Validator: Re-enable yamale validator. (#124)

* Re-enable yamale validator.
* Update relaxed schema and use it for tributaries by using strict_titles arg to control strict schema.
* Allow bare AWS within code-style XML tags.
* Write unit tests for yamale validation.

aws_doc_sdk_examples_tools/config/example_schema.yaml

@@ -1,60 +1,43 @@
 # Yamale Schema for example metadata, which is all .yaml files in the metadata folder
 # with a _metadata.yaml suffix.
 
-map(include('example'), key=example_id())
+map(include('example'), key=str())
 ---
-# An example blocks all the languages together for a single example in a tab list. It is a navigable page on the code examples library. It is the top level "unit" of SoS content. This metadata is merged from tributaries with aws-doc-sdk-examples.
 example:
-  # Human readable title. TODO: Defaults to slug-to-title of the ID if not provided. Overwritten by aws-doc-sdk-example when merging.
-  title: str(upper_start=True, no_end_punc=True, required=False)
-  # Used in the TOC. TODO: Defaults to slug-to-title of the ID if not provided. Overwritten by aws-doc-sdk-example when merging.
-  title_abbrev: str(upper_start=True, no_end_punc=True, required=False)
-  # String label categories. Categories inferred by cross-service with multiple services, and can be whatever else it wants. Controls where in the TOC it appears. Overwritten by aws-doc-sdk-example when merging.
-  category: str(upper_start=True, no_end_punc=True, required=False)
-  # Link to additional topic places. Overwritten by aws-doc-sdk-example when merging.
-  guide_topic: include('guide_topic', required=False) # TODO Make this a list or a single.
-  # TODO how to add a language here and require it in sdks_schema. TODO: Keys merged by aws-doc-sdk-example when merging.
-  languages: map(include('language'), key=enum('Bash', 'C++', 'CLI', 'Go', 'Java', 'JavaScript', 'Kotlin', '.NET', 'PHP', 'Python', 'Ruby', 'Rust', 'SAP ABAP', 'Swift'))
-  # TODO document service_main and services. Not to be used by tributaries. Part of Cross Service.
-  # List of services used by the examples. Lines up with those in services.yaml. Overwritten by aws-doc-sdk-example when merging.
+  title: str(required=False, upper_start=True, no_end_punc=True)
+  title_abbrev: str(required=False, upper_start=True, no_end_punc=True)
+  synopsis: str(required=False)
+  synopsis_list: list(str(upper_start=True), required=False)
+  category: str(required=False, upper_start=True, no_end_punc=True)
+  guide_topic: include('guide_topic', required=False)
+  languages: map(include('language'), key=enum('Bash', 'C++', 'CLI', 'Go', 'Java', 'JavaScript', 'Kotlin', '.NET', 'PHP', 'PowerShell', 'Python', 'Ruby', 'Rust', 'SAP ABAP', 'Swift'))
   service_main: service_name(required=False)
-  services: map(map(key=str(), required=False), key=service_name())
-  synopsis: str(required=False, lower_start=True, end_punc_or_semicolon=True, required=False)
-  synopsis_list: list(str(upper_start=True, end_punc=True), required=False)
+  services: map(key=service_name())
 
-# Used for creating links in the block.
 guide_topic:
   title: str(upper_start=True, no_end_punc=True)
   url: include('doc_url', required=False)
 
-# Language Version configuration. Likely just the single list item.
 language:
   versions: list(include('version'))
 
-# Example for a single Language.
+# Per-language excerpts for the example. Languages and SDK versions are defined in .doc_gen/metadata/sdk_metadata.yaml
 version:
-  sdk_version: int(min=1)
-  # Additional ZonBook XML to include in the tab for this sample.
-  block_content: block_content(required=False)
-  # The specific code samples to include in the example.
-  excerpts: list(include('excerpt'), required=False)
-  # Link to the source code for this example. TODO rename.
+  sdk_version: any(int(min=1), str(check_aws=False))
   github: str(required=False)
+  github_name: str(required=False)
   github_note_at_bottom: bool(required=False)
-  add_services: map(key=service_name(), required=False)
-  # Deprecated. Replace with guide_topic list.
   sdkguide: include('doc_url', required=False)
-  # Link to additional topic places. TODO: Overwritten by aws-doc-sdk-example when merging.
-  more_info: list(include('guide_topic', required=False))
+  excerpts: list(include('excerpt'), required=False)
+  block_content: block_content(required=False)
+  add_services: map(key=service_name(), required=False)
 
-# One language example can have several excerpts, each having a description block and one or more snippets.
-# An excerpt may have either snippet_files OR snippet_tags, but not both.
+# The references to code content that will be included in the example's content.
 excerpt:
-  description: str(required=False, upper_start=True, end_punc=True)
-  # A path within the repo to extract the entire file as a snippet.
-  snippet_files: list(str(), required=False)
-  # Tags embedded in source files to extract as snippets.
+  description: str(required=False)
+  genai: enum('none', 'some', 'most', 'all', required=False)
   snippet_tags: list(str(), required=False)
+  snippet_files: list(str(), required=False)
 
 service_slug_regex: regex('^[-a-z0-9]+$', name='service slug')
 doc_url: regex('^(?!https://docs.aws.amazon.com/).+', name="relative documentation URL")

aws_doc_sdk_examples_tools/config/example_strict_schema.yaml

@@ -4,13 +4,13 @@
 map(include('example'), key=example_id())
 ---
 example:
-  title: str(upper_start=True, no_end_punc=True)
-  title_abbrev: str(upper_start=True, no_end_punc=True)
+  title: str(required=False, upper_start=True, no_end_punc=True)
+  title_abbrev: str(required=False, upper_start=True, no_end_punc=True)
   synopsis: str(required=False, lower_start=True, end_punc_or_semicolon=True)
   synopsis_list: list(str(upper_start=True, end_punc=True), required=False)
   category: str(required=False, upper_start=True, no_end_punc=True)
   guide_topic: include('guide_topic', required=False)
-  languages: map(include('language'), key=enum('Bash', 'C++', 'CLI', 'Go', 'Java', 'JavaScript', 'Kotlin', '.NET', 'PHP', 'Python', 'Ruby', 'Rust', 'SAP ABAP', 'Swift'))
+  languages: map(include('language'), key=enum('Bash', 'C++', 'CLI', 'Go', 'Java', 'JavaScript', 'Kotlin', '.NET', 'PHP', 'PowerShell', 'Python', 'Ruby', 'Rust', 'SAP ABAP', 'Swift'))
   service_main: service_name(required=False)
   services: map(map(key=str(), required=False), key=service_name())
 
@@ -34,6 +34,7 @@ version:
 # The references to code content that will be included in the example's content.
 excerpt:
   description: str(required=False, upper_start=True, end_punc=True)
+  genai: enum('none', 'some', 'most', 'all', required=False)
   snippet_tags: list(str(), required=False)
   snippet_files: list(str(), required=False)
 

aws_doc_sdk_examples_tools/config/sdks_schema.yaml

@@ -1,6 +1,6 @@
 # Yamale Schema for SDK metadata, which is the sdks.yaml file in the metadata folder.
 
-map(include('sdk'), key=enum('Bash', 'C++', 'CLI', 'Go', 'Java', 'JavaScript', 'Kotlin', '.NET', 'PHP', 'Python', 'Ruby', 'Rust', 'SAP ABAP', 'Swift'))
+map(include('sdk'), key=enum('Bash', 'C++', 'CLI', 'Go', 'Java', 'JavaScript', 'Kotlin', '.NET', 'PHP', 'PowerShell', 'Python', 'Ruby', 'Rust', 'SAP ABAP', 'Swift'))
 ---
 sdk:
   property: include('syntax_enum')
@@ -30,6 +30,6 @@ title_override:
   title: str()
   title_abbrev: str()
 
-syntax_enum: enum('bash', 'cli', 'none', 'cpp', 'go', 'java', 'javascript', 'kotlin', 'csharp', 'php', 'python', 'ruby', 'rust', 'sap-abap', 'sh', 'swift')
+syntax_enum: enum('bash', 'cli', 'none', 'cpp', 'go', 'java', 'javascript', 'kotlin', 'csharp', 'php', 'powershell', 'python', 'ruby', 'rust', 'sap-abap', 'sh', 'swift')
 entity_regex: regex('^&[-_a-zA-Z0-9]+;$', name='valid entity')
 entity_with_version_regex: regex('^&[-_a-zA-Z0-9]+;', name='valid entity with version')

aws_doc_sdk_examples_tools/config/services_schema.yaml

@@ -3,8 +3,8 @@
 map(include('service'), key=regex('^[-a-z0-9]+$', name='service slug'))
 ---
 service:
-  long: include('long_entity_regex')
-  short: include('entity_regex')
+  long: str()
+  short: str()
   sort: regex('^[^&]\\w', name='non-entity')
   chapter_override: include('chapter_override', required=False)
   expanded:
@@ -16,7 +16,7 @@ service:
     url: include('doc_url')
   api_client: service_name(required=False)
   api_ref: include('doc_url')
-  version: service_version()
+  version: str()
   caveat: str(required=False, upper_start=True, end_punc=True)
   bundle: service_name(required=False)
   tags: map(key=enum('product_categories'))

aws_doc_sdk_examples_tools/doc_gen.py

@@ -270,7 +270,7 @@ def validate(self):
             service.validate(self.errors)
         for example in self.examples.values():
             example.validate(self.errors, self.root)
-        validate_metadata(self.root, self.errors)
+        validate_metadata(self.root, self.validation.strict_titles, self.errors)
         validate_no_duplicate_api_examples(self.examples.values(), self.errors)
         validate_snippets(
             [*self.examples.values()],

aws_doc_sdk_examples_tools/metadata_validator.py

@@ -12,6 +12,7 @@
 import datetime
 import os
 import re
+import xml.etree.ElementTree as xml_tree
 import yaml
 from dataclasses import dataclass, field
 from pathlib import Path
@@ -122,8 +123,7 @@ def _is_valid(self, value: str):
             return True
         valid = True
         if self.check_aws:
-            # All occurrences of AWS must be entities or within a word.
-            valid = len(re.findall("(?<![&0-9a-zA-Z])AWS(?![;0-9a-zA-Z])", value)) == 0
+            valid = self._validate_aws_entity_usage(value)
             if not valid:
                 self.last_err = 'valid string: it contains a non-entity usage of "AWS"'
         if valid and self.upper_start:
@@ -135,27 +135,51 @@ def _is_valid(self, value: str):
             if not valid:
                 self.last_err = "valid string: it must start with a lowercase letter"
         if valid and self.end_punc:
-            valid = value[-1] in "!.?"
+            valid = value.rstrip()[-1] in "!.?"
             if not valid:
                 self.last_err = "valid sentence or phrase: it must end with punctuation"
         if valid and self.no_end_punc:
-            valid = value[-1] not in "!.?"
+            valid = value.rstrip()[-1] not in "!.?"
             if not valid:
                 self.last_err = "valid string: it must not end with punctuation"
         if valid and self.end_punc_or_colon:
-            valid = value[-1] in "!.?:"
+            valid = value.rstrip()[-1] in "!.?:"
             if not valid:
                 self.last_err = (
                     "valid sentence or phrase: it must end with punctuation or a colon"
                 )
         if valid and self.end_punc_or_semicolon:
-            valid = value[-1] in "!.?;"
+            valid = value.rstrip()[-1] in "!.?;"
             if not valid:
                 self.last_err = "valid sentence or phrase: it must end with punctuation or a semicolon"
         if valid:
             valid = super()._is_valid(value)
         return valid
 
+    @staticmethod
+    def _validate_aws_entity_usage(value: str) -> bool:
+        """
+        All occurrences of AWS must be entities or within a word or within a programlisting or code or noloc block.
+
+        Count all bare AWS occurrences within accepted XML tags.
+        Count all bare AWS occurrences overall.
+        If these counts differ, there's an invalid usage.
+        """
+        xval = value.replace("&", "&amp;")
+        xtree = xml_tree.fromstring(f"<fake><para>{xval}</para></fake>")
+        blocks = (
+            xtree.findall(".//programlisting")
+            + xtree.findall(".//code")
+            + xtree.findall(".//noloc")
+        )
+        aws_in_blocks = 0
+        for element in blocks:
+            aws_in_blocks += len(
+                re.findall("(?<![&0-9a-zA-Z])AWS(?![;0-9a-zA-Z])", str(element.text))
+            )
+        aws_everywhere = len(re.findall("(?<![&0-9a-zA-Z])AWS(?![;0-9a-zA-Z])", value))
+        return aws_everywhere == aws_in_blocks
+
 
 @dataclass
 class ValidateYamaleError(MetadataParseError):
@@ -169,6 +193,7 @@ def validate_files(
     schema_name: Path,
     meta_names: Iterable[Path],
     validators: Dict[str, Validator],
+    strict: bool,
     errors: MetadataErrors,
 ):
     """Iterate a list of files and validate each one against a schema."""
@@ -177,14 +202,16 @@ def validate_files(
     for meta_name in meta_names:
         try:
             data = yamale.make_data(meta_name)
-            yamale.validate(schema, data)
+            yamale.validate(schema, data, strict=strict)
             print(f"{meta_name.resolve()} validation success! 👍")
         except YamaleError as e:
             errors.append(ValidateYamaleError(file=meta_name, yamale_error=e))
     return errors
 
 
-def validate_metadata(doc_gen_root: Path, errors: MetadataErrors) -> MetadataErrors:
+def validate_metadata(
+    doc_gen_root: Path, strict: bool, errors: MetadataErrors
+) -> MetadataErrors:
     config = Path(__file__).parent / "config"
     with open(config / "sdks.yaml") as sdks_file:
         sdks_yaml: Dict[str, Any] = yaml.safe_load(sdks_file)
@@ -206,20 +233,28 @@ def validate_metadata(doc_gen_root: Path, errors: MetadataErrors) -> MetadataErr
     validators[BlockContent.tag] = BlockContent
     validators[String.tag] = StringExtension
 
-    schema_root = Path(__file__).parent / "config"
+    config_root = Path(__file__).parent / "config"
+    if strict:
+        example_schema = "example_strict_schema.yaml"
+    else:
+        example_schema = "example_schema.yaml"
 
     to_validate = [
         # (schema, metadata_glob)
-        ("sdks_schema.yaml", "sdks.yaml"),
-        ("services_schema.yaml", "services.yaml"),
-        # TODO: Switch between strict schema for aws-doc-sdk-examples and loose schema for tributaries
-        ("example_strict_schema.yaml", "*_metadata.yaml"),
+        (config_root / "sdks_schema.yaml", config_root, "sdks.yaml"),
+        (config_root / "services_schema.yaml", config_root, "services.yaml"),
+        (
+            config_root / example_schema,
+            doc_gen_root / ".doc_gen" / "metadata",
+            "*_metadata.yaml",
+        ),
     ]
-    for schema, metadata in to_validate:
+    for schema, meta_root, metadata in to_validate:
         validate_files(
-            schema_root / schema,
-            (doc_gen_root / "metadata").glob(metadata),
+            schema,
+            meta_root.glob(metadata),
             validators,
+            strict,
             errors,
         )
 
@@ -234,9 +269,12 @@ def main():
         help="The folder that contains schema and metadata files.",
         required=False,
     )
+    parser.add_argument(
+        "--strict", default=True, help="Use strict schema.", required=False
+    )
     args = parser.parse_args()
 
-    errors = validate_metadata(Path(args.doc_gen), MetadataErrors())
+    errors = validate_metadata(Path(args.doc_gen), args.strict, MetadataErrors())
 
     if len(errors) == 0:
         print("Validation succeeded! 👍👍👍")

aws_doc_sdk_examples_tools/metadata_validator_test.py

@@ -0,0 +1,31 @@
+#!/usr/bin/env python3
+# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
+# SPDX-License-Identifier: Apache-2.0
+
+from pathlib import Path
+
+import pytest
+
+from .metadata_errors import MetadataErrors
+from .metadata_validator import validate_metadata
+
+
+@pytest.mark.parametrize("strict", [True, False])
+def test_aws_entity_usage(strict):
+    errors = MetadataErrors()
+    validate_metadata(
+        Path(Path(__file__).parent / "test_resources/doc_gen_test"), strict, errors
+    )
+
+    e_str = str(errors)
+    assert "Title has AWS" in e_str
+    assert "Title Abbrev has AWS" in e_str
+    assert "Synopsis has AWS" in e_str
+    assert "Synopsis list has AWS" in e_str
+    assert "Description has AWS" in e_str
+
+    assert "Title has &AWS;" not in e_str
+    assert "Title Abbrev has &AWS;" not in e_str
+    assert "Synopsis programlisting has AWS" not in e_str
+    assert "Synopsis list code has <code>AWS" not in e_str
+    assert "Description programlisting has AWS" not in e_str

aws_doc_sdk_examples_tools/test_resources/doc_gen_test/.doc_gen/metadata/aws_entity_metadata.yaml

@@ -0,0 +1,38 @@
+sns_EntityFailures:
+  title: Title has AWS using an &AWS; SDK
+  title_abbrev: Title Abbrev has AWS in it
+  synopsis: "Synopsis has AWS in it."
+  synopsis_list:
+    - "Synopsis list has AWS in it."
+  category: Cat
+  languages:
+    Java:
+      versions:
+        - sdk_version: 1
+          github: java/example_code/svc_EntityFailures
+          sdkguide:
+          excerpts:
+            - description: Description has AWS in it.
+              snippet_tags:
+                - java.example_code.svc_EntityFailures.Test
+  services:
+    sns:
+sns_EntitySuccesses:
+  title: Title has &AWS; using an &AWS; SDK
+  title_abbrev: Title Abbrev has &AWS; in it
+  synopsis: "this <programlisting>Synopsis programlisting has AWS in it.</programlisting>."
+  synopsis_list:
+    - "Synopsis list code has <code>AWS</code> in it."
+  category: Cat
+  languages:
+    Java:
+      versions:
+        - sdk_version: 1
+          github: java/example_code/svc_EntityFailures
+          sdkguide:
+          excerpts:
+            - description: This <emphasis><programlisting>Description programlisting has AWS in it</programlisting></emphasis> doesn't it.
+              snippet_tags:
+                - java.example_code.svc_EntityFailures.Test
+  services:
+    sns: