ledger-jira-sync

ledger-jira-sync is a tool to synchronize your ledger work logs with JIRA work logs.

Abstract

This tool is useful for people or companies that decided on using ledger (or something compatible with the ledger journal format, like beancount or hledger) to keep track of their work hours, and who want to synchronize this data with JIRA work logs. This program will assume that the ledger journals are the “ground truth”, meaning that nothing will be synchronized from JIRA to ledger. It’s written in Python 2.

Description

The program will assume you have a ledger-compatible journal file, containing accounts that have work logged. For example, say that you’re working for “customer” on a project “moonshot” and you have two issues to work on “planning” and “implementation”, on which you already spent some time. Your ledger file might look like this:

2018-01-12 customer  ; initial planning
	moonshot:usage:planning  1.5h  ; :Sylvester:
	moonshot:quota

2018-01-12 customer  ; initial planning
	moonshot:usage:planning  1.5h  ; :Elmer:
	moonshot:quota

2018-03-12 customer  ; initial implementation
	moonshot:usage:implementation  30m  ; :Elmer:
	moonshot:quota

As you can see, we use a tag to signify who did the work, and then use ordinary units like ’m’ and ’h’. Now, suppose – unsurprisingly – that the customer doesn’t use ledger, but JIRA, and you’re supposed to log your work using the “work log” feature.

As you can imagine, having two databases which you have to manually synchronize is a very bad idea. That’s where ledger-jira-sync comes in! It takes your ledger work hours and transfers them to JIRA work hours. All you have to write in addition to your ledger entries is a file that maps your ledger accounts to JIRA tickets, as such:

moonshot:usage:planning=MYPROJ-128
moonshot:usage:implementation=c:MYPROJ-3

Note the c: after the equals symbol on the second line. While synchronizing, the entry’s comment will be added to the JIRA work log for the issue MYPROJ-3.

Installation, prerequisites

You need:

Python >= 2.7
ledger compiled with Python support

Installation is easy, just do a python setup.py install --user. If you’re using Nix, you can build it via nix-build and then run it via result/bin/ledger-jira-sync (or install it into your profile via nix-env).

If you’re experiencing strange errors calling the program (see below for instructions), you might not have Python support enabled in ledger.

Usage

The program takes the following parameters:

Parameter name	Description
`--server`	JIRA server URL
`--email`	JIRA login e-mail
`--password`	JIRA password in plain text (see below)
`--ledgerfile`	ledger journal file to scan
`--ledgertag`	ledger tag corresponding to your account
`--assocfile`	Association file between ledger and JIRA

The parameters for server, mail, password, ledgerfile should be clear. Note that if you’re afraid of plain text passwords (as we are), consider using pass or gopass and then just inject the password via a subshell, as such:

ledger python ledger-jira-sync -- --password $(pass show customer.atlassian.net)

The ledgertag is the tag that basically corresponds to your name. It’s just used to filter ledger entries, not for the JIRA side of things.

The assocfile is a file that maps your ledger accounts to JIRA tickets. The format is really simple; it’s lines containing: ledgeraccount=jira-ticket. You can specify =c: instead of a plain = if you want to transfer the entry comment to JIRA, too. Note that the JIRA ticket ID is enough, you don’t have to specify the full URL (you do have to specify the complete ledger account, though).

To call the programm, you have to have ledger compiled with Python support. Most binary distributions should have this enabled. If you’re getting weird errors, tough, you might reconsider. The safest way to call it is by using ledger python, so ledger can call the program for you:

ledger python ledger_jira_sync/__main__.py -- \
 --server customer.atlassian.net \
 --email foo@test.com \
 --password yourpassword \
 --ledgerfile foo.ledger \
 --ledgertag person \
 --assocfile foo.assoc

How it works

The program will do the following:

scan ledgerfile for entries tagged with ledgertag
filter out those ledger entries that are associated (via assocfile) with JIRA tickets
log in to JIRA
load the work logs for the JIRA tickets found
compare the ledger entries with the JIRA tickets, compare differences
ask the user if it should proceed
delete those JIRA work logs that don’t appear as ledger entries
add those JIRA work logs that don’t appear in JIRA