Stream: implementers
Topic: Forge 16.5 for STU3
Michel Rutten (Aug 18 2017 at 15:13):
Forge 16.5 for STU3 is now available for download from https://simplifier.net/forge.
This new release fixes some issues reported by the community and also introduces a couple of usability improvements. All users are encouraged to update to the current version.
FHIR API Improvements:
- Update the FHIR .NET API library to 0.93.5-beta6
- Stability improvements for DirectorySource
- Handle bug in SimpleQuantity core datatype definition
Features:
- NEW! Automatically initialize default slicing discriminator for type slices
- NEW! FHIR Path expression validation
Bug fixes:
- EXPERIMENTAL! Improved matching of base element types
- Serialize constraints on meta.security.* and meta.tag.* elements to differential in proper order
- eld-16 sliceName validation
- Explicitly remove old child constraints after updating element type
- Import resources from FHIR server and/or Simplifier
- Fix broken hyperlink to Furore news page on Simplifier
Detailed release notes: https://simplifier.net/ui/OrganizationNews/Show?id=34
If you discover any issues or want to submit a feature request, please contact us.
Happy profiling!
Michel Rutten
Chris Munro (Aug 23 2017 at 11:08):
Forge 16.5 for STU3 is now available for download from https://simplifier.net/forge.
This new release fixes some issues reported by the community and also introduces a couple of usability improvements. All users are encouraged to update to the current version.
FHIR API Improvements:
- Update the FHIR .NET API library to 0.93.5-beta6
- Stability improvements for DirectorySource
- Handle bug in SimpleQuantity core datatype definition
Features:
- NEW! Automatically initialize default slicing discriminator for type slices
- NEW! FHIR Path expression validation
Bug fixes:
- EXPERIMENTAL! Improved matching of base element types
- Serialize constraints on meta.security.* and meta.tag.* elements to differential in proper order
- eld-16 sliceName validation
- Explicitly remove old child constraints after updating element type
- Import resources from FHIR server and/or Simplifier
- Fix broken hyperlink to Furore news page on Simplifier
Detailed release notes: https://simplifier.net/ui/OrganizationNews/Show?id=34
If you discover any issues or want to submit a feature request, please contact us.
Happy profiling!
Michel Rutten
I've just created an extension in forge and run it through the validator and I get an warning
Overall result: SUCCESS
[WARNING] Evaluation of FhirPath for constraint 'eld-16' failed: Invocation of 'binary.or' failed: Invocation of 'matches' failed: parsing "^[a-zA-Z0-9\-\_]+$" - Unrecognized escape sequence \_. (at StructureDefinition.differential[0].element[1])
Validation run took 53 miliseconds
<differential>
<element id="Extension.url">
<path value="Extension.url"/>
<fixedUri value="http://www.eldbdev.org/StructureDefinition/MyExtension"/>
</element>
<element id="Extension.value[x]:valueString">
<path value="Extension.valueString"/>
<sliceName value="valueString"/>
<type>
<code value="string"/>
</type>
</element>
</differential>
Is this an error in the underlying validator rather than the extension? An incorrectly copied regular expression?
Michel Rutten (Aug 23 2017 at 11:30):
Hi @Chris Munro, apparently the current regex on element id rejects the "[" and "]" brackets.
Per convention, the standard element id for polymorphic elements is always generated from the original element name, e.g. "value[x]". We agreed on this format during the Madrid WGM.
So I guess the element id regex should be updated to also accept "[]" characters.
Maybe you can submit an issue to GForge?
Chris Munro (Aug 23 2017 at 12:48):
Thanks @Michel Rutten for the reply- I know that the warning is safe to ignore now.
Michel Rutten (Aug 23 2017 at 12:52):
Did you already submit a GForge issue, or shall I take care of it?
Chris Munro (Aug 23 2017 at 12:54):
Did you already submit a GForge issue, or shall I take care of it?
No, I haven't used GForge before so if you could do that it would be great, thanks
Michel Rutten (Aug 23 2017 at 13:21):
Done: #13769
https://gforge.hl7.org/gf/project/fhir/tracker/?action=TrackerItemEdit&tracker_item_id=13769
Chris Munro (Aug 23 2017 at 13:32):
Thanks @Michel Rutten
I've noticed that even with <sliceName value="Test"/>
I still get a warning of
[WARNING] Evaluation of FhirPath for constraint 'eld-16' failed: Invocation of 'binary.or' failed: Invocation of 'matches' failed: parsing "^[a-zA-Z0-9\-\_]+$" - Unrecognized escape sequence \_. (at StructureDefinition.differential[0].element[6])
This looks to be because in the spec there are escape characters (https://www.hl7.org/fhir/elementdefinition-definitions.html) :
eld-16: sliceName must be composed of proper tokens separated by "/" (expression : sliceName.empty() or sliceName.matches('^[a-zA-Z0-9\\/\\-\\_]+$'), xpath: not(exists(f:sliceName/@value)) or matches(f:sliceName/@value, '^[a-zA-Z0-9\\/\\-\\_]+$'))
and the way java/c# process the escaped character in the regular expression?
Michel Rutten (Aug 23 2017 at 13:46):
Hi @Chris Munro, this has been fixed in release 16.5. I recommend you update to the latest version.
Originally, I literally copied the regex from the spec to C# code. But as you discovered, character escaping rules are different. The current release should fix that.
Michel Rutten (Aug 23 2017 at 14:09):
BTW Are you referring to the .NET Validator or the Java validator?
Chris Munro (Aug 23 2017 at 14:10):
Yes it's fine in forge, but in simplifier there is still an error showing once you load in the profile from forge and the same for the C# API validator I'm not sure about the java one though
Michel Rutten (Aug 23 2017 at 14:25):
OK thanks for the feedback. We'll have to fix this in the API.
Michel Rutten (Aug 23 2017 at 14:28):
https://github.com/ewoutkramer/fhir-net-api/issues/413
Ewout Kramer (Aug 23 2017 at 14:34):
Funny this shows up here and now - it means our unit-test that checks all fhirpath statements must have missed this.
Ewout Kramer (Aug 23 2017 at 15:00):
It seems that on its way from the FHIRPath expression in the StructureDefinition for ElementDefinition to the FHIRPath validator, the double backslashes get unescaped. @Chris Munro, did the message on your screen really have \_ in the pattern or had it \\_ ? Just to make sure Zulip did not mess things up ;-)
Josh Mandel (Aug 23 2017 at 15:01):
(BTW in Zulip you can click a message and "view source" to see how it was initially entered.)
Chris Munro (Aug 23 2017 at 15:01):
Josh Mandel (Aug 23 2017 at 15:03):
And from the FHIR spec:
Ewout Kramer (Aug 24 2017 at 11:02):
Ok, so I think I found the culprit. I was assuming my FHIRPath parser complained about the unrecognized escape sequence, but in fact it is the .NET regex evaluator that is called in the end. It will not accept \_ as an escape sequence, and from what I find (superficially) in Google, it's right not to accept it. \/ and \- are fine, but \_ is not....
Ewout Kramer (Aug 24 2017 at 11:03):
For PCRE, and most other so-called Perl-compatible flavors, escape these outside character classes:
.^$*+?()[{\|
and these inside character classes:
^-]\
For POSIX extended regexes (ERE), escape these outside character classes (same as PCRE):
.^$*+?()[{\|
Escaping any other characters is an error with POSIX ERE.
Ewout Kramer (Aug 24 2017 at 11:03):
(note that last sentence, it's from https://stackoverflow.com/questions/399078/what-special-characters-must-be-escaped-in-regular-expressions)
Ewout Kramer (Aug 24 2017 at 11:17):
Created a tracker item: https://gforge.hl7.org/gf/project/fhir/tracker/?action=TrackerItemEdit&tracker_item_id=13771
Last updated: Apr 12 2022 at 19:14 UTC
