E-books are structured archives pretending to be documents, so this category gives readers and converters something real to work through. The EPUB set ships a valid EPUB 3 with a cover, a navigation document, styled chapters, and a correctly stored OCF mimetype, paired with an intentionally invalid twin whose mimetype is compressed and mislabelled, so you can test how a reader handles a malformed container. There's a FictionBook (FB2) file (a single self-contained XML e-book with full metadata) and a CBZ comic archive built from generated pages. Every book is original content, documented down to its spine and manifest, and safe to open anywhere.
Filter e-books on Browse · 90 files · 24 subcategories
Its OCF mimetype entry is compressed and mislabelled instead of being stored first and uncompressed, so you can test how a reader handles a malformed EPUB container against a correct EPUB 3 twin.
No, every e-book is original generated content with a documented spine and manifest, safe to open and redistribute.
A valid EPUB 3 (cover, navigation document, styled chapters, correct OCF mimetype) and its intentionally-invalid twin, a self-contained FictionBook (FB2) XML e-book, and a CBZ comic archive built from generated pages.
An intentionally invalid EPUB whose META-INF/container.xml contains an empty rootfiles element, so nothing points at the package document - even though OEBPS/content.opf is present and correct. Intentionally invalid and clearly labelled; it separates readers that guess from readers that give up cleanly.
An intentionally invalid EPUB where two manifest items carry the same id, so the spine's references become ambiguous and one chapter can be reached twice while another cannot be reached at all. Intentionally invalid and clearly labelled - an XML ID uniqueness violation, not a container defect.
An intentionally invalid EPUB that declares chapter3.xhtml in the manifest, the spine and the contents, while the file itself is not in the ZIP. Intentionally invalid and clearly labelled - for testing the difference between a missing resource and an unreadable book.
An intentionally invalid EPUB whose spine ends with an itemref pointing at a manifest id that does not exist. Intentionally invalid and clearly labelled - the first three chapters are intact, so a reader can recover if it chooses to.
An intentionally invalid EPUB whose second chapter is declared as text/html instead of application/xhtml+xml, making it a foreign resource in the reading order with no fallback. Intentionally invalid and clearly labelled; the bytes of that file are ordinary, valid XHTML.
A CBZ comic archive: a ZIP of sequentially named JPEG pages, the standard container for digital comics. Built from generated pages; for testing comic readers and CBZ→PDF conversion.
An EPUB 3 carrying the accessibility metadata distributors now require: schema:accessMode, accessModeSufficient, four accessibilityFeature values, a hazard statement, a written summary, and a dcterms:conformsTo link. The claims are true of the file, so a checker can verify rather than take the metadata's word.
An EPUB 3 whose one content document declares properties="scripted" and contains a single inline script that rewrites one paragraph. It makes no network request and reads correctly with scripting switched off, so it tests the property and the fallback rather than the script.
A plain-text index of every EPUB and FB2 fixture in this wave: its validity class, its subcategory, and the scenario and expected result published for it. Generated from the same entry list that builds the catalogue, so it cannot describe a file that is not there.
A plain EPUB 2.0.1 build of a two-chapter book: NCX navigation, an OPF guide with cover and text references, and the meta name="cover" convention. One third of a conversion set that also ships as EPUB 3 and FB2.
The EPUB 3 build of the same two chapters: an XHTML navigation document with an NCX kept for older readers, and the cover declared as a manifest property. The middle member of the EPUB2 / EPUB3 / FB2 conversion set.
The FictionBook build of the same two chapters: one self-contained XML file with a description block, no container and no manifest. The third member of the conversion set, and the one that has to invent an equivalent for everything the EPUB package documents declare.
The same SAMPLE book content with a sans-serif stylesheet: CSS twin of the serif EPUB.
An EPUB 3 sample styled with a serif stylesheet. Pair with the sans CSS twin for reader CSS tests.
The same book obfuscated the other way: Adobe's scheme XORs the first 1024 bytes with the sixteen raw bytes of the identifier UUID. Two schemes, two byte counts, two algorithm URIs - the pair that catches a reader which implements only one and assumes the other.
The same book with its embedded font obfuscated by the IDPF algorithm: the first 1040 bytes are XORed with the SHA-1 of the whitespace-stripped unique identifier, and META-INF/encryption.xml declares it. Paired with the plain-embedding twin so a de-obfuscator can be checked byte for byte.
An intentionally invalid EPUB whose embedded font is IDPF-obfuscated while META-INF/encryption.xml is missing, so nothing in the container says the bytes were scrambled. Intentionally invalid and clearly labelled - this is the real-world failure where a book silently renders in a substitute face.
An EPUB 3 that embeds a TrueType face through @font-face with no obfuscation and no META-INF/encryption.xml. The control member of the obfuscation set: extract the font and it is a working TTF.
A valid EPUB 3 e-book with a cover image, an EPUB navigation document, styled chapters, and a correctly stored OCF mimetype. Original content, for testing EPUB readers and EPUB→other conversion.
An intentionally invalid EPUB: identical to the valid twin except its OCF mimetype entry is compressed and mislabelled as application/zip, a real EPUB spec violation. For testing how readers and validators handle a malformed container. Clearly labelled; harmless content.
A valid EPUB 2.0.1 e-book navigated by an NCX table of contents and an OPF <guide>, with a cover: the older EPUB edition. Paired with an identical EPUB 3 twin for testing EPUB2↔EPUB3 conversion.
The same book as EPUB 3, navigated by an XHTML navigation document (with an NCX fallback for older readers). The modern twin in an EPUB2↔EPUB3 conversion set.
The cover convention that predates every standard way of doing it: a bare meta name="cover" pointing at a manifest id, with no guide reference, no cover XHTML page and nothing in the spine. For testing cover extraction against the messiest of the three EPUB 2 conventions.
An EPUB 2.0.1 that records print pagination the older way: an NCX pageList of pageTarget elements pointing at anchors in the text. The EPUB 2 counterpart of the EPUB 3 page-list nav, and what a converter has to translate between.
A FictionBook whose cover is a base64-encoded PNG in a binary element at the foot of the file, wired to a coverpage in the description by xlink id. For testing FB2 image extraction and the converters that have to turn that blob back into a real file.
A Cyrillic FictionBook stored in windows-1251 and declaring that encoding in its XML declaration - the single-byte legacy encoding a large share of real FB2 files still use. For testing importers that assume every XML file is UTF-8 and produce mojibake instead of an error.
A FictionBook nested three sections deep, with prose directly inside an outer section, an epigraph, a cite block, and a footnote linking into a second body named notes. The FB2 equivalents of a multi-level table of contents and non-linear back matter, in one file.
A FictionBook (FB2) e-book: a single self-contained XML file with full title and document metadata and three chapters. Popular in e-reader ecosystems; for testing FB2 parsers and FB2→EPUB conversion.
A minimal FictionBook (.fb2) SAMPLE with one section, for FB2/XML ebook importers.
A fixed-layout EPUB 3: rendition:layout is pre-paginated, every page declares a 600x800 viewport, and all content is absolutely positioned. Paired with a reflowable twin carrying the same words, so you can see which of the two a reader is actually honouring.
A fixed-layout EPUB 3 that asks for synthetic spreads in landscape and pins each page to a side with rendition:page-spread-left and rendition:page-spread-right. For testing two-up spread rendering and the side assignment readers get wrong first.
A fixed-layout EPUB 3 that requests a portrait orientation lock and suppresses spreads entirely. For testing whether a reader applies rendition:orientation, ignores it, or rotates the content anyway on a landscape device.
An intentionally invalid fixed-layout EPUB: rendition:layout says pre-paginated but no content document declares a viewport meta, so the canvas size is undefined. Intentionally invalid and clearly labelled - for testing validators and the fallback a reader picks when it has no dimensions to scale to.
The reflowable twin of the fixed-layout plate book: rendition:layout is reflowable, nothing is absolutely positioned, and the text is free to rewrap. Use it as the control when checking whether a reader honours the pre-paginated flag or silently reflows everything.
The valid EPUB converted to AZW3 (Kindle KF8) with Calibre: the modern Kindle format. For testing AZW3 readers and EPUB→AZW3 conversion.
The valid EPUB converted to MOBI (Mobipocket) with Calibre: the classic Kindle format. For testing MOBI readers and EPUB→MOBI conversion.
EPUB 3 SAMPLE with landmarks nav for reader TOC/landmark tests.
EPUB 3 SAMPLE with landmarks nav for reader TOC/landmark tests.
EPUB 3 SAMPLE with landmarks nav for reader TOC/landmark tests.
EPUB 3 SAMPLE without landmarks nav for reader TOC/landmark tests.
Additional EPUB 3 SAMPLE volume 5 with landmarks for batch reader tests.
Additional EPUB 3 SAMPLE volume 6 with landmarks for batch reader tests.
Additional EPUB 3 SAMPLE volume 7 with landmarks for batch reader tests.
Additional EPUB 3 SAMPLE volume 8 with landmarks for batch reader tests.
Additional EPUB 3 SAMPLE volume 9 with landmarks for batch reader tests.
Additional EPUB 3 SAMPLE volume 10 with landmarks for batch reader tests.
Plain-text landmark note SAMPLE #1 companion for EPUB nav tests.
Plain-text landmark note SAMPLE #2 companion for EPUB nav tests.
Plain-text landmark note SAMPLE #3 companion for EPUB nav tests.
Plain-text landmark note SAMPLE #4 companion for EPUB nav tests.
Plain-text landmark note SAMPLE #5 companion for EPUB nav tests.
Plain-text landmark note SAMPLE #6 companion for EPUB nav tests.
Plain-text landmark note SAMPLE #7 companion for EPUB nav tests.
Plain-text landmark note SAMPLE #8 companion for EPUB nav tests.
The same equation carrying the fallbacks a reader without a MathML engine needs: an altimg raster, its declared dimensions and baseline offset, and an alttext reading of the formula for speech. For testing fallback selection and what a screen reader announces.
An EPUB 3 with a block MathML equation and the required properties="mathml" on its content document. The baseline for MathML support, paired with a twin that omits the declaration.
An intentionally invalid EPUB: the content document contains MathML while its manifest item omits properties="mathml", so the package understates what a reader needs. Intentionally invalid and clearly labelled - identical to its declared twin in every other byte of markup.
An EPUB 3 with a real media overlay: a SMIL document ties three paragraph ids to clip ranges in an embedded MP3, and the package declares media:duration and media:active-class. The narration track is deliberately silent audio, so the fixture stays small and byte-identical on every run.
Three chapters, three SMIL overlays and three MP3 tracks, with a media:duration refines entry per overlay plus the package total. For testing whether a reader plays continuously across document boundaries and whether it sums the per-document durations correctly.
A working media overlay whose package omits the required media:duration metadata entirely. Everything else is intact, so this fixture isolates one question: does the reader refuse the overlay, or fall back to measuring the audio itself?
An intentionally invalid media overlay: the second par's text src points at a fragment identifier that does not exist in the chapter. Intentionally invalid and clearly labelled - for testing whether playback skips the orphaned segment, stops dead, or throws.
Two dc:title elements distinguished only by refines chains into title-type and display-seq, a creator with file-as and a MARC relator role, and a series declared with belongs-to-collection. The structure a catalogue actually needs, and the one flat metadata readers collapse into a single confused string.
A title containing two characters above the Basic Multilingual Plane - a blue book and a musical clef - plus a Japanese description. Anything that counts characters as UTF-16 code units, or truncates a title to a byte length, splits a surrogate pair here.
A package document whose title and creator carry precomposed accents, a combining ring above a bare letter, and a typographic apostrophe. For testing normalisation, sorting and filename derivation - the combining mark is the one that breaks naive length and comparison code.
Every non-ASCII character in the metadata is written as a hexadecimal numeric character reference, so the package document is pure ASCII on disk while the title is Han and Cyrillic once parsed. For testing tools that read metadata with a regex instead of an XML parser.
The package document is encoded as UTF-16 little-endian with a byte order mark while the content documents stay UTF-8 - a combination EPUB permits and almost nothing tests. For finding the parser that opens the OPF as UTF-8 and sees a file full of null bytes.
The package document, the navigation document and the chapter each begin with a UTF-8 byte order mark before the XML declaration. Legal, and the classic reason a strict parser reports content before the prolog on a file that looks perfectly fine in an editor.
An intentionally invalid EPUB 3 that ships only an NCX: no manifest item carries properties="nav", which EPUB 3 requires. Intentionally invalid and clearly labelled - it separates readers that fall back to the NCX from readers that refuse to open a book with no navigation document.
A book whose two navigation surfaces tell different stories: the NCX reorders the chapters, omits one, and points at a chapter9.xhtml that does not exist, while nav.xhtml is correct. For testing which surface a reader trusts and how it handles a dead NCX target.
The dual-navigation build every commercial EPUB 3 ships: an XHTML navigation document for EPUB 3 readers and a matching NCX for EPUB 2 ones, listing the same three chapters in the same order.
A conforming EPUB 3 navigated solely by an XHTML navigation document, with no NCX fallback at all. The modern baseline, and the twin of the NCX-only book that inverts it.
An EPUB 3 carrying print-page navigation: a hidden page-list nav, matching pagebreak markers in the content with doc-pagebreak roles, and an equivalent NCX pageList. For testing go-to-page features and the page-number reporting that accessibility conformance asks for.
A six-chapter original short novel as EPUB 3, with a cover and navigation: longer, real narrative prose for testing paginators, readers, reflow, and EPUB conversion at a realistic length. Original content, free to use.
DRM-free AZW/MOBI7 e-book source copied byte-for-byte from the catalog's valid reader fixture for legacy import testing. Stable P8 artifact p8-convert-legacy-azw.
Valid CBR comic container with two deterministic PNG pages in lexical reading order for legacy comic conversion. Stable P8 artifact p8-convert-legacy-cbr.
A plain-text multi-chapter SAMPLE extract, for ebook converters that accept .txt sources.
The pathological spine: every itemref is marked linear="no", so the book declares no default reading order at all. For testing what a reader opens to when the sequence it was going to page through is empty.
Three linear chapters followed by endnotes and a colophon marked linear="no": present in the package and linked from the text, but outside the page-through sequence. The single most misread spine attribute in EPUB.
An intentionally invalid spine that references chapter one twice, in first and second position. Intentionally invalid and clearly labelled - it exposes pagination and progress code that assumes spine positions map one-to-one onto documents.
SVG written straight into an XHTML content document, with properties="svg" declared on the manifest item as EPUB 3 requires. The other half of the SVG question: inline vector markup rather than a standalone content document.
An EPUB 3 whose first spine item is an SVG content document rather than XHTML, followed by a chapter that explains what should have appeared. For testing readers that assume every reading-order document is XHTML.
A fixed-layout cover built as an SVG content document that pulls in a PNG through xlink:href and centres it as a single-page spread. For testing cover extraction, which usually looks for a raster and finds vector markup instead.
Exactly the same content and the same set of targets, listed as one flat level. The reference twin for the nested navigation fixture: if a reader renders both identically, it is discarding hierarchy.
A navigation document nested four levels deep with ordered lists inside list items, targeting heading anchors inside two part documents. Paired with a flattened twin so a reader that silently collapses hierarchy is easy to spot.
The same four-level hierarchy expressed as nested NCX navPoints with a depth-first playOrder and a matching dtb:depth, alongside the equivalent nav document. For testing the older navigation path, which nests through element containment rather than through lists.
The left-to-right control for the direction set: identical package structure to the Arabic fixture, differing only in the language tag and the spine direction. A reference point, not a test case on its own.
Left-to-right prose with an Arabic run isolated by dir="rtl" and a Latin identifier isolated by dir="ltr" inside an Arabic phrase. The everyday bidi case that breaks when a renderer applies paragraph direction to every character in the paragraph.
An EPUB 3 in Arabic that declares page-progression-direction="rtl" in the spine, dir="rtl" on the body, and a right-aligned stylesheet. Paired with a left-to-right control carrying the same structure, so a reader's page-turn direction can be checked rather than assumed.
The right-to-left case again in an unconnected script: Hebrew, with the same rtl spine direction and body dir. Arabic shaping and Hebrew ordering fail independently, so a reader that handles one is not proven to handle the other.
A Japanese EPUB 3 set in vertical-rl, declaring writing-mode with the standard property plus the -epub- and -webkit- prefixes readers actually implement, and an rtl spine so pages turn the way a tategaki book does.
