View Source

Python script to validate WIPO ST.92 Priority Document Data Packages (PDDP ZIP files) before REGISTRATION of priority documents in WIPO DAS REST API V2. It works with packages created by the generator (4.- Generating PDDP samples for testing) and with packages produced by any other system.

This is a helper tool for testing. It does not replace the validation performed by WIPO DAS and does not cover every rule of ST.92.

Get the validator here: PDDP_Validator.zip (version of 02.10.2026)

Contents of the package

validatorPddp/
├── validate_pddp.py                    # Validator script
├── requirements.txt                    # Python dependency (lxml)
├── README.md                           # Documentation
├── xsd/                                # Offline fallback copy of the ST.92 V2.0 schemas
│   ├── ST92PDDPIndex_V2_0.xsd
│   ├── ST96_Common_V9_0.xsd
│   ├── ST96_Design_V9_0.xsd
│   └── ST96_Trademark_V9_0.xsd
└── dtd/
    └── ST26SequenceListing_V1_3.dtd    # WIPO ST.26 DTD for sequence listings

By default the validator downloads the schemas directly from WIPO, so it always checks against the published versions:

ST.92 index schema: ST92PDDPIndex_V2_0.xsd (its three ST.96 imports are resolved automatically)
ST.96 V9.0 schemas, used for the supplementary artifacts: ST96XMLSchema_V9_0_Flattened.zip

Downloads are cached under .cache/ and reused on later runs. If a download fails, the validator falls back to the bundled xsd/ copy and reports that it did. Use --offline to force the bundled copy and never touch the network; those files are byte-identical to the published ones, so both modes validate against the same schema.

The ST.26 DTD is only distributed inside the ST.26 Annex II bundle and has no stable standalone URL, so it stays in dtd/.

Installation

The validator needs Python 3.10+ and the lxml library (XSD/DTD validation is not available in the Python standard library).

# Linux / macOS
cd validatorPddp
python3 -m venv .venv
.venv/bin/pip install -r requirements.txt

# Windows (PowerShell)
cd validatorPddp
py -m venv .venv
.venv\Scripts\pip install -r requirements.txt

On Debian / Ubuntu / Raspberry Pi OS you can instead install the system package: sudo apt install python3-lxml.

Usage

# Validate all ZIP files in ../output (default)
.venv/bin/python validate_pddp.py

# Validate specific ZIP files and/or all ZIP files in a folder
.venv/bin/python validate_pddp.py Patent_NO_20210008_20210924.zip ./received/

# Also list every check that passed
.venv/bin/python validate_pddp.py -v ./output

# Report advisory findings as warnings instead of errors
.venv/bin/python validate_pddp.py --lenient ./output

# Use the bundled schemas, without any network access
.venv/bin/python validate_pddp.py --offline ./output

All options

Option	Default	Description
`paths ...`	`../output`	ZIP files and/or folders containing ZIP files
`-v, --verbose`	off	Also list the checks that passed
`--schema URL\|FILE`	the WIPO URL above	ST.92 index schema to validate against
`--offline`	off	Use the bundled `xsd/` copy, never touch the network
`--refresh-schema`	off	Re-download the schemas, ignoring the cache
`--no-fallback`	off	Fail instead of falling back to the bundled schema
`--timeout SECONDS`	`30`	Schema download timeout
`--st96-schema URL`	the WIPO ST.96 zip above	Flattened ST.96 release for supplementary artifacts
`--st96-dir DIR`	-	Use an already-extracted ST.96 schema folder
`--no-st96`	off	Skip ST.96 validation of supplementary artifacts
`--lenient`	off	Report advisory findings as warnings instead of errors
`--st26-dtd FILE`	`dtd/ST26SequenceListing_V1_3.dtd`	ST.26 sequence listing DTD

Strict by default: advisory findings are reported as errors, so nothing is silently accepted. Pass --lenient to keep them as warnings.

Exit code: 0 when no package has errors, 1 otherwise (can be used in scripts / CI).

Checks performed

Group	Check	Result if it fails
Archive	ZIP opens and every entry passes its CRC	ERROR
Archive	No encrypted entries and no duplicate entry names	ERROR
Archive	No absolute paths, `..` traversal, backslashes or control characters in entry names	ERROR
Archive	`PriorityDocumentIndex.xml` present at the ZIP root, with exact casing	ERROR
Index	Well-formed XML, parsed with entity expansion and network access disabled	ERROR
Index	No DOCTYPE and no XML entity references	ERROR
Index	Root element is `pde:PriorityDocumentIndex`	ERROR
Index	Valid against `ST92PDDPIndex_V2_0.xsd`	ERROR
Index	Document categories match `IPRightKindCategory`	ERROR
Index	Mandatory documents present: priority document PDF (all types), certification page and representation (designs, trademarks)	ERROR
Index	Patent package without a separate certification page (it must then be part of the priority document PDF)	advisory
Index	Supplementary document category is one of the values ST.92 recommends for the IP right type. Mandatory categories are a closed enumeration enforced by the schema, but supplementary ones are declared as a union with `xsd:token`, so the schema accepts any value and a typo would otherwise pass unnoticed	advisory
Index	`ApplicationFilingDate` and `DocumentDate` parse and are not in the future	ERROR
Index	ZIP name `<Kind>_<Office>_<Number>_<YYYYMMDD>.zip` matches the index	ERROR / advisory
References	Every file listed in the index exists in the ZIP	ERROR
References	`com:FileName` contains no path component	ERROR
References	`DocumentLocationURI` is relative and does not escape the package	ERROR
References	Mandatory documents live under `MandatoryArtifacts/`, supplementary under `SupplementaryArtifacts/`	ERROR
References	`DocumentLocationURI` is an external URL; file in the ZIP not listed in the index; one file referenced by several entries	advisory
Content	No empty files	ERROR
Content	Declared `DocumentFileFormatCategory` matches both the file extension and the actual bytes	ERROR
Content	File content matches its extension (PDF, PNG, JPEG, WAV, MP3, STL, ZIP, TAR, XML)	ERROR
Content	PDF structure: header, `startxref`, `%%EOF`, `/Root`	ERROR
Content	PNG structure: per-chunk CRCs, IHDR/IDAT/IEND present, non-zero dimensions	ERROR
Content	WAV structure: RIFF size agrees with the file, `fmt` and `data` chunks present and in bounds	ERROR
Content	STL structure: facet / loop / vertex counts agree (ASCII) or size matches the triangle count (binary)	ERROR
Content	Zipped artifacts open, are unencrypted and CRC-clean	ERROR
Content	`PageTotalQuantity` matches the PDF page count; `DocumentSizeQuantity` matches the real byte size	advisory
Content	Internal XML is well-formed	ERROR
Content	Supplementary artifacts valid against the published ST.96 V9.0 schemas	ERROR
Content	ST.26 sequence listings valid against the ST.26 DTD	ERROR
Content	ST.26 cross-checks: `fileName`, `SequenceTotalQuantity`, `sequenceIDNumber` runs 1..n, `INSDSeq_length` matches the sequence	ERROR
Content	ST.26 characters within the Annex IV Basic Latin subset	advisory
Content	ST.25 sequence listings: mandatory numeric identifiers present, `<160>` matches the sequence count, `<210>` runs 1..n, `<212>` is `DNA`/`RNA`/`PRT`, ASCII only	ERROR
Content	ST.25 used only on applications filed before 2022-07-01	ERROR

Checks marked advisory are reported as errors by default and as warnings with --lenient.

Bibliographic and classification data are now validated against the published ST.96 V9.0 schemas, not just checked for well-formedness.

Note that these documents belong to the per-right ST.96 namespaces, not to Common: pat:BibliographicData and pat:PatentClassificationBag for patents, dgn:DesignApplication and com:DesignClassificationBag (Locarno) for designs, tmk:TrademarkApplication and tmk:GoodsServicesBag (Nice) for trademarks.

The trimmed ST96_Common_V9_0.xsd shipped inside the ST.92 Annex I bundle declares only 16 global elements, a subset built for the index schema. It is not a substitute for the full ST.96 Common namespace, which is why the supplementary artifacts are checked against the ST.96 release instead.

pde:*MandatoryDocumentCategory is declared with xsd:restriction, so only these values are valid and xmllint rejects anything else:

Patent	Industrial design	Trademark
Certification page	Certification page	Certification page
Priority document PDF	Priority document PDF	Priority document PDF
Sequence listing	Industrial design representation	Trademark representation

pde:*SupplementaryDocumentCategory is declared as a union of xsd:token and a base type, so the enumerated values are recommended rather than binding and an Office may legitimately use another token. The validator therefore reports an unrecognised supplementary category as advisory, to catch typos without blocking legitimate values:

Patent	Industrial design	Trademark
Abstract	Bibliographic data	Bibliographic data
Application body	Registration certificate	Registration certificate
Bibliographic data	Industrial design representation	Mark representation
Description	Classification data	Classification data
Claims	Intermediate documents	Intermediate documents
Drawings
Classification data
Preconversion document
Sequence listing

Note that com:DocumentName beside these elements is free text (xsd:string union) and carries no controlled vocabulary; the category element is the machine-readable one.

ST.92 paragraph 18 permits ST.26, ST.25 or ST.23, and footnote 2 makes ST.26 mandatory for patent applications filed on or after 1 July 2022. The validator reads pde:ApplicationFilingDate from the index and rejects an ST.25 listing on a later filing:

ERROR  ..._SequenceListing_ST25.txt: ST.25 sequence listing on an application
       filed 2023-03-15. ST.26 (XML) is mandatory for patent applications
       filed on or after 2022-07-01.

ST.25 listings are recognised by their numeric identifiers rather than by file name, so a listing named anything else is still checked. The expected DocumentFileFormatCategory is TEXT for ST.25 and XML for ST.26.

Example output

>>>> .venv/bin/python validate_pddp.py -v ./output/Patent_IB_PCT-IB2026-134592_20260103.zip
ST.92 schema: https://www.wipo.int/standards/XMLSchema/ST92/V2_0/ST92PDDPIndex_V2_0.xsd
ST.96 schema: https://www.wipo.int/standards/en/st96/v9-0/annex-iii/ST96XMLSchema_V9_0_Flattened.zip
Mode:         strict

[OK] Patent_IB_PCT-IB2026-134592_20260103.zip
    ok    ZIP integrity OK
    ok    11 ZIP entr(ies), all names safe and unique
    ok    PriorityDocumentIndex.xml valid against the ST.92 V2.0 schema
    ok    Document categories consistent with 'Patent'
    ok    Mandatory documents present: ['Certification page', 'Priority document PDF', 'Sequence listing']
    ok    ZIP file name consistent with index
    ok    All 7 referenced files resolved
    ok    MandatoryArtifacts/..._PriorityDocument.pdf: PDF OK (2,847 bytes, 3 page(s))
    ok    MandatoryArtifacts/..._SequenceListing_ST26.xml: valid against the ST.26 DTD
    ok    SupplementaryArtifacts/..._BibliographicData.xml: valid against ST.96 V9.0 (BibliographicData_V9_0.xsd)
    ok    SupplementaryArtifacts/..._ClassificationData.xml: valid against ST.96 V9.0 (Patent_V9_0.xsd)

1 package(s) checked: 1 passed, 0 failed.

Without -v only problems are listed:

>>>> .venv/bin/python validate_pddp.py ./received/
[FAIL] Patent_XX_1234567_20260101.zip
    ERROR MandatoryArtifacts/doc.pdf: declared format 'PNG' does not match extension '.pdf' (expected ['.png'])
    ERROR SupplementaryArtifacts/bib.xml ST.96: line 11: Element 'InventionTitle': This element is not expected. Expected is ( InventionTitleBag ).

1 package(s) checked: 0 passed, 1 failed.

Common errors

Message	Cause / fix
`SupplementaryDocumentBag': Missing child element(s)`	An empty `<pde:SupplementaryDocumentBag>` is not allowed. Remove the element if there are no supplementary documents.
`'Text' is not a valid value of the union type ... DocumentFileFormatCategoryType`	File format values are case sensitive: use `TEXT`, `PDF`, `XML`, …
`No declaration for element INSDSeq_organism`	Not part of ST.26. The organism must be given as the `organism` qualifier of the `source` feature.
`<PatentBibliographicData> is not a global element`	There is no such element in the ST.96 Common namespace. Bibliographic data belongs to the per-right namespaces: `pat:BibliographicData`, `dgn:DesignApplication`, `tmk:TrademarkApplication`.
`declared format 'X' does not match extension '.y'`	`DocumentFileFormatCategory` disagrees with the file it points at. Fix whichever is wrong.
`PageTotalQuantity N != M page(s) found in the PDF`	The declared page count does not match the PDF. Advisory, so it only fails the run in strict mode.
`mandatory document stored under SupplementaryArtifacts/`	Mandatory artifacts must sit under `MandatoryArtifacts/` and supplementary ones under `SupplementaryArtifacts/`.
`Path traversal in ZIP entry`	The package contains an entry with `..` or an absolute path. Rebuild it with relative paths only.
`ST.25 sequence listing on an application filed ...`	The package pairs an ST.25 listing with a filing date on or after 1 July 2022. Supply the listing in ST.26 (XML) instead.
`unrecognised DocumentFileFormatCategory 'TXT'`	ST.96 enumerates the plain-text category as `TEXT`, not `TXT`.
`content is not a valid STL/PDF/... file`	The file does not contain what its extension says (e.g. an HTML page saved with a .stl extension).

Python 3.10+ and lxml.

Network access to www.wipo.int is used to fetch the schemas on first run. Behind a proxy or offline, run with --offline to use the bundled copies.