2 44.7. Explicit Subtransactions #
4 44.7.1. Subtransaction Context Managers
6 Recovering from errors caused by database access as described in
7 Section 44.6.2 can lead to an undesirable situation where some
8 operations succeed before one of them fails, and after recovering from
9 that error the data is left in an inconsistent state. PL/Python offers
10 a solution to this problem in the form of explicit subtransactions.
12 44.7.1. Subtransaction Context Managers #
14 Consider a function that implements a transfer between two accounts:
15 CREATE FUNCTION transfer_funds() RETURNS void AS $$
17 plpy.execute("UPDATE accounts SET balance = balance - 100 WHERE account_name
19 plpy.execute("UPDATE accounts SET balance = balance + 100 WHERE account_name
21 except plpy.SPIError as e:
22 result = "error transferring funds: %s" % e.args
24 result = "funds transferred correctly"
25 plan = plpy.prepare("INSERT INTO operations (result) VALUES ($1)", ["text"])
26 plpy.execute(plan, [result])
27 $$ LANGUAGE plpython3u;
29 If the second UPDATE statement results in an exception being raised,
30 this function will report the error, but the result of the first UPDATE
31 will nevertheless be committed. In other words, the funds will be
32 withdrawn from Joe's account, but will not be transferred to Mary's
35 To avoid such issues, you can wrap your plpy.execute calls in an
36 explicit subtransaction. The plpy module provides a helper object to
37 manage explicit subtransactions that gets created with the
38 plpy.subtransaction() function. Objects created by this function
39 implement the context manager interface. Using explicit subtransactions
40 we can rewrite our function as:
41 CREATE FUNCTION transfer_funds2() RETURNS void AS $$
43 with plpy.subtransaction():
44 plpy.execute("UPDATE accounts SET balance = balance - 100 WHERE account_
46 plpy.execute("UPDATE accounts SET balance = balance + 100 WHERE account_
48 except plpy.SPIError as e:
49 result = "error transferring funds: %s" % e.args
51 result = "funds transferred correctly"
52 plan = plpy.prepare("INSERT INTO operations (result) VALUES ($1)", ["text"])
53 plpy.execute(plan, [result])
54 $$ LANGUAGE plpython3u;
56 Note that the use of try/except is still required. Otherwise the
57 exception would propagate to the top of the Python stack and would
58 cause the whole function to abort with a PostgreSQL error, so that the
59 operations table would not have any row inserted into it. The
60 subtransaction context manager does not trap errors, it only assures
61 that all database operations executed inside its scope will be
62 atomically committed or rolled back. A rollback of the subtransaction
63 block occurs on any kind of exception exit, not only ones caused by
64 errors originating from database access. A regular Python exception
65 raised inside an explicit subtransaction block would also cause the
66 subtransaction to be rolled back.