我正在使用sqlalchemy-redshift连接到AWS中的现有Redshift数据库。数据库中的大多数表和视图都有几列，因此我想自动加载其架构并获取Table对象。即我不想在python代码中指定表/视图的架构。我的问题是，尽管我能够自动加载表，但无法自动加载数据库中的视图。
SQLAlchemy docs表明它使用名为reflection的进程来加载视图/表的架构。虽然我能够注意到表的反射是成功的，但是反射视图却失败了。这是一个代码段。
In [1]: from sqlalchemy import inspect, create_engine, Table
In [2]: from sqlalchemy.ext.declarative import declarative_base
In [3]: Base=declarative_base()
In [4]: connection_string = 'redshift+psycopg2://user:password@host:5439/dbname'
In [5]: engine=create_engine(connection_string,
connect_args={'sslmode': 'verify-ca'},
echo=True) # using echo=True only for debugging.
In [6]: insp = inspect(engine)
2018-12-10 11:02:56,544 INFO sqlalchemy.engine.base.Engine select version()
2018-12-10 11:02:56,545 INFO sqlalchemy.engine.base.Engine {}
2018-12-10 11:02:56,620 INFO sqlalchemy.engine.base.Engine select current_schema()
2018-12-10 11:02:56,621 INFO sqlalchemy.engine.base.Engine {}
2018-12-10 11:02:56,693 INFO sqlalchemy.engine.base.Engine SELECT CAST('test plain returns' AS VARCHAR(60)) AS anon_1
2018-12-10 11:02:56,694 INFO sqlalchemy.engine.base.Engine {}
2018-12-10 11:02:56,732 INFO sqlalchemy.engine.base.Engine SELECT CAST('test unicode returns' AS VARCHAR(60)) AS anon_1
2018-12-10 11:02:56,732 INFO sqlalchemy.engine.base.Engine {}
我可以成功反映表（并自动加载它们）。例如，
In [7]: insp.get_table_names()
2018-12-10 11:03:10,475 INFO sqlalchemy.engine.base.Engine
SELECT
c.relkind,
n.oid as "schema_oid",
n.nspname as "schema",
c.oid as "rel_oid",
c.relname,
CASE c.reldiststyle
WHEN 0 THEN 'EVEN' WHEN 1 THEN 'KEY' WHEN 8 THEN 'ALL' END
AS "diststyle",
c.relowner AS "owner_id",
u.usename AS "owner_name",
TRIM(TRAILING ';' FROM pg_catalog.pg_get_viewdef(c.oid, true))
AS "view_definition",
pg_catalog.array_to_string(c.relacl, '
') AS "privileges"
FROM pg_catalog.pg_class c
LEFT JOIN pg_catalog.pg_namespace n ON n.oid = c.relnamespace
JOIN pg_catalog.pg_user u ON u.usesysid = c.relowner
WHERE c.relkind IN ('r', 'v', 'm', 'S', 'f')
AND n.nspname !~ '^pg_'
ORDER BY c.relkind, n.oid, n.nspname;
2018-12-10 11:03:10,475 INFO sqlalchemy.engine.base.Engine {}
Out[7]:
['tab_1',
'tab_2',
'tab_3']
In [8]: load_tab=Table('tab_1',Base.metadata)
...: insp.reflecttable(load_tab, None)
2018-12-10 11:15:00,537 INFO sqlalchemy.engine.base.Engine
SELECT t.typname as "name",
pg_catalog.format_type(t.typbasetype, t.typtypmod) as "attype",
not t.typnotnull as "nullable",
t.typdefault as "default",
pg_catalog.pg_type_is_visible(t.oid) as "visible",
n.nspname as "schema"
FROM pg_catalog.pg_type t
LEFT JOIN pg_catalog.pg_namespace n ON n.oid = t.typnamespace
WHERE t.typtype = 'd'
2018-12-10 11:15:00,538 INFO sqlalchemy.engine.base.Engine {}
2018-12-10 11:15:00,669 INFO sqlalchemy.engine.base.Engine
SELECT
n.nspname as "schema",
c.relname as "table_name",
t.contype,
t.conname,
t.conkey,
a.attnum,
a.attname,
pg_catalog.pg_get_constraintdef(t.oid, true) as condef,
n.oid as "schema_oid",
c.oid as "rel_oid"
FROM pg_catalog.pg_class c
LEFT JOIN pg_catalog.pg_namespace n
ON n.oid = c.relnamespace
JOIN pg_catalog.pg_constraint t
ON t.conrelid = c.oid
JOIN pg_catalog.pg_attribute a
ON t.conrelid = a.attrelid AND a.attnum = ANY(t.conkey)
WHERE n.nspname !~ '^pg_'
ORDER BY n.nspname, c.relname
2018-12-10 11:15:00,671 INFO sqlalchemy.engine.base.Engine {}
2018-12-10 11:15:00,882 INFO sqlalchemy.engine.base.Engine
SELECT c.oid
FROM pg_catalog.pg_class c
LEFT JOIN pg_catalog.pg_namespace n ON n.oid = c.relnamespace
WHERE (pg_catalog.pg_table_is_visible(c.oid))
AND c.relname = %(table_name)s AND c.relkind in ('r', 'v', 'm', 'f')
2018-12-10 11:15:00,883 INFO sqlalchemy.engine.base.Engine {'table_name': 'tab_1'}
2018-12-10 11:15:01,003 INFO sqlalchemy.engine.base.Engine
SELECT
cons.conname as name,
cons.consrc as src
FROM
pg_catalog.pg_constraint cons
WHERE
cons.conrelid = %(table_oid)s AND
cons.contype = 'c'
2018-12-10 11:15:01,005 INFO sqlalchemy.engine.base.Engine {'table_oid': 123456}
2018-12-10 11:15:01,114 INFO sqlalchemy.engine.base.Engine
SELECT
pgd.description as table_comment
FROM
pg_catalog.pg_description pgd
WHERE
pgd.objsubid = 0 AND
pgd.objoid = %(table_oid)s
2018-12-10 11:15:01,116 INFO sqlalchemy.engine.base.Engine {'table_oid': 123456}
但是，如果我尝试反映观点，则会得到KeyError。
In [9]: insp.get_view_names()
Out[9]:
['view_0',
'view_1',
'view_2',
'view_3']
In [8]: view_info=Table('view_1', Base.metadata)
...: insp.reflecttable(view_info, None)
2018-12-10 11:08:41,429 INFO sqlalchemy.engine.base.Engine SHOW search_path
2018-12-10 11:08:41,430 INFO sqlalchemy.engine.base.Engine {}
2018-12-10 11:08:41,505 INFO sqlalchemy.engine.base.Engine
SELECT nspname AS "name"
FROM pg_catalog.pg_namespace n
WHERE nspname !~ '^pg_'
AND nspname <> 'information_schema'
AND n.oid NOT IN
(SELECT esoid FROM pg_catalog.pg_external_schema)
ORDER BY 1
2018-12-10 11:08:41,507 INFO sqlalchemy.engine.base.Engine {}
2018-12-10 11:08:41,546 INFO sqlalchemy.engine.base.Engine SET LOCAL search_path TO "public"
2018-12-10 11:08:41,548 INFO sqlalchemy.engine.base.Engine {}
2018-12-10 11:08:41,585 INFO sqlalchemy.engine.base.Engine
SELECT
n.nspname as "schema",
c.relname as "table_name",
d.column as "name",
encoding as "encode",
type, distkey, sortkey, "notnull", adsrc, attnum,
pg_catalog.format_type(att.atttypid, att.atttypmod),
pg_catalog.pg_get_expr(ad.adbin, ad.adrelid) AS DEFAULT,
n.oid as "schema_oid",
c.oid as "table_oid"
FROM pg_catalog.pg_class c
LEFT JOIN pg_catalog.pg_namespace n
ON n.oid = c.relnamespace
JOIN pg_catalog.pg_table_def d
ON (d.schemaname, d.tablename) = (n.nspname, c.relname)
JOIN pg_catalog.pg_attribute att
ON (att.attrelid, att.attname) = (c.oid, d.column)
LEFT JOIN pg_catalog.pg_attrdef ad
ON (att.attrelid, att.attnum) = (ad.adrelid, ad.adnum)
WHERE n.nspname !~ '^pg_'
ORDER BY n.nspname, c.relname, att.attnum
2018-12-10 11:08:41,586 INFO sqlalchemy.engine.base.Engine {}
2018-12-10 11:08:42,169 INFO sqlalchemy.engine.base.Engine SET LOCAL search_path TO "$user", public
2018-12-10 11:08:42,170 INFO sqlalchemy.engine.base.Engine {}
---------------------------------------------------------------------------
KeyError Traceback (most recent call last)
<ipython-input-15-fb159573ea93> in <module>
1 view_info=Table('view_1',
2 Base.metadata)
----> 3 insp.reflecttable(view_info, None)
~/.pyenv/versions/3.7.0/lib/python3.7/site-packages/sqlalchemy/engine/reflection.py in reflecttable(self, table, include_columns, exclude_columns, _extend_on)
600 # reflect table options, like mysql_engine
601 tbl_opts = self.get_table_options(
--> 602 table_name, schema, **table.dialect_kwargs)
603 if tbl_opts:
604 # add additional kwargs to the Table if the dialect
~/.pyenv/versions/3.7.0/lib/python3.7/site-packages/sqlalchemy/engine/reflection.py in get_table_options(self, table_name, schema, **kw)
310 return self.dialect.get_table_options(
311 self.bind, table_name, schema,
--> 312 info_cache=self.info_cache, **kw)
313 return {}
314
<string> in get_table_options(self, connection, table_name, schema, **kw)
~/.pyenv/versions/3.7.0/lib/python3.7/site-packages/sqlalchemy/engine/reflection.py in cache(fn, self, con, *args, **kw)
52 ret = info_cache.get(key)
53 if ret is None:
---> 54 ret = fn(self, con, *args, **kw)
55 info_cache[key] = ret
56 return ret
~/.pyenv/versions/3.7.0/lib/python3.7/site-packages/sqlalchemy_redshift/dialect.py in get_table_options(self, connection, table_name, schema, **kw)
551 schema, **kw)
552 columns = self._get_redshift_columns(connection, table_name,
--> 553 schema, **kw)
554 sortkey_cols = sorted([col for col in columns if col.sortkey],
555 key=keyfunc)
~/.pyenv/versions/3.7.0/lib/python3.7/site-packages/sqlalchemy_redshift/dialect.py in _get_redshift_columns(self, connection, table_name, schema, **kw)
643 if key not in all_columns.keys():
644 key = key.unquoted()
--> 645 return all_columns[key]
646
647 def _get_redshift_constraints(self, connection, table_name,
KeyError: 'public.view_1'
以下是正在使用的各种软件包的版本：
In [28]: import sqlalchemy_redshift; sqlalchemy_redshift.__version__
Out[28]: '0.7.1'
In [29]: import sqlalchemy; sqlalchemy.__version__
Out[29]: '1.2.14'
In [31]: import sys; sys.version
Out[31]: '3.7.0 (default, Oct 17 2018, 11:19:19) \n[Clang 10.0.0 (clang-1000.10.44.2)]'
任何人都可以分享有关如何解决此问题的见解吗？
答案 0 :(得分：0)
这些视图未得到反映，因为它们是使用no schema binding选项创建的（视图无 no schema binding可以正常工作）。有关此主题的更多讨论，请参见：https://github.com/sqlalchemy-redshift/sqlalchemy-redshift/issues/122