openrxiv 0.0.0

package/dist/cli/src/utils/batches.js

@@ -0,0 +1,61 @@
+/**
+ * Parse batch input to support ranges like "1-10" or "batch-1,batch-2"
+ */
+export function parseBatchInput(batchInput) {
+    // Check if it's a comma-separated list first
+    if (batchInput.includes(',')) {
+        const parts = batchInput
+            .split(',')
+            .map((b) => b.trim())
+            .filter((b) => b.length > 0);
+        const allBatches = [];
+        for (const part of parts) {
+            // Check if this part is a range
+            const rangeMatch = part.match(/^(\d+)-(\d+)$/);
+            if (rangeMatch) {
+                const start = parseInt(rangeMatch[1], 10);
+                const end = parseInt(rangeMatch[2], 10);
+                if (start > end) {
+                    throw new Error(`Invalid batch range: start (${start}) cannot be greater than end (${end})`);
+                }
+                if (end - start >= 100) {
+                    throw new Error(`Batch range too large: ${end - start + 1} batches. Maximum allowed: 100`);
+                }
+                for (let i = start; i <= end; i++) {
+                    allBatches.push(i.toString());
+                }
+            }
+            else {
+                // Single batch
+                allBatches.push(part);
+            }
+        }
+        return allBatches;
+    }
+    // Check if it's a single range (e.g., "1-10")
+    const rangeMatch = batchInput.match(/^(\d+)-(\d+)$/);
+    if (rangeMatch) {
+        const start = parseInt(rangeMatch[1], 10);
+        const end = parseInt(rangeMatch[2], 10);
+        if (start > end) {
+            throw new Error(`Invalid batch range: start (${start}) cannot be greater than end (${end})`);
+        }
+        if (end - start >= 100) {
+            throw new Error(`Batch range too large: ${end - start + 1} batches. Maximum allowed: 100`);
+        }
+        const batches = [];
+        for (let i = start; i <= end; i++) {
+            batches.push(i.toString());
+        }
+        return batches;
+    }
+    // Single batch
+    return [batchInput];
+}
+/**
+ * Validate batch format
+ */
+export function validateBatchFormat(batch) {
+    // Allow numeric batches (1, 2, 3) or named batches (batch-1, Batch_01, etc.)
+    return /^[\w\-_]+$/.test(batch);
+}

package/dist/cli/src/utils/batches.test.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=batches.test.d.ts.map

package/dist/cli/src/utils/batches.test.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"batches.test.d.ts","sourceRoot":"","sources":["../../../../src/utils/batches.test.ts"],"names":[],"mappings":""}

package/dist/cli/src/utils/batches.test.js

@@ -0,0 +1,119 @@
+import { describe, it, expect } from 'vitest';
+import { parseBatchInput, validateBatchFormat } from './batches.js';
+describe('Batch Utilities', () => {
+    describe('parseBatchInput', () => {
+        describe('single batches', () => {
+            it('should parse single numeric batch', () => {
+                expect(parseBatchInput('1')).toEqual(['1']);
+                expect(parseBatchInput('42')).toEqual(['42']);
+            });
+            it('should parse single named batch', () => {
+                expect(parseBatchInput('batch-1')).toEqual(['batch-1']);
+                expect(parseBatchInput('Batch_01')).toEqual(['Batch_01']);
+                expect(parseBatchInput('historical-content')).toEqual(['historical-content']);
+            });
+        });
+        describe('numeric ranges', () => {
+            it('should parse simple ranges', () => {
+                expect(parseBatchInput('1-3')).toEqual(['1', '2', '3']);
+                expect(parseBatchInput('5-10')).toEqual(['5', '6', '7', '8', '9', '10']);
+                expect(parseBatchInput('1-1')).toEqual(['1']);
+            });
+            it('should handle large ranges', () => {
+                const result = parseBatchInput('1-100');
+                expect(result).toHaveLength(100);
+                expect(result[0]).toBe('1');
+                expect(result[99]).toBe('100');
+            });
+            it('should reject invalid ranges', () => {
+                expect(() => parseBatchInput('10-5')).toThrow('Invalid batch range: start (10) cannot be greater than end (5)');
+                expect(() => parseBatchInput('5-5')).not.toThrow(); // Valid single-item range
+            });
+            it('should reject ranges that are too large', () => {
+                expect(() => parseBatchInput('1-102')).toThrow('Batch range too large: 102 batches. Maximum allowed: 100');
+                expect(() => parseBatchInput('1-101')).toThrow('Batch range too large: 101 batches. Maximum allowed: 100');
+                expect(() => parseBatchInput('1-100')).not.toThrow(); // Valid maximum range
+            });
+        });
+        describe('comma-separated lists', () => {
+            it('should parse simple comma-separated lists', () => {
+                expect(parseBatchInput('1,2,3')).toEqual(['1', '2', '3']);
+                expect(parseBatchInput('batch-1,batch-2,batch-3')).toEqual([
+                    'batch-1',
+                    'batch-2',
+                    'batch-3',
+                ]);
+            });
+            it('should handle whitespace in comma-separated lists', () => {
+                expect(parseBatchInput('1, 2, 3')).toEqual(['1', '2', '3']);
+                expect(parseBatchInput(' 1 , 2 , 3 ')).toEqual(['1', '2', '3']);
+            });
+            it('should filter out empty entries', () => {
+                expect(parseBatchInput('1,,2,3')).toEqual(['1', '2', '3']);
+                expect(parseBatchInput('1, ,2,3')).toEqual(['1', '2', '3']);
+            });
+        });
+        describe('mixed formats', () => {
+            it('should handle ranges within comma-separated lists', () => {
+                expect(parseBatchInput('1-3,5,7-9')).toEqual(['1', '2', '3', '5', '7', '8', '9']);
+                expect(parseBatchInput('1-5,10,15-17')).toEqual([
+                    '1',
+                    '2',
+                    '3',
+                    '4',
+                    '5',
+                    '10',
+                    '15',
+                    '16',
+                    '17',
+                ]);
+            });
+            it('should handle complex mixed formats', () => {
+                expect(parseBatchInput('1-3,batch-1,5-7,historical')).toEqual([
+                    '1',
+                    '2',
+                    '3',
+                    'batch-1',
+                    '5',
+                    '6',
+                    '7',
+                    'historical',
+                ]);
+            });
+        });
+        describe('edge cases', () => {
+            it('should handle empty string', () => {
+                expect(parseBatchInput('')).toEqual(['']);
+            });
+            it('should handle single comma', () => {
+                expect(parseBatchInput(',')).toEqual([]);
+            });
+            it('should handle multiple commas', () => {
+                expect(parseBatchInput(',,')).toEqual([]);
+                expect(parseBatchInput('1,,2')).toEqual(['1', '2']);
+            });
+        });
+    });
+    describe('validateBatchFormat', () => {
+        it('should accept valid batch names', () => {
+            expect(validateBatchFormat('1')).toBe(true);
+            expect(validateBatchFormat('42')).toBe(true);
+            expect(validateBatchFormat('batch-1')).toBe(true);
+            expect(validateBatchFormat('Batch_01')).toBe(true);
+            expect(validateBatchFormat('historical-content')).toBe(true);
+            expect(validateBatchFormat('content_2023')).toBe(true);
+        });
+        it('should reject invalid batch names', () => {
+            expect(validateBatchFormat('')).toBe(false);
+            expect(validateBatchFormat('batch 1')).toBe(false); // space not allowed
+            expect(validateBatchFormat('batch.1')).toBe(false); // dot not allowed
+            expect(validateBatchFormat('batch/1')).toBe(false); // slash not allowed
+            expect(validateBatchFormat('batch@1')).toBe(false); // @ not allowed
+        });
+        it('should handle special characters correctly', () => {
+            expect(validateBatchFormat('batch-1')).toBe(true); // hyphen allowed
+            expect(validateBatchFormat('batch_1')).toBe(true); // underscore allowed
+            expect(validateBatchFormat('Batch01')).toBe(true); // alphanumeric allowed
+        });
+    });
+});

package/dist/cli/src/utils/default-server.d.ts

@@ -0,0 +1,3 @@
+export declare function getCliName(): 'openrxiv' | 'biorxiv' | 'medrxiv';
+export declare function getDefaultServer(): 'biorxiv' | 'medrxiv';
+//# sourceMappingURL=default-server.d.ts.map

package/dist/cli/src/utils/default-server.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"default-server.d.ts","sourceRoot":"","sources":["../../../../src/utils/default-server.ts"],"names":[],"mappings":"AAEA,wBAAgB,UAAU,IAAI,UAAU,GAAG,SAAS,GAAG,SAAS,CAY/D;AAED,wBAAgB,gBAAgB,IAAI,SAAS,GAAG,SAAS,CAOxD"}

package/dist/cli/src/utils/default-server.js

@@ -0,0 +1,20 @@
+import path from 'node:path';
+export function getCliName() {
+    // process.argv[1] contains the script path, which includes the alias
+    const scriptPath = process.argv[1];
+    const commandName = path.basename(scriptPath);
+    if (commandName.toLowerCase().includes('biorxiv')) {
+        return 'biorxiv';
+    }
+    if (commandName.toLowerCase().includes('medrxiv')) {
+        return 'medrxiv';
+    }
+    return 'openrxiv';
+}
+export function getDefaultServer() {
+    const cliName = getCliName();
+    if (cliName.toLowerCase().includes('medrxiv')) {
+        return 'medrxiv';
+    }
+    return 'biorxiv';
+}

package/dist/cli/src/utils/index.d.ts

@@ -0,0 +1,5 @@
+export * from './months.js';
+export * from './meca-processor.js';
+export * from './batches.js';
+export * from './default-server.js';
+//# sourceMappingURL=index.d.ts.map

package/dist/cli/src/utils/index.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../../src/utils/index.ts"],"names":[],"mappings":"AACA,cAAc,aAAa,CAAC;AAC5B,cAAc,qBAAqB,CAAC;AACpC,cAAc,cAAc,CAAC;AAC7B,cAAc,qBAAqB,CAAC"}

package/dist/cli/src/utils/index.js

@@ -0,0 +1,5 @@
+// Export all utility functions
+export * from './months.js';
+export * from './meca-processor.js';
+export * from './batches.js';
+export * from './default-server.js';

package/dist/cli/src/utils/meca-processor.d.ts

@@ -0,0 +1,28 @@
+export interface ProcessMecaOptions {
+    batch: string;
+    server: 'biorxiv' | 'medrxiv';
+    apiUrl: string;
+    output?: string;
+    s3Key: string;
+    apiKey?: string;
+    selective?: boolean;
+}
+export interface ProcessMecaResult {
+    success: boolean;
+    paper?: any;
+    error?: string;
+}
+/**
+ * Process a MECA file and extract metadata
+ * @param mecaPath Path to the MECA file (local file path)
+ * @param options Processing options
+ * @returns ProcessMecaResult with success status and extracted paper data
+ */
+export declare function processMecaFile(mecaPath: string, options: ProcessMecaOptions): Promise<ProcessMecaResult>;
+/**
+ * Preprocess XML content to fix common HTML entities that cause parsing errors
+ * @param xmlContent Raw XML content
+ * @returns Preprocessed XML content with entities replaced
+ */
+export declare function preprocessXMLContent(xmlContent: string): string;
+//# sourceMappingURL=meca-processor.d.ts.map

package/dist/cli/src/utils/meca-processor.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"meca-processor.d.ts","sourceRoot":"","sources":["../../../../src/utils/meca-processor.ts"],"names":[],"mappings":"AAqCA,MAAM,WAAW,kBAAkB;IACjC,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,EAAE,SAAS,GAAG,SAAS,CAAC;IAC9B,MAAM,EAAE,MAAM,CAAC;IACf,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,SAAS,CAAC,EAAE,OAAO,CAAC;CACrB;AAED,MAAM,WAAW,iBAAiB;IAChC,OAAO,EAAE,OAAO,CAAC;IACjB,KAAK,CAAC,EAAE,GAAG,CAAC;IACZ,KAAK,CAAC,EAAE,MAAM,CAAC;CAChB;AAED;;;;;GAKG;AACH,wBAAsB,eAAe,CACnC,QAAQ,EAAE,MAAM,EAChB,OAAO,EAAE,kBAAkB,GAC1B,OAAO,CAAC,iBAAiB,CAAC,CAkE5B;AAwQD;;;;GAIG;AACH,wBAAgB,oBAAoB,CAAC,UAAU,EAAE,MAAM,GAAG,MAAM,CAwD/D"}