type validation

[p-schaefer]

I'm having an issue trying to validate the schema for DataStream found here. I'm not sure if the issue is in the schema, or if its in the package, but it seems to be failing at validating. (I'm also struggling a little with the outputs from the functions, so forgive me if there is a better way of compiling the outputs)

Data setup

dt <- tibble::tribble(
  ~DatasetName, ~MonitoringLocationID, ~MonitoringLocationName, ~MonitoringLocationLatitude, ~MonitoringLocationLongitude, ~MonitoringLocationHorizontalCoordinateReferenceSystem, ~MonitoringLocationHorizontalAccuracyMeasure, ~MonitoringLocationHorizontalAccuracyUnit, ~MonitoringLocationVerticalMeasure, ~MonitoringLocationVerticalUnit, ~MonitoringLocationType,   ~ActivityType, ~ActivityMediaName, ~ActivityStartDate, ~ActivityStartTime, ~ActivityEndDate, ~ActivityEndTime, ~ActivityDepthHeightMeasure, ~ActivityDepthHeightUnit, ~SampleCollectionEquipmentName,  ~CharacteristicName, ~MethodSpeciation, ~ResultSampleFraction, ~ResultValue, ~ResultUnit, ~ResultValueType, ~ResultDetectionCondition, ~ResultDetectionQuantitationLimitMeasure, ~ResultDetectionQuantitationLimitUnit, ~ResultDetectionQuantitationLimitType, ~ResultStatusID, ~ResultComment, ~ResultAnalyticalMethodID, ~ResultAnalyticalMethodContext, ~ResultAnalyticalMethodName, ~AnalysisStartDate, ~AnalysisStartTime, ~AnalysisStartTimeZone, ~LaboratoryName, ~LaboratorySampleID,
  "test",             "test-314",               "test-314",                  43.5895361,                  -79.9411775,                                                "NAD83",                                           NA,                                        NA,                                 NA,                              NA,          "River/Stream", "Field Msr/Obs",    "Surface Water",       "2005-07-12",                 NA,               NA,               NA,                          NA,                       NA,                 "Probe/Sensor", "Temperature, water",                NA,                    NA,           18,     "deg C",         "Actual",                        NA,                                       NA,                                    NA,                                    NA,              NA,             NA,                        NA,                             NA,                          NA,                 NA,                 NA,                     NA,              NA,                  NA
)

Validate using jsonvalidate::json_schema$new()$serialise as an input:

dt_json <- jsonlite::toJSON(as.list(dt),digits = 999)

path <- tempfile()
dl <- download.file("https://datastream.org/schema", path,method="curl", quiet =T)

sc <- suppressWarnings(jsonvalidate::json_schema$new(path,strict = T))

json <- sc$serialise(dt_json)
out <- sc$validate(json,verbose=T)
ab <- attr(out,"error")

cbind(ab$schemaPath,ab$keyword,ab$params,ab$message)

and when I validate using jsonvalidate::json_validate() with serialized input

jsonvalidate::json_validate(json, path,verbose = T, strict = T)

and when I validate without serializing:

out <- sc$validate(dt_json,verbose=T)
ab <- attr(out,"error")

bind_cols(ab$schemaPath,ab$keyword,ab$message)

jsonvalidate::json_validate(dt_json, path,verbose = T, strict = T)

I appreciate any assistance you can provide.

[p-schaefer]

I've made some progress. I think part of the issue might be the use of the allOf in the schema. Isolating one of the elements in the allOf gets me further:

dt <- tibble::tribble(
  ~DatasetName, ~MonitoringLocationID, ~MonitoringLocationName, ~MonitoringLocationLatitude, ~MonitoringLocationLongitude, ~MonitoringLocationHorizontalCoordinateReferenceSystem, ~MonitoringLocationHorizontalAccuracyMeasure, ~MonitoringLocationHorizontalAccuracyUnit, ~MonitoringLocationVerticalMeasure, ~MonitoringLocationVerticalUnit, ~MonitoringLocationType,   ~ActivityType, ~ActivityMediaName, ~ActivityStartDate, ~ActivityStartTime, ~ActivityEndDate, ~ActivityEndTime, ~ActivityDepthHeightMeasure, ~ActivityDepthHeightUnit, ~SampleCollectionEquipmentName,  ~CharacteristicName, ~MethodSpeciation, ~ResultSampleFraction, ~ResultValue, ~ResultUnit, ~ResultValueType, ~ResultDetectionCondition, ~ResultDetectionQuantitationLimitMeasure, ~ResultDetectionQuantitationLimitUnit, ~ResultDetectionQuantitationLimitType, ~ResultStatusID, ~ResultComment, ~ResultAnalyticalMethodID, ~ResultAnalyticalMethodContext, ~ResultAnalyticalMethodName, ~AnalysisStartDate, ~AnalysisStartTime, ~AnalysisStartTimeZone, ~LaboratoryName, ~LaboratorySampleID,
  "test",             "test-314",               "test-314",                  43.5895361,                  -79.9411775,                                                "NAD83",                                           NA,                                        NA,                                 NA,                              NA,          "River/Stream", "Field Msr/Obs",    "Surface Water",       "2005-07-12",                 NA,               NA,               NA,                          NA,                       NA,                 "Probe/Sensor", "Temperature, water",                NA,                    NA,           18,     "deg C",         "Actual",                        NA,                                       NA,                                    NA,                                    NA,              NA,             NA,                        NA,                             NA,                          NA,                 NA,                 NA,                     NA,              NA,                  NA
)

path <- tempfile()
dl <- download.file("https://datastream.org/schema", path,method="auto", quiet =T)
scm <- jsonlite::read_json(path)

sc <- jsonvalidate::json_schema$new(
      jsonlite::toJSON(scm$allOf[[1]],auto_unbox = T,digits = 999),
      strict = F)

json <- sc$serialise(as.list(dt[1,1:5]))
out <- sc$validate(json,verbose=T,greedy =T)
ab <- attr(out,"error")

cbind(ab$schemaPath,ab$keyword,ab$message)

it now gets me a pattern error, but I'm not sure I understand why its failing, since the below returns TRUE

stringr::str_detect("test","^[\\p{L}\\p{N}\\p{P}\\p{S} ]+$")

Testing with a simpler schema also doesn't detect the string correctly:

jsonvalidate::json_validate(
    '{"test": "string"}',
    '{
    "$schema": "http://json-schema.org/draft-04/schema#",
    "title": "Test",
    "type": "object",
    "properties": {
      "test": {
        "type": "string",
        "pattern": "^[\\p{L}\\p{N}\\p{P}\\p{S} ]+$"
      }
    }
  }',
    engine = "ajv"
)

# >[1] FALSE