This repository has been archived by the owner on Feb 8, 2018. It is now read-only.

backfill remaining exchanges#3912

Closed

chadwhitacre wants to merge 63 commits intomasterfrom backfill-remaining-exchanges

+849

Commits on Mar 1, 2017

Here's a script for backfilling routes & exchanges
chadwhitacre
committed
Scope ref uniqueness to network
chadwhitacre
committed
Drop unique constraint for now
chadwhitacre
committed
Tweak CSV format
chadwhitacre
committed
Add a matchup script
chadwhitacre
committed
Log network with backfill
chadwhitacre
committed
Update Stripe match script for payments.csv
chadwhitacre
committed
Clean up a couple things in match-stripe.py
chadwhitacre
committed
Emit rows in original order
chadwhitacre
committed
Ignore the test transaction
chadwhitacre
committed
Add some debugging aids
chadwhitacre
committed
Improve match script so we don't exception
chadwhitacre
committed
Stripe's timestamp is only minute granularity
chadwhitacre
committed
Smarten up fuzz guessing
chadwhitacre
committed
Unbuffer output for easier tailing
chadwhitacre
committed
Fix bug in skipping and use immutable ID as able
chadwhitacre
committed
Tweak match criteria based on experience
chadwhitacre
committed
Tighten up linking based on Customer ID
chadwhitacre
committed
Blip logging change
chadwhitacre
committed
Broaden time horizon more
chadwhitacre
committed
Filter out Balanced exchanges
chadwhitacre
committed
Map Stripe status to our status
chadwhitacre
committed
Remember matches across months
chadwhitacre
committed
Output failed exchanges as well as successful ones
chadwhitacre
committed
Enable inserting new exchanges (not just updating)
chadwhitacre
committed
Start a match-balanced.py script
chadwhitacre
committed
Resolve ambiguity w/ Balanced w/ closest timestamp
chadwhitacre
committed
Broaden time range to pick up anomalies
chadwhitacre
committed
Simplify time range check
chadwhitacre
committed
Keep going even if we can't find a match
chadwhitacre
committed
Satisfy pyflakes (fix Travis)
chadwhitacre
committed
Better handle mismatch cases
chadwhitacre
committed
Bring over the tally script from #3807
chadwhitacre
committed
Wrap db access into a class
chadwhitacre
committed
Refactor hail mary for a single db call
chadwhitacre
committed
Start optimizing find; there's a bug!
chadwhitacre
committed
Squish a bug
chadwhitacre
committed
Deprecate hail_mary
chadwhitacre
committed
Start implementing fuzz using in-mem struct
chadwhitacre
committed
Fix a bug and try to fix another
chadwhitacre
committed
Add an arg to tally-backfill.py: filename
chadwhitacre
committed
Enable running *through* a month
chadwhitacre
committed
Fix a bug ... uncover another?
chadwhitacre
committed
Because matches are exchanges, not participants
chadwhitacre
committed
Prune dead code
chadwhitacre
committed
Fix bug in CLI constraint arg
chadwhitacre
committed
Start moving structs onto Matcher
chadwhitacre
committed
Factor out a little function
chadwhitacre
committed
Prune dead code
chadwhitacre
committed
Start simplifying algorithm, to build back up
chadwhitacre
committed
Prune more dead code
chadwhitacre
committed
Start inverting loops on Balanced script
chadwhitacre
committed
Flesh out looping infrastructure
chadwhitacre
committed
We have output!
chadwhitacre
committed
Seems like our new script is kind of working?
chadwhitacre
committed
Show more accurate time remaining
chadwhitacre
committed
Fix off-by-one bug at the end of the script
chadwhitacre
committed
Kill the rest of the old code
chadwhitacre
committed
Rename uncategorized to unmatchable
chadwhitacre
committed
Differentiate error cases
chadwhitacre
committed
Add an option to ctrl-c without dumping
chadwhitacre
committed
Match a bunch of failures
chadwhitacre
committed
Start looking at the exchange side
chadwhitacre
committed