generate index.html automatically based on the jsonld

kuefmz · Jun 13, 2024 · 5fd47ec · 5fd47ec
1 parent d2efbdf
commit 5fd47ec
Show file tree

Hide file tree

Showing 7 changed files with 986 additions and 497 deletions.
diff --git a/.gitignore b/.gitignore
@@ -0,0 +1,2 @@
+__pycache__/
+*.py[cod]
diff --git a/index.html b/index.html
diff --git a/src/generate_doc.py b/src/generate_doc.py
@@ -0,0 +1,144 @@
+import requests
+import json
+from jinja2 import Environment, FileSystemLoader
+from static import namespace_descriptions, codemeta_properties_data
+from static import group_leaders, team_members
+from static import context_desciption
+
+# Get schema.orh jsonld
+schemaorg_url = "https://raw.githubusercontent.com/schemaorg/schemaorg/main/data/releases/26.0/schemaorg-all-http.jsonld"
+schemaorg_response = requests.get(schemaorg_url)
+
+# Get codemeta jsonld
+headers = {
+    "Accept": "application/ld+json"
+}
+codemeta_url = "https://w3id.org/codemeta"
+#codemeta_url = "https://raw.githubusercontent.com/codemeta/codemeta/master/codemeta.jsonld"
+codemeta_response = requests.get(codemeta_url, headers=headers)
+
+
+if schemaorg_response.status_code == 200 and codemeta_response.status_code == 200:
+    schemaorg_data = schemaorg_response.json()
+    codemeta_data = codemeta_response.json()
+
+
+with open('development/context.jsonld', 'r') as f:
+    data = json.load(f)
+
+
+def get_item_by_id(graph, item_id):
+    """Retrieve an item from the graph by its @id."""
+    for item in graph:
+        if item.get('@id') == item_id:
+            return item
+    return None
+
+
+def get_expected_types_and_urls(item):
+    range_includes = item.get('schema:rangeIncludes', [])
+    if not isinstance(range_includes, list):
+        range_includes = [range_includes]
+    expected_types = [rtype.get('@id', '') for rtype in range_includes if isinstance(rtype, dict)]
+    expected_types_url = []
+    for type in expected_types:
+        if 'schema' in type:
+            expected_types_url.append('http://schema.org/' + type.split(':')[-1])
+        else:
+            expected_types_url.append('')
+    return expected_types, expected_types_url
+
+
+def get_description(item, schemaorg_data, codemeta_data):
+    if 'schema:' in item.get('@id', ''):
+        return next((x['rdfs:comment'] for x in schemaorg_data['@graph'] if x['@id'] == item.get('@id', '')), '')
+    elif 'codemeta:' in item.get('@id', ''):
+        return next((x['rdfs:comment'] for x in codemeta_data['@context'] if x['@id'] == item.get('@id', '')), '')
+    return item.get('rdfs:comment', '')
+
+
+def format_property(item, schemaorg_data, codemeta_data):
+    property_name = item.get('@id', '')
+    if property_name.split(':')[-1].isupper():
+        return
+    property_url = ''
+    if 'schema:' in item.get('@id', ''):
+        property_url = 'http://schema.org/' + property_name.split(':')[-1]
+
+    description = get_description(item, schemaorg_data, codemeta_data)
+
+    expected_types, expected_types_url = get_expected_types_and_urls(item)
+    return {
+        'property': property_name,
+        'property_url': property_url,
+        'expected_type_and_urls': list(zip(expected_types, expected_types_url)),
+        'description': description
+    }
+
+# Extract FAIR4ML properties
+fair4ml_properties = [format_property(item, schemaorg_data, codemeta_data) for item in data.get('@graph', []) if item.get('@id', '').startswith('fair4ml:')]
+
+# Extract  schema.org profiles
+schema_profiles = []
+for key, value in data['@context'].items():
+    if type({}) == type(value):
+        item_id = value.get('@id')
+        item = get_item_by_id(schemaorg_data['@graph'], item_id)
+    else:
+        continue
+    if item_id.split(':')[-1][0].isupper():
+        expected_types, expected_types_url = get_expected_types_and_urls(item)
+        profile = {
+            'profile': item_id,
+            'profile_url': 'http://schema.org/' + item_id.split(':')[-1],
+            'description': item.get('rdfs:comment', '')
+        }
+        schema_profiles.append(profile)
+
+# Extract schema.org and codemeta properties
+schema_properties = []
+codemeta_properties = []
+for property, val in data['@context'].items():
+    if not isinstance(val, str):
+        property_name = val['@id'].split(':')[-1]
+        if property_name[0].islower():
+            if 'schema' in val['@id'].split(':')[0]:
+                item = get_item_by_id(schemaorg_data['@graph'], val['@id'])
+                schema_properties.append(format_property(item, schemaorg_data, codemeta_data))
+            else:
+                codemeta_property_info = codemeta_properties_data[property_name]
+                property_json = {
+                    'property': property_name,
+                    'property_url': "https://w3id.org/codemeta/"+property_name,
+                    'expected_type_and_urls': list(zip([codemeta_property_info['type']], "http://schema.org/"+codemeta_property_info['type'])),
+                    'description': codemeta_property_info["desc"]
+                }
+                codemeta_properties.append(property_json)
+
+# Extract namespaces
+namespaces = []
+for prefix, url in data.get('@context', {}).items():
+    if isinstance(url, str):
+        namespaces.append({'prefix': prefix, 'url': url, "description": namespace_descriptions[prefix]})
+
+context = {
+    'description': context_desciption,
+    'fair4ml_properties': fair4ml_properties,
+    'schema_properties': schema_properties,
+    'codemeta_properties': codemeta_properties,
+    'schema_profiles': schema_profiles,
+    'namespaces': namespaces,
+    'group_leaders': group_leaders,
+    'team_members': team_members
+}
+
+env = Environment(loader=FileSystemLoader('.'))
+template = env.get_template('src/template.html')
+
+html_output = template.render(context)
+
+# Save index.html
+with open('index.html', 'w') as f:
+    f.write(html_output)
+
+#print("HTML file has been generated successfully.")
diff --git a/src/requirements.txt b/src/requirements.txt
@@ -0,0 +1,2 @@
+json
+jinja
diff --git a/src/static.py b/src/static.py
@@ -0,0 +1,37 @@
+group_leaders = ['Daniel Garijo', 'Leyla Jael Castro']
+team_members = ['Jenifer Tabita Ciuciu-Kiss', 'Dhwani Solanki', 'Dietrich Rebholz-Schuhmann']
+
+context_desciption = "'In this document we propose a simple way for MLModel to self-describe their genetic variant cardinality service for better integration.'"
+
+namespace_descriptions = {
+    "rdf": "This is the RDF Schema for the RDF vocabulary terms in the RDF Namespace, defined in RDF 1.1 Concepts.",
+    "rdfs": "The RDF Schema vocabulary (RDFS).",
+    "owl": "This ontology partially describes the built-in classes and properties that together form the basis of the RDF/XML syntax of OWL 2.",
+    "schema": "Schema.org is a collaborative, community activity with a mission to create, maintain, and promote schemas for structured data on the Internet, on web pages, in email messages, and beyond.",
+    "codemeta": "Codemeta is a minimal metadata schema for science software and code, in JSON and XML.",
+    "fair4ml": "FAIR4ML metadata schema for machine learning models.",
+    "cr": "Datasets are the basis of machine learning (ML). However, a lack of standardization in the description and semantics of ML datasets has made it increasingly difficult for researchers and practitioners to explore, understand, and use all but a small fraction of popular datasets."
+}
+
+codemeta_properties_data = {
+    "buildInstructions": {
+        "type": "schema:URL",
+        "desc": "Link to installation instructions/documentation."
+    },
+    "developmentStatus": {
+        "type": "schema:Text",
+        "desc": "Description of development status, e.g. Active, inactive, suspended. See repostatus.org ."
+    },
+    "issueTracker": {
+        "type": "schema:URL",
+        "desc": "Link to software bug reporting or issue tracking system."
+    },
+    "readme": {
+        "type": "schema:URL",
+        "desc": "Link to software Readme file."
+    },
+    "referencePublication": {
+        "type": "schema:ScholarlyArticle",
+        "desc": "An academic publication related to the software."
+    }
+}
diff --git a/src/template.html b/src/template.html
@@ -0,0 +1,185 @@
+<!DOCTYPE html>
+<html lang="en">
+<html>
+
+<head>
+    <meta charset="utf-8">
+    <meta http-equiv="X-UA-Compatible" content="chrome=1">
+    <meta name="viewport" content="width=device-width, initial-scale=1">
+    <meta name="msapplication-TileColor" content="#da532c">
+    <meta name="theme-color" content="#ffffff">
+    <title>
+        FAIR4ML
+
+    </title>
+    <meta name="description" content="">
+    <!-- Styling  -->
+    <link href='https://fonts.googleapis.com/css?family=Noto+Sans' rel='stylesheet' type='text/css'>
+    <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/css/bootstrap.min.css"
+        integrity="sha384-Vkoo8x4CGsO3+Hhxv8T/Q5PaXtkKtu6ug5TOeNV6gBiFeWPGFN9MuhOf23Q9Ifjh" crossorigin="anonymous">
+    <!-- JavaScript  -->
+    <script src="https://ajax.googleapis.com/ajax/libs/jquery/3.6.0/jquery.min.js"></script>
+</head>
+
+<body>
+    <div class="container min-vh-100">
+        <main id="main" class="my-5">
+            <h1>FAIR4ML <u></u> Profile</h1>
+
+            <h3><strong>Version:</strong> 0.0.1 (2024.06.04.)</h3>
+            <h2>A convention for FAIR4ML to self-describe. </h2>
+            <br />
+            <p>If you spot any errors or omissions with this type, please file an issue in our <a
+                    href="https://github.com/RDA-FAIR4ML/FAIR4ML-schema" target="_blank">GitHub</a>.</p>
+            <div class="tab-content py-3 mb-3 border-bottom" id="nav-tabContent">
+
+                <div class="tab-pane fade show active" id="nav-description" role="tabpanel" aria-labelledby="nav-description-tab">
+                    <h3>Description</h3>
+                    <p>{{ description }}</p>
+                    <h3>Schema.org hierarchy</h3>
+                    This Profile fits into the schema.org hierarchy as follows:
+                    <br />
+                    <br />
+                    <hr />
+                    <a href="http://schema.org/Thing">Thing</a>
+                    &gt;
+                    <a href="http://schema.org/CreativeWork">CreativeWork</a>
+                    &gt;
+                    <a href="https://github.com/RDA-FAIR4ML/FAIR4ML-schema">MLModel</a>
+                    <hr />
+                </div>
+
+                <div class="table-responsive shadow rounded mt-4 mb-5">
+                    <table class="table table-hover table-borderless mb-0 spec-prof-table">
+                        <thead>
+                            <tr>
+                                <th>Property</th>
+                                <th>Expected Type</th>
+                                <th>Description</th>
+                            </tr>
+                        </thead>
+                        <tbody>
+                        {% for property in fair4ml_properties %}
+                            <tr>
+                                <td class="property-column"><a href="{{ property.property_url }}">{{ property.property }}</a></td>
+                                <td>
+                                    {% for type, url in property.expected_type_and_urls %}
+                                        <a href="{{ url }}">{{ type }}</a><br/>
+                                    {% endfor %}
+                                </td>
+                                <td class="description-column">{{ property.description }}</td>
+                            </tr>
+                        {% endfor %}
+                        </tbody>
+                    </table>
+                </div>
+
+                <h3>Schema.org Profiles</h3>
+                <div class="table-responsive shadow rounded mt-4 mb-5">
+                    <table class="table table-hover table-borderless mb-0 spec-prof-table">
+                        <thead>
+                            <tr>
+                                <th>Profiles</th>
+                                <th>Expected Type</th>
+                                <th>Description</th>
+                            </tr>
+                        </thead>
+                        <tbody>
+                        {% for profile in schema_profiles %}
+                            <tr>
+                                <td class="property-column"><a href="{{ profile.profile_url }}">{{ profile.profile }}</a></td>
+                                <td class="property-column"><a href="{{ profile.profile_url }}">{{ profile.profile }}</a></td>
+                                <td class="description-column">{{ profile.description }}</td>
+                            </tr>
+                        {% endfor %}
+                        </tbody>
+                    </table>
+                </div>
+
+                <h3>Schema.org Properties</h3>
+                <div class="table-responsive shadow rounded mt-4 mb-5">
+                    <table class="table table-hover table-borderless mb-0 spec-prof-table">
+                        <thead>
+                            <tr>
+                                <th>Property</th>
+                                <th>Expected Type</th>
+                                <th>Description</th>
+                            </tr>
+                        </thead>
+                        <tbody>
+                        {% for property in schema_properties %}
+                            <tr>
+                                <td class="property-column"><a href="{{ property.property_url }}">{{ property.property }}</a></td>
+                                <td>
+                                    {% for type, url in property.expected_type_and_urls %}
+                                        <a href="{{ url }}">{{ type }}</a><br/>
+                                    {% endfor %}
+                                </td>
+                                <td class="description-column">{{ property.description }}</td>
+                            </tr>
+                        {% endfor %}
+                        </tbody>
+                    </table>
+                </div>
+
+                <h3>Codemeta Properties</h3>
+                <div class="table-responsive shadow rounded mt-4 mb-5">
+                    <table class="table table-hover table-borderless mb-0 spec-prof-table">
+                        <thead>
+                            <tr>
+                                <th>Property</th>
+                                <th>Expected Type</th>
+                                <th>Description</th>
+                            </tr>
+                        </thead>
+                        <tbody>
+                        {% for property in codemeta_properties %}
+                            <tr>
+                                <td class="property-column"><a href="{{ property.property_url }}">{{ property.property }}</a></td>
+                                <td>
+                                    {% for type, url in property.expected_type_and_urls %}
+                                        <a href="{{ url }}">{{ type }}</a><br/>
+                                    {% endfor %}
+                                </td>
+                                <td class="description-column">{{ property.description }}</td>
+                            </tr>
+                        {% endfor %}
+                        </tbody>
+                    </table>
+                </div>
+
+                <h3>Namespaces</h3>
+                <p>Below are the namespaces used in this document:</p>
+                <ul>
+                {% for namespace in namespaces %}
+                    <li><a href="{{ namespace.url }}" target="_blank" class="external" rel="noopener">{{ namespace.prefix }}</a>: {{ namespace.description }}</li>
+                {% endfor %}
+                </ul>
+
+                <hr/>
+
+                <div>
+                        <h3>Contributors</h3>
+                        <p>The following people have been involved in the creation of this specification document.</p>
+
+                        <h5 class="my-4">Group Leader(s)</h5>
+                        {% for leader in group_leaders %}
+                            <div class="col mt-2">{{ leader }}</div>
+                        {% endfor %}
+
+                        <h5 class="my-4">Other team members</h5>
+                        {% for member in team_members %}
+                            <div class="col mt-2">{{ member }}</div>
+                        {% endfor %}
+                </div>
+
+
+            </div>
+        </main>
+    </div>
+
+    <script src="/assets/js/clipboard.min.js"></script>
+    <script src="/assets/js/scripts.js"></script>
+</body>
+
+</html>