mysteriendrama/lib/python3.11/site-packages/django/db/backends/mysql/introspection.py

350 lines
14 KiB
Python
Raw Normal View History

2023-07-22 12:04:19 +02:00
from collections import namedtuple
import sqlparse
from MySQLdb.constants import FIELD_TYPE
from django.db.backends.base.introspection import BaseDatabaseIntrospection
from django.db.backends.base.introspection import FieldInfo as BaseFieldInfo
from django.db.backends.base.introspection import TableInfo as BaseTableInfo
from django.db.models import Index
from django.utils.datastructures import OrderedSet
FieldInfo = namedtuple(
"FieldInfo",
BaseFieldInfo._fields + ("extra", "is_unsigned", "has_json_constraint", "comment"),
)
InfoLine = namedtuple(
"InfoLine",
"col_name data_type max_len num_prec num_scale extra column_default "
"collation is_unsigned comment",
)
TableInfo = namedtuple("TableInfo", BaseTableInfo._fields + ("comment",))
class DatabaseIntrospection(BaseDatabaseIntrospection):
data_types_reverse = {
FIELD_TYPE.BLOB: "TextField",
FIELD_TYPE.CHAR: "CharField",
FIELD_TYPE.DECIMAL: "DecimalField",
FIELD_TYPE.NEWDECIMAL: "DecimalField",
FIELD_TYPE.DATE: "DateField",
FIELD_TYPE.DATETIME: "DateTimeField",
FIELD_TYPE.DOUBLE: "FloatField",
FIELD_TYPE.FLOAT: "FloatField",
FIELD_TYPE.INT24: "IntegerField",
FIELD_TYPE.JSON: "JSONField",
FIELD_TYPE.LONG: "IntegerField",
FIELD_TYPE.LONGLONG: "BigIntegerField",
FIELD_TYPE.SHORT: "SmallIntegerField",
FIELD_TYPE.STRING: "CharField",
FIELD_TYPE.TIME: "TimeField",
FIELD_TYPE.TIMESTAMP: "DateTimeField",
FIELD_TYPE.TINY: "IntegerField",
FIELD_TYPE.TINY_BLOB: "TextField",
FIELD_TYPE.MEDIUM_BLOB: "TextField",
FIELD_TYPE.LONG_BLOB: "TextField",
FIELD_TYPE.VAR_STRING: "CharField",
}
def get_field_type(self, data_type, description):
field_type = super().get_field_type(data_type, description)
if "auto_increment" in description.extra:
if field_type == "IntegerField":
return "AutoField"
elif field_type == "BigIntegerField":
return "BigAutoField"
elif field_type == "SmallIntegerField":
return "SmallAutoField"
if description.is_unsigned:
if field_type == "BigIntegerField":
return "PositiveBigIntegerField"
elif field_type == "IntegerField":
return "PositiveIntegerField"
elif field_type == "SmallIntegerField":
return "PositiveSmallIntegerField"
# JSON data type is an alias for LONGTEXT in MariaDB, use check
# constraints clauses to introspect JSONField.
if description.has_json_constraint:
return "JSONField"
return field_type
def get_table_list(self, cursor):
"""Return a list of table and view names in the current database."""
cursor.execute(
"""
SELECT
table_name,
table_type,
table_comment
FROM information_schema.tables
WHERE table_schema = DATABASE()
"""
)
return [
TableInfo(row[0], {"BASE TABLE": "t", "VIEW": "v"}.get(row[1]), row[2])
for row in cursor.fetchall()
]
def get_table_description(self, cursor, table_name):
"""
Return a description of the table with the DB-API cursor.description
interface."
"""
json_constraints = {}
if (
self.connection.mysql_is_mariadb
and self.connection.features.can_introspect_json_field
):
# JSON data type is an alias for LONGTEXT in MariaDB, select
# JSON_VALID() constraints to introspect JSONField.
cursor.execute(
"""
SELECT c.constraint_name AS column_name
FROM information_schema.check_constraints AS c
WHERE
c.table_name = %s AND
LOWER(c.check_clause) =
'json_valid(`' + LOWER(c.constraint_name) + '`)' AND
c.constraint_schema = DATABASE()
""",
[table_name],
)
json_constraints = {row[0] for row in cursor.fetchall()}
# A default collation for the given table.
cursor.execute(
"""
SELECT table_collation
FROM information_schema.tables
WHERE table_schema = DATABASE()
AND table_name = %s
""",
[table_name],
)
row = cursor.fetchone()
default_column_collation = row[0] if row else ""
# information_schema database gives more accurate results for some figures:
# - varchar length returned by cursor.description is an internal length,
# not visible length (#5725)
# - precision and scale (for decimal fields) (#5014)
# - auto_increment is not available in cursor.description
cursor.execute(
"""
SELECT
column_name, data_type, character_maximum_length,
numeric_precision, numeric_scale, extra, column_default,
CASE
WHEN collation_name = %s THEN NULL
ELSE collation_name
END AS collation_name,
CASE
WHEN column_type LIKE '%% unsigned' THEN 1
ELSE 0
END AS is_unsigned,
column_comment
FROM information_schema.columns
WHERE table_name = %s AND table_schema = DATABASE()
""",
[default_column_collation, table_name],
)
field_info = {line[0]: InfoLine(*line) for line in cursor.fetchall()}
cursor.execute(
"SELECT * FROM %s LIMIT 1" % self.connection.ops.quote_name(table_name)
)
def to_int(i):
return int(i) if i is not None else i
fields = []
for line in cursor.description:
info = field_info[line[0]]
fields.append(
FieldInfo(
*line[:2],
to_int(info.max_len) or line[2],
to_int(info.max_len) or line[3],
to_int(info.num_prec) or line[4],
to_int(info.num_scale) or line[5],
line[6],
info.column_default,
info.collation,
info.extra,
info.is_unsigned,
line[0] in json_constraints,
info.comment,
)
)
return fields
def get_sequences(self, cursor, table_name, table_fields=()):
for field_info in self.get_table_description(cursor, table_name):
if "auto_increment" in field_info.extra:
# MySQL allows only one auto-increment column per table.
return [{"table": table_name, "column": field_info.name}]
return []
def get_relations(self, cursor, table_name):
"""
Return a dictionary of {field_name: (field_name_other_table, other_table)}
representing all foreign keys in the given table.
"""
cursor.execute(
"""
SELECT column_name, referenced_column_name, referenced_table_name
FROM information_schema.key_column_usage
WHERE table_name = %s
AND table_schema = DATABASE()
AND referenced_table_name IS NOT NULL
AND referenced_column_name IS NOT NULL
""",
[table_name],
)
return {
field_name: (other_field, other_table)
for field_name, other_field, other_table in cursor.fetchall()
}
def get_storage_engine(self, cursor, table_name):
"""
Retrieve the storage engine for a given table. Return the default
storage engine if the table doesn't exist.
"""
cursor.execute(
"""
SELECT engine
FROM information_schema.tables
WHERE
table_name = %s AND
table_schema = DATABASE()
""",
[table_name],
)
result = cursor.fetchone()
if not result:
return self.connection.features._mysql_storage_engine
return result[0]
def _parse_constraint_columns(self, check_clause, columns):
check_columns = OrderedSet()
statement = sqlparse.parse(check_clause)[0]
tokens = (token for token in statement.flatten() if not token.is_whitespace)
for token in tokens:
if (
token.ttype == sqlparse.tokens.Name
and self.connection.ops.quote_name(token.value) == token.value
and token.value[1:-1] in columns
):
check_columns.add(token.value[1:-1])
return check_columns
def get_constraints(self, cursor, table_name):
"""
Retrieve any constraints or keys (unique, pk, fk, check, index) across
one or more columns.
"""
constraints = {}
# Get the actual constraint names and columns
name_query = """
SELECT kc.`constraint_name`, kc.`column_name`,
kc.`referenced_table_name`, kc.`referenced_column_name`,
c.`constraint_type`
FROM
information_schema.key_column_usage AS kc,
information_schema.table_constraints AS c
WHERE
kc.table_schema = DATABASE() AND
c.table_schema = kc.table_schema AND
c.constraint_name = kc.constraint_name AND
c.constraint_type != 'CHECK' AND
kc.table_name = %s
ORDER BY kc.`ordinal_position`
"""
cursor.execute(name_query, [table_name])
for constraint, column, ref_table, ref_column, kind in cursor.fetchall():
if constraint not in constraints:
constraints[constraint] = {
"columns": OrderedSet(),
"primary_key": kind == "PRIMARY KEY",
"unique": kind in {"PRIMARY KEY", "UNIQUE"},
"index": False,
"check": False,
"foreign_key": (ref_table, ref_column) if ref_column else None,
}
if self.connection.features.supports_index_column_ordering:
constraints[constraint]["orders"] = []
constraints[constraint]["columns"].add(column)
# Add check constraints.
if self.connection.features.can_introspect_check_constraints:
unnamed_constraints_index = 0
columns = {
info.name for info in self.get_table_description(cursor, table_name)
}
if self.connection.mysql_is_mariadb:
type_query = """
SELECT c.constraint_name, c.check_clause
FROM information_schema.check_constraints AS c
WHERE
c.constraint_schema = DATABASE() AND
c.table_name = %s
"""
else:
type_query = """
SELECT cc.constraint_name, cc.check_clause
FROM
information_schema.check_constraints AS cc,
information_schema.table_constraints AS tc
WHERE
cc.constraint_schema = DATABASE() AND
tc.table_schema = cc.constraint_schema AND
cc.constraint_name = tc.constraint_name AND
tc.constraint_type = 'CHECK' AND
tc.table_name = %s
"""
cursor.execute(type_query, [table_name])
for constraint, check_clause in cursor.fetchall():
constraint_columns = self._parse_constraint_columns(
check_clause, columns
)
# Ensure uniqueness of unnamed constraints. Unnamed unique
# and check columns constraints have the same name as
# a column.
if set(constraint_columns) == {constraint}:
unnamed_constraints_index += 1
constraint = "__unnamed_constraint_%s__" % unnamed_constraints_index
constraints[constraint] = {
"columns": constraint_columns,
"primary_key": False,
"unique": False,
"index": False,
"check": True,
"foreign_key": None,
}
# Now add in the indexes
cursor.execute(
"SHOW INDEX FROM %s" % self.connection.ops.quote_name(table_name)
)
for table, non_unique, index, colseq, column, order, type_ in [
x[:6] + (x[10],) for x in cursor.fetchall()
]:
if index not in constraints:
constraints[index] = {
"columns": OrderedSet(),
"primary_key": False,
"unique": not non_unique,
"check": False,
"foreign_key": None,
}
if self.connection.features.supports_index_column_ordering:
constraints[index]["orders"] = []
constraints[index]["index"] = True
constraints[index]["type"] = (
Index.suffix if type_ == "BTREE" else type_.lower()
)
constraints[index]["columns"].add(column)
if self.connection.features.supports_index_column_ordering:
constraints[index]["orders"].append("DESC" if order == "D" else "ASC")
# Convert the sorted sets to lists
for constraint in constraints.values():
constraint["columns"] = list(constraint["columns"])
return constraints