ctxgraph-code 0.4.1__tar.gz → 0.5.1__tar.gz

{ctxgraph_code-0.4.1 → ctxgraph_code-0.5.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ctxgraph-code
-Version: 0.4.1
+Version: 0.5.1
 Summary: Code knowledge graph for Claude Code. Build a relationship graph of your Python codebase and query it during coding sessions.
 Author: ctxgraph-code contributors
 License: MIT
@@ -18,6 +18,9 @@ Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 Requires-Dist: typer>=0.9
 Requires-Dist: rich>=13.0
+Provides-Extra: full
+Requires-Dist: tree-sitter>=0.22; extra == "full"
+Requires-Dist: tree-sitter-language-pack>=0.13; extra == "full"
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0; extra == "dev"
 
@@ -27,6 +30,8 @@ Requires-Dist: pytest>=7.0; extra == "dev"
 
 ```bash
 pip install ctxgraph-code
+# For multi-language support (C, Go, Rust, JS, TS, Java, etc.):
+pip install 'ctxgraph-code[full]'
 cd my-project
 ctxgraph-code setup
 ```
@@ -49,10 +54,10 @@ These questions require running multiple `grep` commands or reading dependency c
 ## Quick Start
 
 ```bash
-# Install
-pip install ctxgraph-code
+# Install (add [full] for multi-language support)
+pip install 'ctxgraph-code[full]'
 
-# Navigate to your Python project
+# Navigate to your project
 cd my-project
 
 # One-command setup: init + build + configure Claude Code
@@ -77,7 +82,14 @@ Interactive walkthrough — prompts for:
 Does everything in one step:
 1. Creates `.ctxgraph/config.toml` with your chosen extensions and excludes
 2. Installs the `/ctxgraph-code` slash command globally (works in every Claude Code session)
-3. Builds the knowledge graph from all matching files
+3. Builds the knowledge graph from all matching files — shows live per-graph progress:
+   ```
+   Building 5 graphs with 8 workers...
+     ✔ src/ (42 files, 156 nodes, 34 edges, 0.8s)
+     ✔ api/ (18 files, 73 nodes, 12 edges, 0.4s)
+     ✔ tests/ (31 files, 89 nodes, 0 edges, 0.6s)
+   Built all 5 graphs in 2.1s
+   ```
 
 Non-interactive mode:
 ```bash
@@ -119,6 +131,15 @@ Scans all matching files in the project, runs AST analysis. Extensions are read
 - `--verbose` / `-v` — show per-file progress
 - `--no-summary` — skip docstring extraction for faster builds
 
+Shows live per-graph progress as each completes:
+```
+Building 5 graphs with 8 workers...
+  ✔ src/ (42 files, 156 nodes, 34 edges, 0.8s)
+  ✔ api/ (18 files, 73 nodes, 12 edges, 0.4s)
+  ✔ tests/ (31 files, 89 nodes, 0 edges, 0.6s)
+Built all 5 graphs in 2.1s
+```
+
 Stores graphs in `.ctxgraph/graphs/<dir>.db` (per-directory) or `.ctxgraph/graph.db` (combined).
 
 > The graph is a **static snapshot**. If code changes, run `ctxgraph-code build` again to refresh. Use `--incremental` to only reprocess changed files.
@@ -222,10 +243,10 @@ ctxgraph-code probe "database connection pool"
 ctxgraph-code probe "user authentication" --max 3
 ```
 
-Searches the graph for relevant nodes **and reads the actual source code** inline. Claude gets paths + source in one command, saving 1–2 tool calls. Shows the first 40 lines of matched files with syntax highlighting.
+Searches the graph for relevant nodes **and reads the actual source code** inline. Claude gets paths + source in one command, saving 1–2 tool calls. Shows the first N lines of matched files with automatic syntax highlighting per language.
 
 - `--max` / `-m` — max files to probe (default: 5)
-- `--context` / `-c` — lines of context around matches (default: 3)
+- `--context` / `-c` — lines to show per file (default: 40, use 0 for full file)
 
 ### `install-hooks`
 
@@ -314,7 +335,10 @@ This catches the common case where you edit a file after building the graph and
 | **Multiprocessing** | Combined graphs split files across CPU cores via `multiprocessing.Pool` |
 | **`--jobs`** | Control parallelism level (default: CPU count) |
 | **Incremental builds** | `--incremental` caches file mtimes, only reprocesses changed files |
-| **Trivial file skip** | `_quick_scan()` pre-checks before AST parse for files with no imports/classes/functions |
+| **Trivial file skip** | `_quick_scan()` pre-checks all files (Python and non-Python) — skips full parse for files with no code |
+| **`follow_symlinks` config** | Respects `follow_symlinks = false` setting to avoid duplicate/broken symlinks |
+| **`max_file_size_mb` config** | Skips files exceeding the configured size limit before reading |
+| **Live build progress** | Per-graph status + timing as each completes during parallel builds |
 | **Cached excludes** | `lru_cache` on `should_exclude()` during `os.walk` |
 | **Batch SQLite inserts** | `executemany` instead of per-row `INSERT` statements |
 | **`--no-summary`** | Skips docstring extraction (fastest rebuilds) |
@@ -386,7 +410,7 @@ Built-in default exclusion patterns (always applied): `__pycache__`, `*.pyc`, `.
 
 | Feature | ctxgraph | ctxgraph-code |
 |---------|----------|---------------|
-| CLI commands | 9+ | 16 (init, build, query, deps, usedby, overview, symbols, context, setup, view, info, install-slash, build-status, probe, install-hooks, uninstall-hooks) |
+| CLI commands | 9+ | 17 (init, build, query, deps, usedby, overview, symbols, context, setup, view, info, install-slash, build-status, probe, install-hooks, uninstall-hooks, version) |
 | LLM integration | Built-in (Ollama, Claude, OpenAI, Azure) | None (delegates to Claude Code) |
 | Chat sessions | Yes | No |
 | Visualizer | D3.js HTML + SVG | D3.js HTML (`view` opens in browser, `--tree` for text) |
@@ -401,6 +425,7 @@ Built-in default exclusion patterns (always applied): `__pycache__`, `*.pyc`, `.
 
 - Python 3.10+
 - A Claude Code subscription (for the `/ctxgraph-code` slash command — the graph itself works standalone)
+- For multi-language analysis (C, Go, Rust, JS, TS, Java, etc.): `pip install 'ctxgraph-code[full]'` to install tree-sitter support
 
 ---
 

{ctxgraph_code-0.4.1 → ctxgraph_code-0.5.1}/README.md

@@ -4,6 +4,8 @@
 
 ```bash
 pip install ctxgraph-code
+# For multi-language support (C, Go, Rust, JS, TS, Java, etc.):
+pip install 'ctxgraph-code[full]'
 cd my-project
 ctxgraph-code setup
 ```
@@ -26,10 +28,10 @@ These questions require running multiple `grep` commands or reading dependency c
 ## Quick Start
 
 ```bash
-# Install
-pip install ctxgraph-code
+# Install (add [full] for multi-language support)
+pip install 'ctxgraph-code[full]'
 
-# Navigate to your Python project
+# Navigate to your project
 cd my-project
 
 # One-command setup: init + build + configure Claude Code
@@ -54,7 +56,14 @@ Interactive walkthrough — prompts for:
 Does everything in one step:
 1. Creates `.ctxgraph/config.toml` with your chosen extensions and excludes
 2. Installs the `/ctxgraph-code` slash command globally (works in every Claude Code session)
-3. Builds the knowledge graph from all matching files
+3. Builds the knowledge graph from all matching files — shows live per-graph progress:
+   ```
+   Building 5 graphs with 8 workers...
+     ✔ src/ (42 files, 156 nodes, 34 edges, 0.8s)
+     ✔ api/ (18 files, 73 nodes, 12 edges, 0.4s)
+     ✔ tests/ (31 files, 89 nodes, 0 edges, 0.6s)
+   Built all 5 graphs in 2.1s
+   ```
 
 Non-interactive mode:
 ```bash
@@ -96,6 +105,15 @@ Scans all matching files in the project, runs AST analysis. Extensions are read
 - `--verbose` / `-v` — show per-file progress
 - `--no-summary` — skip docstring extraction for faster builds
 
+Shows live per-graph progress as each completes:
+```
+Building 5 graphs with 8 workers...
+  ✔ src/ (42 files, 156 nodes, 34 edges, 0.8s)
+  ✔ api/ (18 files, 73 nodes, 12 edges, 0.4s)
+  ✔ tests/ (31 files, 89 nodes, 0 edges, 0.6s)
+Built all 5 graphs in 2.1s
+```
+
 Stores graphs in `.ctxgraph/graphs/<dir>.db` (per-directory) or `.ctxgraph/graph.db` (combined).
 
 > The graph is a **static snapshot**. If code changes, run `ctxgraph-code build` again to refresh. Use `--incremental` to only reprocess changed files.
@@ -199,10 +217,10 @@ ctxgraph-code probe "database connection pool"
 ctxgraph-code probe "user authentication" --max 3
 ```
 
-Searches the graph for relevant nodes **and reads the actual source code** inline. Claude gets paths + source in one command, saving 1–2 tool calls. Shows the first 40 lines of matched files with syntax highlighting.
+Searches the graph for relevant nodes **and reads the actual source code** inline. Claude gets paths + source in one command, saving 1–2 tool calls. Shows the first N lines of matched files with automatic syntax highlighting per language.
 
 - `--max` / `-m` — max files to probe (default: 5)
-- `--context` / `-c` — lines of context around matches (default: 3)
+- `--context` / `-c` — lines to show per file (default: 40, use 0 for full file)
 
 ### `install-hooks`
 
@@ -291,7 +309,10 @@ This catches the common case where you edit a file after building the graph and
 | **Multiprocessing** | Combined graphs split files across CPU cores via `multiprocessing.Pool` |
 | **`--jobs`** | Control parallelism level (default: CPU count) |
 | **Incremental builds** | `--incremental` caches file mtimes, only reprocesses changed files |
-| **Trivial file skip** | `_quick_scan()` pre-checks before AST parse for files with no imports/classes/functions |
+| **Trivial file skip** | `_quick_scan()` pre-checks all files (Python and non-Python) — skips full parse for files with no code |
+| **`follow_symlinks` config** | Respects `follow_symlinks = false` setting to avoid duplicate/broken symlinks |
+| **`max_file_size_mb` config** | Skips files exceeding the configured size limit before reading |
+| **Live build progress** | Per-graph status + timing as each completes during parallel builds |
 | **Cached excludes** | `lru_cache` on `should_exclude()` during `os.walk` |
 | **Batch SQLite inserts** | `executemany` instead of per-row `INSERT` statements |
 | **`--no-summary`** | Skips docstring extraction (fastest rebuilds) |
@@ -363,7 +384,7 @@ Built-in default exclusion patterns (always applied): `__pycache__`, `*.pyc`, `.
 
 | Feature | ctxgraph | ctxgraph-code |
 |---------|----------|---------------|
-| CLI commands | 9+ | 16 (init, build, query, deps, usedby, overview, symbols, context, setup, view, info, install-slash, build-status, probe, install-hooks, uninstall-hooks) |
+| CLI commands | 9+ | 17 (init, build, query, deps, usedby, overview, symbols, context, setup, view, info, install-slash, build-status, probe, install-hooks, uninstall-hooks, version) |
 | LLM integration | Built-in (Ollama, Claude, OpenAI, Azure) | None (delegates to Claude Code) |
 | Chat sessions | Yes | No |
 | Visualizer | D3.js HTML + SVG | D3.js HTML (`view` opens in browser, `--tree` for text) |
@@ -378,6 +399,7 @@ Built-in default exclusion patterns (always applied): `__pycache__`, `*.pyc`, `.
 
 - Python 3.10+
 - A Claude Code subscription (for the `/ctxgraph-code` slash command — the graph itself works standalone)
+- For multi-language analysis (C, Go, Rust, JS, TS, Java, etc.): `pip install 'ctxgraph-code[full]'` to install tree-sitter support
 
 ---
 

{ctxgraph_code-0.4.1 → ctxgraph_code-0.5.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "ctxgraph-code"
-version = "0.4.1"
+version = "0.5.1"
 description = "Code knowledge graph for Claude Code. Build a relationship graph of your Python codebase and query it during coding sessions."
 readme = "README.md"
 license = {text = "MIT"}
@@ -30,6 +30,10 @@ dependencies = [
 ]
 
 [project.optional-dependencies]
+full = [
+    "tree-sitter>=0.22",
+    "tree-sitter-language-pack>=0.13",
+]
 dev = [
     "pytest>=7.0",
 ]

{ctxgraph_code-0.4.1 → ctxgraph_code-0.5.1}/src/ctxgraph_code/analyzers/treesitter/analyzer.py

@@ -38,8 +38,16 @@ class TSAnalyzer:
         if not self.can_handle():
             return TSAnalyzerResult()
 
-        import tree_sitter as ts
-        from tree_sitter_language_pack import get_language
+        try:
+            import tree_sitter as ts
+            from tree_sitter_language_pack import get_language
+        except ImportError:
+            import warnings
+            warnings.warn(
+                f"Missing tree-sitter dependency for {self.lang_name} files. "
+                "Install with: pip install 'ctxgraph-code[full]'"
+            )
+            raise
 
         lang = self._get_lang(self.lang_name)
         if not lang:

{ctxgraph_code-0.4.1 → ctxgraph_code-0.5.1}/src/ctxgraph_code/cli.py

@@ -1,4 +1,4 @@
-from __future__ import annotations
+from __future__ import annotations
 
 import os
 import time
@@ -245,12 +245,8 @@ def _write_slash_command(slash_path: Path, path: Path):
     content = SLASH_COMMAND_TEMPLATE.format(build_time=build_label, available=avail_str)
 
     slash_path.parent.mkdir(parents=True, exist_ok=True)
-    if not slash_path.exists():
-        slash_path.write_text(content, encoding="utf-8")
-        console.print(f"[green][OK] Created [bold]{slash_path}[/bold][/green]")
-    else:
-        console.print(f"[yellow]  Skipped (already exists): [bold]{slash_path}[/bold][/yellow]")
-        console.print("[yellow]  To overwrite, delete it and run again.[/yellow]")
+    slash_path.write_text(content, encoding="utf-8")
+    console.print(f"[green][OK] {'Updated' if slash_path.exists() else 'Created'} [bold]{slash_path}[/bold][/green]")
 
 
 # ── slash command template ───────────────────────────────────────────────────
@@ -311,8 +307,7 @@ def _print_build_hint(path: Path):
 
 def _build_single_graph(path, exts, user_patterns, db_path, label,
                          incremental=False, verbose=False, no_summary=False):
-    ctx_dir = path / ".ctxgraph"
-    mark_build_started(ctx_dir, os.getpid())
+    mark_build_started(path, os.getpid())
     start = time.time()
     with console.status(f"Scanning {', '.join(exts)} files for '{label}'..."):
         stats = build_graph(
@@ -324,7 +319,7 @@ def _build_single_graph(path, exts, user_patterns, db_path, label,
             verbose=verbose,
             no_summary=no_summary,
         )
-    mark_build_complete(ctx_dir, time.time() - start)
+    mark_build_complete(path, time.time() - start)
 
     table = Table(title=f"Graph Build: {label}")
     table.add_column("Metric", style="cyan")
@@ -341,8 +336,7 @@ def _build_single_graph(path, exts, user_patterns, db_path, label,
 def _build_dir_worker(path, exts, user_patterns, db_path, label,
                        incremental=False, verbose=False, no_summary=False):
     """Silent build worker for parallel execution. No console output."""
-    ctx_dir = path / ".ctxgraph"
-    mark_build_started(ctx_dir, os.getpid())
+    mark_build_started(path, os.getpid())
     start = time.time()
     stats = build_graph(
         path,
@@ -353,7 +347,7 @@ def _build_dir_worker(path, exts, user_patterns, db_path, label,
         verbose=verbose,
         no_summary=no_summary,
     )
-    mark_build_complete(ctx_dir, time.time() - start)
+    mark_build_complete(path, time.time() - start)
     return label, stats
 
 
@@ -363,9 +357,11 @@ def _build_dirs_parallel(path, exts, user_patterns, top_dirs, graphs_dir, jobs,
     from concurrent.futures import ThreadPoolExecutor, as_completed
 
     n_workers = max(1, jobs) if jobs > 0 else os.cpu_count() or 1
-    console.print(f"[dim]Building {len(top_dirs)} graphs with {n_workers} workers...[/dim]")
+    total = len(top_dirs)
+    console.print(f"[bold]Building {total} graphs with {n_workers} workers...[/bold]")
 
-    futures = []
+    futures = {}
+    _build_progress_start = time.time()
     with ThreadPoolExecutor(max_workers=n_workers) as pool:
         for d in top_dirs:
             db_path = graphs_dir / f"{d.name}.db"
@@ -373,15 +369,37 @@ def _build_dirs_parallel(path, exts, user_patterns, top_dirs, graphs_dir, jobs,
                 _build_dir_worker, path, exts, user_patterns, db_path, d.name,
                 incremental, verbose, no_summary,
             )
-            futures.append(fut)
+            futures[fut] = d.name
 
         results = []
+        completed = 0
+        err_count = 0
+
         for f in as_completed(futures):
+            label = futures[f]
+            completed += 1
             try:
-                label, stats = f.result()
-                results.append((label, stats))
+                lbl, stats = f.result()
+                results.append((lbl, stats))
+                files = stats.get("files_analyzed", 0)
+                nodes = stats.get("total_nodes", 0)
+                edges = stats.get("total_edges", 0)
+                t = stats.get("elapsed_seconds", 0)
+                console.print(
+                    f"  [green]✔[/green] {label}/ "
+                    f"({files} files, {nodes} nodes, {edges} edges, {t}s)"
+                )
             except Exception as e:
-                results.append(("error", str(e)))
+                results.append((label, str(e)))
+                err_count += 1
+                console.print(f"  [red]✘[/red] {label}/ ([red]{e}[/red])")
+
+        elapsed_total = time.time() - _build_progress_start
+        if err_count:
+            console.print(f"\n[yellow]Built {total - err_count}/{total} graphs"
+                          f" ({err_count} failed) in {elapsed_total:.1f}s[/yellow]")
+        else:
+            console.print(f"\n[green]Built all {total} graphs in {elapsed_total:.1f}s[/green]")
 
     results.sort(key=lambda x: x[0] if isinstance(x[0], str) else "")
 
@@ -413,9 +431,8 @@ def _build_dirs_parallel(path, exts, user_patterns, top_dirs, graphs_dir, jobs,
 
     console.print(summary)
     console.print(
-        f"\n[green]Built {len(top_dirs)} graphs: "
-        f"{total_files} files, {total_nodes} nodes, {total_edges} edges "
-        f"in {total_time}s[/green]"
+        f"[green]Total: {total_files} files, {total_nodes} nodes, {total_edges} edges "
+        f"in {elapsed_total:.1f}s[/green]"
     )
     return results
 
@@ -921,7 +938,7 @@ def probe(
         5, "--max", "-m", help="Maximum files to probe"
     ),
     context_lines: int = typer.Option(
-        3, "--context", "-c", help="Lines of context around matches"
+        40, "--context", "-c", help="Number of lines to show per file"
     ),
     dir_name: Optional[str] = typer.Option(
         None, "--dir", "-d", help="Directory graph to query"
@@ -971,10 +988,16 @@ def probe(
 
         if code:
             lines = code.splitlines()
-            snippet_lines = lines[:40]
+            n = context_lines if context_lines > 0 else len(lines)
+            snippet_lines = lines[:n]
             snippet = "\n".join(snippet_lines)
-            extra = f"\n... ({len(lines) - 40} more lines)" if len(lines) > 40 else ""
-            syntax = Syntax(snippet + extra, "python", theme="monokai", line_numbers=True)
+            extra = f"\n... ({len(lines) - n} more lines)" if len(lines) > n else ""
+            lang = "python"
+            if node.path:
+                ext = Path(node.path).suffix.lower()
+                lang_map = {".js": "javascript", ".ts": "typescript", ".tsx": "typescript", ".jsx": "javascript", ".go": "go", ".rs": "rust", ".c": "c", ".h": "c", ".cpp": "cpp", ".java": "java", ".rb": "ruby", ".kt": "kotlin", ".swift": "swift", ".cs": "csharp", ".scala": "scala", ".lua": "lua", ".zig": "zig", ".php": "php", ".sh": "bash", ".ps1": "powershell", ".json": "json", ".yaml": "yaml", ".yml": "yaml"}
+                lang = lang_map.get(ext, "python")
+            syntax = Syntax(snippet + extra, lang, theme="monokai", line_numbers=True)
             panel = Panel(syntax, title=header, border_style="dim")
             console.print(panel)
         else:
@@ -1082,7 +1105,7 @@ def version():
     try:
         ver = _v("ctxgraph-code")
     except Exception:
-        ver = "0.3.0"
+        ver = "0.5.0"
     console.print(f"ctxgraph-code version [bold]{ver}[/bold]")
 
 

{ctxgraph_code-0.4.1 → ctxgraph_code-0.5.1}/src/ctxgraph_code/config/settings.py

@@ -52,6 +52,14 @@ class Settings:
     def exclude_patterns(self) -> list[str]:
         return self._data["graph"].get("exclude", [])
 
+    @property
+    def follow_symlinks(self) -> bool:
+        return self._data["graph"].get("follow_symlinks", False)
+
+    @property
+    def max_file_size_mb(self) -> int:
+        return self._data["graph"].get("max_file_size_mb", 5)
+
     def to_dict(self) -> dict:
         return dict(self._data)
 

{ctxgraph_code-0.4.1 → ctxgraph_code-0.5.1}/src/ctxgraph_code/graph/builder.py

@@ -59,6 +59,11 @@ def build_graph(
     if jobs <= 0:
         jobs = (os.cpu_count() or 1)
 
+    from ctxgraph_code.config.settings import Settings
+    _settings = Settings(repo_path)
+    follow_symlinks = _settings.follow_symlinks
+    max_bytes = _settings.max_file_size_mb * 1024 * 1024
+
     @lru_cache(maxsize=None)
     def _should_exclude(path: Path) -> bool:
         return should_exclude(path, repo_path, exclude_patterns)
@@ -68,7 +73,7 @@ def build_graph(
     scan_files: list[Path] = []
     current_mtimes: dict[str, float] = {}
 
-    for dirpath, dirnames, filenames in os.walk(repo_path):
+    for dirpath, dirnames, filenames in os.walk(repo_path, followlinks=follow_symlinks):
         dirnames[:] = [
             d for d in dirnames
             if not _should_exclude(repo_path / d)
@@ -140,7 +145,7 @@ def build_graph(
 
         with mp.Pool(jobs) as pool:
             results = [
-                pool.apply_async(_worker_batch, (chunk, repo_path, path_index, no_summary))
+                pool.apply_async(_worker_batch, (chunk, repo_path, path_index, no_summary, max_bytes))
                 for chunk in chunks
             ]
             for i, r in enumerate(results):
@@ -159,7 +164,7 @@ def build_graph(
     else:
         for i, file_path in enumerate(changed_files):
             try:
-                nds, eds, fh = _process_file(file_path, repo_path, path_index, no_summary)
+                nds, eds, fh = _process_file(file_path, repo_path, path_index, no_summary, max_bytes)
                 all_nodes.extend(nds)
                 all_edges.extend(eds)
                 file_hashes.update(fh)
@@ -213,14 +218,22 @@ def build_graph(
 # ── worker helpers ───────────────────────────────────────────────────────────
 
 
-def _quick_scan(source: str) -> bool:
-    """Quick pre-check: does this file have imports, classes, functions,
-    or decorators? Returns True if full AST parsing is needed."""
+def _quick_scan(source: str, is_python: bool = True) -> bool:
+    """Quick pre-check: does this file have meaningful code?
+    Returns True if full parsing is needed."""
     for line in source.splitlines():
         line = line.strip()
-        if not line or line.startswith("#"):
+        if not line or line.startswith(("#", "//", "/*", "*", "--", ";")) and is_python:
+            continue
+        if not line:
             continue
-        if any(line.startswith(kw) for kw in ("import ", "from ", "class ", "def ", "@")):
+        if is_python:
+            if any(line.startswith(kw) for kw in ("import ", "from ", "class ", "def ", "@")):
+                return True
+        else:
+            if any(kw in line for kw in ("function", "class", "struct", "trait", "interface", "import ", "include ", "fn ", "def ", "pub ", "export ", "impl ")):
+                return True
+        if any(c in line for c in ("{", "(", "=", ";")):
             return True
     return False
 
@@ -234,9 +247,17 @@ def _process_file(
     root_path: Path,
     path_index: set[str],
     no_summary: bool = False,
+    max_bytes: int = 0,
 ) -> tuple[list[dict], list[dict], dict[str, str]]:
     rel = str(file_path.relative_to(root_path)).replace("\\", "/")
 
+    if max_bytes > 0:
+        try:
+            if file_path.stat().st_size > max_bytes:
+                return [], [], {}
+        except OSError:
+            pass
+
     try:
         source = file_path.read_text(encoding="utf-8", errors="replace")
     except OSError:
@@ -261,7 +282,7 @@ def _process_python(
 ) -> tuple[list[dict], list[dict], dict[str, str]]:
     fhash = fhash or {}
 
-    if not _quick_scan(source):
+    if not _quick_scan(source, is_python=True):
         return [{
             "id": f"{root_path}:{rel}",
             "type": "file",
@@ -320,13 +341,36 @@ def _process_treesitter(
     fhash: dict[str, str] | None = None,
 ) -> tuple[list[dict], list[dict], dict[str, str]]:
     fhash = fhash or {}
+
+    if not _quick_scan(source, is_python=False):
+        return [{
+            "id": f"{root_path}:{rel}",
+            "type": "file",
+            "name": file_path.name,
+            "path": rel,
+            "parent_id": None,
+            "summary": None,
+            "importance": 0.5,
+            "size_bytes": len(source),
+            "lineno": 0,
+        }], [], fhash
+
     from ctxgraph_code.analyzers.treesitter import TSAnalyzer
 
     analyzer = TSAnalyzer(file_path, root_path)
     if not analyzer.can_handle():
         return [], [], fhash
 
-    result = analyzer.analyze(source)
+    import warnings
+    try:
+        result = analyzer.analyze(source)
+    except ImportError:
+        warnings.warn(
+            f"tree-sitter not installed — skipping {rel}. "
+            "Install with: pip install 'ctxgraph-code[full]'"
+        )
+        return [], [], fhash
+
     return result.nodes, result.edges, fhash
 
 
@@ -335,6 +379,7 @@ def _worker_batch(
     root_path: Path,
     path_index: set[str],
     no_summary: bool = False,
+    max_bytes: int = 0,
 ) -> tuple[list[dict], list[dict], int, int, dict[str, str]]:
     nodes: list[dict] = []
     edges: list[dict] = []
@@ -343,7 +388,7 @@ def _worker_batch(
     err = 0
     for fp in file_paths:
         try:
-            nds, eds, fh = _process_file(fp, root_path, path_index, no_summary)
+            nds, eds, fh = _process_file(fp, root_path, path_index, no_summary, max_bytes)
             nodes.extend(nds)
             edges.extend(eds)
             hashes.update(fh)

{ctxgraph_code-0.4.1 → ctxgraph_code-0.5.1}/src/ctxgraph_code.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ctxgraph-code
-Version: 0.4.1
+Version: 0.5.1
 Summary: Code knowledge graph for Claude Code. Build a relationship graph of your Python codebase and query it during coding sessions.
 Author: ctxgraph-code contributors
 License: MIT
@@ -18,6 +18,9 @@ Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 Requires-Dist: typer>=0.9
 Requires-Dist: rich>=13.0
+Provides-Extra: full
+Requires-Dist: tree-sitter>=0.22; extra == "full"
+Requires-Dist: tree-sitter-language-pack>=0.13; extra == "full"
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0; extra == "dev"
 
@@ -27,6 +30,8 @@ Requires-Dist: pytest>=7.0; extra == "dev"
 
 ```bash
 pip install ctxgraph-code
+# For multi-language support (C, Go, Rust, JS, TS, Java, etc.):
+pip install 'ctxgraph-code[full]'
 cd my-project
 ctxgraph-code setup
 ```
@@ -49,10 +54,10 @@ These questions require running multiple `grep` commands or reading dependency c
 ## Quick Start
 
 ```bash
-# Install
-pip install ctxgraph-code
+# Install (add [full] for multi-language support)
+pip install 'ctxgraph-code[full]'
 
-# Navigate to your Python project
+# Navigate to your project
 cd my-project
 
 # One-command setup: init + build + configure Claude Code
@@ -77,7 +82,14 @@ Interactive walkthrough — prompts for:
 Does everything in one step:
 1. Creates `.ctxgraph/config.toml` with your chosen extensions and excludes
 2. Installs the `/ctxgraph-code` slash command globally (works in every Claude Code session)
-3. Builds the knowledge graph from all matching files
+3. Builds the knowledge graph from all matching files — shows live per-graph progress:
+   ```
+   Building 5 graphs with 8 workers...
+     ✔ src/ (42 files, 156 nodes, 34 edges, 0.8s)
+     ✔ api/ (18 files, 73 nodes, 12 edges, 0.4s)
+     ✔ tests/ (31 files, 89 nodes, 0 edges, 0.6s)
+   Built all 5 graphs in 2.1s
+   ```
 
 Non-interactive mode:
 ```bash
@@ -119,6 +131,15 @@ Scans all matching files in the project, runs AST analysis. Extensions are read
 - `--verbose` / `-v` — show per-file progress
 - `--no-summary` — skip docstring extraction for faster builds
 
+Shows live per-graph progress as each completes:
+```
+Building 5 graphs with 8 workers...
+  ✔ src/ (42 files, 156 nodes, 34 edges, 0.8s)
+  ✔ api/ (18 files, 73 nodes, 12 edges, 0.4s)
+  ✔ tests/ (31 files, 89 nodes, 0 edges, 0.6s)
+Built all 5 graphs in 2.1s
+```
+
 Stores graphs in `.ctxgraph/graphs/<dir>.db` (per-directory) or `.ctxgraph/graph.db` (combined).
 
 > The graph is a **static snapshot**. If code changes, run `ctxgraph-code build` again to refresh. Use `--incremental` to only reprocess changed files.
@@ -222,10 +243,10 @@ ctxgraph-code probe "database connection pool"
 ctxgraph-code probe "user authentication" --max 3
 ```
 
-Searches the graph for relevant nodes **and reads the actual source code** inline. Claude gets paths + source in one command, saving 1–2 tool calls. Shows the first 40 lines of matched files with syntax highlighting.
+Searches the graph for relevant nodes **and reads the actual source code** inline. Claude gets paths + source in one command, saving 1–2 tool calls. Shows the first N lines of matched files with automatic syntax highlighting per language.
 
 - `--max` / `-m` — max files to probe (default: 5)
-- `--context` / `-c` — lines of context around matches (default: 3)
+- `--context` / `-c` — lines to show per file (default: 40, use 0 for full file)
 
 ### `install-hooks`
 
@@ -314,7 +335,10 @@ This catches the common case where you edit a file after building the graph and
 | **Multiprocessing** | Combined graphs split files across CPU cores via `multiprocessing.Pool` |
 | **`--jobs`** | Control parallelism level (default: CPU count) |
 | **Incremental builds** | `--incremental` caches file mtimes, only reprocesses changed files |
-| **Trivial file skip** | `_quick_scan()` pre-checks before AST parse for files with no imports/classes/functions |
+| **Trivial file skip** | `_quick_scan()` pre-checks all files (Python and non-Python) — skips full parse for files with no code |
+| **`follow_symlinks` config** | Respects `follow_symlinks = false` setting to avoid duplicate/broken symlinks |
+| **`max_file_size_mb` config** | Skips files exceeding the configured size limit before reading |
+| **Live build progress** | Per-graph status + timing as each completes during parallel builds |
 | **Cached excludes** | `lru_cache` on `should_exclude()` during `os.walk` |
 | **Batch SQLite inserts** | `executemany` instead of per-row `INSERT` statements |
 | **`--no-summary`** | Skips docstring extraction (fastest rebuilds) |
@@ -386,7 +410,7 @@ Built-in default exclusion patterns (always applied): `__pycache__`, `*.pyc`, `.
 
 | Feature | ctxgraph | ctxgraph-code |
 |---------|----------|---------------|
-| CLI commands | 9+ | 16 (init, build, query, deps, usedby, overview, symbols, context, setup, view, info, install-slash, build-status, probe, install-hooks, uninstall-hooks) |
+| CLI commands | 9+ | 17 (init, build, query, deps, usedby, overview, symbols, context, setup, view, info, install-slash, build-status, probe, install-hooks, uninstall-hooks, version) |
 | LLM integration | Built-in (Ollama, Claude, OpenAI, Azure) | None (delegates to Claude Code) |
 | Chat sessions | Yes | No |
 | Visualizer | D3.js HTML + SVG | D3.js HTML (`view` opens in browser, `--tree` for text) |
@@ -401,6 +425,7 @@ Built-in default exclusion patterns (always applied): `__pycache__`, `*.pyc`, `.
 
 - Python 3.10+
 - A Claude Code subscription (for the `/ctxgraph-code` slash command — the graph itself works standalone)
+- For multi-language analysis (C, Go, Rust, JS, TS, Java, etc.): `pip install 'ctxgraph-code[full]'` to install tree-sitter support
 
 ---
 

ctxgraph_code-0.5.1/src/ctxgraph_code.egg-info/requires.txt

@@ -0,0 +1,9 @@
+typer>=0.9
+rich>=13.0
+
+[dev]
+pytest>=7.0
+
+[full]
+tree-sitter>=0.22
+tree-sitter-language-pack>=0.13

ctxgraph_code-0.4.1/src/ctxgraph_code.egg-info/requires.txt

@@ -1,5 +0,0 @@
-typer>=0.9
-rich>=13.0
-
-[dev]
-pytest>=7.0