Stream: committers
Topic: Don't display INFORMATION/WARNING lines on build output?
Jean Duteau (Jun 25 2019 at 17:10):
Is there a setting on the build to not display WARNING and INFORMATION lines? I'm starting with a freshly checked out build and those lines swamp any errors that I see or create myself.
Lloyd McKenzie (Jun 25 2019 at 17:42):
This is the main build?
Jean Duteau (Jun 25 2019 at 17:46):
Yes.
Lloyd McKenzie (Jun 25 2019 at 18:29):
In the root folder, there's a suppressed-messages.txt file in which you can put warning and information messages that have been reviewed and confirmed to not be a problem so they don't bug you anymore.
Jean Duteau (Jun 25 2019 at 18:31):
so no global "don't show any WARNING or INFORMATION messages" flag?
Jean Duteau (Jun 25 2019 at 18:40):
What is the format of a line in that file? Say I want to suppress "Unable to resolve example reference" information messages. What do I need to put in there?
Lloyd McKenzie (Jun 25 2019 at 18:44):
The full text of the warning to suppress. (Suppression works based on string-match.)
Lloyd McKenzie (Jun 25 2019 at 18:46):
The intention is that info and warning messages that haven't been suppressed should be paid the same attention as error. Every single one of them needs to be dealt with - whether that's by fixing the source or by saying "yeah, that's fine" and putting it into the suppression file. (Note: Don't suppress messages that are caused by tooling issues - or no one will ever get around to fixing them.)
Jean Duteau (Jun 25 2019 at 18:48):
so right now, the main build has a lot (i want to say hundreds) of information messages about "Unable to resolve example reference to blah1 in blah2. for my local build, I'd like to suppress those. Is there some sort of pattern matching so I don't have to list every single one of them?
Lloyd McKenzie (Jun 25 2019 at 18:52):
Nope
Jean Duteau (Jun 25 2019 at 18:52):
ugh. when are people going to fix all of these messages then?
Lloyd McKenzie (Jun 25 2019 at 18:59):
When we yell at them more than we have been, I guess...
Jean Duteau (Jun 25 2019 at 21:16):
btw, here is a list of current warning/information and their counts:
423 RIM Mapping is required
249 Unable to resolve example reference
211 A copyright statement should be present for any value set that includes non-HL7 sourced codes
210 An element of type CodeableConcept or Coding must have a binding
112 Short description doesn't add any new content
66 overlaps with name of parent
65 Element has a todo associated with it
22 Need to provide a binding
16 A resource must have w5 mappings
16 A resource must have an 'entered in error' status
10 All resources should have an identifier
10 A resource that contains a status element must have a search parameter
8 Element needs a definition of its own
6 MnM must have confirmed this should not be an Annotation
6 Element names should be singular
5 Bindings for code data types should only use internally defined codes
4 Ignored Column
4 A resource that contains an identifier must have a search parameter
3 Unable to resolve invalid example reference
3 A resource that contains a url element must have a search parameter 'url'
2 The max cardinality of 'comment' must be 1
2 Operation must have an example
1 Short description can't be the same as the name
1 Resource elements are out of order.
1 A resource that contains a subject that can be a patient reference must have a search parameter 'patient'
Brian Postlethwaite (Jun 25 2019 at 21:25):
And in that terror is the list of how guilty each workgroup is too. :smile:
Jean Duteau (Jun 25 2019 at 21:42):
Well, here is a list that might display guilt slightly better:
72 @ CapabilityStatement2
71 @ CapabilityStatement
61 @ Contract
48 @ MeasureReport
43 @ ObservationDefinition
42 @ fhir
42 @ SubstanceDefinition
38 @ pa
37 @ SpecimenDefinition
35 @ RegulatedMedicinalProduct
30 @ SubstanceSourceMaterial
29 @ ClinicalUseIssue
28 @ oo
28 @ AdverseEvent
26 @ pc
23 @ InsurancePlan
22 @ Device
22 @ Binding
21 @ Ingredient
21 @ BiologicallyDerivedProduct
20 @ ResearchStudy
20 @ EffectEvidenceSynthesis
19 @ cds
19 @ SubstanceReferenceInformation
19 @ SubstancePolymer
18 @ Statistic
18 @ DeviceDefinition
17 @ RiskAssessment
17 @ RegulatedAuthorization
17 @ MedicationKnowledge
17 @ Invoice
17 @ CarePlan
16 @ RiskEvidenceSynthesis
15 @ phx
14 @ ResearchDefinition
14 @ RegulatedAdministrableProduct
14 @ Evidence
14 @ CareTeam
13 @ RegulatedPackagedProduct
13 @ Coverage
12 @ ResearchElementDefinition
12 @ RegulatedManufacturedItem
11 @ pher
11 @ Task
11 @ MolecularSequence
11 @ EvidenceVariable
10 @ OrganizationAffiliation
10 @ Goal
10 @ ActivityDefinition
9 @ SubstanceNucleicAcid
9 @ Observation
9 @ ExampleScenario
9 @ EventDefinition
9 @ Consent
9 @ ChargeItem
8 @ VerificationResult
8 @ ResearchSubject
8 @ Group
8 @ CatalogEntry
7 @ vocab
7 @ RequestGroup
6 @ cqi
6 @ SubstanceProtein
6 @ Measure
6 @ GuidanceResponse
6 @ ConditionDefinition
6 @ ClaimResponse
5 @ ServiceRequest
5 @ Population
5 @ PlanDefinition
5 @ MarketingStatus
5 @ ImmunizationEvaluation
5 @ HealthcareService
5 @ FamilyMemberHistory
5 @ ExplanationOfBenefit
5 @ DeviceUseStatement
5 @ DetectedIssue
5 @ ClinicalImpression
4 @ fm
4 @ brr
4 @ TerminologyCapabilities
4 @ SubstanceAmount
4 @ Linkage
4 @ ImplementationGuide
4 @ DeviceMetric
3 @ sec
3 @ cg
3 @ Specimen
3 @ ProductShelfLife
3 @ MedicationAdministration
3 @ Library
3 @ Flag
3 @ Encounter
3 @ CommunicationRequest
2 @ sd
2 @ VisionPrescription
2 @ SupplyRequest
2 @ StructureMap
2 @ ProdCharacteristic
2 @ PaymentReconciliation
2 @ Money
2 @ MessageDefinition
2 @ Expression
2 @ EnrollmentResponse
2 @ EnrollmentRequest
2 @ DeviceRequest
2 @ DataRequirement
2 @ CoverageEligibilityResponse
2 @ CoverageEligibilityRequest
2 @ Communication
2 @ Claim
2 @ ChargeItemDefinition
1 @ inm
1 @ ii
1 @ cbcc
1 @ UsageContext
1 @ TriggerDefinition
1 @ Timing
1 @ TestReport
1 @ SearchParameter
1 @ RelatedArtifact
1 @ Provenance
1 @ Procedure
1 @ Patient
1 @ ParameterDefinition
1 @ OperationDefinition
1 @ NutritionOrder
1 @ MedicationDispense
1 @ ImmunizationRecommendation
1 @ Endpoint
1 @ Dosage
1 @ Contributor
1 @ ContactDetail
1 @ Composition
1 @ BodyStructure
1 @ Account
Brian Postlethwaite (Jun 25 2019 at 21:47):
Thanks. Looks like 2 of PAs fresh resources are a little more guilty than I remembered. I'll get into that. Thanks for the reminder.
Jean Duteau (Jun 25 2019 at 21:47):
And I'm going to ping Melva to help her fix the Pharmacy resources.
Jean Duteau (Jun 25 2019 at 21:47):
Let's all do our part to stop forestFHIRs! :)
Andy Stechishin (Jun 25 2019 at 21:54):
we had an issue with some of the FM resources where the error/warning could not be fixed or explained by Grahame
Grahame Grieve (Jun 25 2019 at 21:55):
which is not a reason not to fix other problems
Andy Stechishin (Jun 25 2019 at 21:56):
when they are your only two then you are done
Andy Stechishin (Jun 25 2019 at 21:57):
We were shooting for a clean sheet
Grahame Grieve (Jun 25 2019 at 21:58):
I'm not sure what you meant but I hope it doesn't mean that you think you're done...
Andy Stechishin (Jun 25 2019 at 21:59):
are we ever done?
Grahame Grieve (Jun 25 2019 at 22:00):
only in other senses of the word
Michelle (Moseman) Miller (Jun 27 2019 at 22:39):
@Lloyd McKenzie you said:
In the root folder, there's a suppressed-messages.txt file in which you can put warning and information messages that have been reviewed and confirmed to not be a problem so they don't bug you anymore.
Is there a process or exemption process (similar to workflow) where we need to get these approved? For example, per GF#22732, Patient Care doesn't want to add a binding to Goal.target.detail -- presumably for the same reasons that OO doesn't have a binding for Observation.value.
Lloyd McKenzie (Jun 27 2019 at 22:51):
Yes. Send an email to MnM.
Eric Haas (Jun 28 2019 at 23:46):
No binding on observation[x] since is a choice. OO proposed was unclear what the binding applied to. So voted at the time could not bind choice element. Not sure if validation would know either.
Lloyd McKenzie (Jun 29 2019 at 02:08):
You can bind a choice element - the binding would apply to whatever elements are allowed to have bindings - which in the case of Observation would be CodeableConcept and Quantity. Because those are so disparate, it's probably reasonable to not have a binding.
Michelle (Moseman) Miller (Jul 11 2019 at 20:51):
Regarding the QA copyright warnings....It reads as if we need copyright for NON-HL7 code systems, but all of the warnings are popping up on HL7 code systems... I don't recall seeing these previously and am unsure of next steps to resolve.
Grahame Grieve (Jul 11 2019 at 20:55):
I'll have a look sometime. main build issues are low on my priority list at the moment
Michelle (Moseman) Miller (Aug 09 2019 at 15:41):
@Lloyd McKenzie you had said:
In the root folder, there's a suppressed-messages.txt file in which you can put warning and information messages that have been reviewed and confirmed to not be a problem so they don't bug you anymore.
What is the syntax? I tried adding as follows, but it still shows up in http://build.fhir.org/qa.html
INFORMATION: Goal.target.detail[x]: Need to provide a binding
John Moehrke (Aug 09 2019 at 15:48):
it is the text that the build tool throws. so likely close to that but less pretty. likely starts with "warning @..."
John Moehrke (Aug 09 2019 at 16:04):
nope, Im wrong... I am confused as what you did seems right.
Lloyd McKenzie (Aug 09 2019 at 16:10):
You don't need the 'INFORMATION: ' bit.
Lloyd McKenzie (Aug 09 2019 at 16:11):
Here's an example from SDC:
Lloyd McKenzie (Aug 09 2019 at 16:11):
John Moehrke (Aug 09 2019 at 16:18):
all those in FHIR core suppressed-messages.txt have that prefix. I wonder if the editor thought they were suppressing and not really working?
Michelle (Moseman) Miller (Aug 09 2019 at 16:30):
@Lloyd McKenzie ALL entries in https://github.com/HL7/fhir/blob/master/suppressed-messages.txt have the prefix (WARNING: or INFORMATION:). Also, other entries in the suppressed-messages.txt are also showing in http://build.fhir.org/qa.html, such as:
INFORMATION: ElementDefinition.definition: Name of child (definition) overlaps with name of parent (ElementDefinition)
The confusing part was that the suppressed-messages has an extra : in most after the element name, but the QA page doesn't have that :
Michelle (Moseman) Miller (Aug 09 2019 at 16:31):
Does that only suppress from the build and not the QA page?
Lloyd McKenzie (Aug 09 2019 at 16:31):
Sorry, I thought you were asking about IGs, not the core spec
Lloyd McKenzie (Aug 09 2019 at 16:32):
There's a text file that's spit out containing all of the warnings. It works best to grab the content from that - not from the HTML
Michelle (Moseman) Miller (Aug 09 2019 at 16:32):
Remind me where that text file is located? Are you talking about the output text after running the publish.bat? I can try that next....
Lloyd McKenzie (Aug 09 2019 at 16:40):
Looks like I lied. (It's been a long time since I tried suppressing anything in the core spec...)
Michelle (Moseman) Miller (Aug 09 2019 at 19:11):
Thanks Lloyd. The publish.bat output was different from QA.html. I think it should work now with the inclusion of "Binding @"
INFORMATION: Binding @ Goal.target.detail[x]: Need to provide a binding
Last updated: Apr 12 2022 at 19:14 UTC
