Skip to content

Commit

Permalink
test: data-driven validation tests (#556)
Browse files Browse the repository at this point in the history
Closes partially #543.
Closes partially #433.

### Summary of Changes

* Port the data-driven framework for validation tests from the old Xtext
implementation to Langium.
* Document how to write validation tests.

---------

Co-authored-by: megalinter-bot <129584137+megalinter-bot@users.noreply.github.com>
  • Loading branch information
lars-reimann and megalinter-bot authored Sep 7, 2023
1 parent befa503 commit f58bf20
Show file tree
Hide file tree
Showing 50 changed files with 1,029 additions and 137 deletions.
59 changes: 59 additions & 0 deletions docs/development/validation-testing.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,59 @@
# Validation Testing

Validation tests are data-driven instead of being specified explicitly. This document explains how to add a new
validation test.

## Adding a validation test

1. Create a new **folder** (not just a file!) in the `tests/resources/validation` directory or any subdirectory. Give
the folder a descriptive name, since the folder name becomes part of the test name.

!!! tip "Skipping a test"

If you want to skip a test, add the prefix `skip-` to the folder name.

2. Add files with the extension `.sdstest`, `.sdspipe`, or `.sdsstub` **directly inside the folder**. All files in a
folder will be loaded into the same workspace, so they can reference each other. Files in different folders are
loaded into different workspaces, so they cannot reference each other.
3. Add the Safe-DS code that you want to test to the file.
4. Specify the expected validation results using test comments (see [below](#format-of-test-comments)) and test
markers (e.g. `fun »F«()`). The test comments are used to specify
* the presence or absence of an issue,
* the severity of the issue, and
* the message of the issue.

The test markers are used to specify the location of the issue. Test comments and test markers are mapped to each
other by their position in the file, i.e. the first test comment corresponds to the first test marker, the second
test comment corresponds to the second test marker, etc. There may be more test comments than test markers, but not
the other way around. Any additional test comments are applied to the entire file.

5. Run the tests. The test runner will automatically pick up the new test.

## Format of test comments

1. As usual, test comments are single-line comments that start with `$TEST$`.
2. Then, you specify whether the issue should be absent by writing `no` or present by writing nothing.
3. Next, you specify the severity of the issue by writing `error`, `warning`, `info`, or `hint`.
4. Finally, you can optionally specify the message of the issue enclosed in double-quotes. You can also add an `r`
before the opening double-quote to indicate that the expected message should be interpreted as a regular expression
that must match the entire actual message.

Here are some examples:

```ts
// $TEST$ error "Incompatible type."
```

We expect an error with the exact message `Incompatible type.`.

```ts
// $TEST$ no warning "Name should be lowerCamelCase."
```

We expect no warning with the exact message `Name should be lowerCamelCase.`.

```ts
// $TEST$ info r".*empty.*"
```

We expect an info with a message that matches the regular expression `.*empty.*`.
1 change: 1 addition & 0 deletions mkdocs.yml
Original file line number Diff line number Diff line change
Expand Up @@ -39,6 +39,7 @@ nav:
- Grammar Testing: development/grammar-testing.md
- Scoping Testing: development/scoping-testing.md
- Formatting Testing: development/formatting-testing.md
- Validation Testing: development/validation-testing.md
- Langium Quickstart: development/langium-quickstart.md

# Configuration of MkDocs & Material for MkDocs --------------------------------
Expand Down
5 changes: 5 additions & 0 deletions src/language/constant/fileExtensions.ts
Original file line number Diff line number Diff line change
Expand Up @@ -25,6 +25,11 @@ export const STUB_FILE_EXTENSION = 'sdsstub';
*/
export const TEST_FILE_EXTENSION = 'sdstest';

/**
* All file extensions that are supported by the Safe-DS language.
*/
export const SAFE_DS_FILE_EXTENSIONS = [PIPELINE_FILE_EXTENSION, STUB_FILE_EXTENSION, TEST_FILE_EXTENSION];

/**
* All file extensions that are supported by the Safe-DS language.
*/
Expand Down
124 changes: 124 additions & 0 deletions src/language/validation/nameConvention.ts
Original file line number Diff line number Diff line change
@@ -0,0 +1,124 @@
import { SdsDeclaration } from '../generated/ast.js';
import { ValidationAcceptor } from 'langium';

const blockLambdaPrefix = '__block_lambda_';

export const nameMustNotStartWithBlockLambdaPrefix = (node: SdsDeclaration, accept: ValidationAcceptor) => {
if (node.name.startsWith(blockLambdaPrefix)) {
accept(
'error',
"Names of declarations must not start with '__block_lambda_'. This is reserved for code generation of block lambdas.",
{
node,
property: 'name',
code: 'nameConvention/blockLambdaPrefix',
},
);
}
};

export const nameShouldHaveCorrectCasing = (node: SdsDeclaration, accept: ValidationAcceptor) => {
switch (node.$type) {
case 'SdsAnnotation':
if (!isUpperCamelCase(node.name)) {
acceptCasingWarning(node, 'annotations', 'UpperCamelCase', accept);
}
return;
case 'SdsAttribute':
if (!isLowerCamelCase(node.name)) {
acceptCasingWarning(node, 'attributes', 'lowerCamelCase', accept);
}
return;
case 'SdsBlockLambdaResult':
if (!isLowerCamelCase(node.name)) {
acceptCasingWarning(node, 'block lambda results', 'lowerCamelCase', accept);
}
return;
case 'SdsClass':
if (!isUpperCamelCase(node.name)) {
acceptCasingWarning(node, 'classes', 'UpperCamelCase', accept);
}
return;
case 'SdsEnum':
if (!isUpperCamelCase(node.name)) {
acceptCasingWarning(node, 'enums', 'UpperCamelCase', accept);
}
return;
case 'SdsEnumVariant':
if (!isUpperCamelCase(node.name)) {
acceptCasingWarning(node, 'enum variants', 'UpperCamelCase', accept);
}
return;
case 'SdsFunction':
if (!isLowerCamelCase(node.name)) {
acceptCasingWarning(node, 'functions', 'lowerCamelCase', accept);
}
return;
case 'SdsModule':
const segments = node.name.split('.');
if (!segments.every(isLowerCamelCase)) {
accept('warning', 'All segments of the qualified name of a package should be lowerCamelCase.', {
node,
property: 'name',
code: 'nameConvention/casing',
});
}
return;
case 'SdsParameter':
if (!isLowerCamelCase(node.name)) {
acceptCasingWarning(node, 'parameters', 'lowerCamelCase', accept);
}
return;
case 'SdsPipeline':
if (!isLowerCamelCase(node.name)) {
acceptCasingWarning(node, 'pipelines', 'lowerCamelCase', accept);
}
return;
case 'SdsPlaceholder':
if (!isLowerCamelCase(node.name)) {
acceptCasingWarning(node, 'placeholders', 'lowerCamelCase', accept);
}
return;
case 'SdsResult':
if (!isLowerCamelCase(node.name)) {
acceptCasingWarning(node, 'results', 'lowerCamelCase', accept);
}
return;
case 'SdsSchema':
if (!isUpperCamelCase(node.name)) {
acceptCasingWarning(node, 'schemas', 'UpperCamelCase', accept);
}
return;
case 'SdsSegment':
if (!isLowerCamelCase(node.name)) {
acceptCasingWarning(node, 'segments', 'lowerCamelCase', accept);
}
return;
case 'SdsTypeParameter':
if (!isUpperCamelCase(node.name)) {
acceptCasingWarning(node, 'type parameters', 'UpperCamelCase', accept);
}
return;
}
};

const isLowerCamelCase = (name: string): boolean => {
return /^[a-z][a-zA-Z0-9]*$/gu.test(name);
};

const isUpperCamelCase = (name: string): boolean => {
return /^[A-Z][a-zA-Z0-9]*$/gu.test(name);
};

const acceptCasingWarning = (
node: SdsDeclaration,
nodeName: string,
expectedCasing: string,
accept: ValidationAcceptor,
) => {
accept('warning', `Names of ${nodeName} should be ${expectedCasing}.`, {
node,
property: 'name',
code: 'nameConvention/casing',
});
};
14 changes: 3 additions & 11 deletions src/language/validation/safe-ds-validator.ts
Original file line number Diff line number Diff line change
@@ -1,6 +1,7 @@
import { ValidationChecks } from 'langium';
import { SafeDsAstType } from '../generated/ast.js';
import type { SafeDsServices } from '../safe-ds-module.js';
import { nameMustNotStartWithBlockLambdaPrefix, nameShouldHaveCorrectCasing } from './nameConvention.js';

/**
* Register custom validation checks.
Expand All @@ -9,21 +10,12 @@ export const registerValidationChecks = function (services: SafeDsServices) {
const registry = services.validation.ValidationRegistry;
const validator = services.validation.SafeDsValidator;
const checks: ValidationChecks<SafeDsAstType> = {
// Person: validator.checkPersonStartsWithCapital
SdsDeclaration: [nameMustNotStartWithBlockLambdaPrefix, nameShouldHaveCorrectCasing],
};
registry.register(checks, validator);
};

/**
* Implementation of custom validations.
*/
export class SafeDsValidator {
// checkPersonStartsWithCapital(person: Person, accept: ValidationAcceptor): void {
// if (person.name) {
// const firstChar = person.name.substring(0, 1);
// if (firstChar.toUpperCase() !== firstChar) {
// accept('warning', 'Person name should start with a capital.', { node: person, property: 'name' });
// }
// }
// }
}
export class SafeDsValidator {}
30 changes: 30 additions & 0 deletions tests/helpers/diagnostics.ts
Original file line number Diff line number Diff line change
@@ -0,0 +1,30 @@
import { validationHelper } from 'langium/test';
import { LangiumServices } from 'langium';
import { Diagnostic, DiagnosticSeverity } from 'vscode-languageserver-types';

/**
* Get syntax errors from a code snippet.
*
* @param services The language services.
* @param code The code snippet to check.
* @returns The syntax errors.
*/
export const getSyntaxErrors = async (services: LangiumServices, code: string): Promise<Diagnostic[]> => {
const validationResult = await validationHelper(services)(code);
return validationResult.diagnostics.filter(
(d) =>
d.severity === DiagnosticSeverity.Error &&
(d.data?.code === 'lexing-error' || d.data?.code === 'parsing-error'),
);
};

/**
* The code contains syntax errors.
*/
export class SyntaxErrorsInCodeError extends Error {
constructor(readonly syntaxErrors: Diagnostic[]) {
const syntaxErrorsAsString = syntaxErrors.map((e) => `- ${e.message}`).join(`\n`);

super(`Code has syntax errors:\n${syntaxErrorsAsString}`);
}
}
49 changes: 43 additions & 6 deletions tests/helpers/testResources.test.ts
Original file line number Diff line number Diff line change
@@ -1,19 +1,56 @@
import { describe, expect, it } from 'vitest';
import { listTestResources } from './testResources.js';
import { listTestResources, listTestsResourcesGroupedByParentDirectory } from './testResources.js';

describe('listTestResources', () => {
it('should yield all Safe-DS files in a directory that are not skipped', () => {
const result = listTestResources('helpers/listTestResources')
.map((path) => path.replace(/\\/gu, '/'))
.sort();
const result = listTestResources('helpers/listTestResources');
const expected = [
'pipeline file.sdspipe',
'stub file.sdsstub',
'test file.sdstest',
'nested/pipeline file.sdspipe',
'nested/stub file.sdsstub',
'nested/test file.sdstest',
].sort();
expect(result).toStrictEqual(expected);
];
expect(normalizePaths(result)).toStrictEqual(normalizePaths(expected));
});
});

describe('listTestResourcesGroupedByParentDirectory', () => {
it('should yield all Safe-DS files in a directory that are not skipped and group them by parent directory', () => {
const result = listTestsResourcesGroupedByParentDirectory('helpers/listTestResources');

const keys = Object.keys(result);
expect(normalizePaths(keys)).toStrictEqual(normalizePaths(['.', 'nested']));

const directlyInRoot = result['.'];
expect(normalizePaths(directlyInRoot)).toStrictEqual(
normalizePaths(['pipeline file.sdspipe', 'stub file.sdsstub', 'test file.sdstest']),
);

const inNested = result.nested;
expect(normalizePaths(inNested)).toStrictEqual(
normalizePaths(['nested/pipeline file.sdspipe', 'nested/stub file.sdsstub', 'nested/test file.sdstest']),
);
});
});

/**
* Normalizes the given paths by replacing backslashes with slashes and sorting them.
*
* @param paths The paths to normalize.
* @return The normalized paths.
*/
const normalizePaths = (paths: string[]): string[] => {
return paths.map(normalizePath).sort();
};

/**
* Normalizes the given path by replacing backslashes with slashes.
*
* @param path The path to normalize.
* @return The normalized path.
*/
const normalizePath = (path: string): string => {
return path.replace(/\\/gu, '/');
};
27 changes: 18 additions & 9 deletions tests/helpers/testResources.ts
Original file line number Diff line number Diff line change
@@ -1,10 +1,7 @@
import path from 'path';
import { globSync } from 'glob';
import {
PIPELINE_FILE_EXTENSION,
STUB_FILE_EXTENSION,
TEST_FILE_EXTENSION,
} from '../../src/language/constant/fileExtensions.js';
import { SAFE_DS_FILE_EXTENSIONS } from '../../src/language/constant/fileExtensions.js';
import { group } from 'radash';

const resourcesPath = path.join(__dirname, '..', 'resources');

Expand All @@ -19,20 +16,32 @@ export const resolvePathRelativeToResources = (pathRelativeToResources: string)
};

/**
* Lists all Safe-DS files in the given directory relative to `tests/resources/` except those that have a name starting
* with 'skip'.
* Lists all Safe-DS files in the given directory relative to `tests/resources/` that are not skipped.
*
* @param pathRelativeToResources The root directory relative to `tests/resources/`.
* @return Paths to the Safe-DS files relative to `pathRelativeToResources`.
*/
export const listTestResources = (pathRelativeToResources: string): string[] => {
const fileExtensions = [PIPELINE_FILE_EXTENSION, STUB_FILE_EXTENSION, TEST_FILE_EXTENSION];
const pattern = `**/*.{${fileExtensions.join(',')}}`;
const pattern = `**/*.{${SAFE_DS_FILE_EXTENSIONS.join(',')}}`;
const cwd = resolvePathRelativeToResources(pathRelativeToResources);

return globSync(pattern, { cwd, nodir: true }).filter(isNotSkipped);
};

/**
* Lists all Safe-DS files in the given directory relative to `tests/resources/` that are not skipped. The result is
* grouped by the parent directory.
*
* @param pathRelativeToResources The root directory relative to `tests/resources/`.
* @return Paths to the Safe-DS files relative to `pathRelativeToResources` grouped by the parent directory.
*/
export const listTestsResourcesGroupedByParentDirectory = (
pathRelativeToResources: string,
): Record<string, string[]> => {
const paths = listTestResources(pathRelativeToResources);
return group(paths, (p) => path.dirname(p)) as Record<string, string[]>;
};

const isNotSkipped = (pathRelativeToResources: string) => {
const segments = pathRelativeToResources.split(path.sep);
return !segments.some((segment) => segment.startsWith('skip'));
Expand Down
Loading

0 comments on commit f58bf20

Please sign in to comment.