadelenium
Addict
Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.Kayadelenium ought to be getting tired of karma fortunes by now.
 
Kayadelenium's Avatar
 
Posts: 277
Karma: 364036
Join Date: Jul 2026
Location: Planet Earth
Device: Kobo Forma
epubveri 0.44.0 - a small release: one new advisory, and CSS urls inside other functions are now checked like any other url.

https://github.com/veripublica/epubv...es/tag/v0.44.0

If you use one of the plugins, there is nothing to do: all three fetch the latest release on their own. They also turn advisories on, so the new one will show up there. epubveri still never writes to a book.

New advisory: a link whose spelling differs from the file name (ADV-025). This came from the kepubverto thread. In one of my books the manifest names a chapter Agatha'nin_Anahtari_split_000.html, and the table of contents links to Agatha%27nin_Anahtari_split_000.html. Both tools find the file, because they decode %27 back to an apostrophe before they look. A reading system that compares the two names as written sees two different files. My Kobo did exactly that: it counted the chapter twice, and the book opened at page 89 of 178. epubveri now reports such a link when you pass --advisory. It is usage level, so it never changes the verdict. In EPUB 3 it checks the way the URL Standard compares URLs, and in EPUB 2 it reports an encoded apostrophe, bracket, ampersand and the like.

Why only an advisory? Read literally, EPUB 3.3 section 4.2.5 makes this an error: a link must equal the file's URL after parsing, and %27 is not an apostrophe to a URL parser. But epubcheck's own test for that section calls content%26001.xhtml a correct link to content&001.xhtml, and the editors described it the same way in an earlier issue. The text and its own test disagree, so I asked the EPUB Working Group which one is meant: w3c/epub-specs#3116. Until they answer, epubcheck's verdict stands.

What can now fail: a url() inside another CSS function, such as image-set(url(a.png) 1x), gets the same checks a url() gets on its own. A space or a stray character in it is an invalid URL, and a ?query is not allowed. A remote file is allowed only for fonts, audio and video, and then the document needs the remote-resources property. Its target was already checked. epubcheck 5.4.0 never looks inside the function. One check stays at the top level on purpose: a function like image-set() can list its own fallback, so a nested url is not reported for a missing fallback.

Where this goes against epubcheck, each case is listed with its clause and the epubcheck issue under "Known differences from epubcheck" in AGENTS.md.

How I checked. epub-test-books now has 626 published sets. On all 626, epubveri 0.44.0 gives the verdict the specification gives, and epubcheck 5.4.0 does on 547. epub-test-books also built 138 books for the spelling cases above. They are not published yet.

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
Kayadelenium is offline   Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
kepubify v4: A fast, standalone EPUB to Kobo EPUB converter geek1011 Kobo Reader 50 11-08-2026 08:21 AM
[Plugin] epubveri - Simple epubveri wrapper Doitsu Plugins 59 09-26-2026 07:36 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


All times are GMT -4. The time now is 08:47 PM.


MobileRead.com is a privately owned, operated and funded community.