Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Fix script that processes feeds #310

Merged
merged 10 commits into from
Nov 7, 2023
Merged

Fix script that processes feeds #310

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
10 commits
Select commit Hold shift + click to select a range
e20bd4a
fix: switch to column E for schedule, H for RT
fredericsimard Oct 24, 2023
0437c3b
Fix for data type for real time
fredericsimard Oct 25, 2023
8a6f7f4
fix: missing realtime type when column G is empty
fredericsimard Oct 25, 2023
e22d231
fix: added logic when provider cell empty
fredericsimard Oct 25, 2023
3be05f3
fix: Used constant instead of plain string
fredericsimard Oct 25, 2023
a2e46b0
fix: `license_url` between quotes
fredericsimard Oct 28, 2023
1feb2fe
fix: if license_url is invalid, provide context
fredericsimard Oct 28, 2023
c65b290
Fix: add pip upgrade to remove warning in log
fredericsimard Nov 7, 2023
34919e1
fix: changes to correct errors in Python script
fredericsimard Nov 7, 2023
d22ba4d
Merge branch 'main' into fix-script-process-csv
fredericsimard Nov 7, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions .github/workflows/add_new_or_updated_feeds.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -87,6 +87,7 @@ jobs:
run: |
python -m venv env
source env/bin/activate
pip install --upgrade pip
pip install virtualenv --quiet
pip install gtfs_kit --quiet
pip install unidecode --quiet
Expand Down
162 changes: 113 additions & 49 deletions scripts/process_csv_in_github_action.swift
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,39 +4,40 @@ import Foundation
#endif

enum column : Int, CaseIterable {
case timestamp = 0
case provider = 1
case regioncity = 2
case currenturl = 3
case updatednewsourceurl = 4
case datatype1 = 5
case request = 6
case downloadurl = 7
case country = 8
case subdivision_name = 9
case municipality = 10
case name = 11
case yournameorg = 12
case license_url = 13
case tripupdatesurl = 14
case servicealertsurl = 15
case genunknownrturl = 16
case authentication_type = 17
case authentication_info_url = 18
case api_key_parameter_name = 19
case note = 20
case gtfsschedulefeatures = 21
case gtfsschedulestatus = 22
case gtfsrealtimestatus = 23
case youremail = 24
case dataproduceremail = 25
case realtimefeatures = 26
case isocountrycode = 27
case feedupdatestatus = 28
case timestamp = 0 // A
case provider = 1 // B
case regioncity = 2 // C
case currenturl = 3 // D
case updatednewsourceurl = 4 // E
case datatype = 5 // F
case request = 6 // G
case downloadurl = 7 // H
case country = 8 // I
case subdivision_name = 9 // J
case municipality = 10 // K
case name = 11 // L
case yournameorg = 12 // M
case license_url = 13 // N
case tripupdatesurl = 14 // O
case servicealertsurl = 15 // P
case genunknownrturl = 16 // Q
case authentication_type = 17 // R
case authentication_info_url = 18 // S
case api_key_parameter_name = 19 // T
case note = 20 // U
case gtfsschedulefeatures = 21 // W
case gtfsschedulestatus = 22 // Y
case gtfsrealtimestatus = 23 // Z
case youremail = 24 // AA
case dataproduceremail = 25 // AB
case realtimefeatures = 26 // AC
case isocountrycode = 27 // AB
case feedupdatestatus = 28 // AC
}

enum defaults: String {
case date = "01/01/1970"
case toBeProvided = "TO_BE_PROVIDED"
}

enum requestType: String {
Expand All @@ -50,6 +51,20 @@ enum dataType: String {
case realtime = "Realtime"
}

enum realtimeDataType: String {
case vehiclePositions = "Vehicle Positions"
case tripUpdates = "Trip Updates"
case serviceAlerts = "Service Alerts"
case unknown = "general / unknown"
}

enum realtimeDataTypeCode: String {
case vehiclePositions = "vp"
case tripUpdates = "tu"
case serviceAlerts = "sa"
case unknown = "gu"
}

let arguments : [String] = CommandLine.arguments

if CommandLine.argc == 5 {
Expand Down Expand Up @@ -79,6 +94,7 @@ if CommandLine.argc == 5 {
}

var PYTHON_SCRIPT_OUTPUT : String = ""
var lastKnownProvider : String = defaults.toBeProvided.rawValue
let dateFormatAsRegex : Regex<AnyRegexOutput> = try Regex(dateFormatGREPArg)

for line : [String] in csvArray {
Expand All @@ -89,14 +105,15 @@ if CommandLine.argc == 5 {

let timestamp : String = line[column.timestamp.rawValue].trimmingCharacters(in: .whitespacesAndNewlines)
let provider : String = line[column.provider.rawValue]
let datatype1 : String = line[column.datatype1.rawValue]
let datatype : String = line[column.datatype.rawValue]
let request : String = line[column.request.rawValue]
let country : String = line[column.country.rawValue]
let subdivision_name : String = line[column.subdivision_name.rawValue]
let municipality : String = line[column.municipality.rawValue]
let name : String = line[column.name.rawValue]
let license_url : String = line[column.license_url.rawValue]
var license_url : String = line[column.license_url.rawValue]
let downloadURL : String = line[column.downloadurl.rawValue]
let updatednewsourceurl : String = line[column.updatednewsourceurl.rawValue]
let authentication_type : String = line[column.authentication_type.rawValue]
let authentication_info_url : String = line[column.authentication_info_url.rawValue]
let api_key_parameter_name : String = line[column.api_key_parameter_name.rawValue]
Expand All @@ -106,50 +123,76 @@ if CommandLine.argc == 5 {
let gtfsrealtimestatus : String = line[column.gtfsrealtimestatus.rawValue]
let realtimefeatures : String = line[column.realtimefeatures.rawValue]

// Check if provider is empty, suggest last known if true.
if provider.count > 0 { lastKnownProvider = provider }
let finalProvider : String = provider.isEmpty ? "\(defaults.toBeProvided.rawValue) (\(lastKnownProvider) ?)" : provider

// Check if license URL is valid
let urlPresent : Bool = isURLPresent(in: license_url)
if ( urlPresent == false && license_url.count > 0 ) { license_url = "INVALID_OR_NO_URL_PROVIDED" }

let dateFromCurrentLine : String = extractDate(from: timestamp, usingGREP: dateFormatAsRegex, desiredDateFormat: dateFormatDesiredArg)

if dateFromCurrentLine == dateToFind { // ...the row has been added on the date we're looking for, process it.

if request.contains(requestType.isAddNewFeed.rawValue) { // add new feed

if datatype1.contains(dataType.schedule.rawValue) { // add_gtfs_schedule_source

PYTHON_SCRIPT_ARGS_TEMP = "add_gtfs_schedule_source(provider=\(provider), country_code=\(country), direct_download_url=\(downloadURL), authentication_type=\(authentication_type), authentication_info_url=\(authentication_info_url), api_key_parameter_name=\(api_key_parameter_name), subdivision_name=\(subdivision_name), municipality=\(municipality), license_url=\(license_url), name=\(name), status=\(gtfsschedulestatus), features=\(gtfsschedulefeatures))"
if datatype.contains(dataType.schedule.rawValue) { // add_gtfs_schedule_source
let authType : Int = authenticationType(for: authentication_type)
PYTHON_SCRIPT_ARGS_TEMP = "add_gtfs_schedule_source(provider=\"\(finalProvider)\", country_code=\"\(country)\", direct_download_url=\"\(updatednewsourceurl.isEmpty ? downloadURL : updatednewsourceurl)\", authentication_type=\(authType), authentication_info_url=\"\(authentication_info_url)\", api_key_parameter_name=\"\(api_key_parameter_name)\", subdivision_name=\"\(subdivision_name)\", municipality=\"\(municipality)\", license_url=\"\(license_url)\", name=\"\(name)\", status=\"\(gtfsschedulestatus)\", features=\"\(gtfsschedulefeatures)\")"

} else if datatype1.contains(dataType.realtime.rawValue) { // add_gtfs_realtime_source
} else if datatype.contains(dataType.realtime.rawValue) { // add_gtfs_realtime_source
// Emma: entity_type matches the realtime Data type options of Vehicle Positions, Trip Updates, or Service Alerts. If one of those three are selected, add it. If not, omit it.

PYTHON_SCRIPT_ARGS_TEMP = "add_gtfs_realtime_source(entity_type=\(datatype1), provider=\(provider), direct_download_url=\(downloadURL), authentication_type=\(authentication_type), authentication_info_url=\(authentication_info_url), api_key_parameter_name=\(api_key_parameter_name), license_url=\(license_url), name=\(name), static_reference=\"TO_BE_PROVIDED\", note=\(note), status=\(gtfsrealtimestatus), features=\(realtimefeatures))"
let authType : Int = authenticationType(for: authentication_type)
let realtimecode : String = realtimeCode(for:datatype)
PYTHON_SCRIPT_ARGS_TEMP = "add_gtfs_realtime_source(entity_type=\"\(realtimecode)\", provider=\"\(finalProvider)\", direct_download_url=\"\(downloadURL.isEmpty ? updatednewsourceurl : downloadURL)\", authentication_type=\(authType), authentication_info_url=\"\(authentication_info_url)\", api_key_parameter_name=\"\(api_key_parameter_name)\", license_url=\"\(license_url)\", name=\"\(name)\", note=\"\(note)\", status=\"\(gtfsrealtimestatus)\", features=\"\(realtimefeatures)\")"

}

} else if request.contains(requestType.isUpdateExistingFeed.rawValue) { // update existing feed

if datatype1.contains(dataType.schedule.rawValue) { // update_gtfs_schedule_source
if datatype.contains(dataType.schedule.rawValue) { // update_gtfs_schedule_source

PYTHON_SCRIPT_ARGS_TEMP = "update_gtfs_schedule_source(mdb_source_id=\"\", provider=\(provider), name=\(name), country_code=\(country), subdivision_name=\(subdivision_name), municipality=\(municipality), direct_download_url=\(downloadURL), authentication_type=\(authentication_type), authentication_info_url=\(authentication_info_url), api_key_parameter_name=\(api_key_parameter_name), license_url=\(license_url), status=\(gtfsschedulestatus), features=\(gtfsschedulefeatures))"
let authType : Int = authenticationType(for: authentication_type)
PYTHON_SCRIPT_ARGS_TEMP = "update_gtfs_schedule_source(mdb_source_id=\"\", provider=\"\(finalProvider)\", name=\"\(name)\", country_code=\"\(country)\", subdivision_name=\"\(subdivision_name)\", municipality=\"\(municipality)\", direct_download_url=\"\(updatednewsourceurl.isEmpty ? downloadURL : updatednewsourceurl)\", authentication_type=\(authType), authentication_info_url=\"\(authentication_info_url)\", api_key_parameter_name=\"\(api_key_parameter_name)\", license_url=\"\(license_url)\", status=\"\(gtfsschedulestatus)\", features=\"\(gtfsschedulefeatures)\")"

} else if datatype1.contains(dataType.realtime.rawValue) { // update_gtfs_realtime_source

PYTHON_SCRIPT_ARGS_TEMP = "update_gtfs_realtime_source(mdb_source_id=\"\", entity_type=\(datatype1), provider=\(provider), direct_download_url=\(downloadURL), authentication_type=\(authentication_type), authentication_info_url=\(authentication_info_url), api_key_parameter_name=\(api_key_parameter_name), license_url=\(license_url), name=\(name), static_reference=\"TO_BE_PROVIDED\", note=\(note), status=\(gtfsrealtimestatus), features=\(realtimefeatures))"
} else if datatype.contains(dataType.realtime.rawValue) { // update_gtfs_realtime_source

let authType : Int = authenticationType(for: authentication_type)
let realtimecode : String = realtimeCode(for:datatype)
PYTHON_SCRIPT_ARGS_TEMP = "update_gtfs_realtime_source(mdb_source_id=\"\", entity_type=\"\(realtimecode)\", provider=\"\(finalProvider)\", direct_download_url=\"\(downloadURL.isEmpty ? updatednewsourceurl : downloadURL)\", authentication_type=\(authType), authentication_info_url=\"\(authentication_info_url)\", api_key_parameter_name=\"\(api_key_parameter_name)\", license_url=\"\(license_url)\", name=\"\(name)\", note=\"\(note)\", status=\"\(gtfsrealtimestatus)\", features=\"\(realtimefeatures)\")"
}

} else if request.contains(requestType.isToRemoveFeed.rawValue) { // remove feed

if datatype1.contains(dataType.schedule.rawValue) { // update_gtfs_schedule_source

PYTHON_SCRIPT_ARGS_TEMP = "update_gtfs_schedule_source(mdb_source_id=\"\", provider=\(provider), name=\"**** Requested for removal ****\", country_code=\(country), subdivision_name=\(subdivision_name), municipality=\(municipality), direct_download_url=\(downloadURL), authentication_type=\(authentication_type), authentication_info_url=\(authentication_info_url), api_key_parameter_name=\(api_key_parameter_name), license_url=\(license_url), status=\(gtfsschedulestatus), features=\(gtfsschedulefeatures))"
if datatype.contains(dataType.schedule.rawValue) { // update_gtfs_schedule_source

} else if datatype1.contains(dataType.realtime.rawValue) { // update_gtfs_realtime_source
let authType : Int = authenticationType(for: authentication_type)
PYTHON_SCRIPT_ARGS_TEMP = "update_gtfs_schedule_source(mdb_source_id=\"\", provider=\"\(finalProvider)\", name=\"**** Requested for removal ****\", country_code=\"\(country)\", subdivision_name=\"\(subdivision_name)\", municipality=\"\(municipality)\", direct_download_url=\"\(updatednewsourceurl.isEmpty ? downloadURL : updatednewsourceurl)\", authentication_type=\(authType), authentication_info_url=\"\(authentication_info_url)\", api_key_parameter_name=\"\(api_key_parameter_name)\", license_url=\"\(license_url)\", status=\"\(gtfsschedulestatus)\", features=\"\(gtfsschedulefeatures)\")"

PYTHON_SCRIPT_ARGS_TEMP = "update_gtfs_realtime_source(mdb_source_id=\"\", entity_type=\(datatype1), provider=\(provider), direct_download_url=\(downloadURL), authentication_type=\(authentication_type), authentication_info_url=\(authentication_info_url), api_key_parameter_name=\(api_key_parameter_name), license_url=\(license_url), name=\"**** Requested for removal ****\", static_reference=\"TO_BE_PROVIDED\", note=\(note), status=\(gtfsrealtimestatus), features=\(realtimefeatures))"
} else if datatype.contains(dataType.realtime.rawValue) { // update_gtfs_realtime_source

let authType : Int = authenticationType(for: authentication_type)
let realtimecode : String = realtimeCode(for:datatype)
PYTHON_SCRIPT_ARGS_TEMP = "update_gtfs_realtime_source(mdb_source_id=\"\", entity_type=\"[\(realtimecode)]\", provider=\"\(finalProvider)\", direct_download_url=\"\(downloadURL.isEmpty ? updatednewsourceurl : downloadURL)\", authentication_type=\(authType), authentication_info_url=\"\(authentication_info_url)\", api_key_parameter_name=\"\(api_key_parameter_name)\", license_url=\"\(license_url)\", name=\"**** Requested for removal ****\", note=\"\(note)\", status=\"\(gtfsrealtimestatus)\", features=\"\(realtimefeatures)\")"

}

} else { // ... assume this is a new feed by default :: add_gtfs_schedule_source

PYTHON_SCRIPT_ARGS_TEMP = "add_gtfs_schedule_source(provider=\(provider), country_code=\(country), direct_download_url=\(downloadURL), authentication_type=\(authentication_type), authentication_info_url=\(authentication_info_url), api_key_parameter_name=\(api_key_parameter_name), subdivision_name=\(subdivision_name), municipality=\(municipality), license_url=\(license_url), name=\(name), status=\(gtfsschedulestatus), features=\(gtfsschedulefeatures))"


if datatype.contains(dataType.schedule.rawValue) { // update_gtfs_schedule_source

let authType : Int = authenticationType(for: authentication_type)
PYTHON_SCRIPT_ARGS_TEMP = "add_gtfs_schedule_source(provider=\"\(finalProvider)\", country_code=\"\(country)\", direct_download_url=\"\(updatednewsourceurl.isEmpty ? downloadURL : updatednewsourceurl)\", authentication_type=\(authType), authentication_info_url=\"\(authentication_info_url)\", api_key_parameter_name=\"\(api_key_parameter_name)\", subdivision_name=\"\(subdivision_name)\", municipality=\"\(municipality)\", license_url=\"\(license_url)\", name=\"\(name)\", status=\"\(gtfsschedulestatus)\", features=\"\(gtfsschedulefeatures)\")"

} else if datatype.contains(dataType.realtime.rawValue) { // update_gtfs_realtime_source

let authType : Int = authenticationType(for: authentication_type)
let realtimecode : String = realtimeCode(for:datatype)
PYTHON_SCRIPT_ARGS_TEMP = "add_gtfs_realtime_source(entity_type=\"\(realtimecode)\", provider=\"\(finalProvider)\", direct_download_url=\"\(downloadURL.isEmpty ? updatednewsourceurl : downloadURL)\", authentication_type=\(authType), authentication_info_url=\"\(authentication_info_url)\", api_key_parameter_name=\"\(api_key_parameter_name)\", license_url=\"\(license_url)\", name=\"\(name)\", note=\"\(note)\", status=\"\(gtfsrealtimestatus)\", features=\"\(realtimefeatures)\")"

}
}

}
Expand Down Expand Up @@ -188,4 +231,25 @@ func extractDate(from theDateToConvert: String, usingGREP dateFormatAsGREP: Rege

// return default date
return defaults.date.rawValue
}

func authenticationType(for authString: String) -> Int {
if authString.contains("0") { return 0 }
if authString.contains("1") { return 1 }
if authString.contains("2") { return 2 }
return 0
}

func realtimeCode(for theDataType: String) -> String {
if theDataType.contains(realtimeDataType.vehiclePositions.rawValue) { return realtimeDataTypeCode.vehiclePositions.rawValue }
if theDataType.contains(realtimeDataType.tripUpdates.rawValue) { return realtimeDataTypeCode.tripUpdates.rawValue }
if theDataType.contains(realtimeDataType.serviceAlerts.rawValue) { return realtimeDataTypeCode.serviceAlerts.rawValue }
return realtimeDataTypeCode.tripUpdates.rawValue
}

func isURLPresent(in string: String) -> Bool {
let pattern : String = #"https?:\/\/(www\.)?[-a-zA-Z0-9@:%._\+~#=]{1,256}\.[a-zA-Z0-9()]{1,6}\b([-a-zA-Z0-9()@:%_\+.~#?&//=]*)"#
let range = string.range(of: pattern, options: .regularExpression)
if range != nil { return true }
return false
}
Loading