ottrec

README

Ottawa Rec Schedules

City of Ottawa drop-in recreation schedule data scraper.

I will be making a website for this soon.

Usage

The following links contain the most recent data, updated daily. See data.ottrec.ca for more information.

Format	Data	Schema	Notes
JSON (simplified)	data.ottrec.ca/export/latest.json	schema.json	Recommended for most uses, easiest to use.
CSV (simplified)	data.ottrec.ca/export/latest.csv.zip	schema.csv	Recommended for tools which require CSV.
Protobuf (raw)	data.ottrec.ca/v1/latest/pb	schema.proto	Best for advanced use cases, least lossy.
TextPB (raw)	data.ottrec.ca/v1/latest/textpb	schema.proto	Best for manual inspection when testing.
JSON (raw)	data.ottrec.ca/v1/latest/json	schema.proto	Not recommended.

About

Simplified data

Features

• Has facility longitude/latitude.
• Has parsed start/end date/time (if it was able to be parsed unambiguously).
• Has normalized activity names for easy filtering.
• Marks if reservation is likely required for an activity, and has a list of links.
• Error messages are included.
• Has fields with raw text from the website for validation or as a fallback if parsing fails.

Limitations

• Special schedules may overlap with a subset of the dates of the regular schedule, so the start/end dates are only good for determining that a scheduled activity does not apply to a specific date.
• Exceptions and notifications are included as raw HTML since they're freeform.

Raw data

Features / Limitations

• Only basic facility information and schedule information are scraped. This helps keep the scraper reliable and ensures the schema can be kept stable long-term.
• Facility addresses are geocoded using geocodio (which has better results than pelias/geocode.earth and nominatim).
• Schedule changes and facility notifications are scraped on a best-effort basis without additional parsing since these fields are inherently free-form. This helps keep the scraper reliable and reduces the likelihood of accidentally missing important information.
• Scraped fields have minimal processing. This helps keep the scraper reliable and reduces the likelihood of accidentally missing important information.
• Optional fields are available which contain best-effort parsing and normalization of scraped fields (to assist with usage), including:
  • Normalized schedule group name.
  • Normalized schedule name (facility and date range stripped).
  • Raw schedule date range (if stripped from the normalized schedule name).
  • Parsed schedule date range.
  • Normalized schedule activity name.
  • Activity time range and weekday as an integer.
  • Explicit reservation requirement in activity names as a boolean (typically, this is used as an exception to the default based on whether the schedule group has reservation links).
• Overlapping schedules (e.g., holiday schedules) are not merged. These schedules are not consistently formatted as they are manually named and created, so although I attempt to parse time ranges, I don't use them to merge schedules. This helps keep the scraper reliable and reduces the likelihood of accidentally missing important information.
• Any potential parsing problems are included in an array of error messages for each facility.
• A protobuf schema is used for maintainability, but it may be changed in backwards-incompatible ways if needed.

Changes

• 2025-10-07: Initial stable release.