@webpieces/dev-config 0.2.95 → 0.2.97

package/architecture/executors/validate-dtos/executor.ts

@@ -1,689 +0,0 @@
-/**
- * Validate DTOs Executor
- *
- * Validates that every non-deprecated field in a XxxDto class/interface exists
- * in the corresponding XxxDbo Prisma model. This catches AI agents inventing
- * field names that don't match the database schema.
- *
- * ============================================================================
- * MODES
- * ============================================================================
- * - OFF:            Skip validation entirely
- * - MODIFIED_CLASS: Only validate Dto classes that have changed lines in the diff
- * - MODIFIED_FILES: Validate ALL Dto classes in files that were modified
- *
- * ============================================================================
- * SKIP CONDITIONS
- * ============================================================================
- * - If schema.prisma itself is modified, validation is skipped (schema in flux)
- * - Dto classes ending with "JoinDto" are skipped (they compose other Dtos)
- * - Fields marked @deprecated in a comment are exempt
- *
- * ============================================================================
- * MATCHING
- * ============================================================================
- * - UserDto matches UserDbo by case-insensitive prefix ("user")
- * - Dbo field names are converted from snake_case to camelCase for comparison
- * - Dto fields must be a subset of Dbo fields
- * - Extra Dbo fields are allowed (e.g., password)
- */
-
-import type { ExecutorContext } from '@nx/devkit';
-import { execSync } from 'child_process';
-import * as fs from 'fs';
-import * as path from 'path';
-import * as ts from 'typescript';
-
-export type ValidateDtosMode = 'OFF' | 'MODIFIED_CLASS' | 'MODIFIED_FILES';
-
-export interface ValidateDtosOptions {
-    mode?: ValidateDtosMode;
-    disableAllowed?: boolean;
-    prismaSchemaPath?: string;
-    dtoSourcePaths?: string[];
-    ignoreModifiedUntilEpoch?: number;
-}
-
-export interface ExecutorResult {
-    success: boolean;
-}
-
-interface DtoFieldInfo {
-    name: string;
-    line: number;
-    deprecated: boolean;
-}
-
-interface DtoInfo {
-    name: string;
-    file: string;
-    startLine: number;
-    endLine: number;
-    fields: DtoFieldInfo[];
-}
-
-interface DtoViolation {
-    file: string;
-    line: number;
-    dtoName: string;
-    fieldName: string;
-    dboName: string;
-    availableFields: string[];
-}
-
-interface DboEntry {
-    name: string;
-    fields: Set<string>;
-}
-
-/**
- * Auto-detect the base branch by finding the merge-base with origin/main.
- */
-function detectBase(workspaceRoot: string): string | null {
-    try {
-        const mergeBase = execSync('git merge-base HEAD origin/main', {
-            cwd: workspaceRoot,
-            encoding: 'utf-8',
-            stdio: ['pipe', 'pipe', 'pipe'],
-        }).trim();
-
-        if (mergeBase) {
-            return mergeBase;
-        }
-    } catch {
-        try {
-            const mergeBase = execSync('git merge-base HEAD main', {
-                cwd: workspaceRoot,
-                encoding: 'utf-8',
-                stdio: ['pipe', 'pipe', 'pipe'],
-            }).trim();
-
-            if (mergeBase) {
-                return mergeBase;
-            }
-        } catch {
-            // Ignore
-        }
-    }
-    return null;
-}
-
-/**
- * Get changed files between base and head (or working tree if head not specified).
- */
-// webpieces-disable max-lines-new-methods -- Git command handling with untracked files requires multiple code paths
-function getChangedFiles(workspaceRoot: string, base: string, head?: string): string[] {
-    try {
-        const diffTarget = head ? `${base} ${head}` : base;
-        const output = execSync(`git diff --name-only ${diffTarget}`, {
-            cwd: workspaceRoot,
-            encoding: 'utf-8',
-        });
-        const changedFiles = output
-            .trim()
-            .split('\n')
-            .filter((f) => f.length > 0);
-
-        if (!head) {
-            try {
-                const untrackedOutput = execSync('git ls-files --others --exclude-standard', {
-                    cwd: workspaceRoot,
-                    encoding: 'utf-8',
-                });
-                const untrackedFiles = untrackedOutput
-                    .trim()
-                    .split('\n')
-                    .filter((f) => f.length > 0);
-                const allFiles = new Set([...changedFiles, ...untrackedFiles]);
-                return Array.from(allFiles);
-            } catch {
-                return changedFiles;
-            }
-        }
-
-        return changedFiles;
-    } catch {
-        return [];
-    }
-}
-
-/**
- * Get the diff content for a specific file.
- */
-function getFileDiff(workspaceRoot: string, file: string, base: string, head?: string): string {
-    try {
-        const diffTarget = head ? `${base} ${head}` : base;
-        const diff = execSync(`git diff ${diffTarget} -- "${file}"`, {
-            cwd: workspaceRoot,
-            encoding: 'utf-8',
-        });
-
-        if (!diff && !head) {
-            const fullPath = path.join(workspaceRoot, file);
-            if (fs.existsSync(fullPath)) {
-                const isUntracked = execSync(`git ls-files --others --exclude-standard "${file}"`, {
-                    cwd: workspaceRoot,
-                    encoding: 'utf-8',
-                }).trim();
-
-                if (isUntracked) {
-                    const content = fs.readFileSync(fullPath, 'utf-8');
-                    const lines = content.split('\n');
-                    return lines.map((line) => `+${line}`).join('\n');
-                }
-            }
-        }
-
-        return diff;
-    } catch {
-        return '';
-    }
-}
-
-/**
- * Parse diff to extract changed line numbers (additions only - lines starting with +).
- */
-function getChangedLineNumbers(diffContent: string): Set<number> {
-    const changedLines = new Set<number>();
-    const lines = diffContent.split('\n');
-    let currentLine = 0;
-
-    for (const line of lines) {
-        const hunkMatch = line.match(/^@@ -\d+(?:,\d+)? \+(\d+)(?:,\d+)? @@/);
-        if (hunkMatch) {
-            currentLine = parseInt(hunkMatch[1], 10);
-            continue;
-        }
-
-        if (line.startsWith('+') && !line.startsWith('+++')) {
-            changedLines.add(currentLine);
-            currentLine++;
-        } else if (line.startsWith('-') && !line.startsWith('---')) {
-            // Deletions don't increment line number
-        } else {
-            currentLine++;
-        }
-    }
-
-    return changedLines;
-}
-
-/**
- * Convert a snake_case string to camelCase.
- * e.g., "version_number" -> "versionNumber", "id" -> "id"
- */
-function snakeToCamel(s: string): string {
-    return s.replace(/_([a-z])/g, (_, letter: string) => letter.toUpperCase());
-}
-
-/**
- * Parse schema.prisma to build a map of Dbo model name -> set of field names (camelCase).
- * Only models whose name ends with "Dbo" are included.
- * Field names are converted from snake_case to camelCase since Dto fields use camelCase.
- */
-function parsePrismaSchema(schemaPath: string): Map<string, Set<string>> {
-    const models = new Map<string, Set<string>>();
-
-    if (!fs.existsSync(schemaPath)) {
-        return models;
-    }
-
-    const content = fs.readFileSync(schemaPath, 'utf-8');
-    const lines = content.split('\n');
-
-    let currentModel: string | null = null;
-    let currentFields: Set<string> | null = null;
-
-    for (const line of lines) {
-        const trimmed = line.trim();
-
-        // Match model declaration: model XxxDbo {
-        const modelMatch = trimmed.match(/^model\s+(\w+Dbo)\s*\{/);
-        if (modelMatch) {
-            currentModel = modelMatch[1];
-            currentFields = new Set<string>();
-            continue;
-        }
-
-        // End of model block
-        if (currentModel && trimmed === '}') {
-            models.set(currentModel, currentFields!);
-            currentModel = null;
-            currentFields = null;
-            continue;
-        }
-
-        // Inside a model block - extract field names
-        if (currentModel && currentFields) {
-            // Skip empty lines, comments, and model-level attributes (@@)
-            if (!trimmed || trimmed.startsWith('//') || trimmed.startsWith('@@')) {
-                continue;
-            }
-
-            // Field name is the first word on the line, converted to camelCase
-            const fieldMatch = trimmed.match(/^(\w+)\s/);
-            if (fieldMatch) {
-                currentFields.add(snakeToCamel(fieldMatch[1]));
-            }
-        }
-    }
-
-    return models;
-}
-
-/**
- * Check if a field has @deprecated in a comment above it (within 3 lines).
- */
-function isFieldDeprecated(fileLines: string[], fieldLine: number): boolean {
-    const start = Math.max(0, fieldLine - 4);
-    for (let i = start; i <= fieldLine - 1; i++) {
-        const line = fileLines[i]?.trim() ?? '';
-        if (line.includes('@deprecated')) return true;
-    }
-    return false;
-}
-
-/**
- * Parse a TypeScript file to find Dto class/interface declarations and their fields.
- * Skips classes ending with "JoinDto" since they compose other Dtos.
- */
-// webpieces-disable max-lines-new-methods -- AST traversal for both class and interface Dto detection with field extraction
-function findDtosInFile(filePath: string, workspaceRoot: string): DtoInfo[] {
-    const fullPath = path.join(workspaceRoot, filePath);
-    if (!fs.existsSync(fullPath)) return [];
-
-    const content = fs.readFileSync(fullPath, 'utf-8');
-    const fileLines = content.split('\n');
-    const sourceFile = ts.createSourceFile(filePath, content, ts.ScriptTarget.Latest, true);
-
-    const dtos: DtoInfo[] = [];
-
-    function visit(node: ts.Node): void {
-        const isClass = ts.isClassDeclaration(node);
-        const isInterface = ts.isInterfaceDeclaration(node);
-
-        if ((isClass || isInterface) && node.name) {
-            const name = node.name.text;
-
-            // Must end with Dto but NOT with JoinDto
-            if (name.endsWith('Dto') && !name.endsWith('JoinDto')) {
-                const fields: DtoFieldInfo[] = [];
-                const nodeStart = sourceFile.getLineAndCharacterOfPosition(node.getStart(sourceFile));
-                const nodeEnd = sourceFile.getLineAndCharacterOfPosition(node.getEnd());
-
-                for (const member of node.members) {
-                    if (ts.isPropertyDeclaration(member) || ts.isPropertySignature(member)) {
-                        if (member.name && ts.isIdentifier(member.name)) {
-                            const fieldName = member.name.text;
-                            const startPos = member.getStart(sourceFile);
-                            const pos = sourceFile.getLineAndCharacterOfPosition(startPos);
-                            const line = pos.line + 1;
-                            const deprecated = isFieldDeprecated(fileLines, line);
-
-                            fields.push({ name: fieldName, line, deprecated });
-                        }
-                    }
-                }
-
-                dtos.push({
-                    name,
-                    file: filePath,
-                    startLine: nodeStart.line + 1,
-                    endLine: nodeEnd.line + 1,
-                    fields,
-                });
-            }
-        }
-
-        ts.forEachChild(node, visit);
-    }
-
-    visit(sourceFile);
-    return dtos;
-}
-
-/**
- * Extract the prefix from a Dto/Dbo name by removing the suffix.
- * e.g., "UserDto" -> "user", "UserDbo" -> "user"
- */
-function extractPrefix(name: string, suffix: string): string {
-    return name.slice(0, -suffix.length).toLowerCase();
-}
-
-/**
- * Find violations: Dto fields that don't exist in the corresponding Dbo.
- */
-function findViolations(
-    dtos: DtoInfo[],
-    dboModels: Map<string, Set<string>>
-): DtoViolation[] {
-    const violations: DtoViolation[] = [];
-
-    // Build a lowercase prefix -> Dbo info map
-    const dboByPrefix = new Map<string, DboEntry>();
-    for (const [dboName, fields] of dboModels) {
-        const prefix = extractPrefix(dboName, 'Dbo');
-        dboByPrefix.set(prefix, { name: dboName, fields });
-    }
-
-    for (const dto of dtos) {
-        const prefix = extractPrefix(dto.name, 'Dto');
-        const dbo = dboByPrefix.get(prefix);
-
-        if (!dbo) {
-            // No matching Dbo found - skip (might be a Dto without a DB table)
-            continue;
-        }
-
-        for (const field of dto.fields) {
-            if (field.deprecated) continue;
-
-            if (!dbo.fields.has(field.name)) {
-                violations.push({
-                    file: dto.file,
-                    line: field.line,
-                    dtoName: dto.name,
-                    fieldName: field.name,
-                    dboName: dbo.name,
-                    availableFields: Array.from(dbo.fields).sort(),
-                });
-            }
-        }
-    }
-
-    return violations;
-}
-
-/**
- * Compute similarity between two strings using longest common subsequence ratio.
- * Returns a value between 0 and 1, where 1 is an exact match.
- */
-function similarity(a: string, b: string): number {
-    const al = a.toLowerCase();
-    const bl = b.toLowerCase();
-    if (al === bl) return 1;
-
-    const m = al.length;
-    const n = bl.length;
-    const prev = new Array<number>(n + 1).fill(0);
-    const curr = new Array<number>(n + 1).fill(0);
-
-    for (let i = 1; i <= m; i++) {
-        for (let j = 1; j <= n; j++) {
-            if (al[i - 1] === bl[j - 1]) {
-                curr[j] = prev[j - 1] + 1;
-            } else {
-                curr[j] = Math.max(prev[j], curr[j - 1]);
-            }
-        }
-        for (let j = 0; j <= n; j++) {
-            prev[j] = curr[j];
-            curr[j] = 0;
-        }
-    }
-
-    const lcsLen = prev[n];
-    return (2 * lcsLen) / (m + n);
-}
-
-/**
- * Find the closest matching Dbo field for a given Dto field name.
- * Returns the best match or null if nothing scores above the threshold.
- */
-function suggestClosestField(fieldName: string, availableFields: string[]): string | null {
-    let bestMatch: string | null = null;
-    let bestScore = 0.4; // minimum threshold
-
-    for (const candidate of availableFields) {
-        const score = similarity(fieldName, candidate);
-        if (score > bestScore) {
-            bestScore = score;
-            bestMatch = candidate;
-        }
-    }
-
-    return bestMatch;
-}
-
-/**
- * Report violations to console with prominent AI warning and rename suggestions.
- */
-function reportViolations(violations: DtoViolation[]): void {
-    console.error('');
-    console.error('❌❌❌ DTO FIELD MISMATCH - AI AGENT: READ THIS CAREFULLY ❌❌❌');
-    console.error('');
-    console.error('🛑 DO NOT attempt find/replace or manual renaming. Instead:');
-    console.error('   1. IDE rename is 10x faster than AI renaming PLUS 100% correct vs. AI\'s 80%');
-    console.error('   2. Based on errors below, suggest renames to the user');
-    console.error('   3. Ask the human to use IDE rename-refactoring (F2 / right-click → Rename Symbol)');
-    console.error('');
-    console.error('📚 Every non-deprecated field in a Dto must exist in the corresponding Dbo.');
-    console.error('   Dbo can have extra fields (e.g., password) — Dto fields cannot be invented.');
-    console.error('');
-    console.error('   When needing fields from multiple tables (e.g., a join), use a XxxJoinDto that');
-    console.error('   contains YYDto and ZZDto fields from the other tables instead of flattening.');
-    console.error('   Fields marked @deprecated in the Dto are exempt from this check.');
-    console.error('');
-
-    for (const v of violations) {
-        console.error(`  ❌ ${v.file}:${v.line}`);
-        console.error(`     ${v.dtoName}.${v.fieldName} does not exist in ${v.dboName}`);
-
-        const suggestion = suggestClosestField(v.fieldName, v.availableFields);
-        if (suggestion) {
-            console.error(`     💡 Suggested rename: ${v.fieldName} → ${suggestion}   (Use IDE rename on ${v.dtoName}.${v.fieldName})`);
-        } else {
-            const preview = v.availableFields.slice(0, 8).join(', ');
-            const ellipsis = v.availableFields.length > 8 ? ', ...' : '';
-            console.error(`     💡 No close match found. Available: ${preview}${ellipsis}`);
-        }
-    }
-    console.error('');
-
-    console.error('❌❌❌ AI AGENT: Suggest to the human which are the most likely renames needed and list them ❌❌❌');
-    console.error('');
-}
-
-/**
- * Filter changed files to only TypeScript Dto source files within configured paths.
- */
-function filterDtoFiles(changedFiles: string[], dtoSourcePaths: string[]): string[] {
-    return changedFiles.filter((f) => {
-        if (!f.endsWith('.ts') && !f.endsWith('.tsx')) return false;
-        if (f.includes('.spec.ts') || f.includes('.test.ts')) return false;
-        return dtoSourcePaths.some((srcPath) => f.startsWith(srcPath));
-    });
-}
-
-/**
- * Collect all Dto definitions from the given files.
- */
-function collectDtos(dtoFiles: string[], workspaceRoot: string): DtoInfo[] {
-    const allDtos: DtoInfo[] = [];
-    for (const file of dtoFiles) {
-        const dtos = findDtosInFile(file, workspaceRoot);
-        allDtos.push(...dtos);
-    }
-    return allDtos;
-}
-
-/**
- * Check if a Dto class overlaps with any changed lines in the diff.
- */
-function isDtoTouched(dto: DtoInfo, changedLines: Set<number>): boolean {
-    for (let line = dto.startLine; line <= dto.endLine; line++) {
-        if (changedLines.has(line)) return true;
-    }
-    return false;
-}
-
-/**
- * Filter Dtos to only those that have changed lines in the diff (MODIFIED_CLASS mode).
- */
-function filterTouchedDtos(
-    dtos: DtoInfo[],
-    workspaceRoot: string,
-    base: string,
-    head?: string
-): DtoInfo[] {
-    // Group dtos by file to avoid re-fetching diffs
-    const byFile = new Map<string, DtoInfo[]>();
-    for (const dto of dtos) {
-        const list = byFile.get(dto.file) ?? [];
-        list.push(dto);
-        byFile.set(dto.file, list);
-    }
-
-    const touched: DtoInfo[] = [];
-    for (const [file, fileDtos] of byFile) {
-        const diff = getFileDiff(workspaceRoot, file, base, head);
-        const changedLines = getChangedLineNumbers(diff);
-        for (const dto of fileDtos) {
-            if (isDtoTouched(dto, changedLines)) {
-                touched.push(dto);
-            }
-        }
-    }
-    return touched;
-}
-
-/**
- * Resolve git base ref from env vars or auto-detection.
- */
-function resolveBase(workspaceRoot: string): string | undefined {
-    const envBase = process.env['NX_BASE'];
-    if (envBase) return envBase;
-    return detectBase(workspaceRoot) ?? undefined;
-}
-
-/**
- * Run the core validation after early-exit checks have passed.
- */
-// webpieces-disable max-lines-new-methods -- Core validation orchestration with multiple early-exit checks
-function validateDtoFiles(
-    workspaceRoot: string,
-    prismaSchemaPath: string,
-    changedFiles: string[],
-    dtoSourcePaths: string[],
-    mode: ValidateDtosMode,
-    base: string,
-    head?: string
-): ExecutorResult {
-    if (changedFiles.some((f) => f.endsWith(prismaSchemaPath))) {
-        console.log('⏭️  Skipping validate-dtos (schema.prisma is modified - schema in flux)');
-        console.log('');
-        return { success: true };
-    }
-
-    const dtoFiles = filterDtoFiles(changedFiles, dtoSourcePaths);
-
-    if (dtoFiles.length === 0) {
-        console.log('✅ No Dto files changed');
-        return { success: true };
-    }
-
-    console.log(`📂 Checking ${dtoFiles.length} changed file(s) for Dto definitions...`);
-
-    const fullSchemaPath = path.join(workspaceRoot, prismaSchemaPath);
-    const dboModels = parsePrismaSchema(fullSchemaPath);
-
-    if (dboModels.size === 0) {
-        console.log('⏭️  No Dbo models found in schema.prisma');
-        console.log('');
-        return { success: true };
-    }
-
-    console.log(`   Found ${dboModels.size} Dbo model(s) in schema.prisma`);
-
-    let allDtos = collectDtos(dtoFiles, workspaceRoot);
-
-    if (allDtos.length === 0) {
-        console.log('✅ No Dto definitions found in changed files');
-        return { success: true };
-    }
-
-    // In MODIFIED_CLASS mode, narrow to only Dtos with changed lines
-    if (mode === 'MODIFIED_CLASS') {
-        allDtos = filterTouchedDtos(allDtos, workspaceRoot, base, head);
-        if (allDtos.length === 0) {
-            console.log('✅ No Dto classes were modified');
-            return { success: true };
-        }
-    }
-
-    console.log(`   Validating ${allDtos.length} Dto definition(s)`);
-
-    const violations = findViolations(allDtos, dboModels);
-
-    if (violations.length === 0) {
-        console.log('✅ All Dto fields match their Dbo models');
-        return { success: true };
-    }
-
-    reportViolations(violations);
-    return { success: false };
-}
-
-/**
- * Resolve mode considering ignoreModifiedUntilEpoch override.
- * When active, downgrades to OFF. When expired, logs a warning.
- */
-function resolveMode(normalMode: ValidateDtosMode, epoch: number | undefined): ValidateDtosMode {
-    if (epoch === undefined || normalMode === 'OFF') {
-        return normalMode;
-    }
-    const nowSeconds = Date.now() / 1000;
-    if (nowSeconds < epoch) {
-        const expiresDate = new Date(epoch * 1000).toISOString().split('T')[0];
-        console.log(`\n⏭️  Skipping validate-dtos (ignoreModifiedUntilEpoch active, expires: ${expiresDate})`);
-        console.log('');
-        return 'OFF';
-    }
-    return normalMode;
-}
-
-export default async function runExecutor(
-    options: ValidateDtosOptions,
-    context: ExecutorContext
-): Promise<ExecutorResult> {
-    const workspaceRoot = context.root;
-    const mode = resolveMode(options.mode ?? 'OFF', options.ignoreModifiedUntilEpoch);
-
-    if (mode === 'OFF') {
-        console.log('\n⏭️  Skipping validate-dtos (mode: OFF)');
-        console.log('');
-        return { success: true };
-    }
-
-    const prismaSchemaPath = options.prismaSchemaPath;
-    const dtoSourcePaths = options.dtoSourcePaths ?? [];
-
-    if (!prismaSchemaPath || dtoSourcePaths.length === 0) {
-        const reason = !prismaSchemaPath ? 'no prismaSchemaPath configured' : 'no dtoSourcePaths configured';
-        console.log(`\n⏭️  Skipping validate-dtos (${reason})`);
-        console.log('');
-        return { success: true };
-    }
-
-    console.log('\n📏 Validating DTOs match Prisma Dbo models\n');
-    console.log(`   Mode: ${mode}`);
-    console.log(`   Schema: ${prismaSchemaPath}`);
-    console.log(`   Dto paths: ${dtoSourcePaths.join(', ')}`);
-
-    const base = resolveBase(workspaceRoot);
-    const head = process.env['NX_HEAD'];
-
-    if (!base) {
-        console.log('\n⏭️  Skipping validate-dtos (could not detect base branch)');
-        console.log('');
-        return { success: true };
-    }
-
-    console.log(`   Base: ${base}`);
-    console.log(`   Head: ${head ?? 'working tree (includes uncommitted changes)'}`);
-    console.log('');
-
-    const changedFiles = getChangedFiles(workspaceRoot, base, head);
-
-    return validateDtoFiles(workspaceRoot, prismaSchemaPath, changedFiles, dtoSourcePaths, mode, base, head);
-}