feat: adds odata metadata to entity type #74

snebjorn · 2020-03-29T19:52:34Z

fixes #60

lib/csdl2openapi.js

tools/V4-CSDL-to-OpenAPI.xsl

fixes oasis-tcs#60

examples/miscellaneous.openapi3.json

snebjorn · 2020-04-01T21:09:08Z

I have no more outstanding issues/questions

snebjorn · 2020-04-09T13:24:06Z

@ralfhandl can we move forward with this?

ralfhandl

Please adapt

lib/csdl2openapi.js

tools/V4-CSDL-to-OpenAPI.xsl

also removed @odata.id as it's typically not in the response. Made @odata.type only show up when type had a basetype

they were accidentally committed

tools/V4-CSDL-to-OpenAPI.xsl

ralfhandl · 2020-05-01T06:08:36Z

Good that the @id is gone, and the @type is there only for derived types.

@etag actually is only there if the entity set uses optimistic concurrency control, in which case there also is an ETag response header for GET and an If-Match request header for POST, PATCH and DELETE. Just having @etag is incomplete and thus confusing.

Maybe split the @etag change off into a separate pull request?

snebjorn · 2020-05-01T06:23:31Z

It's only the etag I'm interested in, so I'd prefer it was there. But if we can get this merged quickly I can do an etag PR

# Conflicts: # tools/V4-CSDL-to-OpenAPI.xsl # tools/tests/Northwind-key-as-segment.openapi3.json # tools/tests/Northwind-key-as-segment.swagger.json # tools/tests/TripPin.openapi3.json # tools/tests/TripPin.swagger.json # tools/tests/annotations.openapi3.json # tools/tests/annotations.swagger.json # tools/tests/authorization.openapi3.json # tools/tests/authorization.swagger.json # tools/tests/containment.openapi3.json # tools/tests/containment.swagger.json # tools/tests/csdl-16.1.openapi3.json # tools/tests/csdl-16.1.swagger.json # tools/tests/descriptions.openapi3.json # tools/tests/descriptions.swagger.json # tools/tests/odata-rw-v2.openapi3.json # tools/tests/odata-rw-v2.swagger.json # tools/tests/odata-rw-v3.openapi3.json # tools/tests/odata-rw-v3.swagger.json

snebjorn · 2020-05-03T12:13:31Z

examples/TripPin.openapi3.json

                "title": "Event (for update)",
                "type": "object",
                "properties": {
+                    "@odata.type": {
+                        "$ref": "#/components/schemas/type"
+                    },
                    "ConfirmationCode": {
                        "type": "string",
                        "nullable": true


@ralfhandl I'm having an odd issue with the test claiming stuff isn't there when it is

1) Examples TripPin: AssertionError [ERR_ASSERTION]: OpenAPI document + expected - actual "type": "object" } "Microsoft.OData.SampleService.Models.TripPin.Event-update": { "properties": { + "@odata.type": { + "$ref": "#/components/schemas/type" + } "ConfirmationCode": { "nullable": true "type": "string" }

as you can see in the diff, it's quite clearly there. No idea what going on :(

The unit test npm run test compares the current in-memory output with the "expected" output from the examples folder.

If you are satisified with the diff, you can run npm run build to refresh the "expected" output. Use with care 😄

part of the problem here was the + expected - actual output not having enough context to identify where the issue was.

I ended up running npm test > report and counting how many lines it wrote to identify if I removed or added the @odata.type in the correct place.

It wasn't fun :(

snebjorn · 2020-05-03T12:16:44Z

examples/TripPin.openapi3.json

@@ -3301,6 +3301,9 @@
                "title": "Location",


@ralfhandl Also need a bit of help here. When I run test.cmd I get this error. It makes no sense to me.

odata-rw-v3 lexical error: invalid char in json text. {"swagger":"2.0","info":{"titl (right here) ------^ lexical error: invalid char in json text. {"openapi":"3.0.0","info":{"ti (right here) ------^ TripPin lexical error: invalid char in json text. {"swagger":"2.0","info":{"titl (right here) ------^ lexical error: invalid char in json text. {"openapi":"3.0.0","info":{"ti (right here) ------^

As a result the json files are empty

@ralfhandl do you have any clue why this is happening?

The only cause I can think of is that test.cmd only works if called from within the tools folder.

Note that the XSLT and the JS versions are not in sync and have their own set of test inputs and expected outputs.

I have a suspisseion this has something to do with CRLF and LF line endings.

I changed autocrlf = true to autocrlf = input a while ago and this problem started happening when I merged master in.
Also the CI doesn't seem to have this problem.

I have autocrlf = false

FML..... it was caused by a missing " in a xsl:text block

… with @count not being prefix in v3.0

ralfhandl · 2020-05-22T18:44:27Z

examples/annotations.openapi3.json

@@ -3061,6 +3061,9 @@
                ],
                "description": "The number of entities in the collection. Available when using the [$count](http://docs.oasis-open.org/odata/odata/v4.01/odata-v4.01-part1-protocol.html#sec_SystemQueryOptioncount) query option."
            },
+            "type": {
+                "type": "string"
+            },


Probably better to not have this „reuse schema“, or only have it when it isn’t needed.

I found this function deleteUnreferencedSchemas which has the desired effect. But it also removes a lot of other unused schemas causing a bunch of tests to fail. Fixable though.

But I assume deleteUnreferencedSchemas wasn't used for a reason?

Reason is that there can be CSDL files with schemas for reusable types and no entity container. The types are only referenced from other files, not from within the same file.

Consequently "delete unreferenced schemas" is not on by default.

Easiest fix here is not to define a reuse schema and generate @type directly with type: "string".

ralfhandl · 2020-05-22T18:49:04Z

tools/tests/annotations-v2.swagger.json

@@ -12,6 +12,9 @@
        "Supported.Annotations.SinglePartKey": {
            "type": "object",
            "properties": {
+                "__metadata": {
+                    "$ref": "#/definitions/metadata"


Don’t see this in definitions: lost?

Good find. This is strange.

I'm running test.cmd to update the json files. I double checked and It doesn't generate the metadata definition.
But it doesn't generate any of the reusable schemas. Hmmm the paths property of the OAS is empty. Perhaps that's part of the issue

ralfhandl · 2020-07-19T09:09:06Z

Hi @snebjorn, is this PR still active/desired?

It seems to be almost ready for merging..

snebjorn · 2020-07-20T13:50:32Z

Hi @ralfhandl, sorry. It's still desired. I've just been lazy. Lets hope for really bad weather - then I'll get around to doing it :D

ralfhandl · 2021-03-10T12:47:27Z

Hi @snebjorn, I hope the weather is getting worse daily 😄 .

I added prettier and eslint, so the code base is now formatted differently. Best you first copy over the .vscode settings, install the Prettier plugin, and reformat all source files before merging master into your branch.

Sorry for the hassle!

snebjorn · 2021-03-10T12:57:49Z

Hi @snebjorn, I hope the weather is getting worse daily 😄 .

I added prettier and eslint, so the code base is now formatted differently. Best you first copy over the .vscode settings, install the Prettier plugin, and reformat all source files before merging master into your branch.

Sorry for the hassle!

No problem at all. I was thinking of doing that at some point as well. I had to do ctrl+shift+p and save without formatting cause I had format on save on :)
So this is a welcome change 👍

Status on the PR. I tried to fix it a few weeks ago. But man this is hard getting back into. I'll need to get some more top shelf whiskey 😎

ralfhandl · 2021-03-10T13:03:04Z

Whiskey, not whisky? What's your recommendation?

My favorite is https://www.obanwhisky.com/whiskies/14-years-old.

snebjorn · 2021-03-10T13:13:15Z

Oh yes Whisky, I thought it looked a little funny.

I've had some good times with https://www.aberlour.com/en/product/abunadh-scotch-whisky/

sterlp · 2023-02-24T14:19:15Z

@snebjorn how far are you with the PR? I might need it too, as our gateway will most likely otherwise fail the schema validation on the response path. As the fields are where without beeing mentioned in the schema.

snebjorn · 2023-02-24T14:57:19Z

how far are you with the PR?

This PR is quite out of date now. I'm probably not going to finish this any time soon. My workarounds have been working, so getting this fixed is very low on my todo list.
You're welcome to continue the work here :)

snebjorn · 2023-11-15T13:21:28Z

sorry, it's unlikely I'll finish this

ralfhandl · 2023-11-15T13:41:43Z

Sorry for making things more complicated by refactoring while you were at it.

snebjorn · 2023-11-15T13:45:21Z

that was not the reason. Priorities switched and I just never got back to this

ralfhandl · 2023-11-15T14:06:18Z

That can happen to the best 😄

snebjorn commented Mar 29, 2020

View reviewed changes

lib/csdl2openapi.js Outdated Show resolved Hide resolved

snebjorn commented Mar 29, 2020

View reviewed changes

tools/V4-CSDL-to-OpenAPI.xsl Outdated Show resolved Hide resolved

feat: adds odata metadata to entity type

f609f64

fixes oasis-tcs#60

snebjorn force-pushed the feature/metadata branch from 6bb870d to f609f64 Compare March 31, 2020 21:51

snebjorn commented Mar 31, 2020

View reviewed changes

examples/miscellaneous.openapi3.json Outdated Show resolved Hide resolved

snebjorn marked this pull request as ready for review March 31, 2020 21:58

Merge branch 'master' into feature/metadata

e7fbc4e

ralfhandl requested changes Apr 10, 2020

View reviewed changes

lib/csdl2openapi.js Outdated Show resolved Hide resolved

tools/V4-CSDL-to-OpenAPI.xsl Outdated Show resolved Hide resolved

snebjorn added 2 commits April 27, 2020 00:17

moved metadata into reuseable schemas

5d5a74f

also removed @odata.id as it's typically not in the response. Made @odata.type only show up when type had a basetype

removed temp files

491334b

they were accidentally committed

ralfhandl requested changes May 1, 2020

View reviewed changes

snebjorn commented May 3, 2020

View reviewed changes

removed @id and @etag. @type is conditionally added. Also fixed issue…

cffcaa1

… with @count not being prefix in v3.0

snebjorn force-pushed the feature/metadata branch from 262ca65 to cffcaa1 Compare May 19, 2020 21:37

snebjorn requested a review from ralfhandl May 19, 2020 21:43

ralfhandl requested changes May 22, 2020

View reviewed changes

ralfhandl marked this pull request as draft January 12, 2023 12:06

snebjorn closed this Nov 15, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat: adds odata metadata to entity type #74

feat: adds odata metadata to entity type #74

snebjorn commented Mar 29, 2020

snebjorn commented Apr 1, 2020

snebjorn commented Apr 9, 2020

ralfhandl left a comment

ralfhandl commented May 1, 2020

snebjorn commented May 1, 2020

snebjorn May 3, 2020

ralfhandl May 19, 2020

snebjorn May 19, 2020 •

edited

Loading

snebjorn May 3, 2020 •

edited

Loading

snebjorn May 19, 2020

ralfhandl May 19, 2020

snebjorn May 19, 2020

ralfhandl May 19, 2020

snebjorn May 19, 2020

ralfhandl May 22, 2020

snebjorn May 25, 2020

ralfhandl May 25, 2020

ralfhandl May 22, 2020

snebjorn May 25, 2020 •

edited

Loading

ralfhandl commented Jul 19, 2020

snebjorn commented Jul 20, 2020

ralfhandl commented Mar 10, 2021

snebjorn commented Mar 10, 2021

ralfhandl commented Mar 10, 2021

snebjorn commented Mar 10, 2021

sterlp commented Feb 24, 2023

snebjorn commented Feb 24, 2023

snebjorn commented Nov 15, 2023

ralfhandl commented Nov 15, 2023

snebjorn commented Nov 15, 2023

ralfhandl commented Nov 15, 2023

feat: adds odata metadata to entity type #74

feat: adds odata metadata to entity type #74

Conversation

snebjorn commented Mar 29, 2020

snebjorn commented Apr 1, 2020

snebjorn commented Apr 9, 2020

ralfhandl left a comment

Choose a reason for hiding this comment

ralfhandl commented May 1, 2020

snebjorn commented May 1, 2020

Choose a reason for hiding this comment

Choose a reason for hiding this comment

snebjorn May 19, 2020 • edited Loading

Choose a reason for hiding this comment

snebjorn May 3, 2020 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

snebjorn May 25, 2020 • edited Loading

Choose a reason for hiding this comment

ralfhandl commented Jul 19, 2020

snebjorn commented Jul 20, 2020

ralfhandl commented Mar 10, 2021

snebjorn commented Mar 10, 2021

ralfhandl commented Mar 10, 2021

snebjorn commented Mar 10, 2021

sterlp commented Feb 24, 2023

snebjorn commented Feb 24, 2023

snebjorn commented Nov 15, 2023

ralfhandl commented Nov 15, 2023

snebjorn commented Nov 15, 2023

ralfhandl commented Nov 15, 2023

snebjorn May 19, 2020 •

edited

Loading

snebjorn May 3, 2020 •

edited

Loading

snebjorn May 25, 2020 •

edited

Loading