ctxora 6.2.1 → 6.2.2

package/README.md

@@ -67,6 +67,7 @@ Codex · Claude Code · Cursor · Copilot · MCP clients
 ### Local context engine
 
 - AST-aware chunking for Python, JavaScript, and TypeScript, with bounded fallback chunking for other text formats.
+- Framework-agnostic setup, indexing, retrieval, and MCP startup verified for Python, TypeScript, Flutter, Go, Rust, Java, Kotlin, Swift, PHP, and Ruby repositories.
 - Hybrid lexical and local semantic retrieval using BM25, TF-IDF/LSA, keyword overlap, symbols, and paths.
 - Code dependency graph traversal and graph-augmented context selection.
 - CAG, RAG, hybrid CAG/RAG, long-context, and graph-augmented planning strategies.

package/README.vi.md

@@ -67,6 +67,7 @@ Codex · Claude Code · Cursor · Copilot · MCP clients
 ### Local context engine
 
 - AST-aware chunking cho Python, JavaScript và TypeScript; fallback chunking có giới hạn cho định dạng text khác.
+- Luồng setup, indexing, retrieval và MCP startup độc lập framework, được kiểm thử với Python, TypeScript, Flutter, Go, Rust, Java, Kotlin, Swift, PHP và Ruby.
 - Hybrid lexical và local semantic retrieval bằng BM25, TF-IDF/LSA, keyword overlap, symbol và path.
 - Code dependency graph và graph-augmented context selection.
 - Các strategy CAG, RAG, hybrid CAG/RAG, long context và graph augmented.

package/bin/ctxora.mjs

@@ -6,7 +6,7 @@ import { dirname, join, resolve } from "node:path";
 import { fileURLToPath } from "node:url";
 import { spawnSync } from "node:child_process";
 
-const PACKAGE_VERSION = "6.2.1";
+const PACKAGE_VERSION = "6.2.2";
 const PYTHON_RANGE = "3.10-3.13";
 const PACKAGE_ROOT = resolve(dirname(fileURLToPath(import.meta.url)), "..");
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "ctxora",
-  "version": "6.2.1",
+  "version": "6.2.2",
   "description": "Local-first context engine for coding agents.",
   "type": "module",
   "bin": {

package/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "ctxora-engine"
-version = "6.2.1"
+version = "6.2.2"
 description = "Local-first context engine for coding agents."
 readme = "README.md"
 requires-python = ">=3.10"

package/src/harness_context/workspace/roots.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import fnmatch
+import os
 from pathlib import Path
 
 from harness_context.schemas import HarnessError
@@ -9,8 +10,11 @@ from harness_context.workspace.policy import WorkspacePolicy
 
 DEFAULT_DENY = (
     ".git", ".hg", ".svn", ".ctxora", ".harness", ".ecc", "node_modules", "vendor", ".venv", "venv",
-    "dist", "build", "__pycache__", ".env", ".env.*", "*.pem", "*.key", "id_rsa", "id_ed25519",
-    "*.sqlite", "*.sqlite3", "*.db", "*.egg-info",
+    "dist", "build", "out", "target", "coverage", ".dart_tool", ".gradle", "Pods", "DerivedData",
+    ".build", "obj", ".next", ".nuxt", ".svelte-kit", ".angular", ".turbo", ".parcel-cache",
+    ".expo", ".wrangler", ".serverless", ".terraform", "oh_modules", ".cache", ".pytest_cache",
+    ".mypy_cache", ".ruff_cache", ".tox", ".nox", "__pycache__", ".env", ".env.*",
+    "*.pem", "*.key", "id_rsa", "id_ed25519", "*.sqlite", "*.sqlite3", "*.db", "*.egg-info",
 )
 
 
@@ -53,9 +57,37 @@ class WorkspaceRegistry:
         for name in (".gitignore", ".ctxoraignore", ".harnessignore"):
             ignore = root / name
             if ignore.is_file():
-                patterns.extend(line.strip().lstrip("/") for line in ignore.read_text("utf-8").splitlines() if line.strip() and not line.lstrip().startswith("#") and not line.startswith("!"))
+                for line in ignore.read_text("utf-8").splitlines():
+                    pattern = line.strip()
+                    if not pattern or pattern.startswith("#") or pattern.startswith("!"):
+                        continue
+                    patterns.append(pattern.lstrip("/").rstrip("/"))
         return patterns
 
+    @staticmethod
+    def _ignored(relative: str, path: Path, patterns: list[str]) -> bool:
+        parts = path.parts
+        for pattern in patterns:
+            if relative == pattern or relative.startswith(f"{pattern}/"):
+                return True
+            if fnmatch.fnmatch(relative, pattern) or fnmatch.fnmatch(relative, f"{pattern}/*"):
+                return True
+            if any(fnmatch.fnmatch(part, pattern) for part in parts):
+                return True
+        return False
+
+    @classmethod
+    def _candidate_files(cls, root: Path, patterns: list[str]):
+        for current, directories, files in os.walk(root, topdown=True, followlinks=False):
+            current_path = Path(current)
+            directories[:] = [
+                name for name in directories
+                if not (current_path / name).is_symlink()
+                and not cls._ignored((current_path / name).relative_to(root).as_posix(), current_path / name, patterns)
+            ]
+            for name in files:
+                yield current_path / name
+
     def files(self, workspace_id: str, paths: list[str]) -> list[Path]:
         policy = self.get(workspace_id)
         selected, total = [], 0
@@ -64,12 +96,12 @@ class WorkspaceRegistry:
                 continue
             base = source if source.is_dir() else source.parent
             patterns = self._patterns(base)
-            candidates = [source] if source.is_file() else source.rglob("*")
+            candidates = [source] if source.is_file() else self._candidate_files(source, patterns)
             for path in candidates:
                 if not path.is_file() or path.is_symlink():
                     continue
                 relative = path.relative_to(base).as_posix()
-                if any(fnmatch.fnmatch(relative, pattern) or any(fnmatch.fnmatch(part, pattern) for part in path.parts) for pattern in patterns):
+                if self._ignored(relative, path, patterns):
                     continue
                 size = path.stat().st_size
                 if size > policy.max_file_bytes:
@@ -77,6 +109,10 @@ class WorkspaceRegistry:
                 content = path.read_bytes()
                 if b"\x00" in content[:4096] or contains_secret(content):
                     continue
+                try:
+                    content.decode("utf-8")
+                except UnicodeDecodeError:
+                    continue
                 total += size
                 if total > policy.max_total_bytes or len(selected) >= policy.max_files:
                     raise HarnessError("workspace_limit", "workspace indexing limits exceeded")