project2txt 1.0.0

package/README.md

@@ -0,0 +1,62 @@
+# project2txt
+
+Concatenates any project (local folder or GitHub repo) into a single text file, ready for AI consumption.
+
+## Install
+
+```bash
+npm i -g project2txt        # global install (run anywhere)
+# or
+npx project2txt <args>      # no install required
+```
+
+## Usage
+
+```bash
+project2txt <path_or_repo> [flags]
+```
+
+## Флаги
+
+- `--out NAME` — output file name (default: project.txt)
+- `--dir FOLDER` — output folder (default: output)
+- `--maxMb N` — max total size in MB (default: 5)
+- `--maxLines N` — max lines per chunk; 0 = unlimited
+- `--keep` — keep temporary clone folder (GitHub mode only)
+
+## Examples
+
+```bash
+# local project
+project2txt ./myProject --out bundle.txt --maxLines 5000
+
+# short repo syntax
+project2txt facebook/react --dir scans --name react.txt
+
+# full URL
+project2txt https://github.com/vercel/next.js --dir output --name next.txt --maxMb 10
+
+# keep clone folder
+project2txt user/repo --keep
+```
+
+## What it does
+
+
+
+- Clones the repo if a GitHub URL is provided.
+- Ignores:  
+  `.git`, `node_modules`, `dist`, `build`, `*.log`, `*.yaml`,  
+  `readme*`, `eslint.config.js`, `manifest.json`.
+- Collects:  
+  `ts`, `tsx`, `js`, `jsx`, `css`, `scss`, `sass`, `less`, `json`, `html`, `md`.
+- Splits output only at file boundaries  
+  (`----- file: … -----`).
+- If a single file exceeds `maxLines`, it is split into `part-N` pieces.
+- Saves the result to the specified label; on subsequent runs, `-N` is appended to the name.
+
+## License
+
+MIT License  
+© da-b1rmuda  
+Web2Bizz

package/index.js

@@ -0,0 +1,158 @@
+#!/usr/bin/env node
+/**
+ * Склеивает проект в один текстовый файл для ИИ
+ * node project2singletxt.js [путь_или_репо] [--out имя] [--dir папка] [--maxMb 5] [--maxLines 5000] [--keep]
+ */
+const fs   = require('fs');
+const path = require('path');
+const { glob } = require('glob');
+const ignore = require('ignore');
+const pino   = require('pino');
+const { execSync } = require('child_process');
+const tmp = require('tmp');
+tmp.setGracefulCleanup();
+
+const logger = pino({ level: 'info' });
+
+/* ========== CLI ========== */
+const argv = process.argv.slice(2);
+let projectRoot = argv[0] || '.';
+let outFile     = 'project.txt';
+let maxSizeMb   = 5;
+let outDir      = 'output';
+let maxLines    = 0;               // 0 = без лимита строк
+const keepFlag  = argv.includes('--keep');
+
+const outFlag  = argv.indexOf('--out');
+const dirFlag  = argv.indexOf('--dir');
+const maxMbFlag= argv.indexOf('--maxMb');
+const maxLFlag = argv.indexOf('--maxLines');
+
+if (outFlag   !== -1 && argv[outFlag   + 1]) outFile  = argv[outFlag   + 1];
+if (dirFlag   !== -1 && argv[dirFlag   + 1]) outDir   = argv[dirFlag   + 1];
+if (maxMbFlag !== -1 && argv[maxMbFlag + 1]) maxSizeMb= Number(argv[maxMbFlag + 1]);
+if (maxLFlag  !== -1 && argv[maxLFlag  + 1]) maxLines = Number(argv[maxLFlag  + 1]);
+
+/* ========== clone GitHub repo if needed ========== */
+function cloneIfNeeded(input) {
+  if (fs.existsSync(input)) return path.resolve(input);
+  const m = input.match(/(?:https?:\/\/github\.com\/|^)([^/]+\/[^/]+?)(?:\.git)?$/i);
+  if (!m) throw new Error('Unsupported URL or path: ' + input);
+  const repo = m[1];
+  const tmpDir = tmp.dirSync({ unsafeCleanup: !keepFlag });
+  const clonePath = path.join(tmpDir.name, repo.replace('/', '_'));
+  logger.info({ repo, clonePath }, 'cloning…');
+  execSync(`git clone --depth 1 https://github.com/${repo}.git "${clonePath}"`, { stdio: 'inherit' });
+  return clonePath;
+}
+projectRoot = cloneIfNeeded(projectRoot);
+outDir      = path.resolve(projectRoot, outDir);
+fs.mkdirSync(outDir, { recursive: true });
+
+const maxBytes = maxSizeMb * 1024 * 1024;
+
+/* ========== ignore ========== */
+const ig = ignore();
+try { ig.add(fs.readFileSync(path.join(projectRoot, '.gitignore'), 'utf8')); } catch (e) {}
+ig.add(['.git', 'node_modules', 'dist', 'build', 'out', '*.log', '*.yaml',
+        'readme*', 'eslint.config.js', 'manifest.json']);
+
+/* ========== extensions ========== */
+const EXT = ['ts','tsx','js','jsx','css','scss','sass','less','json','html','md'];
+const pattern = `**/*.{${EXT.join(',')}}`;
+
+/* ========== utils ========== */
+function uniqueName(folder, name) {
+  const ext = path.extname(name);
+  const base = path.basename(name, ext);
+  let counter = 0;
+  let candidate = path.join(folder, name);
+  while (fs.existsSync(candidate)) {
+    counter++;
+    candidate = path.join(folder, `${base}-${counter}${ext}`);
+  }
+  return candidate;
+}
+
+function splitBigFile(fileBlock, maxLines) {
+  const lines = fileBlock.split(/\r?\n/);
+  const header = lines[0];
+  const body = lines.slice(1);
+  const parts = [];
+  for (let i = 0; i < body.length; i += maxLines - 1) {
+    parts.push([header, ...body.slice(i, i + maxLines - 1)].join('\n'));
+  }
+  return parts;
+}
+
+function splitRespectingBoundaries(fullText, maxLines) {
+  if (!maxLines) return [fullText];
+  const chunks = [];
+  let current = '';
+  let currentLines = 0;
+
+  const re = /(----- file:[^-]+ -----[\s\S]*?)(?=----- file:|$)/g;
+  let m;
+  while ((m = re.exec(fullText)) !== null) {
+    const block = m[1];
+    const lines = block.split(/\r?\n/).length;
+    if (lines > maxLines) {
+      splitBigFile(block, maxLines).forEach(ch => chunks.push(ch));
+      continue;
+    }
+    if (currentLines + lines > maxLines) {
+      chunks.push(current);
+      current = block;
+      currentLines = lines;
+    } else {
+      current += block;
+      currentLines += lines;
+    }
+  }
+  if (current) chunks.push(current);
+  return chunks;
+}
+
+/* ========== main ========== */
+(async () => {
+  const files = await glob(pattern, {
+    cwd: projectRoot,
+    dot: true,
+    markDirectories: true,
+    ignore: ['**/node_modules/**', '**/dist/**', '**/.git/**']
+  });
+
+  const filtered = files.sort().filter(f => !ig.ignores(f));
+  if (!filtered.length) { logger.warn('no files found'); process.exit(0); }
+
+  let fullText = '';
+  for (const rel of filtered) {
+    const head = `----- file: ${rel} -----\n`;
+    const tail = '-----------------------------------------\n';
+    fullText += head + fs.readFileSync(path.join(projectRoot, rel), 'utf8') + tail;
+  }
+
+  const chunks = splitRespectingBoundaries(fullText, maxLines);
+  let writtenBytes = 0;
+
+  chunks.forEach((chunk, idx) => {
+    const ext = path.extname(outFile);
+    const base = path.basename(outFile, ext);
+    const name = `${base}-${idx + 1}${ext}`;
+    const target = uniqueName(outDir, name);
+
+    const buffer = Buffer.from(chunk);
+    if (writtenBytes + buffer.length > maxBytes) {
+      fs.writeFileSync(target, '\n[truncated by size limit]\n');
+      logger.warn('reached size limit');
+      return;
+    }
+    fs.writeFileSync(target, chunk);
+    writtenBytes += buffer.length;
+    logger.info({ chunk: target, lines: chunk.split(/\r?\n/).length }, 'written');
+  });
+  logger.info({ outDir, totalMb: (writtenBytes / 1024 / 1024).toFixed(2) }, 'done');
+})().catch(err => {
+  logger.error(err);
+  process.exit(1);
+});

package/package.json

@@ -0,0 +1,32 @@
+{
+  "name": "project2txt",
+  "version": "1.0.0",
+  "description": "Concatenate any project (local or GitHub repo) into a single text file for AI",
+  "main": "index.js",
+  "bin": {
+    "project2txt": "./index.js"
+  },
+  "preferGlobal": true,
+  "keywords": ["cli", "ai", "source", "concat"],
+  "author": {
+		"name": "da-b1rmuda",
+		"url": "https://github.com/da-b1rmuda"
+	},
+	"license": "MIT",
+	"homepage": "https://github.com/da-b1rmuda/project2txt#readme",
+	"repository": {
+		"type": "git",
+		"url": "https://github.com/da-b1rmuda/project2txt.git"
+	},
+	"bugs": {
+		"url": "https://github.com/da-b1rmuda/project2txt/issues"
+	},
+	"engines": {
+		"node": ">=18.0.0"
+	},
+  "dependencies": {
+    "glob": "^13.0.0",
+    "ignore": "^7.0.5",
+    "pino": "^10.1.1" 
+  }
+}