cluxion-agentplugin-preprocessing 0.3.18__tar.gz → 0.3.20__tar.gz

{cluxion_agentplugin_preprocessing-0.3.18 → cluxion_agentplugin_preprocessing-0.3.20}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cluxion-agentplugin-preprocessing
-Version: 0.3.18
+Version: 0.3.20
 Summary: Universal agent plugin for Cluxion preprocessing, honesty contracts, clarification, Rust work queue, and resource-aware harness handoff.
 Project-URL: Homepage, https://github.com/cluxion/cluxion-Agentplugin-preprocessing
 Project-URL: Repository, https://github.com/cluxion/cluxion-Agentplugin-preprocessing

{cluxion_agentplugin_preprocessing-0.3.18 → cluxion_agentplugin_preprocessing-0.3.20}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "cluxion-agentplugin-preprocessing"
-version = "0.3.18"
+version = "0.3.20"
 description = "Universal agent plugin for Cluxion preprocessing, honesty contracts, clarification, Rust work queue, and resource-aware harness handoff."
 readme = "README.md"
 requires-python = ">=3.11"

{cluxion_agentplugin_preprocessing-0.3.18 → cluxion_agentplugin_preprocessing-0.3.20}/src/cluxion_agentplugin_preprocessing/plugin.yaml

@@ -1,5 +1,5 @@
 name: cluxion-agentplugin-preprocessing
-version: 0.3.18
+version: 0.3.20
 description: "Universal agent preprocessing plugin: honesty contracts, clarification, Rust work queue, resource-aware harness handoff. Connected AI calls cluxion tools; plugin does not own models."
 author: cluxion
 kind: standalone

{cluxion_agentplugin_preprocessing-0.3.18 → cluxion_agentplugin_preprocessing-0.3.20}/src/cluxion_runtime/core/hybrid_forget.py

@@ -6,9 +6,11 @@ is removed permanently. Stage 3 is Python-only; the Rust mirror does not run thi
 
 from __future__ import annotations
 
+import os
 import re
 import shutil
 import subprocess
+import sys
 import uuid
 from dataclasses import dataclass
 from typing import TYPE_CHECKING, Protocol
@@ -18,7 +20,14 @@ from cluxion_runtime.core.preprocess import estimate_tokens
 if TYPE_CHECKING:
     from collections.abc import Sequence
 
-_FORGETFORGE_BIN = "forgetforge"
+def _resolve_bin(name: str) -> str:
+    candidate = os.path.join(os.path.dirname(sys.executable), name)
+    if os.path.isfile(candidate) and os.access(candidate, os.X_OK):
+        return candidate
+    return shutil.which(name) or name
+
+
+_FORGETFORGE_BIN = _resolve_bin("forgetforge")
 _JUNK_MARKERS = ("[cluxion: duplicate", "[cluxion digest]")
 _IMPORTANCE_KEYWORDS = (
     "decide",
@@ -75,7 +84,9 @@ class ForgetResult:
 
 
 def forgetforge_available() -> bool:
-    return shutil.which(_FORGETFORGE_BIN) is not None
+    return (os.path.isfile(_FORGETFORGE_BIN) and os.access(_FORGETFORGE_BIN, os.X_OK)) or (
+        shutil.which(_FORGETFORGE_BIN) is not None
+    )
 
 
 def apply_hybrid_forget(

{cluxion_agentplugin_preprocessing-0.3.18 → cluxion_agentplugin_preprocessing-0.3.20}/src/cluxion_runtime/core/llm_compress.py

@@ -13,13 +13,23 @@ import os
 import re
 import shutil
 import subprocess
+import sys
 from typing import TYPE_CHECKING, Protocol
 
 if TYPE_CHECKING:
     from collections.abc import Mapping, Sequence
 
 DEFAULT_TIMEOUT_S = 120.0
-_HERMES_BIN = "hermes"
+
+
+def _resolve_bin(name: str) -> str:
+    candidate = os.path.join(os.path.dirname(sys.executable), name)
+    if os.path.isfile(candidate) and os.access(candidate, os.X_OK):
+        return candidate
+    return shutil.which(name) or name
+
+
+_HERMES_BIN = _resolve_bin("hermes")
 logger = logging.getLogger(__name__)
 
 _SUMMARY_INSTRUCTIONS = (
@@ -54,7 +64,9 @@ class _MessageLike(Protocol):
 
 
 def hermes_available() -> bool:
-    return shutil.which(_HERMES_BIN) is not None
+    return (os.path.isfile(_HERMES_BIN) and os.access(_HERMES_BIN, os.X_OK)) or (
+        shutil.which(_HERMES_BIN) is not None
+    )
 
 
 def summarize_messages(

{cluxion_agentplugin_preprocessing-0.3.18 → cluxion_agentplugin_preprocessing-0.3.20}/src/cluxion_runtime/core/preprocess.py

@@ -172,8 +172,8 @@ def preprocess_work(
 
 def estimate_tokens(text: str) -> int:
     """Estimate tokens conservatively without depending on the old Cluxion OS package."""
-    cjk = sum(1 for ch in text if ord(ch) > 127)
-    ascii_chars = max(0, len(text) - cjk)
+    ascii_chars = len(text.encode("ascii", "ignore"))
+    cjk = len(text) - ascii_chars
     return max(1, cjk + ascii_chars // 4)
 
 

cluxion_agentplugin_preprocessing-0.3.20/tests/runtime/test_bin_resolution.py

@@ -0,0 +1,103 @@
+"""PATH-independent resolution for forgetforge and hermes console scripts."""
+
+from __future__ import annotations
+
+import os
+import stat
+import sys
+from pathlib import Path
+
+import pytest
+
+from cluxion_runtime.core import hybrid_forget, llm_compress
+
+
+def _write_executable(path: Path) -> None:
+    path.write_text("#!/bin/sh\necho ok\n")
+    path.chmod(path.stat().st_mode | stat.S_IXUSR)
+
+
+@pytest.fixture
+def isolated_bin_dir(tmp_path: Path, monkeypatch: pytest.MonkeyPatch) -> Path:
+    """Temp venv bin dir; PATH excludes it so only sys.executable dir can resolve bins."""
+    bin_dir = tmp_path / "bin"
+    bin_dir.mkdir()
+    monkeypatch.setenv("PATH", str(tmp_path / "empty"))
+    monkeypatch.setattr(sys, "executable", str(bin_dir / "python"))
+    return bin_dir
+
+
+def test_forgetforge_available_via_venv_bin_without_path(
+    isolated_bin_dir: Path,
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    forgetforge = isolated_bin_dir / "forgetforge"
+    _write_executable(forgetforge)
+    monkeypatch.setattr(hybrid_forget, "_FORGETFORGE_BIN", hybrid_forget._resolve_bin("forgetforge"))
+
+    assert hybrid_forget.forgetforge_available() is True
+    assert os.path.isabs(hybrid_forget._FORGETFORGE_BIN)
+    assert str(forgetforge) == hybrid_forget._FORGETFORGE_BIN
+
+
+def test_forgetforge_unavailable_when_missing_everywhere(
+    isolated_bin_dir: Path,
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    monkeypatch.setattr(hybrid_forget, "_FORGETFORGE_BIN", hybrid_forget._resolve_bin("forgetforge"))
+    assert hybrid_forget.forgetforge_available() is False
+
+
+def test_forgetforge_available_via_path_fallback(
+    tmp_path: Path,
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    bin_dir = tmp_path / "on-path"
+    bin_dir.mkdir()
+    forgetforge = bin_dir / "forgetforge"
+    _write_executable(forgetforge)
+
+    monkeypatch.setenv("PATH", str(bin_dir))
+    monkeypatch.setattr(sys, "executable", str(tmp_path / "venv" / "bin" / "python"))
+    monkeypatch.setattr(hybrid_forget, "_FORGETFORGE_BIN", hybrid_forget._resolve_bin("forgetforge"))
+
+    assert hybrid_forget.forgetforge_available() is True
+    assert str(forgetforge) == hybrid_forget._FORGETFORGE_BIN
+
+
+def test_hermes_available_via_venv_bin_without_path(
+    isolated_bin_dir: Path,
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    hermes = isolated_bin_dir / "hermes"
+    _write_executable(hermes)
+    monkeypatch.setattr(llm_compress, "_HERMES_BIN", llm_compress._resolve_bin("hermes"))
+
+    assert llm_compress.hermes_available() is True
+    assert os.path.isabs(llm_compress._HERMES_BIN)
+    assert str(hermes) == llm_compress._HERMES_BIN
+
+
+def test_hermes_unavailable_when_missing_everywhere(
+    isolated_bin_dir: Path,
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    monkeypatch.setattr(llm_compress, "_HERMES_BIN", llm_compress._resolve_bin("hermes"))
+    assert llm_compress.hermes_available() is False
+
+
+def test_hermes_available_via_path_fallback(
+    tmp_path: Path,
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    bin_dir = tmp_path / "on-path"
+    bin_dir.mkdir()
+    hermes = bin_dir / "hermes"
+    _write_executable(hermes)
+
+    monkeypatch.setenv("PATH", str(bin_dir))
+    monkeypatch.setattr(sys, "executable", str(tmp_path / "venv" / "bin" / "python"))
+    monkeypatch.setattr(llm_compress, "_HERMES_BIN", llm_compress._resolve_bin("hermes"))
+
+    assert llm_compress.hermes_available() is True
+    assert str(hermes) == llm_compress._HERMES_BIN

{cluxion_agentplugin_preprocessing-0.3.18 → cluxion_agentplugin_preprocessing-0.3.20}/tests/runtime/test_context_compress_llm_forget.py

@@ -9,7 +9,11 @@ import pytest
 
 from cluxion_runtime.core import context_compress
 from cluxion_runtime.core.context_compress import _Msg
-from cluxion_runtime.core.hybrid_forget import _cold_demote, apply_hybrid_forget
+from cluxion_runtime.core.hybrid_forget import (
+    _FORGETFORGE_BIN,
+    _cold_demote,
+    apply_hybrid_forget,
+)
 from cluxion_runtime.core.preprocess import estimate_tokens
 
 
@@ -95,7 +99,13 @@ def test_cold_demote_returns_true_on_successful_store_only(monkeypatch) -> None:
 
     assert _cold_demote("recoverable content", "sess-abc123") is True
     assert len(calls) == 1
-    assert calls[0] == ["forgetforge", "store", "sess-abc123", "--content", "recoverable content"]
+    assert calls[0] == [
+        _FORGETFORGE_BIN,
+        "store",
+        "sess-abc123",
+        "--content",
+        "recoverable content",
+    ]
 
 
 def test_cold_demote_returns_false_when_store_fails(monkeypatch) -> None:

cluxion_agentplugin_preprocessing-0.3.20/tests/runtime/test_estimate_tokens.py

@@ -0,0 +1,52 @@
+"""Byte-identical parity tests for estimate_tokens optimization."""
+
+from __future__ import annotations
+
+import random
+
+from cluxion_runtime.core.preprocess import estimate_tokens
+
+
+def _estimate_tokens_legacy(text: str) -> int:
+    cjk = sum(1 for ch in text if ord(ch) > 127)
+    ascii_chars = max(0, len(text) - cjk)
+    return max(1, cjk + ascii_chars // 4)
+
+
+def test_estimate_tokens_matches_legacy_fixed_cases() -> None:
+    cases = [
+        ("", 1),
+        ("abcd", 1),
+        ("한글텍스트", 5),
+        ("ab한", 1),
+    ]
+    for text, expected in cases:
+        assert estimate_tokens(text) == expected
+        assert estimate_tokens(text) == _estimate_tokens_legacy(text)
+
+
+def test_estimate_tokens_matches_legacy_random_unicode() -> None:
+    rng = random.Random(0)
+    chars: list[str] = []
+    for _ in range(10_000):
+        roll = rng.random()
+        if roll < 0.4:
+            chars.append(chr(rng.randint(0, 127)))
+        elif roll < 0.7:
+            chars.append(chr(rng.randint(128, 0xFFFF)))
+        else:
+            chars.append(chr(rng.randint(0x10000, 0x10FFFF)))
+    text = "".join(chars)
+    assert estimate_tokens(text) == _estimate_tokens_legacy(text)
+
+
+def test_estimate_tokens_lone_surrogate_does_not_raise() -> None:
+    text = "prefix\ud800suffix"
+    assert estimate_tokens(text) == _estimate_tokens_legacy(text)
+
+
+def test_estimate_tokens_legacy_reference_cases() -> None:
+    assert _estimate_tokens_legacy("") == 1
+    assert _estimate_tokens_legacy("abcd") == 1
+    assert _estimate_tokens_legacy("한글텍스트") == 5
+    assert _estimate_tokens_legacy("ab한") == 1