Project tooling to generate example property descriptions (#104)

* Project tooling to generate example property descriptions

* Exclude type_descriptions from ajv-cli

* Fix generate-comments when value is null

* validate-examples after generate-descriptions

* Generate descriptions and check for diff in build

* maybeDebugLog -> debug

* Add type descriptions for new types

* Skip copying type_descriptions.yaml in make validator-copy-schema

* Update type_descriptions.yaml with stream include/exclude

.github/workflows/build-check.yaml

@@ -21,3 +21,21 @@ jobs:
 
     - name: validate example
       run: make validate-examples
+
+    - name: generate descriptions
+      run: make generate-descriptions
+
+    - name: check for diff
+      run: |
+        # need to "git add" to detect any changes to descriptions which are not checked in
+        # select files from both /examples
+        git add examples**
+        if git diff --cached --quiet
+        then 
+          echo "No diff detected."
+        else 
+          echo "Diff detected - did you run 'make generate-descriptions'?"
+          echo $(git diff --cached --name-only)
+          echo $(git diff --cached)
+          exit 1
+        fi

CONTRIBUTING.md

@@ -27,6 +27,67 @@ You can perform all checks locally using this command:
 make all
 ```
 
+## Description generation
+
+The [./examples](./examples) directory contains a variety of examples, which
+include important comments describing the semantics of the configuration
+properties. In order to keep these comments consistent across examples, we have
+tooling which automatically generates comments for each property.
+
+How it works:
+
+* The [./schema/type_descriptions.yaml](./schema/type_descriptions.yaml) file
+  defines descriptions for each of the properties of each type defines in
+  the [JSON schema](./schema) data model.
+* The [./scripts/generate-descriptions.js](./scripts/generate-comments.js) is a
+  script which for a given input configuration file will:
+  * Parse the YAML.
+  * Walk through each key / value pair, and for each:
+    * Compute the JSON dot notation location of the current key / value pair.
+    * Find the first matching rule
+      in [type_description.yaml](./schema/type_descriptions.yaml). Iterate
+      through the rules and evaluate the key / value pair dot notation location
+      against each of the rule's `path_patterns`.
+    * Inject / overwrite comments for its properties according
+      to `type_descriptions.yaml`.
+  * Write the resulting content to specified output file or to the console.
+
+The `make generate-descriptions` command runs this process against each file
+in `./examples` and overwrites each file in the process. 
+
+**NOTE:** The [build](./.github/workflows/build-check.yaml) will fail
+if `make generate-descriptions` produces any changes which are not checked into
+version control.
+
+To run against a single file:
+
+- Install the latest LTS release of **[Node](https://nodejs.org/)**.
+  For example, using **[nvm][]** under Linux run:
+
+  ```bash
+  nvm install --lts
+  ```
+
+- Install tooling packages:
+
+  ```bash
+  npm install
+  ```
+
+- Run the script:
+
+```shell
+npm run-script generate-descriptions -- /absolute/path/to/input/file.yaml /absolute/path/to/output/file.yaml
+```
+
+Optionally, include the `--debug` parameter. This prints out information about
+each key value pair, including the computed dot notation location, the matched
+rule, the previous description, the new description, etc.
+
+```shell
+npm run-script generate-comments -- /absolute/path/to/input/file.yaml /absolute/path/to/output/file.yaml --debug
+```
+
 ## Pull requests
 
 A PR is ready to merge when:

Makefile

@@ -3,15 +3,15 @@ EXAMPLE_FILES := $(shell find . -path './examples/*.yaml' -exec basename {} \; |
 $(shell mkdir -p out)
 
 .PHONY: all
-all: install-tools compile-schema validate-examples
+all: install-tools compile-schema generate-descriptions validate-examples
 
 include validator/Makefile
 
 .PHONY: compile-schema
 compile-schema:
 	@if ! npm ls ajv-cli; then npm install; fi
 	@for f in $(SCHEMA_FILES); do \
-	    npx --no ajv-cli compile --spec=draft2020 --allow-matching-properties -s ./schema/$$f -r "./schema/!($$f)" \
+	    npx --no ajv-cli compile --spec=draft2020 --allow-matching-properties -s ./schema/$$f -r "./schema/!($$f|*.yaml)" \
 	        || exit 1; \
 	done
 
@@ -20,10 +20,17 @@ validate-examples:
 	@if ! npm ls ajv-cli; then npm install; fi
 	@for f in $(EXAMPLE_FILES); do \
 	    npx envsub ./examples/$$f ./out/$$f || exit 1; \
-		npx --no ajv-cli validate --spec=draft2020 --allow-matching-properties --errors=text -s ./schema/opentelemetry_configuration.json -r "./schema/!(opentelemetry_configuration.json)" -d ./out/$$f \
+		npx --no ajv-cli validate --spec=draft2020 --allow-matching-properties --errors=text -s ./schema/opentelemetry_configuration.json -r "./schema/!(opentelemetry_configuration.json|*.yaml)" -d ./out/$$f \
 		    || exit 1; \
 	done
 
+.PHONY: generate-descriptions
+generate-descriptions:
+	@if ! npm ls minimatch yaml; then npm install; fi
+	@for f in $(EXAMPLE_FILES); do \
+	    npm run-script generate-descriptions -- $(shell pwd)/examples/$$f $(shell pwd)/examples/$$f || exit 1; \
+	done
+
 .PHONY: install-tools
 install-tools:
 	npm install

examples/anchors.yaml

@@ -1,5 +1,6 @@
 # anchors.yaml demonstrates anchor substitution to reuse OTLP exporter configuration across signals.
 
+# The file format version.
 file_format: "0.1"
 exporters:
   otlp: &otlp-exporter
@@ -13,33 +14,52 @@ exporters:
     compression: gzip
     timeout: 10000
 
+# Configure logger provider.
 logger_provider:
+  # Configure log record processors.
   processors:
-    - batch:
+    - # Configure a batch log record processor.
+      batch:
+        # Configure exporter.
         exporter:
+          # Configure exporter to be OTLP.
           otlp:
-            # expand the otlp-exporter anchor
             <<: *otlp-exporter
+            # Configure endpoint.
             endpoint: http://localhost:4318/v1/logs
 
+# Configure meter provider.
 meter_provider:
+  # Configure metric readers.
   readers:
-    - periodic:
+    - # Configure a periodic metric reader.
+      periodic:
+        # Configure delay interval (in milliseconds) between start of two consecutive exports.
         interval: 5000
+        # Configure maximum allowed time (in milliseconds) to export data.
         timeout: 30000
+        # Configure exporter.
         exporter:
+          # Configure exporter to be OTLP.
           otlp:
-            # expand the otlp-exporter anchor and add metric specific configuration
             <<: *otlp-exporter
+            # Configure endpoint.
             endpoint: http://localhost:4318/v1/metrics
+            # Configure temporality preference.
             temporality_preference: delta
+            # Configure default histogram aggregation.
             default_histogram_aggregation: base2_exponential_bucket_histogram
 
+# Configure tracer provider.
 tracer_provider:
+  # Configure span processors.
   processors:
-    - batch:
+    - # Configure a batch span processor.
+      batch:
+        # Configure exporter.
         exporter:
+          # Configure exporter to be OTLP.
           otlp:
-            # expand the otlp-exporter anchor
             <<: *otlp-exporter
+            # Configure endpoint.
             endpoint: http://localhost:4318/v1/traces