当我最近搜索使用 Django 2.0 和我正在使用的 Oracle 11g 遗留数据库整理 inspectdb 时出现了这个问题，我已经开始修复内省，到目前为止我已经设法得到修改/lib/python3.6/site-packages/django/db/backends/oracle/introspection.py后基本表的输出：
（本质上将所有 user_* 表替换为 all_* 表）
我目前的工作解决方案是下面的文件内容 (introspection.py)。
import warnings
from collections import namedtuple
import cx_Oracle
from django.db import models
from django.db.backends.base.introspection import (
BaseDatabaseIntrospection, FieldInfo as BaseFieldInfo, TableInfo,
)
from django.utils.deprecation import RemovedInDjango21Warning
FieldInfo = namedtuple('FieldInfo', BaseFieldInfo._fields + ('is_autofield',))
class DatabaseIntrospection(BaseDatabaseIntrospection):
# Maps type objects to Django Field types.
data_types_reverse = {
cx_Oracle.BLOB: 'BinaryField',
cx_Oracle.CLOB: 'TextField',
cx_Oracle.DATETIME: 'DateField',
cx_Oracle.FIXED_CHAR: 'CharField',
cx_Oracle.FIXED_NCHAR: 'CharField',
cx_Oracle.NATIVE_FLOAT: 'FloatField',
cx_Oracle.NCHAR: 'CharField',
cx_Oracle.NCLOB: 'TextField',
cx_Oracle.NUMBER: 'DecimalField',
cx_Oracle.STRING: 'CharField',
cx_Oracle.TIMESTAMP: 'DateTimeField',
}
cache_bust_counter = 1
def get_field_type(self, data_type, description):
if data_type == cx_Oracle.NUMBER:
precision, scale = description[4:6]
if scale == 0:
if precision > 11:
return 'BigAutoField' if description.is_autofield else 'BigIntegerField'
elif precision == 1:
return 'BooleanField'
elif description.is_autofield:
return 'AutoField'
else:
return 'IntegerField'
elif scale == -127:
return 'FloatField'
return super().get_field_type(data_type, description)
def get_table_list(self, cursor):
"""Return a list of table and view names in the current database."""
# cursor.execute("SELECT TABLE_NAME, 't' FROM USER_TABLES UNION ALL "
# "SELECT VIEW_NAME, 'v' FROM USER_VIEWS")
cursor.execute("SELECT TABLE_NAME, 't' FROM ALL_TABLES WHERE OWNER = 'V500' ")
return [TableInfo(row[0].lower(), row[1]) 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.
"""
cursor.execute("""
SELECT
column_name,
data_default,
CASE
WHEN char_used IS NULL THEN data_length
ELSE char_length
END as internal_size,
0 as is_autofield
FROM ALL_TAB_COLUMNS
WHERE table_name = UPPER(%s)""", [table_name])
field_map = {
column: (internal_size, default if default != 'NULL' else None, is_autofield)
for column, default, internal_size, is_autofield in cursor.fetchall()
}
self.cache_bust_counter += 1
cursor.execute("SELECT * FROM {} WHERE ROWNUM < 2 AND {} > 0".format(
self.connection.ops.quote_name(table_name),
self.cache_bust_counter))
description = []
for desc in cursor.description:
name = desc[0]
internal_size, default, is_autofield = field_map[name]
name = name % {} # cx_Oracle, for some reason, doubles percent signs.
description.append(FieldInfo(*(
(name.lower(),) +
desc[1:3] +
(internal_size, desc[4] or 0, desc[5] or 0) +
desc[6:] +
(default, is_autofield)
)))
return description
def table_name_converter(self, name):
"""Table name comparison is case insensitive under Oracle."""
return name.lower()
def get_sequences(self, cursor, table_name, table_fields=()):
# Tables don't exist in 11g (this function added in django 2
# cursor.execute("""
# SELECT
# user_tab_identity_cols.sequence_name,
# user_tab_identity_cols.column_name
# FROM
# user_tab_identity_cols,
# user_constraints,
# user_cons_columns cols
# WHERE
# user_constraints.constraint_name = cols.constraint_name
# AND user_constraints.table_name = user_tab_identity_cols.table_name
# AND cols.column_name = user_tab_identity_cols.column_name
# AND user_constraints.constraint_type = 'P'
# AND user_tab_identity_cols.table_name = UPPER(%s)
# """, [table_name])
# # Oracle allows only one identity column per table.
# row = cursor.fetchone()
# if row:
# return [{'name': row[0].lower(), 'table': table_name, 'column': row[1].lower()}]
# # To keep backward compatibility for AutoFields that aren't Oracle
# # identity columns.
# for f in table_fields:
# if isinstance(f, models.AutoField):
# return [{'table': table_name, 'column': f.column}]
return []
def get_relations(self, cursor, table_name):
"""
Return a dictionary of {field_name: (field_name_other_table, other_table)}
representing all relationships to the given table.
"""
table_name = table_name.upper()
cursor.execute("""
SELECT ca.column_name, cb.table_name, cb.column_name
FROM ALL_CONSTRAINTS, ALL_CONS_COLUMNS ca, ALL_CONS_COLUMNS cb
WHERE ALL_CONSTRAINTS.table_name = %s AND
ALL_CONSTRAINTS.constraint_name = ca.constraint_name AND
ALL_CONSTRAINTS.r_constraint_name = cb.constraint_name AND
ca.position = cb.position""", [table_name])
relations = {}
for row in cursor.fetchall():
relations[row[0].lower()] = (row[2].lower(), row[1].lower())
return relations
def get_key_columns(self, cursor, table_name):
cursor.execute("""
SELECT ccol.column_name, rcol.table_name AS referenced_table, rcol.column_name AS referenced_column
FROM ALL_CONSTRAINTS c
JOIN ALL_CONS_COLUMNS ccol
ON ccol.constraint_name = c.constraint_name
JOIN ALL_CONS_COLUMNS rcol
ON rcol.constraint_name = c.r_constraint_name
WHERE c.table_name = %s AND c.constraint_type = 'R'""", [table_name.upper()])
return [tuple(cell.lower() for cell in row)
for row in cursor.fetchall()]
def get_indexes(self, cursor, table_name):
warnings.warn(
"get_indexes() is deprecated in favor of get_constraints().",
RemovedInDjango21Warning, stacklevel=2
)
sql = """
SELECT LOWER(uic1.column_name) AS column_name,
CASE ALL_CONSTRAINTS.constraint_type
WHEN 'P' THEN 1 ELSE 0
END AS is_primary_key,
CASE ALL_INDEXES.uniqueness
WHEN 'UNIQUE' THEN 1 ELSE 0
END AS is_unique
FROM ALL_CONSTRAINTS, ALL_INDEXES, ALL_IND_COLUMNS uic1
WHERE ALL_CONSTRAINTS.constraint_type (+) = 'P'
AND ALL_CONSTRAINTS.index_name (+) = uic1.index_name
AND ALL_INDEXES.uniqueness (+) = 'UNIQUE'
AND ALL_INDEXES.index_name (+) = uic1.index_name
AND uic1.table_name = UPPER(%s)
AND uic1.column_position = 1
AND NOT EXISTS (
SELECT 1
FROM ALL_IND_COLUMNS uic2
WHERE uic2.index_name = uic1.index_name
AND uic2.column_position = 2
)
"""
cursor.execute(sql, [table_name])
indexes = {}
for row in cursor.fetchall():
indexes[row[0]] = {'primary_key': bool(row[1]),
'unique': bool(row[2])}
return indexes
def get_constraints(self, cursor, table_name):
"""
Retrieve any constraints or keys (unique, pk, fk, check, index) across
one or more columns.
"""
constraints = {}
# Loop over the constraints, getting PKs, uniques, and checks
cursor.execute("""
SELECT
ALL_CONSTRAINTS.constraint_name,
LISTAGG(LOWER(cols.column_name), ',') WITHIN GROUP (ORDER BY cols.position),
CASE ALL_CONSTRAINTS.constraint_type
WHEN 'P' THEN 1
ELSE 0
END AS is_primary_key,
CASE
WHEN ALL_CONSTRAINTS.constraint_type IN ('P', 'U') THEN 1
ELSE 0
END AS is_unique,
CASE ALL_CONSTRAINTS.constraint_type
WHEN 'C' THEN 1
ELSE 0
END AS is_check_constraint
FROM
ALL_CONSTRAINTS
LEFT OUTER JOIN
ALL_CONS_COLUMNS cols ON ALL_CONSTRAINTS.constraint_name = cols.constraint_name
WHERE
ALL_CONSTRAINTS.constraint_type = ANY('P', 'U', 'C')
AND ALL_CONSTRAINTS.table_name = UPPER(%s)
GROUP BY ALL_CONSTRAINTS.constraint_name, ALL_CONSTRAINTS.constraint_type
""", [table_name])
for constraint, columns, pk, unique, check in cursor.fetchall():
constraints[constraint] = {
'columns': columns.split(','),
'primary_key': pk,
'unique': unique,
'foreign_key': None,
'check': check,
'index': unique, # All uniques come with an index
}
# Foreign key constraints
cursor.execute("""
SELECT
cons.constraint_name,
LISTAGG(LOWER(cols.column_name), ',') WITHIN GROUP (ORDER BY cols.position),
LOWER(rcols.table_name),
LOWER(rcols.column_name)
FROM
ALL_CONSTRAINTS cons
INNER JOIN
ALL_CONS_COLUMNS rcols ON rcols.constraint_name = cons.r_constraint_name AND rcols.position = 1
LEFT OUTER JOIN
ALL_CONS_COLUMNS cols ON cons.constraint_name = cols.constraint_name
WHERE
cons.constraint_type = 'R' AND
cons.table_name = UPPER(%s)
GROUP BY cons.constraint_name, rcols.table_name, rcols.column_name
""", [table_name])
for constraint, columns, other_table, other_column in cursor.fetchall():
constraints[constraint] = {
'primary_key': False,
'unique': False,
'foreign_key': (other_table, other_column),
'check': False,
'index': False,
'columns': columns.split(','),
}
# Now get indexes
cursor.execute("""
SELECT
ind.index_name,
LOWER(ind.index_type),
LISTAGG(LOWER(cols.column_name), ',') WITHIN GROUP (ORDER BY cols.column_position),
LISTAGG(cols.descend, ',') WITHIN GROUP (ORDER BY cols.column_position)
FROM
ALL_IND_COLUMNS cols, ALL_INDEXES ind
WHERE
cols.table_name = UPPER(%s) AND
NOT EXISTS (
SELECT 1
FROM ALL_CONSTRAINTS cons
WHERE ind.index_name = cons.index_name
) AND cols.index_name = ind.index_name
GROUP BY ind.index_name, ind.index_type
""", [table_name])
for constraint, type_, columns, orders in cursor.fetchall():
constraints[constraint] = {
'primary_key': False,
'unique': False,
'foreign_key': None,
'check': False,
'index': True,
'type': 'idx' if type_ == 'normal' else type_,
'columns': columns.split(','),
'orders': orders.split(','),
}
return constraints
注意到 Django 1.11 是最后一个正式支持 11g 的版本，这是我从 How to make Django 2.0 to use Oracle 11g syntax instead of 12c? 那里得到的，这促使我查看 1.11 代码库以了解发生了什么变化 (https://github.com/django/django/blob/stable/1.11.x/django/db/backends/oracle/introspection.py)
目前我正在从只读数据库连接读取数据，因此修复迁移不是我目前的动力。在编写inspectdb 函数时，将6000 多个表作为python 代码输出，没有任何重大问题。