This is an automated email from the ASF dual-hosted git repository.
janhoy pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/solr-orbit-workloads.git
The following commit(s) were added to refs/heads/main by this push:
new a4684f1 nyc_taxis: fix the two value-source defects that query
randomization makes reachable (#20)
a4684f1 is described below
commit a4684f18c05fe2b2103aa2f39c5c87604cdc2345
Author: Serhiy Bzhezytskyy <[email protected]>
AuthorDate: Fri Sep 4 10:25:32 2026 +0300
nyc_taxis: fix the two value-source defects that query randomization makes
reachable (#20)
* nyc_taxis: compute the money bounds in whole cents
* nyc_taxis: let a whole-day date range cover its last day
Co-Authored-By: Claude Opus 5 <[email protected]>
---
nyc_taxis/workload.py | 12 ++++++++----
1 file changed, 8 insertions(+), 4 deletions(-)
diff --git a/nyc_taxis/workload.py b/nyc_taxis/workload.py
index e58c04a..c0d793e 100644
--- a/nyc_taxis/workload.py
+++ b/nyc_taxis/workload.py
@@ -30,14 +30,15 @@ import datetime
# Common helper functions
def random_money_values(max_value):
- gte_cents = random.randrange(0, max_value*100)
- lte_cents = random.randrange(gte_cents, max_value*100)
+ max_cents = round(max_value * 100)
+ gte_cents = random.randrange(0, max_cents)
+ lte_cents = random.randrange(gte_cents, max_cents)
return {
"gte":gte_cents/100,
"lte":lte_cents/100
}
-def random_dates(min_value, max_value, format_string):
+def random_dates(min_value, max_value, format_string, whole_days=False):
# arguments are datetime objects
min_timestamp = datetime.datetime.timestamp(min_value)
max_timestamp = datetime.datetime.timestamp(max_value)
@@ -47,6 +48,9 @@ def random_dates(min_value, max_value, format_string):
gte_date = datetime.datetime.fromtimestamp(min_timestamp +
int(gte_fraction * diff))
lte_date = datetime.datetime.fromtimestamp(min_timestamp +
int(lte_fraction * diff))
+ if whole_days:
+ gte_date = gte_date.replace(hour=0, minute=0, second=0, microsecond=0)
+ lte_date = lte_date.replace(hour=0, minute=0, second=0, microsecond=0)
+ datetime.timedelta(days=1)
return {
"gte":gte_date.strftime(format_string),
"lte":lte_date.strftime(format_string)
@@ -63,7 +67,7 @@ def date_source_with_hours():
return random_dates(start_date, end_date,
format_string="%Y-%m-%dT%H:%M:%SZ")
def date_source_without_hours():
- return random_dates(start_date, end_date,
format_string="%Y-%m-%dT00:00:00Z")
+ return random_dates(start_date, end_date,
format_string="%Y-%m-%dT%H:%M:%SZ", whole_days=True)
def trip_distance_source():
gte = random.randint(0, 10)
