mcp-code-indexer 2.4.0__py3-none-any.whl → 3.0.2__py3-none-any.whl

mcp_code_indexer/git_hook_handler.py

@@ -307,17 +307,12 @@ Return ONLY a JSON object:
             except subprocess.CalledProcessError:
                 pass  # No upstream remote
             
-            # Get current branch
-            branch_result = await self._run_git_command(["rev-parse", "--abbrev-ref", "HEAD"])
-            branch = branch_result.strip() if branch_result else "main"
-            
             # Extract project name from remote URL or use directory name
             project_name = self._extract_project_name(remote_origin, project_root)
             
             return {
                 "projectName": project_name,
                 "folderPath": str(project_root),
-                "branch": branch,
                 "remoteOrigin": remote_origin,
                 "upstreamOrigin": upstream_origin
             }
@@ -515,7 +510,7 @@ Return ONLY a JSON object:
             
             if project:
                 overview = await self.db_manager.get_project_overview(
-                    project.id, project_info["branch"]
+                    project.id
                 )
                 return overview.overview if overview else ""
             
@@ -538,7 +533,7 @@ Return ONLY a JSON object:
             
             if project:
                 descriptions = await self.db_manager.get_all_file_descriptions(
-                    project.id, project_info["branch"]
+                    project.id
                 )
                 return {desc.file_path: desc.description for desc in descriptions}
             
@@ -883,7 +878,6 @@ Return ONLY a JSON object:
                 
                 file_desc = FileDescription(
                     project_id=project.id,
-                    branch=project_info["branch"],
                     file_path=file_path,
                     description=description,
                     file_hash=None,
@@ -901,7 +895,6 @@ Return ONLY a JSON object:
                 
                 overview = ProjectOverview(
                     project_id=project.id,
-                    branch=project_info["branch"],
                     overview=overview_update,
                     last_modified=datetime.utcnow(),
                     total_files=len(file_updates),

mcp_code_indexer/migrations/001_initial.sql

@@ -0,0 +1,100 @@
+-- Initial database schema for MCP Code Indexer
+-- Creates tables for projects, file descriptions, and search functionality
+
+-- Enable WAL mode for better concurrency
+PRAGMA journal_mode=WAL;
+
+-- Enable foreign key support
+PRAGMA foreign_keys=ON;
+
+-- Projects table: stores project metadata and identification
+CREATE TABLE IF NOT EXISTS projects (
+    id TEXT PRIMARY KEY,
+    name TEXT NOT NULL,
+    remote_origin TEXT,
+    upstream_origin TEXT,
+    aliases TEXT DEFAULT '[]', -- JSON array of aliases
+    created DATETIME DEFAULT CURRENT_TIMESTAMP,
+    last_accessed DATETIME DEFAULT CURRENT_TIMESTAMP
+);
+
+-- Index for project lookups by various identifiers
+CREATE INDEX IF NOT EXISTS idx_projects_remote_origin ON projects(remote_origin);
+CREATE INDEX IF NOT EXISTS idx_projects_upstream_origin ON projects(upstream_origin);
+CREATE INDEX IF NOT EXISTS idx_projects_name ON projects(name);
+
+-- File descriptions table: stores file content descriptions
+CREATE TABLE IF NOT EXISTS file_descriptions (
+    project_id TEXT NOT NULL,
+    branch TEXT NOT NULL,
+    file_path TEXT NOT NULL,
+    description TEXT NOT NULL,
+    file_hash TEXT,
+    last_modified DATETIME DEFAULT CURRENT_TIMESTAMP,
+    version INTEGER DEFAULT 1,
+    source_project_id TEXT,
+    PRIMARY KEY (project_id, branch, file_path),
+    FOREIGN KEY (project_id) REFERENCES projects(id) ON DELETE CASCADE,
+    FOREIGN KEY (source_project_id) REFERENCES projects(id) ON DELETE SET NULL
+);
+
+-- Indexes for file description queries
+CREATE INDEX IF NOT EXISTS idx_file_descriptions_project_branch ON file_descriptions(project_id, branch);
+CREATE INDEX IF NOT EXISTS idx_file_descriptions_file_hash ON file_descriptions(file_hash);
+CREATE INDEX IF NOT EXISTS idx_file_descriptions_last_modified ON file_descriptions(last_modified);
+
+-- Full-text search table for file descriptions
+CREATE VIRTUAL TABLE IF NOT EXISTS file_descriptions_fts USING fts5(
+    project_id,
+    branch,
+    file_path,
+    description,
+    content='file_descriptions',
+    content_rowid='rowid'
+);
+
+-- Triggers to keep FTS table in sync with file_descriptions
+CREATE TRIGGER IF NOT EXISTS file_descriptions_ai AFTER INSERT ON file_descriptions BEGIN
+  INSERT INTO file_descriptions_fts(rowid, project_id, branch, file_path, description)
+  VALUES (new.rowid, new.project_id, new.branch, new.file_path, new.description);
+END;
+
+CREATE TRIGGER IF NOT EXISTS file_descriptions_ad AFTER DELETE ON file_descriptions BEGIN
+  INSERT INTO file_descriptions_fts(file_descriptions_fts, rowid, project_id, branch, file_path, description)
+  VALUES ('delete', old.rowid, old.project_id, old.branch, old.file_path, old.description);
+END;
+
+CREATE TRIGGER IF NOT EXISTS file_descriptions_au AFTER UPDATE ON file_descriptions BEGIN
+  INSERT INTO file_descriptions_fts(file_descriptions_fts, rowid, project_id, branch, file_path, description)
+  VALUES ('delete', old.rowid, old.project_id, old.branch, old.file_path, old.description);
+  INSERT INTO file_descriptions_fts(rowid, project_id, branch, file_path, description)
+  VALUES (new.rowid, new.project_id, new.branch, new.file_path, new.description);
+END;
+
+-- Merge conflicts table: temporary storage for merge conflicts
+CREATE TABLE IF NOT EXISTS merge_conflicts (
+    id INTEGER PRIMARY KEY AUTOINCREMENT,
+    project_id TEXT NOT NULL,
+    file_path TEXT NOT NULL,
+    source_branch TEXT NOT NULL,
+    target_branch TEXT NOT NULL,
+    source_description TEXT NOT NULL,
+    target_description TEXT NOT NULL,
+    resolution TEXT,
+    created DATETIME DEFAULT CURRENT_TIMESTAMP,
+    FOREIGN KEY (project_id) REFERENCES projects(id) ON DELETE CASCADE
+);
+
+-- Index for merge conflict lookups
+CREATE INDEX IF NOT EXISTS idx_merge_conflicts_project ON merge_conflicts(project_id, source_branch, target_branch);
+
+-- Token cache table: cache token counts for performance
+CREATE TABLE IF NOT EXISTS token_cache (
+    cache_key TEXT PRIMARY KEY,
+    token_count INTEGER NOT NULL,
+    created DATETIME DEFAULT CURRENT_TIMESTAMP,
+    expires DATETIME
+);
+
+-- Index for token cache cleanup
+CREATE INDEX IF NOT EXISTS idx_token_cache_expires ON token_cache(expires);

mcp_code_indexer/migrations/002_performance_indexes.sql

@@ -0,0 +1,61 @@
+-- Performance optimization migration for MCP Code Indexer
+-- Adds additional indexes and optimizations for common query patterns
+
+-- Additional composite indexes for performance optimization
+CREATE INDEX IF NOT EXISTS idx_file_descriptions_project_branch_path ON file_descriptions(project_id, branch, file_path);
+CREATE INDEX IF NOT EXISTS idx_file_descriptions_project_modified ON file_descriptions(project_id, last_modified DESC);
+CREATE INDEX IF NOT EXISTS idx_file_descriptions_source_project ON file_descriptions(source_project_id) WHERE source_project_id IS NOT NULL;
+
+-- Index for merge conflicts by project and branches
+CREATE INDEX IF NOT EXISTS idx_merge_conflicts_project_branches ON merge_conflicts(project_id, source_branch, target_branch, created DESC);
+
+-- Index for token cache expiration cleanup
+CREATE INDEX IF NOT EXISTS idx_token_cache_created ON token_cache(created DESC);
+
+-- Optimize FTS5 for prefix searches and ranking
+DROP TRIGGER IF EXISTS file_descriptions_ai;
+DROP TRIGGER IF EXISTS file_descriptions_ad; 
+DROP TRIGGER IF EXISTS file_descriptions_au;
+DROP TABLE IF EXISTS file_descriptions_fts;
+
+-- Recreate FTS table with better configuration
+CREATE VIRTUAL TABLE file_descriptions_fts USING fts5(
+    project_id UNINDEXED,
+    branch UNINDEXED,
+    file_path,
+    description,
+    content='file_descriptions',
+    content_rowid='rowid',
+    prefix='2 3 4'  -- Enable prefix searching for 2-4 character prefixes
+);
+
+-- Recreate triggers for FTS sync
+CREATE TRIGGER file_descriptions_ai AFTER INSERT ON file_descriptions BEGIN
+  INSERT INTO file_descriptions_fts(rowid, project_id, branch, file_path, description)
+  VALUES (new.rowid, new.project_id, new.branch, new.file_path, new.description);
+END;
+
+CREATE TRIGGER file_descriptions_ad AFTER DELETE ON file_descriptions BEGIN
+  INSERT INTO file_descriptions_fts(file_descriptions_fts, rowid, project_id, branch, file_path, description)
+  VALUES ('delete', old.rowid, old.project_id, old.branch, old.file_path, old.description);
+END;
+
+CREATE TRIGGER file_descriptions_au AFTER UPDATE ON file_descriptions BEGIN
+  INSERT INTO file_descriptions_fts(file_descriptions_fts, rowid, project_id, branch, file_path, description)
+  VALUES ('delete', old.rowid, old.project_id, old.branch, old.file_path, old.description);
+  INSERT INTO file_descriptions_fts(rowid, project_id, branch, file_path, description)
+  VALUES (new.rowid, new.project_id, new.branch, new.file_path, new.description);
+END;
+
+-- Rebuild FTS index if there's existing data
+INSERT INTO file_descriptions_fts(file_descriptions_fts) VALUES('rebuild');
+
+-- Set SQLite optimizations
+PRAGMA optimize;
+PRAGMA auto_vacuum = INCREMENTAL;
+PRAGMA temp_store = MEMORY;
+PRAGMA mmap_size = 268435456; -- 256MB memory mapping
+PRAGMA cache_size = -64000;   -- 64MB cache
+
+-- Update statistics for query planner
+ANALYZE;

mcp_code_indexer/migrations/003_project_overviews.sql

@@ -0,0 +1,20 @@
+-- Migration 003: Add project_overviews table for condensed codebase overviews
+-- This table stores comprehensive narrative overviews of entire codebases
+-- as an alternative to file-by-file descriptions
+
+CREATE TABLE IF NOT EXISTS project_overviews (
+    project_id TEXT NOT NULL,
+    branch TEXT NOT NULL,
+    overview TEXT NOT NULL,
+    last_modified TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
+    total_files INTEGER NOT NULL DEFAULT 0,
+    total_tokens INTEGER NOT NULL DEFAULT 0,
+    PRIMARY KEY (project_id, branch),
+    FOREIGN KEY (project_id) REFERENCES projects(id) ON DELETE CASCADE
+) WITHOUT ROWID;
+
+-- Index for temporal queries and overview management
+CREATE INDEX IF NOT EXISTS idx_project_overviews_last_modified ON project_overviews(last_modified);
+
+-- Index for project-based queries
+CREATE INDEX IF NOT EXISTS idx_project_overviews_project_id ON project_overviews(project_id);

mcp_code_indexer/migrations/004_remove_branch_dependency.sql

@@ -0,0 +1,166 @@
+-- Migration 004: Remove branch dependency from database schema
+-- This migration consolidates multi-branch data and simplifies the schema
+-- by removing branch columns from file_descriptions and project_overviews tables
+
+-- Ensure WAL mode is enabled for safe migrations
+PRAGMA journal_mode=WAL;
+
+-- Enable foreign key support
+PRAGMA foreign_keys=ON;
+
+-- Start transaction for atomic migration
+BEGIN TRANSACTION;
+
+-- Create new file_descriptions table without branch dependency
+CREATE TABLE file_descriptions_new (
+    id INTEGER PRIMARY KEY AUTOINCREMENT,
+    project_id TEXT NOT NULL,
+    file_path TEXT NOT NULL,
+    description TEXT NOT NULL,
+    file_hash TEXT,
+    last_modified DATETIME DEFAULT CURRENT_TIMESTAMP,
+    version INTEGER DEFAULT 1,
+    source_project_id TEXT,
+    to_be_cleaned INTEGER DEFAULT NULL, -- UNIX timestamp for cleanup, NULL = active
+    UNIQUE(project_id, file_path),
+    FOREIGN KEY (project_id) REFERENCES projects(id) ON DELETE CASCADE,
+    FOREIGN KEY (source_project_id) REFERENCES projects(id) ON DELETE SET NULL
+);
+
+-- Create indexes for the new table
+CREATE INDEX idx_file_descriptions_new_project_id ON file_descriptions_new(project_id);
+CREATE INDEX idx_file_descriptions_new_file_hash ON file_descriptions_new(file_hash);
+CREATE INDEX idx_file_descriptions_new_last_modified ON file_descriptions_new(last_modified);
+CREATE INDEX idx_file_descriptions_new_to_be_cleaned ON file_descriptions_new(to_be_cleaned);
+
+-- Consolidate data from old table - keep most recent description per file
+-- This handles multi-branch scenarios by selecting the newest data
+INSERT INTO file_descriptions_new (
+    project_id, file_path, description, file_hash, last_modified, version, source_project_id
+)
+SELECT 
+    project_id,
+    file_path,
+    description,
+    file_hash,
+    last_modified,
+    version,
+    source_project_id
+FROM (
+    SELECT 
+        project_id,
+        file_path,
+        description,
+        file_hash,
+        last_modified,
+        version,
+        source_project_id,
+        ROW_NUMBER() OVER (
+            PARTITION BY project_id, file_path 
+            ORDER BY last_modified DESC
+        ) as rn
+    FROM file_descriptions
+) ranked_descriptions
+WHERE rn = 1;
+
+-- Create new project_overviews table without branch dependency
+CREATE TABLE project_overviews_new (
+    project_id TEXT PRIMARY KEY,
+    overview TEXT NOT NULL,
+    last_modified TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
+    total_files INTEGER NOT NULL DEFAULT 0,
+    total_tokens INTEGER NOT NULL DEFAULT 0,
+    FOREIGN KEY (project_id) REFERENCES projects(id) ON DELETE CASCADE
+);
+
+-- Create indexes for the new table
+CREATE INDEX idx_project_overviews_new_last_modified ON project_overviews_new(last_modified);
+
+-- Consolidate project overviews - keep the one with most tokens (most comprehensive)
+INSERT INTO project_overviews_new (
+    project_id, overview, last_modified, total_files, total_tokens
+)
+SELECT 
+    project_id,
+    overview,
+    last_modified,
+    total_files,
+    total_tokens
+FROM (
+    SELECT 
+        project_id,
+        overview,
+        last_modified,
+        total_files,
+        total_tokens,
+        ROW_NUMBER() OVER (
+            PARTITION BY project_id 
+            ORDER BY total_tokens DESC, last_modified DESC
+        ) as rn
+    FROM project_overviews
+) ranked_overviews
+WHERE rn = 1;
+
+-- Drop FTS5 triggers for old table
+DROP TRIGGER IF EXISTS file_descriptions_ai;
+DROP TRIGGER IF EXISTS file_descriptions_ad;
+DROP TRIGGER IF EXISTS file_descriptions_au;
+
+-- Drop FTS5 virtual table
+DROP TABLE IF EXISTS file_descriptions_fts;
+
+-- Drop old tables
+DROP TABLE file_descriptions;
+DROP TABLE project_overviews;
+
+-- Rename new tables to original names
+ALTER TABLE file_descriptions_new RENAME TO file_descriptions;
+ALTER TABLE project_overviews_new RENAME TO project_overviews;
+
+-- Create new FTS5 virtual table without branch column
+CREATE VIRTUAL TABLE file_descriptions_fts USING fts5(
+    project_id,
+    file_path,
+    description,
+    content='file_descriptions',
+    content_rowid='id'
+);
+
+-- Populate FTS5 table with existing data (only active records)
+INSERT INTO file_descriptions_fts(rowid, project_id, file_path, description)
+SELECT id, project_id, file_path, description 
+FROM file_descriptions 
+WHERE to_be_cleaned IS NULL;
+
+-- Create new FTS5 triggers for the updated schema
+CREATE TRIGGER file_descriptions_ai AFTER INSERT ON file_descriptions BEGIN
+  -- Only index active records (not marked for cleanup)
+  INSERT INTO file_descriptions_fts(rowid, project_id, file_path, description)
+  SELECT new.id, new.project_id, new.file_path, new.description
+  WHERE new.to_be_cleaned IS NULL;
+END;
+
+CREATE TRIGGER file_descriptions_ad AFTER DELETE ON file_descriptions BEGIN
+  INSERT INTO file_descriptions_fts(file_descriptions_fts, rowid, project_id, file_path, description)
+  VALUES ('delete', old.id, old.project_id, old.file_path, old.description);
+END;
+
+CREATE TRIGGER file_descriptions_au AFTER UPDATE ON file_descriptions BEGIN
+  -- Remove old record from FTS
+  INSERT INTO file_descriptions_fts(file_descriptions_fts, rowid, project_id, file_path, description)
+  VALUES ('delete', old.id, old.project_id, old.file_path, old.description);
+  
+  -- Add new record only if it's active (not marked for cleanup)
+  INSERT INTO file_descriptions_fts(rowid, project_id, file_path, description)
+  SELECT new.id, new.project_id, new.file_path, new.description
+  WHERE new.to_be_cleaned IS NULL;
+END;
+
+-- Update merge_conflicts table to remove branch references (optional cleanup)
+-- This table structure can remain as-is since it's used for temporary conflict resolution
+-- but we'll remove unused indexes that reference branches
+DROP INDEX IF EXISTS idx_merge_conflicts_project;
+CREATE INDEX idx_merge_conflicts_project ON merge_conflicts(project_id, created);
+
+-- Commit the migration
+COMMIT;