Validate query schema (#6)

.github/workflows/pre-commit.yml

@@ -7,7 +7,7 @@ on:
 
 jobs:
   pre-commit:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     steps:
       - uses: actions/checkout@v2
       - uses: actions/setup-python@v2

.github/workflows/tox.yml

@@ -28,6 +28,7 @@ jobs:
           - ubuntu-20.04
         env:
           - TOXENV: lint
+          - TOXENV: py38
     env:
       TOX_PARALLEL_NO_SPINNER: 1
       FORCE_COLOR: 1

README.md

@@ -4,22 +4,10 @@ Hosts reusable log queries which are built into a single queries.json file.
 
 ## Query database structure
 
-```yaml
-queries:
-  pip-no-matching-distribution:  # also known as query "id", (sort key)
-    pattern: "ERROR: No matching distribution found for"  # required, can be list!
-    # optionals:
-    name: "PIP failed to find package"  # optional (aka 'reason'), id is used instead
-    files: []  # list of glob patterns, narrows down searching
-    regex: false  # optional, default is false
-    url: ... # str or list[str], issue links to lp, bz,...
-    category: ... # see Categories
-    tags: []  # see Tags
-    suppress-graph: false  # used for elastic-recheck
-    # uncommon fields
-    multiline: true  # https://opendev.org/openstack/ansible-role-collect-logs/src/branch/master/vars/sova-patterns.yml#L47
-  another-query-name: ...
-```
+Queries are defined using the data model from src/model.py which builds
+a JSON Validation schema, making easy to validate the file.
+
+One example of file can be seen at [queries-example.yml](https://github.com/rdo-infra/queries/blob/main/src/data/queries-example.yml)
 
 Both [elastic-search](https://www.elastic.co/guide/en/elasticsearch/reference/current/term-level-queries.html) and [artcl](https://opendev.org/openstack/ansible-role-collect-logs) can make use of `regex` searches.
 

output/queries-schema.json

@@ -0,0 +1,112 @@
+{
+  "title": "Queries",
+  "type": "object",
+  "properties": {
+    "queries": {
+      "title": "Queries",
+      "type": "array",
+      "items": {
+        "$ref": "#/definitions/Query"
+      }
+    }
+  },
+  "required": [
+    "queries"
+  ],
+  "definitions": {
+    "Query": {
+      "title": "Query",
+      "type": "object",
+      "properties": {
+        "id": {
+          "title": "Id",
+          "type": "string"
+        },
+        "name": {
+          "title": "Name",
+          "type": "string"
+        },
+        "pattern": {
+          "title": "Pattern",
+          "anyOf": [
+            {
+              "type": "array",
+              "items": {
+                "type": "string"
+              }
+            },
+            {
+              "type": "string"
+            }
+          ]
+        },
+        "category": {
+          "title": "Category",
+          "type": "string"
+        },
+        "url": {
+          "title": "Url",
+          "anyOf": [
+            {
+              "type": "array",
+              "items": {
+                "type": "string",
+                "minLength": 1,
+                "maxLength": 2083,
+                "format": "uri"
+              }
+            },
+            {
+              "type": "string",
+              "minLength": 1,
+              "maxLength": 2083,
+              "format": "uri"
+            }
+          ]
+        },
+        "tags": {
+          "title": "Tags",
+          "anyOf": [
+            {
+              "type": "array",
+              "items": {
+                "type": "string"
+              }
+            },
+            {
+              "type": "string"
+            }
+          ]
+        },
+        "suppress-graph": {
+          "title": "Suppress-Graph",
+          "description": "Used for elastic-recheck",
+          "type": "boolean"
+        },
+        "regex": {
+          "title": "Regex",
+          "default": false,
+          "type": "boolean"
+        },
+        "multiline": {
+          "title": "Multiline",
+          "default": false,
+          "type": "boolean"
+        },
+        "files": {
+          "title": "Files",
+          "description": "List of glob patterns, narrows down searching",
+          "type": "array",
+          "items": {
+            "type": "string"
+          }
+        }
+      },
+      "required": [
+        "id",
+        "pattern"
+      ],
+      "additionalProperties": false
+    }
+  }
+}

requirements.txt

@@ -0,0 +1,3 @@
+pydantic>=1.7.4 # MIT
+yq  # Apache
+jsonschema  # MIT

src/data/queries-example.yml

@@ -0,0 +1,14 @@
+# yaml-language-server: $schema=https://raw.githubusercontent.com/rdo-infra/queries/main/output/queries-schema.json
+queries:
+  - id: pip-no-matching-distribution  # sort key
+    name: "PIP failed to find package"  # optional (aka 'reason'), id is used instead
+    pattern: "ERROR: No matching distribution found for"  # required, can be list!
+    # optionals:
+    files: []  # list of glob patterns, narrows down searching
+    regex: false  # optional, default is false
+    url: ...  # str or list[str], issue links to lp, bz,...
+    category: ...  # see Categories
+    tags: []  # see Tags
+    suppress-graph: false  # used for elastic-recheck
+    # uncommon fields
+    multiline: true  # https://opendev.org/openstack/ansible-role-collect-logs/src/branch/master/vars/sova-patterns.yml#L47

src/data/queries.yml

@@ -1,6 +1,7 @@
-# Keep all keys sorted (enforced via CI)
+# yaml-language-server: $schema=https://raw.githubusercontent.com/rdo-infra/queries/main/output/queries-schema.json
+# Keep all entries sorted by id
 queries:
-  java_io_exception_remote_call:
+  - id: java_io_exception_remote_call
     pattern:
       - java.io.IOException
       - Remote call on
@@ -10,10 +11,10 @@ queries:
       - console.html
       - job-output.txt
     url: https://bugs.launchpad.net/openstack-ci/+bug/1260654
-  overcloud_create_failed:
+  - id: overcloud_create_failed
     name: "Overcloud stack: FAILED."
     pattern: "Stack overcloud CREATE_FAILED"
   # from https://opendev.org/opendev/elastic-recheck/src/branch/master/queries/1260654.yaml
-  timeout_re:
+  - id: timeout_re
     pattern: Killed\s+timeout -s 9
     regex: true

src/model.py

@@ -0,0 +1,41 @@
+from pathlib import Path
+from typing import List, Optional, Union
+
+from pydantic import BaseModel, Field, Extra, HttpUrl
+
+
+class Query(BaseModel):
+    id: str
+    name: Optional[str]
+    pattern: Union[List[str], str]
+
+    category: Optional[str]
+    url: Optional[Union[List[HttpUrl], HttpUrl]]
+
+    # elastic-search specific fields
+    tags: Optional[Union[List[str], str]]
+    suppress_graph: Optional[bool] = Field(
+        alias='suppress-graph',
+        description="Used for elastic-recheck")
+
+    # artcl/sove specific fields
+    regex: Optional[bool] = False
+    # https://opendev.org/openstack/ansible-role-collect-logs/src/branch/master/vars/sova-patterns.yml#L47
+    multiline: Optional[bool] = False
+    files: Optional[List[str]] = Field(
+        description="List of glob patterns, narrows down searching")
+
+    class Config:
+        extra = Extra.forbid
+
+
+class Queries(BaseModel):
+    queries: List[Query]
+
+
+# this is equivalent to json.dumps(MainModel.schema(), indent=2):
+my_dir = Path(__file__).resolve().parents[1]
+output_file = my_dir / "output" / "queries-schema.json"
+with open(output_file, "w") as f:
+    f.write(Queries.schema_json(indent=2))
+    f.write("\n")

tox.ini

@@ -1,6 +1,6 @@
 [tox]
 minversion = 3.16.1
-envlist = lint
+envlist = lint,py38
 isolated_build = true
 requires =
   setuptools >= 41.4.0
@@ -10,8 +10,12 @@ skip_missing_interpreters = True
 usedevelop = false
 
 [testenv]
+deps =
+  yq
+  pydantic
+  jsonschema
 commands =
-  true
+  bash -c "python3 src/model.py && cat src/data/queries.yml | yq | jsonschema -i /dev/stdin output/queries-schema.json"
 passenv =
   CURL_CA_BUNDLE  # https proxies, https://github.com/tox-dev/tox/issues/1437
   FORCE_COLOR
@@ -25,6 +29,9 @@ passenv =
 setenv =
   PIP_DISABLE_PIP_VERSION_CHECK = 1
   PRE_COMMIT_COLOR = always
+skip_install = true
+allowlist_externals =
+  bash
 
 [testenv:lint]
 description = Run all linters