heads up: 1.32 has a broken import command

[Simon Michael]

Please be aware of https://github.com/simonmichael/hledger/issues/2125 . Sorry!

In 1.32, `hledger import FILE1 FILE2 ...` seems to write all files' new transaction dates to all of the `.latest*` files, corrupting them and preventing or confusing subsequent imports. 

This is a fairly serious regression that could be tedious to recover from if you import from many/active CSV files (or other import source). Please avoid importing from multiple files at once with hledger-1.32 for now. 

Importing from one file at a time is safe, I believe (but won't interleave transactions by date). 

# bad: hledger import a.csv b.csv

# good: hledger import a.csv; hledger import b.csv

[Simon Michael]

A bugfix release - 1.32 users should upgrade to this to fix the import command.

hledger 1.32.1

• Fixed: import with multiple files now updates .latest files correctly. (#2125)

• Fixed: print --round=hard now properly pads/rounds amounts with inferred costs. (#2123)

• CSV matcher syntax: mention that ! and & can't be used in the same line yet. (#2088)

• Drop the "a difference of ..." line from balance assertion failure output. I feel it made the message harder to read and isn't really necessary.

• Declaring the empty payee name with payee "" now works, to let hledger check payees accept payee-less transactions. (#2119)

• Built-in tags with special meaning like type: and t: are now implicitly declared, so using type: in account declarations or generating t: with timedot letters won't cause hledger check tags to fail. (#2119)

--
You received this message because you are subscribed to the Google Groups "hledger" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hledger+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hledger/939C1435-0FEA-40AA-BA64-71511468592F%40joyful.com.