client-go/util/jsonpath: resolve #16707 by outputting json for non-primitive types #89660

pjferrell · 2020-03-30T18:17:53Z

What type of PR is this?
/kind bug

What this PR does / why we need it:
When printing JSONPATH results, this changes the display format of non-primitive types from Go-syntax to JSON. This improves readability for maps/slices and simplifies processing outputs in other tools.

Which issue(s) this PR fixes:

Fixes #16707
Fixes #67268

Special notes for your reviewer:

Does this PR introduce a user-facing change?:

jsonpath support in kubectl / client-go serializes complex types (maps / slices / structs) as json instead of Go-syntax.

Additional documentation e.g., KEPs (Kubernetes Enhancement Proposals), usage docs, etc.:

k8s-ci-robot · 2020-03-30T18:17:56Z

Thanks for your pull request. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

📝 Please follow instructions at https://git.k8s.io/community/CLA.md#the-contributor-license-agreement to sign the CLA.

It may take a couple minutes for the CLA signature to be fully registered; after that, please reply here with a new comment and we'll verify. Thanks.

If you've already signed a CLA, it's possible we don't have your GitHub username or you're using a different email address. Check your existing CLA data and verify that your email is set on your git commits.
If you signed the CLA as a corporation, please sign in with your organization's credentials at https://identity.linuxfoundation.org/projects/cncf to be authorized.
If you have done the above and are still having issues with the CLA being reported as unsigned, please log a ticket with the Linux Foundation Helpdesk: https://support.linuxfoundation.org/
Should you encounter any issues with the Linux Foundation Helpdesk, send a message to the backup e-mail support address at: login-issues@jira.linuxfoundation.org

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. I understand the commands that are listed here.

k8s-ci-robot · 2020-03-30T18:18:01Z

Welcome @pjferrell!

It looks like this is your first PR to kubernetes/kubernetes 🎉. Please refer to our pull request process documentation to help your PR have a smooth ride to approval.

You will be prompted by a bot to use commands during the review process. Do not be afraid to follow the prompts! It is okay to experiment. Here is the bot commands documentation.

You can also check if kubernetes/kubernetes has its own contribution guidelines.

You may want to refer to our testing guide if you run into trouble with your tests not passing.

If you are having difficulty getting your pull request seen, please follow the recommended escalation practices. Also, for tips and tricks in the contribution process you may want to read the Kubernetes contributor cheat sheet. We want to make sure your contribution gets all the attention it needs!

Thank you, and welcome to Kubernetes. 😃

k8s-ci-robot · 2020-03-30T18:18:02Z

Hi @pjferrell. Thanks for your PR.

I'm waiting for a kubernetes member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

pjferrell · 2020-03-30T18:29:50Z

/assign @smarterclayton

brianpursley · 2020-04-12T20:36:45Z

/ok-to-test

brianpursley · 2020-04-12T20:41:19Z

Hi @pjferrell
Thanks for the PR.

if you edit your original message to change Fixes # 16707 to Fixes #16707 (remove the space) it will turn that into a link to the issue which will help the reviewers.

brianpursley · 2020-04-12T20:47:19Z

staging/src/k8s.io/client-go/util/jsonpath/jsonpath.go

+		var text []byte
+		var err error
+		if r.Kind() == reflect.Interface && (r.Elem().Kind() == reflect.Map ||
+			r.Elem().Kind() == reflect.Slice || r.Elem().Kind() == reflect.Struct) {


Do you want this change to apply for slice and struct? I thought the issue was only about map. I guess I would need to see some unit tests around those types to understand how the output would be different. (By the way, I'm not saying this is wrong, just unsure what the effect is)

Based on the comments in #16707 (@thockin last comment and others), it appeared favorable to address all non-primitive types which could be represented by json. Go-style output of structs does not contain attribute/key names and slices are printed without comma separators, which makes passing these outputs to other tools more difficult.

The current test failure is an example jsonpath output where the slice was previously output is Go-style instead of json.

Expected was the old output, and actual is the proper json representation of the slice.

--- FAIL: TestTokenOutput (0.00s) --- PASS: TestTokenOutput/JSON_output (0.00s) --- PASS: TestTokenOutput/YAML_output (0.00s) --- PASS: TestTokenOutput/Go_template_output (0.00s) --- PASS: TestTokenOutput/text_output (0.00s) --- FAIL: TestTokenOutput/jsonpath_output (0.00s) token_test.go:449: failed TestTokenOutput: expected: abcdef.1234567890123456 [system:bootstrappers:kubeadm:default-node-token] actual: abcdef.1234567890123456 ["system:bootstrappers:kubeadm:default-node-token"]

This test case was fixed and added to commit.

Can you add a unit test to cover the slice and struct scenarios?
Something like this might work to cover both (slice with multiple structs):

{range .items[1]}{.status.addresses} {end}

I'm just thinking it will be good to have this to show the expected output and prevent any future regression.

I've added several test cases for slices of each primitive type.

brianpursley · 2020-04-12T22:34:34Z

/retest

neolit123 · 2020-04-15T23:46:29Z

/approve
for cmd/kubeadm

/area kubectl
/priority backlog
(backlog, given this has been an issue for a while)

…th output of non-primitive type

lavalamp · 2020-04-28T20:30:15Z

/assign

lavalamp · 2020-04-28T20:34:46Z

As long as this is a client-side change, which it seems to be, I'm fine with it (thanks for all the tests). I think someone from SIG CLI should sign off on it, then I'm happy to approve for the api machinery OWNERs files.

/assign @pwittrock @seans3 @apelisse

pjferrell · 2020-04-28T20:54:37Z

@lavalamp It does change the default output behavior for complex objects to json instead of go-syntax. IMO it would be a positive change, but it could break a script expecting go-style output ("1 2 3" vs "[1,2,3]"). I fixed all kubernetes test cases (3) that used the go-style output.

CRD AdditionalPrinter columns would be affected if the path specified by user points to a struct/map/slice.

The cli flag provides an output option to display a json slice of results (consistent with how other jsonpath implementations output).

lavalamp · 2020-04-28T20:58:56Z

Yeah, I think that makes sense, we just need SIG CLI to agree.

…

On Tue, Apr 28, 2020 at 1:54 PM Phillip Ferrell ***@***.***> wrote: @lavalamp <https://github.com/lavalamp> It does change the default output behavior for complex objects to json instead of go-syntax. IMO it would be a positive change, but it could break a script expecting go-style output ("1 2 3" vs "[1,2,3]"). I fixed all kubernetes test cases (3) that used the go-style output. The cli flag provides an output option to display a json slice of results (consistent with how other jsonpath implementations output). — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#89660 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAE6BFX4S3KB66UIPT7EXA3RO47B5ANCNFSM4LWZNMLQ> .

apelisse · 2020-04-28T21:06:30Z

I see a couple of comments from informed people that make sense to me:

From #33890 (comment):

Single values are used to provide lists of arguments. Changing the format would break all scripts using it.

The printing of complex values is a bug, and can be fixed.

From #33890 (comment):

The go type dumps are borderline useless. We can either say JSON or we can
define our own format.

lavalamp · 2020-04-28T21:14:34Z

/approve
/lgtm

OK that's good enough for me

k8s-ci-robot · 2020-04-28T21:15:06Z

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: lavalamp, neolit123, pjferrell

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

~~cmd/kubeadm/OWNERS~~ [lavalamp,neolit123]
~~staging/src/k8s.io/apiextensions-apiserver/OWNERS~~ [lavalamp]
~~staging/src/k8s.io/cli-runtime/OWNERS~~ [lavalamp]
~~staging/src/k8s.io/client-go/OWNERS~~ [lavalamp]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

pjferrell · 2020-04-29T05:25:32Z

/retest

thockin · 2020-05-15T02:31:23Z

This is awesome!

Now how do I format a JSON list without a trailing comma?

kubectl get nodes -o jsonpath='{"{"}{range .items[*]}"{.metadata.name}": {.metadata.labels},{end}{"}"}' | jq .
parse error: Expected another key-value pair at line 1, column 1582

:P

brianpursley · 2020-05-15T05:30:13Z

Its probably not what you're looking for, but you could use sed to remove it...

kubectl get nodes -o jsonpath='{"{"}{range .items[*]}"{.metadata.name}": {.metadata.labels},{end}{"}"}' | sed 's/\(.*\),/\1/' | jq

Are you thinking there should be some kind of conditional rendering? ...maybe some new thing like {separator ","} that conditionally renders only when it is a separating range iterations?

Is jsonpath an actual spec?

Also, if it were able to render \b, then you could add a \b after the {end} to "backspace away" the last comma. A little hacky but might be a simple way to do it if it actually worked.

lavalamp · 2020-05-15T18:37:33Z

IMO just use jq if you want to do that, it's much simpler.

To do it without jq, I would do something exceedingly stupid, e.g.

$ echo "[x,y,]" | (head -c -3; echo "]")
[x,y]

lavalamp · 2020-05-15T18:41:50Z

I haven't tested it but a jq script should look something like this:

jq '[.items[*] | {.metadata.name, .metadata.labels}]'

(Yes, I feel guilty for being this nerdy about jq)

k8s-ci-robot added release-note do-not-merge/invalid-commit-message kind/bug size/S labels Mar 30, 2020

k8s-ci-robot added cncf-cla: no needs-sig needs-priority labels Mar 30, 2020

k8s-ci-robot added the needs-ok-to-test label Mar 30, 2020

k8s-ci-robot requested review from sttts and yliaog March 30, 2020 18:21

k8s-ci-robot added sig/api-machinery and removed needs-sig labels Mar 30, 2020

pjferrell force-pushed the kubectl-jsonpath-nonprimitive-types branch from b4d3a0b to 631cd15 Compare March 30, 2020 18:22

k8s-ci-robot added cncf-cla: yes and removed cncf-cla: no labels Mar 30, 2020

pjferrell force-pushed the kubectl-jsonpath-nonprimitive-types branch from 631cd15 to eac9368 Compare March 30, 2020 18:27

k8s-ci-robot removed the do-not-merge/invalid-commit-message label Mar 30, 2020

k8s-ci-robot assigned smarterclayton Mar 30, 2020

k8s-ci-robot added ok-to-test and removed needs-ok-to-test labels Apr 12, 2020

brianpursley reviewed Apr 12, 2020

View reviewed changes

pjferrell force-pushed the kubectl-jsonpath-nonprimitive-types branch from eac9368 to 1c6352a Compare April 13, 2020 06:20

k8s-ci-robot added the area/kubeadm label Apr 13, 2020

k8s-ci-robot assigned seans3 and timothysc Apr 15, 2020

k8s-ci-robot added area/kubectl priority/backlog and removed needs-priority labels Apr 15, 2020

apiextensions-apiserver: update test for CRD custom columns to jsonpa…

f092c38

…th output of non-primitive type

pjferrell force-pushed the kubectl-jsonpath-nonprimitive-types branch from d657a1c to f092c38 Compare April 16, 2020 10:21

k8s-ci-robot assigned lavalamp Apr 28, 2020

k8s-ci-robot assigned apelisse and pwittrock Apr 28, 2020

k8s-ci-robot added the lgtm label Apr 28, 2020

k8s-ci-robot added the approved label Apr 28, 2020

brianpursley approved these changes Apr 29, 2020

View reviewed changes

k8s-ci-robot merged commit 0c3c2cd into kubernetes:master Apr 29, 2020

k8s-ci-robot added this to the v1.19 milestone Apr 29, 2020

brianpursley mentioned this pull request May 1, 2020

Added unit tests to document current behavior of TableConverter #90531

Merged

glyn mentioned this pull request Jun 11, 2020

Allow dot notation without root in Proposal A cburgmer/json-path-comparison#45

Open

client-go/util/jsonpath: resolve #16707 by outputting json for non-primitive types #89660

client-go/util/jsonpath: resolve #16707 by outputting json for non-primitive types #89660

Conversation

pjferrell commented Mar 30, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

k8s-ci-robot commented Mar 30, 2020

Uh oh!

k8s-ci-robot commented Mar 30, 2020

Uh oh!

k8s-ci-robot commented Mar 30, 2020

Uh oh!

pjferrell commented Mar 30, 2020

Uh oh!

brianpursley commented Apr 12, 2020

Uh oh!

brianpursley commented Apr 12, 2020

Uh oh!

brianpursley Apr 12, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

pjferrell Apr 13, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

pjferrell Apr 13, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

brianpursley Apr 13, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

pjferrell Apr 15, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

brianpursley commented Apr 12, 2020

Uh oh!

neolit123 commented Apr 15, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lavalamp commented Apr 28, 2020

Uh oh!

lavalamp commented Apr 28, 2020

Uh oh!

pjferrell commented Apr 28, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lavalamp commented Apr 28, 2020 via email

Uh oh!

apelisse commented Apr 28, 2020

Uh oh!

lavalamp commented Apr 28, 2020

Uh oh!

k8s-ci-robot commented Apr 28, 2020

Uh oh!

pjferrell commented Apr 29, 2020

Uh oh!

thockin commented May 15, 2020

Uh oh!

brianpursley commented May 15, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lavalamp commented May 15, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lavalamp commented May 15, 2020

Uh oh!

pjferrell commented Mar 30, 2020 •

edited

Loading

brianpursley Apr 12, 2020 •

edited

Loading

pjferrell Apr 13, 2020 •

edited

Loading

pjferrell Apr 13, 2020 •

edited

Loading

brianpursley Apr 13, 2020 •

edited

Loading

pjferrell Apr 15, 2020 •

edited

Loading

neolit123 commented Apr 15, 2020 •

edited

Loading

pjferrell commented Apr 28, 2020 •

edited

Loading

brianpursley commented May 15, 2020 •

edited

Loading

lavalamp commented May 15, 2020 •

edited

Loading