Repository navigation
Expand file tree
/
Copy pathserver.py
More file actions
264 lines (207 loc) · 8.39 KB
/
Copy pathserver.py
File metadata and controls
264 lines (207 loc) · 8.39 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
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
"""MCP server for source code navigation — five tools.
Tool 1: read_symbol(qualified_name)
- Lookup by qualified name, returns identity + code/signature
Tool 2: search_fts_tool(keyword, path_filter)
- FTS5 search, returns located blocks with code preview
Tool 3: read_file_range(file_path, start_line, end_line)
- Read by position, returns code with symbol metadata
Tool 4: get_directory_structure()
- Module/file counts from index
Tool 5: list_databases()
- List available databases and their stats
"""
from __future__ import annotations
import os
import sqlite3
import threading
from pathlib import Path
from typing import Any
from mcp.server.fastmcp import FastMCP
from .db import (
connect,
initialize_schema,
)
from .db import (
get_directory_structure as db_get_directory_structure,
)
from .db import (
read_file_range as db_read_file_range,
)
from .db import (
read_symbol as db_read_symbol,
)
from .db import (
search_fts as db_search_fts,
)
mcp = FastMCP("code-source-sql")
# ── 多数据库注册表 ──────────────────────────────────────────────────
def _parse_db_registry() -> dict[str, str]:
"""从环境变量解析数据库注册表。
CODE_SOURCE_DBS 格式（冒号分隔的路径列表）：
"/data/unreal.db:/data/mygame.db:/data/lib.db"
别名从路径的文件主干名（不含扩展名）自动生成：
"unreal", "mygame", "lib"
CODE_SOURCE_DB 仍然作为主库（默认库），同时加入注册表。
"""
registry: dict[str, str] = {}
# 主库 — CODE_SOURCE_DB
primary = os.environ.get("CODE_SOURCE_DB", "code_source.db")
primary_alias = Path(primary).stem # e.g. "unreal" from "unreal.db"
registry[primary_alias] = primary
registry[""] = primary # 空字符串 → 主库
# 附加库 — CODE_SOURCE_DBS
extra = os.environ.get("CODE_SOURCE_DBS", "")
if extra:
for path_str in extra.split(":"):
path_str = path_str.strip()
if not path_str:
continue
alias = Path(path_str).stem
# 别名冲突时后者覆盖（概率极低，用户自行保证不重名）
registry[alias] = path_str
return registry
_DB_REGISTRY = _parse_db_registry()
# ── 连接缓存 ────────────────────────────────────────────────────────
_conn_cache: dict[str, sqlite3.Connection] = {}
_conn_lock = threading.Lock()
def _get_conn(db_alias: str = "") -> sqlite3.Connection:
"""按别名获取数据库连接，带缓存。"""
db_path = _DB_REGISTRY.get(db_alias)
if db_path is None:
# 别名不存在 → 回退主库
db_path = _DB_REGISTRY[""]
with _conn_lock:
conn = _conn_cache.get(db_path)
if conn is not None:
try:
conn.execute("SELECT 1")
return conn
except Exception:
# 连接已失效，移除缓存
del _conn_cache[db_path]
conn = connect(db_path)
initialize_schema(conn)
_conn_cache[db_path] = conn
return conn
# ── 工具实现 ─────────────────────────────────────────────────────────
@mcp.tool()
def list_databases() -> dict[str, Any]:
"""List available databases and their stats.
Returns the registry of databases the server can access.
Use the 'alias' value as the 'db' parameter in other tools.
The first entry is the default database (used when db is omitted).
"""
results = []
seen_paths: set[str] = set()
for alias, path in _DB_REGISTRY.items():
if alias == "" or path in seen_paths:
continue
seen_paths.add(path)
entry: dict[str, Any] = {"alias": alias, "path": path}
try:
conn = _get_conn(alias)
total = conn.execute(
"SELECT COUNT(*) AS c FROM file_content"
).fetchone()["c"]
sym_total = conn.execute(
"SELECT COUNT(*) AS c FROM symbol_index"
).fetchone()["c"]
entry["total_files"] = total
entry["total_symbols"] = sym_total
except Exception as e:
entry["error"] = str(e)
results.append(entry)
return {"default": _DB_REGISTRY.get("", ""), "databases": results}
@mcp.tool()
def read_symbol(
qualified_name: str,
view: str = "full",
expand_item: list[str] | None = None,
db: str = "",
) -> dict[str, Any]:
"""Read a symbol's source code by qualified name.
Accepts qualified names like 'ClassName::MethodName' or short names.
Supports fuzzy matching: 'MethodName' matches 'ClassName::MethodName'.
view: "full" (default) = complete code, "signature" = summary, "meta" = identity only.
expand_item: optional ranking hints for ambiguous matches.
db: database alias (from list_databases). Default: primary database.
Returns dict with {qn, type, file, range, code?, alt?} or {error, query, fts?}.
"""
conn = _get_conn(db)
entries = db_read_symbol(conn, qualified_name, view=view, expand_item=expand_item)
if not entries:
fts_results = db_search_fts(conn, qualified_name, limit=5)
if fts_results:
return {
"error": "not_found",
"query": qualified_name,
"fts": [
{"file": r["file"], "line": r["line"], "block": r.get("block")}
for r in fts_results[:5]
],
}
return {"error": "not_found", "query": qualified_name}
result = entries[0]
if len(entries) > 1:
result["alt"] = entries[1:5]
return result
@mcp.tool()
def search_fts_tool(
keyword: str = "",
path_filter: str = "",
expand_item: list[str] | None = None,
raw_query: str = "",
db: str = "",
) -> list[dict[str, Any]]:
"""Locate code blocks by keyword or raw FTS5 query.
Two query modes (use one):
- keyword: auto-escaped AND of tokens. Simple, safe. Example: "AddDynamic"
- raw_query: full FTS5 MATCH expression. Column filters, OR, NOT.
Example: '(file_path : "Character.h") AND "BeginPlay"'
FTS5 columns: file_path, module_name, content (all trigram, ≥3 chars).
raw_query operators:
AND → '"AddDynamic" AND "UObject"'
OR → '"AActor" OR "APawn"'
NOT → '"Update" NOT "Test"'
Column filter → '(file_path : "Shader.h") AND "FShaderType"'
Module filter → '(module_name : "Renderer") AND "VirtualTexture"'
Each result includes file, line, and optionally block QN + block_type.
For full code, use read_symbol with the block QN, or read_file_range with file+line.
db: database alias (from list_databases). Default: primary database.
"""
conn = _get_conn(db)
return db_search_fts(conn, keyword, path_filter, expand_item=expand_item, raw_query=raw_query)
@mcp.tool()
def read_file_range(
file_path: str,
start_line: int,
end_line: int,
view: str = "full",
expand_item: list[str] | None = None,
db: str = "",
) -> dict[str, Any]:
"""Read source code by file path and line range.
Use when search_fts returns a location but read_symbol cannot resolve it,
or when reading code outside symbol boundaries.
view: "full" = complete code, "signature" = summary, "meta" = symbols only.
Returns dict with {file, range, code?, symbols?} or {error, file}.
db: database alias (from list_databases). Default: primary database.
"""
conn = _get_conn(db)
entry = db_read_file_range(conn, file_path, start_line, end_line, view=view, expand_item=expand_item)
if not entry:
return {"error": "not_found", "file": file_path}
return entry
@mcp.tool()
def get_directory_structure(db: str = "") -> dict[str, Any]:
"""Get the directory structure summary from the indexed database.
Returns total files, total modules, and module breakdown.
Use module names as path_filter in search_fts_tool.
db: database alias (from list_databases). Default: primary database.
"""
conn = _get_conn(db)
return db_get_directory_structure(conn)
def main() -> None:
mcp.run(transport="stdio")
if __name__ == "__main__":
main()