This post is also available in: Français
The 16th Chouette Sprint has been delivered! With 50 tasks completed and 24 bugs fixed. Even if we have been busy with many performance and user interface fixes, we have taken a close interest in GTFS import, by offering a much better management of unorthodox files: by testing files extracted from OpenData platforms, we have found that many public files were quite poorly written, mostly not validated by Google’s own tools. By optimizing Chouette, we can now import with much better performance poorly optimized GTFS files.
An authorized user of a Working Group can initiate the aggregation of the merged data provided by each Workspace in the group.
A specific validation can be set up to ensure that the aggregated data complies with the desired rules.
A lot of performance efforts have been made on the GTFS import, notably by testing many very large and poorly optimized files.
Maintaining the import despite an invalid object
From now on, when importing a GTFS file, if an object is not valid, it will be presented as an error, and the import will no longer be blocked.
Filtering of non-commercial judgments
The GTFS export ignores non-commercial stops (depots, borders, etc.) because they do not affect the passenger.
NeTEx IDFM Import
Error message on complete Routing Constraint Zones
The NeTEx IDFM import now presents an error message for Routing Constraint Zones spanning over all stop places of a route (complete Routing Constraint Zone). Functionally, a complete Routing Constraint Zone is not necessary because it means that a VehicleJourney does not go to any stop.
A failed import must generate an incorrect dataset
Optimization of merge operations
Optimizations for the merge steps were performed on the writing phases, using a transaction system.
The launch of a merge operation should only generate one Referential
Attached file indicator
The submission of a file as an attachment to a stop (via a custom field) is now confirmed by an indicator to see if the download of the document went well.
Data set management
Cleaning after restricting the perimeter of a data set
When a user removes one or more lines from the perimeter of a dataset, Chouette is now able to trigger automatic data cleaning.
This cleaning will remove routes, VehicleJourney and JourneyPatterns related to lines removed from the perimeter of a dataset.
This mechanism existed on the duplication of a dataset, it is now extended to the simple editing of the definition of a dataset.
Search by line name
When creating and editing a Data Set, the search by line name is more efficient and consistent.
Journey pattern management
The withdrawal of a stop place in a journey pattern must result in the withdrawal of the passage schedule in the associated Vehicle Journeys.
A user must select at least 2 stops to create a journey pattern
Passing times management
On a passing time with a service time, the modification of a start time of a Vehicle Journey must automatically generate the arrival time with the same time lag
The “Opposing passing times” button should not be displayed if no return route is defined
For vehicle journeys with a loop (twice the same stop), the selection of a stop in a journey pattern must not activate the same stop
Automatic calculation of mileage/time
Optimization of API calls
When calculating mileages/time is required in massive operations (such as imports), Chouette now optimizes the calculation by concentrating them at the end of the operation.
Global view of strokes
Taking into account calendar colors
The color assigned to a sales calendar is used in the Vehicle Journey overview when it is unfolded.
The display of the transit times of a trip in the consolidated offer must be identical to that defined in the timetable edition
Assistance in the use of API
Chouette now displays the workspace ID when viewing API keys. This information is necessary for the construction of the API request, for example for NeTEx IDFM or GTFS files imports.
Security update from nokogiri >= 1.8.5
Some unused parts of the Chouette code have been removed.
Cleaning of test organization
A task to clean up Cucumber test accounts has been implemented, which allows organizations and their data created by automatic scenarios to be automatically deleted.
Switch to winter time (in progress)
An extensive quality control has been launched to verify the non-impact of winter time on Chouette.
Some graphic defects were noted and corrected, including truncated titles, and missing logos on two pages.