Add schema generation and validation infrastructure

Co-authored-by: estruyf <2900833+estruyf@users.noreply.github.com>

src/commands/StatusListener.ts

@@ -7,7 +7,7 @@ import {
   SETTING_SEO_TITLE_LENGTH
 } from './../constants';
 import * as vscode from 'vscode';
-import { ArticleHelper, Notifications, SeoHelper, Settings } from '../helpers';
+import { ArticleHelper, Notifications, SeoHelper, Settings, FrontMatterValidator } from '../helpers';
 import { PanelProvider } from '../panelWebView/PanelProvider';
 import { ContentType } from '../helpers/ContentType';
 import { DataListener } from '../listeners/panel';
@@ -20,6 +20,7 @@ import { i18n } from './i18n';
 import { getDescriptionField, getTitleField } from '../utils';
 
 export class StatusListener {
+  private static validator: FrontMatterValidator = new FrontMatterValidator();
   /**
    * Update the text of the status bar
    *
@@ -70,6 +71,9 @@ export class StatusListener {
           // Check the required fields
           if (editor) {
             StatusListener.verifyRequiredFields(editor, article, collection);
+            
+            // Schema validation
+            await StatusListener.verifySchemaValidation(editor, article, collection);
           }
         }
 
@@ -173,6 +177,71 @@ export class StatusListener {
     }
   }
 
+  /**
+   * Verify schema validation
+   * @param editor Text editor
+   * @param article Parsed front matter
+   * @param collection Diagnostic collection
+   */
+  private static async verifySchemaValidation(
+    editor: vscode.TextEditor,
+    article: ParsedFrontMatter,
+    collection: vscode.DiagnosticCollection
+  ) {
+    try {
+      const contentType = await ArticleHelper.getContentType(article);
+      if (!contentType || !contentType.fields || contentType.fields.length === 0) {
+        return;
+      }
+
+      // Validate against schema
+      const errors = StatusListener.validator.validate(article.data, contentType);
+
+      if (errors.length === 0) {
+        return;
+      }
+
+      const text = editor.document.getText();
+      const schemaDiagnostics: vscode.Diagnostic[] = [];
+
+      for (const error of errors) {
+        // Find the field in the document
+        const fieldPath = error.field.split('.');
+        const fieldName = fieldPath[fieldPath.length - 1];
+
+        // Try to find the field location in the document
+        const fieldIdx = text.indexOf(fieldName);
+
+        if (fieldIdx !== -1) {
+          const posStart = editor.document.positionAt(fieldIdx);
+          const posEnd = editor.document.positionAt(fieldIdx + fieldName.length);
+
+          const diagnostic: vscode.Diagnostic = {
+            code: '',
+            message: error.message,
+            range: new vscode.Range(posStart, posEnd),
+            severity: vscode.DiagnosticSeverity.Warning,
+            source: EXTENSION_NAME
+          };
+
+          schemaDiagnostics.push(diagnostic);
+        }
+      }
+
+      if (schemaDiagnostics.length > 0) {
+        if (collection.has(editor.document.uri)) {
+          const otherDiag = collection.get(editor.document.uri) || [];
+          collection.set(editor.document.uri, [...otherDiag, ...schemaDiagnostics]);
+        } else {
+          collection.set(editor.document.uri, [...schemaDiagnostics]);
+        }
+      }
+    } catch (error) {
+      // Silently fail validation errors to not disrupt the user experience
+      console.error('Schema validation error:', error);
+    }
+  }
+
   /**
    * Find the line of the field
    * @param text

src/helpers/ContentTypeSchemaGenerator.ts

@@ -0,0 +1,294 @@
+import { ContentType, Field, FieldType } from '../models';
+import { Settings } from '../helpers/SettingsHelper';
+import { SETTING_TAXONOMY_FIELD_GROUPS } from '../constants';
+
+/**
+ * JSON Schema type definition
+ */
+export interface JSONSchema {
+  $schema?: string;
+  type?: string | string[];
+  properties?: { [key: string]: JSONSchema };
+  required?: string[];
+  items?: JSONSchema;
+  enum?: any[];
+  format?: string;
+  anyOf?: JSONSchema[];
+  oneOf?: JSONSchema[];
+  allOf?: JSONSchema[];
+  description?: string;
+  default?: any;
+  minimum?: number;
+  maximum?: number;
+}
+
+/**
+ * Generates JSON Schema from Front Matter Content Type definitions
+ */
+export class ContentTypeSchemaGenerator {
+  /**
+   * Generate JSON Schema from a content type
+   * @param contentType The content type to generate schema from
+   * @returns JSON Schema object
+   */
+  public static generateSchema(contentType: ContentType): JSONSchema {
+    const schema: JSONSchema = {
+      $schema: 'http://json-schema.org/draft-07/schema#',
+      type: 'object',
+      properties: {},
+      required: []
+    };
+
+    if (!contentType.fields || contentType.fields.length === 0) {
+      return schema;
+    }
+
+    // Process each field in the content type
+    for (const field of contentType.fields) {
+      const fieldSchema = this.generateFieldSchema(field);
+      if (fieldSchema && schema.properties) {
+        schema.properties[field.name] = fieldSchema;
+
+        // Add to required array if field is required
+        if (field.required && schema.required) {
+          schema.required.push(field.name);
+        }
+      }
+    }
+
+    // Remove required array if empty
+    if (schema.required && schema.required.length === 0) {
+      delete schema.required;
+    }
+
+    return schema;
+  }
+
+  /**
+   * Generate JSON Schema for a single field
+   * @param field The field to generate schema from
+   * @returns JSON Schema object for the field
+   */
+  private static generateFieldSchema(field: Field): JSONSchema | null {
+    // Skip divider and heading fields as they are UI-only
+    if (field.type === 'divider' || field.type === 'heading') {
+      return null;
+    }
+
+    const schema: JSONSchema = {};
+
+    // Add description if available
+    if (field.description) {
+      schema.description = field.description;
+    }
+
+    // Add default value if specified
+    if (field.default !== undefined && field.default !== null) {
+      schema.default = field.default;
+    }
+
+    // Map field type to JSON Schema type
+    switch (field.type) {
+      case 'string':
+      case 'slug':
+      case 'image':
+      case 'file':
+      case 'customField':
+        schema.type = 'string';
+        break;
+
+      case 'number':
+        schema.type = 'number';
+        if (field.numberOptions) {
+          if (field.numberOptions.min !== undefined) {
+            schema.minimum = field.numberOptions.min;
+          }
+          if (field.numberOptions.max !== undefined) {
+            schema.maximum = field.numberOptions.max;
+          }
+        }
+        break;
+
+      case 'boolean':
+      case 'draft':
+        schema.type = 'boolean';
+        break;
+
+      case 'datetime':
+        schema.type = 'string';
+        schema.format = 'date-time';
+        break;
+
+      case 'choice':
+        if (field.multiple) {
+          schema.type = 'array';
+          schema.items = {
+            type: 'string'
+          };
+          if (field.choices && field.choices.length > 0) {
+            schema.items.enum = this.extractChoiceValues(field.choices);
+          }
+        } else {
+          schema.type = 'string';
+          if (field.choices && field.choices.length > 0) {
+            schema.enum = this.extractChoiceValues(field.choices);
+          }
+        }
+        break;
+
+      case 'tags':
+      case 'categories':
+      case 'taxonomy':
+      case 'list':
+        schema.type = 'array';
+        schema.items = {
+          type: 'string'
+        };
+        break;
+
+      case 'fields':
+        schema.type = 'object';
+        schema.properties = {};
+        schema.required = [];
+
+        if (field.fields && field.fields.length > 0) {
+          for (const subField of field.fields) {
+            const subFieldSchema = this.generateFieldSchema(subField);
+            if (subFieldSchema && schema.properties) {
+              schema.properties[subField.name] = subFieldSchema;
+
+              if (subField.required && schema.required) {
+                schema.required.push(subField.name);
+              }
+            }
+          }
+        }
+
+        // Remove required array if empty
+        if (schema.required && schema.required.length === 0) {
+          delete schema.required;
+        }
+        break;
+
+      case 'block': {
+        // Block fields can contain different field groups
+        schema.type = 'array';
+        schema.items = {
+          type: 'object'
+        };
+
+        // Try to get the field group schemas
+        const blockSchemas = this.getBlockFieldGroupSchemas(field);
+        if (blockSchemas.length > 0) {
+          schema.items = {
+            oneOf: blockSchemas
+          };
+        }
+        break;
+      }
+
+      case 'json':
+        // JSON fields can be any valid JSON
+        schema.type = ['object', 'array', 'string', 'number', 'boolean', 'null'];
+        break;
+
+      case 'dataFile':
+        // Data file references are typically strings (IDs or keys)
+        schema.type = 'string';
+        break;
+
+      case 'contentRelationship':
+        // Content relationships can be a string (slug/path) or array of strings
+        if (field.multiple) {
+          schema.type = 'array';
+          schema.items = {
+            type: 'string'
+          };
+        } else {
+          schema.type = 'string';
+        }
+        break;
+
+      case 'fieldCollection':
+        // Field collections reference field groups, handle similarly to blocks
+        schema.type = 'array';
+        schema.items = {
+          type: 'object'
+        };
+        break;
+
+      default:
+        // Unknown field type, default to string
+        schema.type = 'string';
+        break;
+    }
+
+    return schema;
+  }
+
+  /**
+   * Extract choice values from field choices
+   * @param choices Array of choice strings or objects
+   * @returns Array of choice values
+   */
+  private static extractChoiceValues(choices: (string | { id?: string | null; title: string })[]): string[] {
+    return choices.map((choice) => {
+      if (typeof choice === 'string') {
+        return choice;
+      } else {
+        return choice.id || choice.title;
+      }
+    });
+  }
+
+  /**
+   * Get schemas for block field groups
+   * @param field The block field
+   * @returns Array of JSON Schemas for each field group
+   */
+  private static getBlockFieldGroupSchemas(field: Field): JSONSchema[] {
+    const schemas: JSONSchema[] = [];
+
+    if (!field.fieldGroup) {
+      return schemas;
+    }
+
+    const fieldGroupIds = Array.isArray(field.fieldGroup) ? field.fieldGroup : [field.fieldGroup];
+    const fieldGroups = Settings.get(SETTING_TAXONOMY_FIELD_GROUPS) as any[];
+
+    if (!fieldGroups || fieldGroups.length === 0) {
+      return schemas;
+    }
+
+    for (const groupId of fieldGroupIds) {
+      const fieldGroup = fieldGroups.find((fg) => fg.id === groupId);
+      if (fieldGroup && fieldGroup.fields) {
+        const groupSchema: JSONSchema = {
+          type: 'object',
+          properties: {},
+          required: []
+        };
+
+        for (const groupField of fieldGroup.fields) {
+          const fieldSchema = this.generateFieldSchema(groupField);
+          if (fieldSchema && groupSchema.properties) {
+            groupSchema.properties[groupField.name] = fieldSchema;
+
+            if (groupField.required && groupSchema.required) {
+              groupSchema.required.push(groupField.name);
+            }
+          }
+        }
+
+        // Remove required array if empty
+        if (groupSchema.required && groupSchema.required.length === 0) {
+          delete groupSchema.required;
+        }
+
+        schemas.push(groupSchema);
+      }
+    }
+
+    return schemas;
+  }
+}

src/helpers/FrontMatterValidator.ts

@@ -0,0 +1,196 @@
+import Ajv, { ErrorObject } from 'ajv';
+import { ContentType } from '../models';
+import { ContentTypeSchemaGenerator, JSONSchema } from './ContentTypeSchemaGenerator';
+
+/**
+ * Validation error with location information
+ */
+export interface ValidationError {
+  field: string;
+  message: string;
+  keyword?: string;
+  params?: Record<string, any>;
+}
+
+/**
+ * Validates front matter data against content type schemas
+ */
+export class FrontMatterValidator {
+  private ajv: Ajv;
+  private schemaCache: Map<string, JSONSchema>;
+
+  constructor() {
+    this.ajv = new Ajv({
+      allErrors: true,
+      verbose: true,
+      strict: false,
+      allowUnionTypes: true
+    });
+    this.schemaCache = new Map();
+  }
+
+  /**
+   * Validate front matter data against a content type
+   * @param data The front matter data to validate
+   * @param contentType The content type to validate against
+   * @returns Array of validation errors (empty if valid)
+   */
+  public validate(data: any, contentType: ContentType): ValidationError[] {
+    if (!contentType || !contentType.fields || contentType.fields.length === 0) {
+      return [];
+    }
+
+    // Get or generate schema
+    const schema = this.getSchema(contentType);
+    if (!schema) {
+      return [];
+    }
+
+    // Compile and validate
+    const validate = this.ajv.compile(schema);
+    const valid = validate(data);
+
+    if (valid) {
+      return [];
+    }
+
+    // Convert AJV errors to our format
+    return this.convertAjvErrors(validate.errors || []);
+  }
+
+  /**
+   * Get or generate schema for a content type
+   * @param contentType The content type
+   * @returns JSON Schema
+   */
+  private getSchema(contentType: ContentType): JSONSchema | null {
+    // Check cache first
+    const cacheKey = contentType.name;
+    if (this.schemaCache.has(cacheKey)) {
+      return this.schemaCache.get(cacheKey) || null;
+    }
+
+    // Generate new schema
+    const schema = ContentTypeSchemaGenerator.generateSchema(contentType);
+    this.schemaCache.set(cacheKey, schema);
+
+    return schema;
+  }
+
+  /**
+   * Clear the schema cache
+   */
+  public clearCache(): void {
+    this.schemaCache.clear();
+  }
+
+  /**
+   * Convert AJV errors to validation errors
+   * @param ajvErrors AJV error objects
+   * @returns Array of validation errors
+   */
+  private convertAjvErrors(ajvErrors: ErrorObject[]): ValidationError[] {
+    const errors: ValidationError[] = [];
+
+    for (const error of ajvErrors) {
+      const field = this.extractFieldName(error.instancePath);
+      const message = this.formatErrorMessage(error, field);
+
+      errors.push({
+        field,
+        message,
+        keyword: error.keyword,
+        params: error.params
+      });
+    }
+
+    return errors;
+  }
+
+  /**
+   * Extract field name from instance path
+   * @param instancePath The JSON pointer path
+   * @returns Field name
+   */
+  private extractFieldName(instancePath: string): string {
+    if (!instancePath || instancePath === '') {
+      return 'root';
+    }
+
+    // Remove leading slash and convert to dot notation
+    return instancePath
+      .replace(/^\//, '')
+      .replace(/\//g, '.')
+      .replace(/~1/g, '/')
+      .replace(/~0/g, '~');
+  }
+
+  /**
+   * Format error message for display
+   * @param error AJV error object
+   * @param field Field name
+   * @returns Formatted error message
+   */
+  private formatErrorMessage(error: ErrorObject, field: string): string {
+    const displayField = field === 'root' ? 'The document' : `Field '${field}'`;
+
+    switch (error.keyword) {
+      case 'required': {
+        const missingProperty = error.params?.missingProperty;
+        return `Missing required field '${missingProperty}'`;
+      }
+
+      case 'type': {
+        const expectedType = error.params?.type;
+        return `${displayField} must be of type ${expectedType}`;
+      }
+
+      case 'enum': {
+        const allowedValues = error.params?.allowedValues;
+        if (allowedValues && Array.isArray(allowedValues)) {
+          return `${displayField} must be one of: ${allowedValues.join(', ')}`;
+        }
+        return `${displayField} has an invalid value`;
+      }
+
+      case 'format': {
+        const format = error.params?.format;
+        return `${displayField} must be in ${format} format`;
+      }
+
+      case 'minimum': {
+        const minimum = error.params?.limit;
+        return `${displayField} must be greater than or equal to ${minimum}`;
+      }
+
+      case 'maximum': {
+        const maximum = error.params?.limit;
+        return `${displayField} must be less than or equal to ${maximum}`;
+      }
+
+      case 'minItems': {
+        const minItems = error.params?.limit;
+        return `${displayField} must have at least ${minItems} items`;
+      }
+
+      case 'maxItems': {
+        const maxItems = error.params?.limit;
+        return `${displayField} must have at most ${maxItems} items`;
+      }
+
+      case 'additionalProperties': {
+        const additionalProperty = error.params?.additionalProperty;
+        return `Unexpected field '${additionalProperty}' is not allowed`;
+      }
+
+      case 'oneOf':
+        return `${displayField} must match exactly one of the allowed schemas`;
+
+      case 'anyOf':
+        return `${displayField} must match at least one of the allowed schemas`;
+
+      default:
+        return error.message || `${displayField} is invalid`;
+    }
+  }
+}

src/helpers/index.ts

@@ -38,3 +38,5 @@ export * from './processFmPlaceholders';
 export * from './processI18nPlaceholders';
 export * from './processPathPlaceholders';
 export * from './processTimePlaceholders';
+export * from './ContentTypeSchemaGenerator';
+export * from './FrontMatterValidator';

src/utils/index.ts

@@ -28,3 +28,4 @@ export * from './sleep';
 export * from './sortPages';
 export * from './unlinkAsync';
 export * from './writeFileAsync';
+export * from '../helpers/ContentTypeSchemaGenerator';