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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265
|
#!/usr/bin/env python
"""
Update the maps of builtin types and names.
This script updates some of the files in psycopg source code with data read
from a database catalog.
Hint: use docker to upgrade types from a new version in isolation. Run:
docker run --rm -p 11111:5432 --name pg -e POSTGRES_PASSWORD=password postgres:TAG
with a specified version tag, and then query it using:
%(prog)s "host=localhost port=11111 user=postgres password=password"
"""
from __future__ import annotations
import re
import argparse
import subprocess as sp
from typing import TypeAlias
from pathlib import Path
import psycopg
from psycopg.pq import version_pretty
from psycopg.crdb import CrdbConnection
from psycopg.rows import TupleRow
Connection: TypeAlias = psycopg.Connection[TupleRow]
ROOT = Path(__file__).parent.parent
def main() -> None:
opt = parse_cmdline()
if CrdbConnection.is_crdb(conn := psycopg.connect(opt.dsn, autocommit=True)):
conn = CrdbConnection.connect(opt.dsn, autocommit=True)
update_crdb_python_oids(conn)
else:
update_python_oids(conn)
update_python_types(conn)
update_cython_oids(conn)
def update_python_types(conn: Connection) -> None:
fn = ROOT / "psycopg/psycopg/postgres.py"
lines = []
lines.extend(get_version_comment(conn))
lines.extend(get_py_types(conn))
lines.extend(get_py_ranges(conn))
lines.extend(get_py_multiranges(conn))
update_file(fn, lines)
sp.check_call(["black", "-q", fn])
def update_python_oids(conn: Connection) -> None:
fn = ROOT / "psycopg/psycopg/_oids.py"
lines = []
lines.extend(get_version_comment(conn))
lines.extend(get_py_oids(conn))
update_file(fn, lines)
sp.check_call(["black", "-q", fn])
def update_cython_oids(conn: Connection) -> None:
fn = ROOT / "psycopg_c/psycopg_c/_psycopg/oids.pxd"
lines = []
lines.extend(get_version_comment(conn))
lines.extend(get_cython_oids(conn))
update_file(fn, lines)
def update_crdb_python_oids(conn: Connection) -> None:
fn = ROOT / "psycopg/psycopg/crdb/_types.py"
lines = []
lines.extend(get_version_comment(conn))
lines.extend(get_py_types(conn))
update_file(fn, lines)
sp.check_call(["black", "-q", fn])
def get_version_comment(conn: Connection) -> list[str]:
if conn.info.vendor == "PostgreSQL":
version = version_pretty(conn.info.server_version)
elif conn.info.vendor == "CockroachDB":
assert isinstance(conn, CrdbConnection)
version = version_pretty(conn.info.server_version)
else:
raise NotImplementedError(f"unexpected vendor: {conn.info.vendor}")
return ["", f" # Generated from {conn.info.vendor} {version}", ""]
def get_py_oids(conn: Connection) -> list[str]:
lines = []
for typname, oid in conn.execute(
"""
select typname, oid
from pg_type
where
oid < 10000
and (typtype = any('{b,r,m}') or typname = 'record')
and (typname !~ '^(_|pg_)' or typname = 'pg_lsn')
order by typname
"""
):
const_name = typname.upper() + "_OID"
lines.append(f"{const_name} = {oid}")
return lines
typemods = {
"char": "CharTypeModifier",
"bpchar": "CharTypeModifier",
"varchar": "CharTypeModifier",
"numeric": "NumericTypeModifier",
"time": "TimeTypeModifier",
"timetz": "TimeTypeModifier",
"timestamp": "TimeTypeModifier",
"timestamptz": "TimeTypeModifier",
"interval": "TimeTypeModifier",
"bit": "BitTypeModifier",
"varbit": "BitTypeModifier",
}
def get_py_types(conn: Connection) -> list[str]:
# Note: "record" is a pseudotype but still a useful one to have.
# "pg_lsn" is a documented public type and useful in streaming replication
lines = []
for typname, oid, typarray, regtype, typdelim in conn.execute(
"""
select typname, oid, typarray,
-- CRDB might have quotes in the regtype representation
replace(typname::regtype::text, '''', '') as regtype,
typdelim
from pg_type t
where
oid < 10000
and oid != '"char"'::regtype
and (typtype = 'b' or typname = 'record')
and (typname !~ '^(_|pg_)' or typname = 'pg_lsn')
order by typname
"""
):
typemod = typemods.get(typname)
# Weird legacy type in postgres catalog
if typname == "char":
typname = regtype = '"char"'
# https://github.com/cockroachdb/cockroach/issues/81645
if typname == "int4" and conn.info.vendor == "CockroachDB":
regtype = typname
params = [repr(typname), str(oid), str(typarray)]
if regtype != typname:
params.append(f"regtype={regtype!r}")
if typemod:
params.append(f"typemod={typemod}")
if typdelim != ",":
params.append(f"delimiter={typdelim!r}")
lines.append(f"TypeInfo({','.join(params)}),")
return lines
def get_py_ranges(conn: Connection) -> list[str]:
lines = []
for typname, oid, typarray, rngsubtype in conn.execute(
"""
select typname, oid, typarray, rngsubtype
from
pg_type t
join pg_range r on t.oid = rngtypid
where
oid < 10000
and typtype = 'r'
order by typname
"""
):
params = [f"{typname!r}, {oid}, {typarray}, subtype_oid={rngsubtype}"]
lines.append(f"RangeInfo({','.join(params)}),")
return lines
def get_py_multiranges(conn: Connection) -> list[str]:
lines = []
for typname, oid, typarray, rngtypid, rngsubtype in conn.execute(
"""
select typname, oid, typarray, rngtypid, rngsubtype
from
pg_type t
join pg_range r on t.oid = rngmultitypid
where
oid < 10000
and typtype = 'm'
order by typname
"""
):
params = [
f"{typname!r}, {oid}, {typarray},"
f" range_oid={rngtypid}, subtype_oid={rngsubtype}"
]
lines.append(f"MultirangeInfo({','.join(params)}),")
return lines
def get_cython_oids(conn: Connection) -> list[str]:
lines = []
for typname, oid in conn.execute(
"""
select typname, oid
from pg_type
where
oid < 10000
and (typtype = any('{b,r,m}') or typname = 'record')
and (typname !~ '^(_|pg_)' or typname = 'pg_lsn')
order by typname
"""
):
const_name = typname.upper() + "_OID"
lines.append(f" {const_name} = {oid}")
return lines
def update_file(fn: Path, new: list[str]) -> None:
with fn.open("r") as f:
lines = f.read().splitlines()
istart, iend = (
i
for i, line in enumerate(lines)
if re.match(r"\s*#\s*autogenerated:\s+(start|end)", line)
)
lines[istart + 1 : iend] = new + [""]
with fn.open("w") as f:
f.write("\n".join(lines))
f.write("\n")
def parse_cmdline() -> argparse.Namespace:
parser = argparse.ArgumentParser(
description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter
)
parser.add_argument("dsn", help="where to connect to")
opt = parser.parse_args()
return opt
if __name__ == "__main__":
main()
|