Add an XSD

.gitmodules

@@ -0,0 +1,3 @@
+[submodule "xml-schemas/ebu-tt-m-xsd"]
+	path = xml-schemas/ebu-tt-m-xsd
+	url = https://github.com/ebu/ebu-tt-m-xsd.git

examples/non-dapt-divs.xml

@@ -5,11 +5,11 @@
     <div> <!-- This cannot be a Script Event because it has no xml:id -->
         <p> <!-- Would be a Text if its parent were a Script Event --> </p>
     </div>
-    <div> <!-- div parent of another div -->
+    <div xml:id="d2_1"> <!-- div parent of another div -->
         <div xml:id="d2"> <!-- Possibly a Script Event --></div>
     </div>
-    <div> <!-- double layer of nesting -->
-        <div>
+    <div xml:id="d3_1"> <!-- double layer of nesting -->
+        <div xml:id="d3_1_1">
             <div xml:id="d3" begin="..." end="..." xml:lang="ja" foo:bar="baz">
                 <!-- A Script Event with possibly unexpected attributes -->
             </div>

schema-validator/README.md

@@ -0,0 +1,44 @@
+# DAPT XSD Validator
+
+Basic command line utility for validating DAPT documents using
+the XML Schema Definition in the w3c/dapt repository.
+
+This script uses the MIT licensed [`xmlschema`](https://github.com/sissaschool/xmlschema) library.
+
+This script is provided as-is with no warranties of any kind.
+The repository's `LICENSE.md` applies, with the contents of this folder
+being considered a _code example_.
+
+## Build
+
+1. Install poetry - [installation instructions](https://python-poetry.org/docs/#installing-with-the-official-installer)
+2. Ensure you have a version of Python greater than or equal to 3.11 available
+for example with the command `poetry env use 3.11`
+3. Install the dependencies by running `poetry install`
+
+## Usage
+
+```sh
+poetry run validate -dapt_in path/to/dapt_file.ttml
+```
+
+or pass the document for validating in via stdin, e.g.:
+
+```sh
+poetry run validate < path/to/dapt_file.ttml
+```
+
+### Validating without pruning unrecognised vocabulary
+
+By default this script prunes unrecognised vocabulary before
+XSD validation, as required by
+[DAPT §5.2.1 Unrecognised vocabulary](https://www.w3.org/TR/dapt/#unrecognised-elements-and-attributes);
+this behaviour can be disabled via the command line parameter
+`-noprune`.
+
+
+## Tests
+
+```sh
+poetry run python -m unittest
+```

schema-validator/pyproject.toml

@@ -0,0 +1,24 @@
+[tool.poetry]
+name = "dapt-xsd-val"
+version = "0.1.0"
+description = "Thin wrapper around xmlschema to support XSD validation using the supplied DAPT XSD"
+authors = ["Nigel Megitt <[email protected]>"]
+readme = "README.md"
+packages = [
+    { include = "src" },
+    { include = "test" },
+]
+
+[tool.poetry.dependencies]
+python = ">=3.11"
+xmlschema = "^3.4.3"
+
+[tool.poetry.group.dev.dependencies]
+flake8 = "^7.1.2"
+
+[build-system]
+requires = ["poetry-core"]
+build-backend = "poetry.core.masonry.api"
+
+[tool.poetry.scripts]
+validate = "src.validate:main"

schema-validator/src/validate.py

@@ -0,0 +1,175 @@
+import os
+import sys
+import argparse
+import xmlschema
+import xml.etree.ElementTree as ElementTree
+from xml.etree.ElementTree import Element as Element
+import logging
+
+logging.getLogger().setLevel(logging.INFO)
+
+
+schema_path = os.path.normpath(
+    os.path.join(
+        os.path.dirname(__file__),
+        '../../xml-schemas/dapt.xsd')
+)
+metadata_items_schema_path = os.path.normpath(
+    os.path.join(
+        os.path.dirname(__file__),
+        '../../xml-schemas/ttml2-metadata-items.xsd')
+)
+
+recognised_namespaces = set([
+    'http://www.w3.org/XML/1998/namespace', # xml
+    'http://www.w3.org/ns/ttml', # tt
+    'http://www.w3.org/ns/ttml#parameter', # ttp
+    'http://www.w3.org/ns/ttml#audio', # tta
+    'http://www.w3.org/ns/ttml#metadata', # ttm
+    'http://www.w3.org/ns/ttml/feature/',
+    'http://www.w3.org/ns/ttml/profile/dapt#metadata', # daptm
+    'http://www.w3.org/ns/ttml/profile/dapt/extension/',
+    'urn:ebu:tt:metadata', # ebuttm
+])
+
+# We will not prune attributes in no namespace if they are
+# defined on any element in TTML or DAPT, even if they are
+# not defined on the specific element on which they occur.
+known_no_ns_attributes = set([
+    'agent',
+    'animate',
+    'begin',
+    'calcMode',
+    'clipBegin',
+    'clipEnd',
+    'condition',
+    'dur',
+    'encoding',
+    'end',
+    'family',
+    'fill',
+    'format',
+    'keySplines',
+    'keyTimes',
+    'length',
+    'name',
+    'range',
+    'region',
+    'repeatCount',
+    'src',
+    'style',
+    'timeContainer',
+    'type',
+    'weight',
+])
+
+
+def get_namespace(tag: str) -> str:
+    if (len(tag) == 0 or tag[0] != '{'):
+        return ''
+
+    if '}' not in tag:
+        raise ValueError('No closing brace found')
+
+    return tag.split('{', 1)[1].split('}', 1)[0]
+
+
+def get_unqualified_name(tag: str) -> str:
+    if '}' not in tag:
+        return tag
+
+    return tag.split('}', 1)[1]
+
+
+def prune_unrecognised_vocabulary(el: Element):
+    to_remove = []
+    for child in el:
+        if get_namespace(child.tag) not in recognised_namespaces:
+            logging.debug('pruning element {}'.format(child.tag))
+            to_remove.append(child)
+        else:
+            prune_unrecognised_vocabulary(child)
+
+    for e in to_remove:
+        el.remove(e)
+
+    for attr_key in el.keys():
+        attr_ns = get_namespace(attr_key)
+
+        attr_name = get_unqualified_name(attr_key)
+        if (attr_ns and attr_ns not in recognised_namespaces) \
+           or \
+           (not attr_ns and attr_name not in known_no_ns_attributes):
+            logging.debug('pruning {}@{}'.format(el.tag, attr_key))
+            el.attrib.pop(attr_key)
+
+    return el
+
+
+def validate_dapt(args):
+    # xmlschema gets baffled following the import of metadata_items,
+    # so make it load it explicitly instead, which seems to work.
+    schema_paths = [schema_path, metadata_items_schema_path]
+    logging.info('Creating schema from XSDs at {}'.format(schema_paths))
+    schema = xmlschema.XMLSchema(schema_paths)
+    schema.build()
+    if schema.validity:
+        logging.info('Schemas are valid')
+    else:
+        logging.error('Schemas are not valid, exiting early')
+        return -1
+
+    try:
+        logging.info('Validating document at {}'.format(args.dapt_in.name))
+        dapt_in_bytes = args.dapt_in.read()
+        in_xml_str = str(dapt_in_bytes, encoding='utf-8', errors='strict')
+        root = ElementTree.fromstring(in_xml_str)
+
+        if not args.noprune:
+            logging.info('Pruning unrecognised vocabulary')
+            prune_unrecognised_vocabulary(el=root)
+        schema.validate(source=root)
+    except xmlschema.XMLSchemaValidationError as valex:
+        logging.error(str(valex))
+        logging.error(
+            'Document is not valid {} pruning unrecognised vocabulary.'.format(
+                'before' if args.noprune else 'after'
+            ))
+        return -1
+
+    logging.info(
+        'Document is syntactically valid with respect to the '
+        'DAPT XML Schema Definition{}; '
+        'this does not check all semantic requirements of the '
+        'DAPT specification.'.format(
+            '' if args.noprune else ' after pruning unrecognised vocabulary'
+        ))
+    return 0
+
+
+def main():
+    parser = argparse.ArgumentParser()
+
+    parser.add_argument(
+        '-dapt_in',
+        type=argparse.FileType('rb'),
+        default=sys.stdin,
+        nargs='?',
+        action='store',
+        help='Input DAPT file to validate')
+    parser.add_argument(
+        '-noprune',
+        default=False,
+        required=False,
+        action='store_true',
+        help='If set, attempts to validate without '
+             'pruning unrecognised vocabulary.')
+    parser.set_defaults(func=validate_dapt)
+
+    args = parser.parse_args()
+    return args.func(args)
+
+
+if __name__ == "__main__":
+    # execute only if run as a script
+    sys.exit(main())

schema-validator/test/fixtures/valid_dapt.ttml

@@ -0,0 +1,34 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<tt xmlns="http://www.w3.org/ns/ttml"
+    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+    xmlns:tta="http://www.w3.org/ns/ttml#audio"
+    xmlns:ttm="http://www.w3.org/ns/ttml#metadata"
+    xmlns:ttp="http://www.w3.org/ns/ttml#parameter"
+    xmlns:daptm="http://www.w3.org/ns/ttml/profile/dapt#metadata"
+    xmlns:ebuttm="urn:ebu:tt:metadata"
+    ttp:contentProfiles="http://www.w3.org/ns/ttml/profile/dapt1.0/content"
+    daptm:scriptRepresents="audio"
+    daptm:scriptType="originalTranscript"
+    xml:lang="en">
+    <head>
+        <metadata xmlns:otherns="urn:some:other:namespace">
+            <ttm:agent xmlns:nttm="http://www.netflix.com/ns/ttml#metadata" 
+                nttm:voice="en-US-Wavenet-B" foo="bar" otherns:test="x" type="person" xml:id="actor_A" id="huh">
+                <ttm:name type="full">Matthias Schoenaerts</ttm:name>
+            </ttm:agent>
+            <ttm:agent type="character" xml:id="character_2">
+                <ttm:name type="alias">BOOKER</ttm:name>
+                <ttm:actor agent="actor_A"/>
+            </ttm:agent>
+            <otherns:x/>
+            <otherns:y/>
+        </metadata>
+    </head>
+    <body>
+        <div xml:id="se1" begin="3s" end="10s" ttm:agent="character_2" daptm:represents="audio.dialogue" daptm:onScreen="ON">
+            <ttm:desc daptm:descType="scene">high mountain valley</ttm:desc>
+            <metadata></metadata>
+            <p daptm:langSrc="en"><span>Look at this beautiful valley.</span></p>
+        </div>
+    </body>
+</tt>