cecli/cecli/utils.py at main · cecli-dev/cecli · GitHub

Repository navigation

History

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

474

475

476

477

478

479

480

481

482

483

484

485

486

487

488

489

490

491

492

493

494

495

496

497

498

499

500

501

502

503

504

505

506

507

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

530

531

532

533

534

535

536

537

538

539

540

541

542

543

544

545

546

547

548

549

550

551

552

553

554

555

556

557

558

559

560

561

562

563

564

565

566

567

568

569

570

571

572

573

574

575

576

577

578

579

580

581

582

583

584

585

586

587

588

589

590

591

592

593

594

595

596

597

598

599

600

601

602

603

604

605

606

607

608

609

610

611

612

613

614

615

616

617

618

619

620

621

622

623

624

625

626

627

628

629

630

631

632

633

634

import glob

import json

import os

import platform

import shlex

import shutil

import ssl

import subprocess

import sys

import tempfile

from pathlib import Path

from cecli.dump import dump # noqa: F401

IMAGE_EXTENSIONS = {".png", ".jpg", ".jpeg", ".gif", ".bmp", ".tiff", ".webp", ".pdf"}

def expand_glob_patterns(patterns):

"""Expand glob patterns in a list of file paths."""

expanded_files = []

for pattern in patterns:

# Check if the pattern contains glob characters

if any(c in pattern for c in "*?[]"):

# Use glob to expand the pattern

matches = glob.glob(pattern, recursive=True)

if matches:

expanded_files.extend(matches)

else:

# If no matches, keep the original pattern

expanded_files.append(pattern)

else:

# Not a glob pattern, keep as is

expanded_files.append(pattern)

return expanded_files

def _execute_fzf(input_data, multi=False):

"""

Runs fzf as a subprocess, feeding it input_data.

Returns the selected items.

"""

if not shutil.which("fzf"):

return [] # fzf not available

fzf_command = ["fzf", "--read0"]

if multi:

fzf_command.append("--multi")

# Recommended flags for a good experience

fzf_command.extend(["--height", "80%", "--reverse"])

process = subprocess.Popen(

fzf_command,

stdin=subprocess.PIPE,

stdout=subprocess.PIPE,

text=False, # Use binary mode for null character handling

)

# fzf expects a null-separated list of strings for multi-line items

# Join with null character instead of newline

input_bytes = "\0".join(input_data).encode("utf-8")

stdout, _ = process.communicate(input_bytes)

if process.returncode == 0:

# fzf returns selected items null-separated when using --read0

output = stdout.decode("utf-8").rstrip("\0\n")

return output.split("\0") if output else []

else:

# User cancelled (e.g., pressed Esc)

return []

def run_fzf(input_data, multi=False, coder=None):

"""

Runs fzf as a subprocess, feeding it input_data.

Returns the selected items.

"""

if not shutil.which("fzf"):

return [] # fzf not available

tui = None

if coder is not None and coder.tui:

tui = coder.tui()

result = []

if tui:

result = tui.run_obstructive(_execute_fzf, input_data, multi=multi)

else:

result = _execute_fzf(input_data, multi=multi)

return result

class IgnorantTemporaryDirectory:

def __init__(self):

if sys.version_info >= (3, 10):

self.temp_dir = tempfile.TemporaryDirectory(ignore_cleanup_errors=True)

else:

self.temp_dir = tempfile.TemporaryDirectory()

def __enter__(self):

return self.temp_dir.__enter__()

def __exit__(self, exc_type, exc_val, exc_tb):

self.cleanup()

def cleanup(self):

try:

self.temp_dir.cleanup()

except (OSError, PermissionError, RecursionError):

pass # Ignore errors (Windows and potential recursion)

def __getattr__(self, item):

return getattr(self.temp_dir, item)

class ChdirTemporaryDirectory(IgnorantTemporaryDirectory):

def __init__(self):

try:

self.cwd = os.getcwd()

except FileNotFoundError:

self.cwd = None

super().__init__()

def __enter__(self):

res = super().__enter__()

os.chdir(Path(self.temp_dir.name).resolve())

return str(Path(res).resolve())

def __exit__(self, exc_type, exc_val, exc_tb):

if self.cwd:

try:

os.chdir(self.cwd)

except FileNotFoundError:

pass

super().__exit__(exc_type, exc_val, exc_tb)

class GitTemporaryDirectory(ChdirTemporaryDirectory):

def __enter__(self):

dname = super().__enter__()

self.repo = make_repo(dname)

return dname

def __exit__(self, exc_type, exc_val, exc_tb):

del self.repo

super().__exit__(exc_type, exc_val, exc_tb)

def make_repo(path=None):

import git

if not path:

path = "."

repo = git.Repo.init(path)

repo.config_writer().set_value("user", "name", "Test User").release()

repo.config_writer().set_value("user", "email", "testuser@example.com").release()

return repo

def is_image_file(file_name):

"""

Check if the given file name has an image file extension.

:param file_name: The name of the file to check.

:return: True if the file is an image, False otherwise.

"""

file_name = str(file_name) # Convert file_name to string

return any(file_name.endswith(ext) for ext in IMAGE_EXTENSIONS)

def safe_abs_path(res):

"Gives an abs path, which safely returns a full (not 8.3) windows path"

res = Path(res).resolve()

return str(res)

def safe_relpath(path, start=None):

"""Return ``os.path.relpath(path, start)``, falling back to *path* on cross-drive.

On Windows, ``os.path.relpath`` raises ``ValueError`` when *path* and *start*

(or the implicit CWD) are on different drives. Keep the absolute path in that

case instead of crashing.

"""

try:

return os.path.relpath(path, start)

except ValueError:

return str(path)

def format_content(role, content):

formatted_lines = []

for line in content.splitlines():

formatted_lines.append(f"{role} {line}")

return "\n".join(formatted_lines)

def format_messages(messages, title=None):

output = []

if title:

output.append(f"{title.upper()} {'*' * 50}")

for msg in messages:

output.append("-------")

role = msg["role"].upper()

content = msg.get("content")

if isinstance(content, list): # Handle list content (e.g., image messages)

for item in content:

if isinstance(item, dict):

for key, value in item.items():

if isinstance(value, dict) and "url" in value:

output.append(f"{role} {key.capitalize()} URL: {value['url']}")

else:

output.append(f"{role} {key}: {value}")

else:

output.append(f"{role} {item}")

elif isinstance(content, str): # Handle string content

# For large content, especially with many files, use a truncated display approach

if len(content) > 5000:

# Count the number of code blocks (approximation)

fence_count = content.count("```") // 2

if fence_count > 5:

# Show truncated content with file count for large files to improve performance

first_line = content.split("\n", 1)[0]

output.append(

f"{role} {first_line} [content with ~{fence_count} files truncated]"

)

else:

output.append(format_content(role, content))

else:

output.append(format_content(role, content))

function_call = msg.get("function_call")

if function_call:

output.append(f"{role} Function Call: {function_call}")

return "\n".join(output)

def show_messages(messages, title=None, functions=None):

formatted_output = format_messages(messages, title)

print(formatted_output)

if functions:

dump(functions)

def split_chat_history_markdown(text, include_tool=False):

messages = []

user = []

assistant = []

tool = []

lines = text.splitlines(keepends=True)

def append_msg(role, lines):

lines = "".join(lines)

if lines.strip():

messages.append(dict(role=role, content=lines))

for line in lines:

if line.startswith("# "):

continue

if line.startswith("> "):

append_msg("assistant", assistant)

assistant = []

append_msg("user", user)

user = []

tool.append(line[2:])

continue

# if line.startswith("#### /"):

# continue

if line.startswith("#### "):

append_msg("assistant", assistant)

assistant = []

append_msg("tool", tool)

tool = []

content = line[5:]

user.append(content)

continue

append_msg("user", user)

user = []

append_msg("tool", tool)

tool = []

assistant.append(line)

append_msg("assistant", assistant)

append_msg("user", user)

if not include_tool:

messages = [m for m in messages if m["role"] != "tool"]

return messages

def get_pip_install(args):

cmd = [

sys.executable,

"-m",

"pip",

"install",

"--upgrade",

"--upgrade-strategy",

"only-if-needed",

]

cmd += args

return cmd

def get_self_upgrade_command():

"""Return the command used to upgrade cecli for the current install method."""

# pipx sets these environment variables inside each tool's venv

if os.environ.get("PIPX_VENV_DIR") or os.environ.get("PIPX_PACKAGE_DIR"):

return ["pipx", "upgrade", "cecli-dev"]

if _is_uv_tool_env():

return ["uv", "tool", "upgrade", "cecli-dev"]

return get_pip_install(["cecli-dev"])

def _is_pip_install_command(cmd):

"""Return True if cmd is a `python -m pip install ...` invocation."""

try:

i = cmd.index("-m")

except ValueError:

return False

return i + 2 < len(cmd) and cmd[i + 1] == "pip" and cmd[i + 2] == "install"

def _is_uv_tool_env():

"""Return True if running inside a `uv tool install` environment."""

cfg = Path(sys.prefix) / "pyvenv.cfg"

try:

if not (cfg.exists() and "uv = " in cfg.read_text(errors="replace")):

return False

except OSError:

return False

prefix = Path(sys.prefix)

tool_dir = os.environ.get("UV_TOOL_DIR")

if tool_dir:

try:

if prefix.is_relative_to(tool_dir):

return True

except (ValueError, OSError):

pass

# Default uv tool installs live under ".../uv/tools/<name>".

# A plain `uv venv` project venv does not have the uv/tools path parts.

return "uv" in prefix.parts and "tools" in prefix.parts

def run_install(cmd):

print()

print("Installing:", printable_shell_command(cmd))

if _is_pip_install_command(cmd):

# First ensure pip is available

ensurepip_cmd = [sys.executable, "-m", "ensurepip", "--upgrade"]

try:

subprocess.run(ensurepip_cmd, capture_output=True, check=False)

except Exception:

pass # Continue even if ensurepip fails

try:

from cecli.waiting import Spinner

output = []

process = subprocess.Popen(

cmd,

stdout=subprocess.PIPE,

stderr=subprocess.STDOUT,

text=True,

bufsize=1,

universal_newlines=True,

encoding=getattr(sys.stdout, "encoding", None) or "utf-8",

errors="replace",

)

spinner = Spinner("Installing...")

while True:

char = process.stdout.read(1)

if not char:

break

output.append(char)

spinner.step()

spinner.end()

return_code = process.wait()

output = "".join(output)

if return_code == 0:

print("Installation complete.")

print()

return True, output

except subprocess.CalledProcessError as e:

print(f"\nError running pip install: {e}")

print("\nInstallation failed.\n")

return False, output

def find_common_root(abs_fnames):

try:

if len(abs_fnames) == 1:

return safe_abs_path(os.path.dirname(list(abs_fnames)[0]))

elif abs_fnames:

return safe_abs_path(os.path.commonpath(list(abs_fnames)))

except (OSError, ValueError):

# ValueError: cross-drive commonpath on Windows (e.g. C: vs E:).

pass

# Restore the original safe fallback: callers assign this straight to

# Coder.root, and Path(False) (or the implicit None for empty input)

# would TypeError on the next Path(root) call. "" resolves as the CWD,

# matching the pre-existing behavior.

return ""

def format_tokens(count):

if count < 1000:

return f"{count}"

elif count < 10000:

return f"{count / 1000:.1f}k"

else:

return f"{round(count / 1000)}k"

def touch_file(fname):

fname = Path(fname)

try:

fname.parent.mkdir(parents=True, exist_ok=True)

fname.touch()

return True

except OSError:

return False

async def check_pip_install_extra(

io, module, prompt, pip_install_cmd=None, self_update=False, cmd=None

if module:

for _attempt in range(2):

try:

__import__(module)

return True

except ssl.SSLError:

# OpenSSL CONF module lazy-init flake (WSL2 + OpenSSL 3.5 + Py3.14):

# the very first SSL context creation can fail with

# ``[CONF: MODULE_INITIALIZATION_ERROR]`` / "unknown error (0x0)",

# and a second attempt succeeds. Mirrors llms.runtime.make_client.

if _attempt == 0:

continue

raise

except (ImportError, ModuleNotFoundError, RuntimeError):

break

if cmd is None:

cmd = get_pip_install(pip_install_cmd or [])

if prompt:

io.tool_warning(prompt)

if self_update and platform.system() == "Windows":

io.tool_output("Run this command to update:")

print()

print(printable_shell_command(cmd)) # plain print so it doesn't line-wrap

return

run_prompt = "Run pip install?"

if cmd and cmd[0] == "uv":

run_prompt = "Run uv tool upgrade?"

elif cmd and cmd[0] == "pipx":

run_prompt = "Run pipx upgrade?"

if not await io.confirm_ask(run_prompt, default="y", subject=printable_shell_command(cmd)):

return

success, output = run_install(cmd)

if success:

if not module:

return True

try:

__import__(module)

return True

except (ImportError, ModuleNotFoundError, RuntimeError) as err:

io.tool_error(str(err))

pass

io.tool_error(output)

print()

print("Install failed, try running this command manually:")

print(printable_shell_command(cmd))

def printable_shell_command(cmd_list):

"""

Convert a list of command arguments to a properly shell-escaped string.

Args:

cmd_list (list): List of command arguments.

Returns:

str: Shell-escaped command string.

"""

return shlex.join(cmd_list)

def split_concatenated_json(s: str) -> list[str]:

"""

Splits a string containing one or more concatenated JSON objects

and returns them as a list of raw strings.

"""

res = []

decoder = json.JSONDecoder()

idx = 0

s_len = len(s)

while idx < s_len:

# 1. Use Regex-free "find" to jump to the next potential JSON start

# This replaces your manual 'while s[i].isspace()' loop

brace_idx = s.find("{", idx)

bracket_idx = s.find("[", idx)

# Determine the earliest starting point

if brace_idx == -1 and bracket_idx == -1:

# No more JSON documents found, but check for trailing text

remainder = s[idx:].strip()

if remainder:

res.append(s[idx:])

break

# Set idx to the first '{' or '[' found

start_index = (

min(brace_idx, bracket_idx)

if (brace_idx != -1 and bracket_idx != -1)

else max(brace_idx, bracket_idx)

)

try:

# 2. Let the C-optimized parser find the end of the object

_, end_idx = decoder.raw_decode(s, start_index)

# 3. Slice the original string and add to results

res.append(s[start_index:end_idx])

# Move our pointer to the end of the last document

idx = end_idx

except json.JSONDecodeError:

# If it looks like JSON but fails (e.g. malformed),

# we skip this character and try to find the next valid start

idx = start_index + 1

return res

def parse_concatenated_json(s: str) -> list:

objs = []

decoder = json.JSONDecoder()

idx = 0

s_len = len(s)

while idx < s_len:

# Jump to the next potential start of a JSON object or array

# This skips whitespace, commas, or "noise" between documents instantly

brace_idx = s.find("{", idx)

bracket_idx = s.find("[", idx)

# Determine which one comes first

if brace_idx == -1 and bracket_idx == -1:

break

elif brace_idx == -1:

idx = bracket_idx

elif bracket_idx == -1:

idx = brace_idx

else:

idx = min(brace_idx, bracket_idx)

try:

# raw_decode attempts to parse starting exactly at idx

obj, end_idx = decoder.raw_decode(s, idx)

objs.append(obj)

idx = end_idx

except json.JSONDecodeError:

# If it's a false start (like a { inside a non-JSON string),

# skip it and keep looking

idx += 1

return objs

def copy_tool_call(tool_call):

"""

Copies a tool call whether it's a Pydantic model, SimpleNamespace, or dict.

"""

from types import SimpleNamespace

if hasattr(tool_call, "model_copy"):

return tool_call.model_copy(deep=True)

if isinstance(tool_call, SimpleNamespace):

import copy

return copy.deepcopy(tool_call)

if isinstance(tool_call, dict):

import copy

return copy.deepcopy(tool_call)

return tool_call

def tool_call_to_dict(tool_call):

"""

Converts any tool-call representation to a dict.

"""

if hasattr(tool_call, "model_dump"):

return tool_call.model_dump()

if hasattr(tool_call, "__dict__"):

res = dict(tool_call.__dict__)

if "function" in res and hasattr(res["function"], "__dict__"):

res["function"] = dict(res["function"].__dict__)

return res

if isinstance(tool_call, dict):

return tool_call

return {}

cecli/cecli/utils.py at main · cecli-dev/cecli · GitHub

Repository navigation

Expand file tree

Breadcrumbs

utils.py

Latest commit

History

Breadcrumbs

utils.py

File metadata and controls