49 Commits

Author SHA1 Message Date
Doug Smith
18630fde0b Merge pull request #1424 from nirdothan/mount-multus-conf-dir
Thick client: mount multus-conf-dir
2025-05-13 08:32:26 -04:00
Nir Dothan
19f9283db4 Thick client: mount multus-conf-dir
Currently, the default CNI config dir is /etc/cni/multus/net.d [1].
However, pods do not mount this hostPath.
This issue only occurs when a delegate network-attachment-definition
has no spec, and it therefore needs to be loaded from disk[2].
For example, when deploying Istio-cni in multus mode,
the deployment creates an empty Istio-cni NAD in default namespace, while
the actual config is deployed on disk.
CmdAdd fails with the following error message:
GetNetworkDelegates: failed getting the delegate: GetCNIConfig:
err in GetCNIConfigFromFile: No networks found in /etc/cni/multus/net.d

[[1]](https://github.com/k8snetworkplumbingwg/multus-cni/blob/v4.2.0/pkg/types/conf.go#L38)
[[2]](https://github.com/k8snetworkplumbingwg/multus-cni/blob/v4.2.0/pkg/k8sclient/k8sclient.go#L506)

Signed-off-by: Nir Dothan <ndothan@redhat.com>
2025-05-05 17:45:52 +03:00
Doug Smith
1655d540cb Merge pull request #1418 from rollandf/baseimage
chore: update Dockerfile base image
2025-04-24 15:43:05 +02:00
Ben Pickard
4517063b79 Merge pull request #1419 from dougbtv/rebase5-cni-subdirectory-chain
Safe subdirectory based CNI chain configuration loading
2025-04-16 16:04:18 -04:00
dougbtv
4104fea90d Subdirectory CNI chain loading e2e tests
Adds a test for plain subdirectory chaining and also using passthru CNI with auxiliaryCNIChainName
2025-04-15 15:53:18 -04:00
dougbtv
528d4f150c Functionality for Aux CNI Chain using subdirectory based CNI configuration loading.
Removes the it `fails to execute confListDel given no 'plugins' key"` test.

This test no longer fails after libcni version 1.2.3.
It probably shouldn't failduring a DEL action as it is, we want the least error prone path.

The GC test now uses both cni.dev attachment formats.

Uses both attachment formats as per https://github.com/containernetworking/cni/issues/1101 for GC's cni.dev/valid-attachments & cni.dev/attachments
2025-04-15 15:53:00 -04:00
dougbtv
fa3c7cfee3 [bump] Bumps to libcni v1.3.0 2025-04-09 14:42:47 -04:00
Fred Rolland
96bfb26dac chore: update Dockerfile base image
- Fix CVEs

Signed-off-by: Fred Rolland <frolland@nvidia.com>
2025-04-06 16:31:47 +03:00
Doug Smith
55ef3b1f0b Merge pull request #1370 from thomasferrandiz/add-trivy
Add trivy vulnerability scanner in build step
2025-04-03 15:35:40 +02:00
Ben Pickard
41321963b8 Merge pull request #1374 from buroa/master
fix: dockerfile change cmd to entrypoint
2025-03-31 16:15:02 -04:00
Thomas Ferrandiz
ef8f01b299 Use cross-compilation for thick plugin build 2025-03-31 15:39:41 +00:00
Thomas Ferrandiz
51752f1a6e Add trivy vulnerability scanner in build step
Signed-off-by: Thomas Ferrandiz <thomas.ferrandiz@suse.com>
2025-03-31 15:39:41 +00:00
Ben Pickard
1821311479 Merge pull request #1412 from dougbtv/fix-empty-cni-result
Properly structure empty CNI result
2025-03-25 17:42:43 -04:00
dougbtv
ccfd8f5fea When returning an empty CNI result, it must be properly structured
For a previous fix of returning an empty CNI result when pods are not found, the CNI result wasn't properly structured. This fixes the structuring.
2025-03-25 14:45:04 -04:00
Doug Smith
2a91646eaf Merge pull request #1409 from maiqueb/bump-net-attach-def-client-lib-1.7.6
build: consume net-attach-def-client lib 1.7.6
2025-03-24 19:22:12 +01:00
Ben Pickard
47e5153714 Merge pull request #1408 from dougbtv/pod-not-found-on-add
handle pod not found during CNI ADD gracefully
2025-03-24 14:11:03 -04:00
Miguel Duarte Barroso
21f7282088 build: consume net-attach-def-client lib 1.7.6
This consumes the latest release of the network-attachment-definition-client
library which fixes a regression affecting CNI plugins that do not specify
interfaces in their CNI ADD result. This was fixed in [0].

[0] - https://github.com/k8snetworkplumbingwg/network-attachment-definition-client/pull/77

Signed-off-by: Miguel Duarte Barroso <mdbarroso@redhat.com>
2025-03-24 15:51:11 +00:00
dougbtv
641f6a3b63 handle pod not found in CNI ADD gracefully
sometimes pods get deleted super fast (like jobs or CI) and they come back as not found.

instead of erroring, just return an empty CNI result so things don't blow up.

adds a sentinel errPodNotFound and skips the rest of CmdAdd when we hit it.

shouts to race conditions.
2025-03-24 09:58:38 -04:00
Ben Pickard
e156e815ad Merge pull request #1407 from dougbtv/pod-deleted-on-add
Tolerate issues writing network status annotation on CNI ADD.
2025-03-20 15:44:34 -04:00
dougbtv
5892d705da Tolerate issues writing network status annotation on CNI ADD.
This change adds toleration for such errors like:

```
failed to [query/update] the pod pod-name-here in out of cluster comm: pod "pod-name-here" not found
```

During CNI ADD. While this change is a trade off in terms of debugability for RBAC, it's potentially noisy in scaled clusters when it is working properly.
2025-03-20 14:20:00 -04:00
Ben Pickard
431a735eca Merge pull request #1404 from dougbtv/fix-e2e-dra
The e2e kind config should use api/beta for the runtimeConfig
2025-03-20 12:44:42 -04:00
dougbtv
99d72d14a3 The e2e kind config should use api/beta for the runtimeConfig
Otherwise, the latest changes to DRA (which is beta in K8s 1.32) are incompatible.

Additionally, this:

* Bumps kind version to 0.27.0
* Changes `loglevel` flag to `v` verbosity flag for `kind export logs`
* fixes lint in the Dockerfile.
* adds a couple notes in the docs.
2025-03-20 11:21:03 -04:00
Doug Smith
4a0b5073af Merge pull request #1273 from s1061123/cni110
CNI 1.1.0 support
2025-01-16 09:40:09 -05:00
Steven Kreitzer
5216844263 fix: dockerfile change cmd to entrypoint 2024-12-29 10:41:06 -07:00
Tomofumi Hayashi
7eb9673a1a Call GC command with valid attachments from multus cache
This code changes CNI's GC command argument. Previously it just
passes from parent CNI runtime, however, it may causes unexpected
resource deletion if one CNI plugin is used in both cluster
network and net-attach-def. This change generates valid attachments
from multus CNI cache and passed to delegate CNI plugin.
2024-12-20 11:28:41 +09:00
Tomofumi Hayashi
a439f91721 Support GC and STATUS command for cluster network
This change supports up to date CNI 1.1 command, GC and STATUS for
cluster network.
2024-12-20 11:28:41 +09:00
Tomofumi Hayashi
6d3d800226 Update vendor packages, including CNI v1.2.0 2024-12-20 11:28:38 +09:00
Doug Smith
fba1fea81e Merge pull request #1373 from dougbtv/livequery-context
adds context to GetPodAPILiveQuery
2024-12-19 14:48:14 -05:00
dougbtv
f186370654 adds context to GetPodAPILiveQuery 2024-12-19 14:41:32 -05:00
Doug Smith
fc72ddbd24 Merge pull request #1332 from dougbtv/getlivepod
always attempt a live pod get on miss to confirm its really not there
2024-12-19 14:13:10 -05:00
dougbtv
fb03b0f754 This makes sure that stale caches never result in NotFound errors.
It was explained to me that informers are almost always are more efficient, and in most cases will work, but a live lookup is appropriate after a number of failures.

This happens only on the retry portion, so we're still getting the benefits of informers, but, on a retry situation, we don't get a cache miss.

Additionally, changes out use of cache get on this, since it already bails out before it on CNI DEL.
2024-12-19 13:57:57 -05:00
Doug Smith
5338017bf6 Merge pull request #1372 from pmtk/dont-wait-too-long-for-apiserver
Thin plugin: don't wait too long for an answer from API Server
2024-12-19 10:55:40 -05:00
Patryk Matuszak
4ff141c18d Don't wait too long for an answer from API Server
If Multus plugin gets a DEL request, but the API Server is down (e.g.
via 'crictl rmp'), the call takes so long, it actually never finishes.
This prevents CRI-O from deleting the Pods.
2024-12-19 16:13:38 +01:00
Doug Smith
4fc16b3bb8 Merge pull request #1355 from Nordix/fix-cve-moshiur
Support go 1.22 and 1.23 build to fix CVE
2024-12-06 16:56:48 -05:00
smoshiur1237
ddbcd2c4ef Support go 1.22 to fix CVE
Signed-off-by: smoshiur1237 <moshiur.rahman@est.tech>
2024-12-05 17:03:03 +02:00
Tomofumi Hayashi
781ecdaecd Merge pull request #1353 from xrstf/master
clean up go.mod, get rid of client-go v1.5.2
2024-11-08 01:53:15 +09:00
Christoph Mewes
808185b10f clean up go.mod, get rid of client-go v1.5.2 2024-11-06 12:51:30 +01:00
Doug Smith
e1a0d2a3fd Merge pull request #1345 from dougbtv/net-attach-def-lib-175
Update net-attach-def client library to 1.7.5 for cri-o functionality
2024-10-15 12:05:24 -04:00
dougbtv
ecf5854ca9 Update net-attach-def client library to 1.7.5 for cri-o functionality
From the release notes:

> This release contains a fix related to the determination of the default interface, e.g. setting the default parameter to true in the network-status annotation based on the presence of a gateway in the CNI ADD success result ips.gateway and makes the determination of the default based on the first interface that has an associated value of gateway (using the interface index in the ips element in the CNI ADD success result).

> This provides flexibility especially in CRI-O which uses the first interface and IP addresses for the pod.IP in Kubernetes, therefore. Containerd functionality is unchanged in that it uses the value for the IP addresses specifically

> It's worth noting that CNI ADD success results which do not contain any interfaces will be discarded in this determination of the default, therefore it's recommended to set one with an associated gateway if aiming to have it be noted as the default.

See also:
https://github.com/k8snetworkplumbingwg/network-attachment-definition-client/releases/tag/v1.7.5
https://github.com/k8snetworkplumbingwg/network-attachment-definition-client/pull/73
2024-10-15 11:37:32 -04:00
Doug Smith
adfb270991 Merge pull request #1341 from dougbtv/net-attach-def-client-v174
Updates net-attach-def client library to v1.7.4
2024-10-01 09:36:43 -04:00
dougbtv
b171bb702b Updates net-attach-def client library to v1.7.4
Which improves backwards compatibility for network-status in latest updates to the client library, especially related to Calico.

See also: https://github.com/k8snetworkplumbingwg/network-attachment-definition-client/pull/72
2024-09-30 15:57:48 -04:00
Doug Smith
f1e887e239 Merge pull request #1336 from dougbtv/net-attach-def-client-v173
Bumps net-attach-def client to v1.7.3
2024-09-13 12:03:57 -04:00
dougbtv
100766d1a4 Bumps net-attach-def client to v1.7.3
Previous version didn't account for accounts for the sandox interfaces when reporting the interfaces in the network-status annotation when calculating the default:true interface
2024-09-13 09:56:52 -04:00
Doug Smith
e074c2a56b Merge pull request #1335 from dougbtv/net-attach-def-client-v172
Bumps net-attach-def client library to v1.7.2
2024-09-12 13:35:46 -04:00
dougbtv
38d03eb816 Bumps net-attach-def client library to v1.7.2
This fixes the default:true for multiple interface returns from CNI for cluster default network, where all interfaces in that return were marked as default:true in the network-status
2024-09-12 11:25:26 -04:00
Doug Smith
b554c96160 Merge pull request #1334 from dougbtv/disable-dra-e2e-temporarily
Disabled DRA test temporarily
2024-09-12 10:55:16 -04:00
dougbtv
92ff1b1ee8 Disabled DRA test temporarily 2024-09-12 10:06:17 -04:00
Doug Smith
31e77aafab Merge pull request #1321 from ah8ad3/update-install-readme
Doc: change install from file to url in readme, how-to-use
2024-08-29 09:41:23 -04:00
Ahmad Zolfaghari
dec0607a94 doc: change install from file to url in readme, how-to-use
Signed-off-by: Ahmad Zolfaghari <ah8ad3@gmail.com>
2024-08-09 16:32:07 +03:30
356 changed files with 17614 additions and 29672 deletions

View File

@@ -4,7 +4,7 @@ jobs:
build:
strategy:
matrix:
go-version: [1.20.x, 1.21.x]
go-version: [1.22.x, 1.23.x]
goarch: [386, amd64, arm, arm64, ppc64le, s390x]
os: [ubuntu-latest] #, macos-latest, windows-latest]
runs-on: ${{ matrix.os }}

View File

@@ -13,7 +13,7 @@ jobs:
# note: disable sbom/provenance for now (gchr.io does not managed well yet)
- name: Build container image
uses: docker/build-push-action@v5
uses: docker/build-push-action@v6
with:
context: .
push: false
@@ -25,7 +25,7 @@ jobs:
# note: disable sbom/provenance for now (gchr.io does not managed well yet)
- name: Build container debug image
uses: docker/build-push-action@v5
uses: docker/build-push-action@v6
with:
context: .
push: false
@@ -46,7 +46,7 @@ jobs:
uses: docker/setup-buildx-action@v3
- name: Build container image
uses: docker/build-push-action@v5
uses: docker/build-push-action@v6
with:
context: .
push: false
@@ -56,6 +56,22 @@ jobs:
sbom: false
provenance: false
- name: Run Trivy vulnerability scanner
uses: aquasecurity/trivy-action@0.29.0
with:
image-ref: ghcr.io/${{ github.repository }}:latest-thick
ignore-unfixed: true
vuln-type: 'os,library'
severity: 'CRITICAL,HIGH'
format: 'sarif'
output: 'trivy-results.sarif'
- name: Upload Trivy scan results to GitHub Security tab
uses: github/codeql-action/upload-sarif@v3
if: always()
with:
sarif_file: 'trivy-results.sarif'
build-origin:
name: Image build/origin
runs-on: ubuntu-latest

View File

@@ -85,15 +85,25 @@ jobs:
working-directory: ./e2e
run: ./test-default-route1.sh
- name: Test DRA integration
# - name: Test DRA integration
# working-directory: ./e2e
# run: ./test-dra-integration.sh
- name: Test subdirectory CNI chaining
if: ${{ matrix.multus-manifest == 'multus-daemonset-thick.yml' }}
working-directory: ./e2e
run: ./test-dra-integration.sh
run: ./test-subdirectory-chaining.sh
- name: Test subdirectory CNI chaining with passthru CNI / auxiliaryCNIChainName
if: ${{ matrix.multus-manifest == 'multus-daemonset-thick.yml' }}
working-directory: ./e2e
run: ./test-subdirectory-chaining-passthru.sh
- name: Export kind logs
if: always()
run: |
mkdir -p /tmp/kind/logs
kind export logs --loglevel=debug /tmp/kind/logs
kind export logs /tmp/kind/logs -v 2147483647
- name: Upload kind logs
if: always()

View File

@@ -15,7 +15,7 @@ jobs:
- name: Set up Go
uses: actions/setup-go@v5
with:
go-version: 1.19.x
go-version: 1.22.x
- name: Run GoReleaser
uses: goreleaser/goreleaser-action@v5

View File

@@ -4,7 +4,7 @@ jobs:
test:
strategy:
matrix:
go-version: [1.20.x, 1.21.x]
go-version: [1.22.x, 1.23.x]
os: [ubuntu-latest]
runs-on: ${{ matrix.os }}
steps:

View File

@@ -24,10 +24,10 @@ Here's an illustration of the network interfaces attached to a pod, as provision
The quickstart installation method for Multus requires that you have first installed a Kubernetes CNI plugin to serve as your pod-to-pod network, which we refer to as your "default network" (a network interface that every pod will be created with). Each network attachment created by Multus will be in addition to this default network interface. For more detail on installing a default network CNI plugin, refer to our [quick-start guide](docs/quickstart.md).
Clone this GitHub repository, and apply a daemonset which installs Multus using `kubectl`. From the root directory of the clone, apply the daemonset YAML file:
To use latest features try command below which applies a daemonset and installs thick Multus using `kubectl`:
```
cat ./deployments/multus-daemonset-thick.yml | kubectl apply -f -
kubectl apply -f https://raw.githubusercontent.com/k8snetworkplumbingwg/multus-cni/master/deployments/multus-daemonset-thick.yml
```
This will configure your systems to be ready to use Multus CNI, but, to get started with adding additional interfaces to your pods, refer to our complete [quick-start guide](docs/quickstart.md)
@@ -39,7 +39,7 @@ With the multus 4.0 release, we introduce a new client/server-style plugin deplo
We recommend using the thick plugin in most environments, but if you wish to run the thin plugin, or are in a resource-constrained environment, you may do so with:
```
cat ./deployments/multus-daemonset.yml | kubectl apply -f -
kubectl apply -f https://raw.githubusercontent.com/k8snetworkplumbingwg/multus-cni/master/deployments/multus-daemonset.yml
```
## Additional Installation Options

View File

@@ -53,4 +53,15 @@ func main() {
}
fmt.Printf("multus %s copy succeeded!\n", multusFileName)
// Copy the passthru CNI
passthruPath := "/usr/src/multus-cni/bin/passthru"
err = cmdutils.CopyFileAtomic(passthruPath, *destDir, fmt.Sprintf("%s.temp", "passthru"), "passthru")
if err != nil {
fmt.Fprintf(os.Stderr, "failed to copy file %s: %v\n", multusFileName, err)
os.Exit(1)
}
fmt.Printf("passthru cni %s copy succeeded!\n", passthruPath)
}

View File

@@ -44,15 +44,23 @@ func main() {
return
}
skel.PluginMain(
func(args *skel.CmdArgs) error {
return api.CmdAdd(args)
},
func(args *skel.CmdArgs) error {
return api.CmdCheck(args)
},
func(args *skel.CmdArgs) error {
return api.CmdDel(args)
skel.PluginMainFuncs(
skel.CNIFuncs{
Add: func(args *skel.CmdArgs) error {
return api.CmdAdd(args)
},
Check: func(args *skel.CmdArgs) error {
return api.CmdCheck(args)
},
Del: func(args *skel.CmdArgs) error {
return api.CmdDel(args)
},
GC: func(args *skel.CmdArgs) error {
return api.CmdGC(args)
},
Status: func(args *skel.CmdArgs) error {
return api.CmdStatus(args)
},
},
cniversion.All, "meta-plugin that delegates to other CNI plugins")
}

View File

@@ -43,17 +43,27 @@ func main() {
return
}
skel.PluginMain(
func(args *skel.CmdArgs) error {
result, err := multus.CmdAdd(args, nil, nil)
if err != nil {
return err
}
return result.Print()
skel.PluginMainFuncs(
skel.CNIFuncs{
Add: func(args *skel.CmdArgs) error {
result, err := multus.CmdAdd(args, nil, nil)
if err != nil {
return err
}
return result.Print()
},
Del: func(args *skel.CmdArgs) error {
return multus.CmdDel(args, nil, nil)
},
Check: func(args *skel.CmdArgs) error {
return multus.CmdCheck(args, nil, nil)
},
GC: func(args *skel.CmdArgs) error {
return multus.CmdGC(args, nil, nil)
},
Status: func(args *skel.CmdArgs) error {
return multus.CmdStatus(args, nil, nil)
},
},
func(args *skel.CmdArgs) error {
return multus.CmdCheck(args, nil, nil)
},
func(args *skel.CmdArgs) error { return multus.CmdDel(args, nil, nil) },
cniversion.All, "meta-plugin that delegates to other CNI plugins")
}

58
cmd/passthru-cni/main.go Normal file
View File

@@ -0,0 +1,58 @@
// Package: passthru-cni
package main
import (
"encoding/json"
"fmt"
"github.com/containernetworking/cni/pkg/skel"
cniTypes "github.com/containernetworking/cni/pkg/types"
current "github.com/containernetworking/cni/pkg/types/100"
cniVersion "github.com/containernetworking/cni/pkg/version"
)
// NetConf is a CNI configuration structure
type NetConf struct {
cniTypes.NetConf
}
func main() {
skel.PluginMain(
cmdAdd,
nil,
cmdDel,
cniVersion.PluginSupports("0.3.0", "0.3.1", "0.4.0", "1.0.0", "1.1.0"),
"Passthrough CNI Plugin v1.0",
)
}
func cmdAdd(args *skel.CmdArgs) error {
n, err := loadNetConf(args.StdinData)
if err != nil {
return fmt.Errorf("passthru cni: error parsing CNI configuration: %s", err)
}
// Create an empty but valid CNI result
result := &current.Result{
CNIVersion: n.CNIVersion,
Interfaces: []*current.Interface{},
IPs: []*current.IPConfig{},
Routes: []*cniTypes.Route{},
DNS: cniTypes.DNS{},
}
return cniTypes.PrintResult(result, n.CNIVersion)
}
func cmdDel(_ *skel.CmdArgs) error {
// Nothing to do for DEL command, just return nil
return nil
}
func loadNetConf(bytes []byte) (*NetConf, error) {
n := &NetConf{}
if err := json.Unmarshal(bytes, n); err != nil {
return nil, fmt.Errorf("passthru cni: failed to load netconf: %s", err)
}
return n, nil
}

View File

@@ -192,6 +192,8 @@ spec:
- name: hostroot
mountPath: /hostroot
mountPropagation: HostToContainer
- mountPath: /etc/cni/multus/net.d
name: multus-conf-dir
env:
- name: MULTUS_NODE_NAME
valueFrom:
@@ -201,9 +203,9 @@ spec:
- name: install-multus-binary
image: ghcr.io/k8snetworkplumbingwg/multus-cni:snapshot-thick
command:
- "cp"
- "/usr/src/multus-cni/bin/multus-shim"
- "/host/opt/cni/bin/multus-shim"
- "sh"
- "-c"
- "cp /usr/src/multus-cni/bin/multus-shim /host/opt/cni/bin/multus-shim && cp /usr/src/multus-cni/bin/passthru /host/opt/cni/bin/passthru"
resources:
requests:
cpu: "10m"
@@ -247,3 +249,6 @@ spec:
- name: host-run-netns
hostPath:
path: /run/netns/
- name: multus-conf-dir
hostPath:
path: /etc/cni/multus/net.d

View File

@@ -37,6 +37,7 @@ Example configuration using `clusterNetwork` (see also [using delegates](#using-
"defaultNetworks": ["sidecarCRD", "exampleNetwork"],
"systemNamespaces": ["kube-system", "admin"],
"multusNamespace": "kube-system",
"auxiliaryCNIChainName": "cni-chain-config",
allowTryDeleteOnErr: false
}
```
@@ -63,6 +64,7 @@ message to next when some missing error. Defaults to false.
* `systemNamespaces` ([]string, optional): list of namespaces for Kubernetes system (namespaces listed here will not have `defaultNetworks` added)
* `multusNamespace` (string, optional): namespace for `clusterNetwork`/`defaultNetworks` (the default value is `kube-system`)
* `retryDeleteOnError` (bool, optional): Enable or disable delegate DEL
* [`auxiliaryCNIChainName`](#auxiliaryCNIChainName) (string, optional): Enable loading CNI configurations from disk as chained plugins in an auxiliary CNI chain
### Using `clusterNetwork`
@@ -380,3 +382,47 @@ annotations:
v1.multus-cni.io/default-network: calico-conf
...
```
### `auxiliaryCNIChainName`
`auxiliaryCNIChainName` (of value string) is used to express the name of an additional auxiliary CNI chain that will execute in order to composably execute chained CNI plugins from configurations on the host's disk in a subdirectory of the CNI configuration directory.
**NOTE**: The path used to determine the base for the subdirectory is the pathname of the `clusterNetwork` value, which must be set to a file in order to use this functionality.
When this string is set, Multus will execute an additional CNI chain, outside of the default network, on its own independent CNI chain (as to not interfere with default network functionality that might be hampered by CNI chaining and to otherwise isolate this execution) and will load CNI configurations from a subdirectory of the same name in the CNI configuration directory.
This feature is based on [improvements made to libcni for "safe subdirectory-based plugin conf loading"](https://github.com/containernetworking/cni/pull/1052).
`auxiliaryCNIChainName` is meant to be set as a CNI configuration name, this name is arbitrary but must match the subdirectory name.
Consider this [daemon configuration](https://github.com/k8snetworkplumbingwg/multus-cni/blob/master/deployments/multus-daemonset-thick.yml#L113):
```
{
"cniConfigDir": "/host/etc/cni/net.d",
"multusAutoconfigDir": "/host/etc/cni/net.d",
"multusConfigFile": "auto",
"socketDir": "/host/run/multus/",
"auxiliaryCNIChainName": "cni-chain-config"
}
```
Here we have set `"auxiliaryCNIChainName": "cni-chain-config"`, and we have expressed that our CNI configurations are on `/etc/cni/net.d/` on the host.
In this case, we would also have a directory named in `/etc/cni/net.d/cni-chain-config`
One could add any number of CNI configurations to be used as part of this chain, consider this example if we added a tuning CNI configuration called `/etc/cni/net.d/cni-chain-config/mytuning.conf` with these contents:
```
{
"name": "mytuning",
"type": "tuning",
"sysctl": {
"net.ipv4.conf.IFNAME.arp_filter": "1"
}
}
```
With the given configuration, plus this configuration, this would be executed for every pod launched by Multus CNI.
If this is unset, no auxiliary chain will be executed. However, if the default network CNI configuration is loaded from disk and is a conflist format, the libcni functionality for loading from a subdirectory will still apply.

View File

@@ -39,7 +39,7 @@ cd multus-cni
./hack/build-go.sh
```
## How do I run CI tests?
## How do I run the unit tests?
Multus has go unit tests (based on ginkgo framework).The following commands drive CI tests manually in your environment:
@@ -47,6 +47,10 @@ Multus has go unit tests (based on ginkgo framework).The following commands driv
sudo ./hack/test-go.sh
```
## How do I run the e2e tests?
Check the `README.md` in the `./e2e/` folder.
## What are the best practices for logging?
The following are the best practices for multus logging:

View File

@@ -19,13 +19,13 @@ You may acquire the Multus binary via compilation (see the [developer guide](dev
*Via Daemonset method*
As a [quickstart](quickstart.md), you may apply these YAML files (included in the clone of this repository). Run this command (typically you would run this on the master, or wherever you have access to the `kubectl` command to manage your cluster).
As a [quickstart](quickstart.md), you may apply these YAML files. Run this command (typically you would run this on the master, or wherever you have access to the `kubectl` command to manage your cluster).
cat ./deployments/multus-daemonset.yml | kubectl apply -f - # thin deployment
kubectl apply -f https://raw.githubusercontent.com/k8snetworkplumbingwg/multus-cni/master/deployments/multus-daemonset.yml # thin deployment
or
cat ./deployments/multus-daemonset-thick.yml | kubectl apply -f - # thick (client/server) deployment
kubectl apply -f https://raw.githubusercontent.com/k8snetworkplumbingwg/multus-cni/master/deployments/multus-daemonset-thick.yml # thick (client/server) deployment
If you need more comprehensive detail, continue along with this guide, otherwise, you may wish to either [follow the quickstart guide]() or skip to the ['Create network attachment definition'](#create-network-attachment-definition) section.

View File

@@ -72,6 +72,7 @@ is provided.
- `"logLevel"`: the logging level for the multus daemon logs.
- `"logToStderr"`: enable this to have the daemon multus logs echoed to stderr
as well. By default, it is disabled.
- `"auxiliaryCNIChainName"`: set a value to execute chained cni configurations from disk in an auxiliary CNI chain (see details in [configuration.md](configuration.md))
In addition, you can add any configuration which is in [configuration reference](https://github.com/k8snetworkplumbingwg/multus-cni/blob/master/docs/configuration.md#multus-cni-configuration-reference). Server configuration override multus CNI configuration (e.g. `/etc/cni/net.d/00-multus.conf`)

View File

@@ -5,7 +5,7 @@
To run the e2e test, you need the following components:
- curl
- jinjanator
- jinjanator (optional)
- docker
### How to test e2e
@@ -14,7 +14,23 @@ To run the e2e test, you need the following components:
$ git clone https://github.com/k8snetworkplumbingwg/multus-cni.git
$ cd multus-cni/e2e
$ ./get_tools.sh
```
If you have `jinjanator` you can generate the YAML with:
```
$ ./generate_yamls.sh
```
Alternatively, if you have trouble with it, use the `sed` script.
```
$ ./e2e/sed_generate_yaml.sh
```
Then, setup the cluster
```
$ ./setup_cluster.sh
$ ./test-simple-macvlan1.sh
```

View File

@@ -5,7 +5,7 @@ if [ ! -d bin ]; then
mkdir bin
fi
curl -Lo ./bin/kind "https://github.com/kubernetes-sigs/kind/releases/download/v0.22.0/kind-$(uname)-amd64"
curl -Lo ./bin/kind "https://github.com/kubernetes-sigs/kind/releases/download/v0.27.0/kind-$(uname)-amd64"
chmod +x ./bin/kind
curl -Lo ./bin/kubectl https://storage.googleapis.com/kubernetes-release/release/`curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt`/bin/linux/amd64/kubectl
chmod +x ./bin/kubectl

17
e2e/sed_generate_yaml.sh Executable file
View File

@@ -0,0 +1,17 @@
#!/bin/sh
if [ ! -d yamls ]; then
mkdir yamls
fi
# specify CNI version (default: 0.4.0)
CNI_VERSION=${CNI_VERSION:-0.4.0}
templates_dir="$(dirname $(readlink -f $0))/templates"
# generate yaml files based on templates/*.j2 to yamls directory
for i in `ls ${templates_dir}/*.j2`; do
echo "Processing $i..."
# Use sed to replace the placeholder with the CNI_VERSION variable
sed "s/{{ CNI_VERSION }}/$CNI_VERSION/g" $i > yamls/$(basename ${i%.j2})
done

View File

@@ -34,14 +34,16 @@ nodes:
nodeRegistration:
kubeletExtraArgs:
pod-manifest-path: "/etc/kubernetes/manifests/"
feature-gates: "DynamicResourceAllocation=true,KubeletPodResourcesDynamicResources=true"
feature-gates: "DynamicResourceAllocation=true,DRAResourceClaimDeviceStatus=true,KubeletPodResourcesDynamicResources=true"
- role: worker
# Required by DRA Integration
##
featureGates:
DynamicResourceAllocation: true
DRAResourceClaimDeviceStatus: true
KubeletPodResourcesDynamicResources: true
runtimeConfig:
"api/alpha": "true"
"api/beta": "true"
containerdConfigPatches:
# Enable CDI as described in
# https://github.com/container-orchestrated-devices/container-device-interface#containerd-configuration

View File

@@ -170,9 +170,9 @@ spec:
- name: install-multus-shim
image: localhost:5000/multus:e2e
command:
- "cp"
- "/usr/src/multus-cni/bin/multus-shim"
- "/host/opt/cni/bin/multus-shim"
- "sh"
- "-c"
- "cp /usr/src/multus-cni/bin/multus-shim /host/opt/cni/bin/multus-shim && cp /usr/src/multus-cni/bin/passthru /host/opt/cni/bin/passthru"
resources:
requests:
cpu: "10m"

View File

@@ -0,0 +1,26 @@
---
kind: ConfigMap
apiVersion: v1
metadata:
name: multus-daemon-config
namespace: kube-system
labels:
tier: node
app: multus
data:
daemon-config.json: |
{
"confDir": "/host/etc/cni/net.d",
"logToStderr": true,
"logLevel": "debug",
"logFile": "/tmp/multus.log",
"binDir": "/host/opt/cni/bin",
"cniDir": "/var/lib/cni/multus",
"socketDir": "/host/run/multus",
"cniVersion": "{{ CNI_VERSION }}",
"cniConfigDir": "/host/etc/cni/net.d",
"multusConfigFile": "auto",
"forceCNIVersion": true,
"multusAutoconfigDir": "/host/etc/cni/net.d",
"auxiliaryCNIChainName": "vendor-cni-chain"
}

View File

@@ -0,0 +1,94 @@
---
apiVersion: v1
kind: ConfigMap
metadata:
name: cni-setup-script
namespace: default
data:
setup.sh: |
#!/bin/bash
set -euxo pipefail
DEFAULT_NETWORK_CNI_NAME="vendor-cni-chain"
cleanup() {
echo "Cleaning up..."
rm -f /host/etc/cni/net.d/${DEFAULT_NETWORK_CNI_NAME}/sysctltwiddle.conf
if [ $? -ne 0 ]; then
echo "Failed to remove sysctltwiddle.conf" >&2
exit 1
fi
echo "Cleanup completed successfully"
}
trap cleanup EXIT
# Create the chained CNI directory if it doesn't exist
mkdir -p /host/etc/cni/net.d/${DEFAULT_NETWORK_CNI_NAME}
if [ $? -ne 0 ]; then
echo "Failed to create directory /host/etc/cni/net.d/${DEFAULT_NETWORK_CNI_NAME}" >&2
exit 1
fi
# Write the chained tuning CNI config
cat <<EOF > /host/etc/cni/net.d/${DEFAULT_NETWORK_CNI_NAME}/sysctltwiddle.conf
{
"cniVersion": "{{ CNI_VERSION }}",
"name": "sysctltwiddle",
"type": "tuning",
"sysctl": {
"net.ipv4.conf.eth0.arp_filter": "1"
}
}
EOF
if [ $? -ne 0 ]; then
echo "Failed to create chained CNI config" >&2
exit 1
fi
echo "CNI chained setup completed successfully."
sleep infinity
---
apiVersion: apps/v1
kind: DaemonSet
metadata:
name: cni-setup-daemonset
namespace: default
labels:
app: cni-setup
spec:
selector:
matchLabels:
app: cni-setup
template:
metadata:
labels:
app: cni-setup
spec:
tolerations:
- operator: Exists
effect: NoSchedule
- operator: Exists
effect: NoExecute
containers:
- name: setup
image: quay.io/fedora/fedora:40
securityContext:
privileged: true
volumeMounts:
- name: cni-config
mountPath: /host/etc/cni/net.d
- name: script-volume
mountPath: /scripts
command: ["/bin/bash", "/scripts/setup.sh"]
volumes:
- name: cni-config
hostPath:
path: /etc/cni/net.d
type: Directory
- name: script-volume
configMap:
name: cni-setup-script
items:
- key: setup.sh
path: setup.sh

View File

@@ -0,0 +1,11 @@
apiVersion: v1
kind: Pod
metadata:
name: sysctl-modified
spec:
containers:
- name: sysctl
image: quay.io/dosmith/fedora-procps
command: ["/bin/bash", "-c", "trap : TERM INT; sleep infinity & wait"]
securityContext:
privileged: true

View File

@@ -0,0 +1,95 @@
---
apiVersion: v1
kind: ConfigMap
metadata:
name: cni-setup-script
namespace: default
data:
setup.sh: |
#!/bin/bash
set -euxo pipefail
DEFAULT_NETWORK_CNI_NAME="kindnet"
cleanup() {
echo "Cleaning up..."
rm -f /host/etc/cni/net.d/${DEFAULT_NETWORK_CNI_NAME}/sysctltwiddle.conf
if [ $? -ne 0 ]; then
echo "Failed to remove sysctltwiddle.conf" >&2
exit 1
fi
echo "Cleanup completed successfully"
}
trap cleanup EXIT
# Create the chained CNI directory if it doesn't exist
mkdir -p /host/etc/cni/net.d/${DEFAULT_NETWORK_CNI_NAME}
if [ $? -ne 0 ]; then
echo "Failed to create directory /host/etc/cni/net.d/${DEFAULT_NETWORK_CNI_NAME}" >&2
exit 1
fi
# Write the chained tuning CNI config
cat <<EOF > /host/etc/cni/net.d/${DEFAULT_NETWORK_CNI_NAME}/sysctltwiddle.conf
{
"cniVersion": "{{ CNI_VERSION }}",
"name": "sysctltwiddle",
"type": "tuning",
"sysctl": {
"net.ipv4.conf.IFNAME.arp_filter": "1"
}
}
EOF
if [ $? -ne 0 ]; then
echo "Failed to create chained CNI config" >&2
exit 1
fi
echo "CNI chained setup completed successfully."
sleep infinity
---
apiVersion: apps/v1
kind: DaemonSet
metadata:
name: cni-setup-daemonset
namespace: default
labels:
app: cni-setup
spec:
selector:
matchLabels:
app: cni-setup
template:
metadata:
labels:
app: cni-setup
spec:
hostNetwork: true
tolerations:
- operator: Exists
effect: NoSchedule
- operator: Exists
effect: NoExecute
containers:
- name: setup
image: quay.io/fedora/fedora:40
securityContext:
privileged: true
volumeMounts:
- name: cni-config
mountPath: /host/etc/cni/net.d
- name: script-volume
mountPath: /scripts
command: ["/bin/bash", "/scripts/setup.sh"]
volumes:
- name: cni-config
hostPath:
path: /etc/cni/net.d
type: Directory
- name: script-volume
configMap:
name: cni-setup-script
items:
- key: setup.sh
path: setup.sh

View File

@@ -0,0 +1,81 @@
#!/bin/bash
set -o errexit
set -o nounset
set -o pipefail
export PATH=${PATH}:./bin
TEST_POD_NAME="sysctl-modified"
EXPECTED_BINARIES="${EXPECTED_BINARIES:-/opt/cni/bin/ptp /opt/cni/bin/portmap /opt/cni/bin/tuning}"
EXPECTED_CNI_DIR="/etc/cni/net.d"
# Reconfigure multus
echo "Applying subdirectory chain passthru config..."
kubectl apply -f yamls/subdirectory-chain-passthru-configupdate.yml
# Restart the multus daemonset to pick up the new config
echo "Restarting Multus DaemonSet..."
kubectl rollout restart daemonset kube-multus-ds-amd64 -n kube-system
kubectl rollout status daemonset/kube-multus-ds-amd64 -n kube-system
# Debug: show CNI configs and binaries inside each Kind node
echo "Checking CNI configs and binaries on nodes..."
for node in $(kubectl get nodes --no-headers | awk '{print $1}'); do
container_name=$(docker ps --format '{{.Names}}' | grep "^${node}$")
echo "------"
echo "Node: ${node} (container: ${container_name})"
echo "Listing /opt/cni/bin contents..."
docker exec "${container_name}" ls -l /opt/cni/bin || echo "WARNING: /opt/cni/bin missing!"
echo "Checking expected binaries..."
for bin in $EXPECTED_BINARIES; do
echo "Checking for ${bin}..."
if docker exec "${container_name}" test -f "${bin}"; then
echo "SUCCESS: ${bin} found."
else
echo "FAIL: ${bin} NOT found!"
fi
done
echo "Listing /etc/cni/net.d configs..."
docker exec "${container_name}" ls -l ${EXPECTED_CNI_DIR} || echo "WARNING: ${EXPECTED_CNI_DIR} missing!"
done
echo "------"
# Deploy the daemonset that will lay down the chained CNI config
echo "Applying CNI setup DaemonSet..."
kubectl apply -f yamls/subdirectory-chaining-passthru.yml
# Wait for the daemonset pods to be ready (make sure they set up CNI config)
echo "Waiting for CNI setup DaemonSet to be Ready..."
kubectl rollout status daemonset/cni-setup-daemonset --timeout=300s
# Deploy a test pod that will get chained CNI applied
echo "Applying test pod..."
kubectl apply -f yamls/subdirectory-chaining-pod.yml
# Wait for the pod to be Ready
echo "Waiting for test pod to be Ready..."
kubectl wait --for=condition=ready pod/${TEST_POD_NAME} --timeout=300s
# Check that the sysctl got set
echo "Verifying sysctl arp_filter is set to 1 on eth0..."
SYSCTL_VALUE=$(kubectl exec ${TEST_POD_NAME} -- sysctl -n net.ipv4.conf.eth0.arp_filter)
if [ "$SYSCTL_VALUE" != "1" ]; then
echo "FAIL: net.ipv4.conf.eth0.arp_filter is not set to 1, got ${SYSCTL_VALUE}" >&2
exit 1
else
echo "SUCCESS: net.ipv4.conf.eth0.arp_filter is set correctly."
fi
# Cleanup
echo "Cleaning up test resources..."
kubectl delete -f yamls/subdirectory-chaining-pod.yml
kubectl delete -f yamls/subdirectory-chaining-passthru.yml
echo "Test completed successfully."
exit 0

View File

@@ -0,0 +1,37 @@
#!/bin/sh
set -o errexit
export PATH=${PATH}:./bin
TEST_POD_NAME="sysctl-modified"
# Deploy the daemonset that will lay down the chained CNI config
kubectl apply -f yamls/subdirectory-chaining.yml
# Wait for the daemonset pods to be ready (we need the config to be laid down)
kubectl rollout status daemonset/cni-setup-daemonset
# Deploy a test pod that will get chained CNI applied
kubectl apply -f yamls/subdirectory-chaining-pod.yml
# Wait for the pod to be Ready
kubectl wait --for=condition=ready pod/sysctl-modified --timeout=300s
# Check that the sysctl got set properly inside the pod's eth0 interface
echo "Verifying sysctl arp_filter is set to 1 on eth0"
SYSCTL_VALUE=$(kubectl exec sysctl-modified -- sysctl -n net.ipv4.conf.eth0.arp_filter)
if [ "$SYSCTL_VALUE" != "1" ]; then
echo "FAIL: net.ipv4.conf.eth0.arp_filter is not set to 1, got ${SYSCTL_VALUE}" >&2
exit 1
else
echo "SUCCESS: net.ipv4.conf.eth0.arp_filter is set correctly."
fi
# 6. Clean up
echo "Cleaning up test resources"
kubectl delete -f yamls/subdirectory-chaining-pod.yml
kubectl delete -f yamls/subdirectory-chaining.yml
exit 0

76
go.mod
View File

@@ -4,53 +4,46 @@ go 1.21
require (
github.com/blang/semver v3.5.1+incompatible
github.com/containernetworking/cni v1.2.0-rc1
github.com/containernetworking/cni v1.3.0
github.com/containernetworking/plugins v1.1.0
github.com/fsnotify/fsnotify v1.6.0
github.com/go-logr/logr v1.3.0 // indirect
github.com/k8snetworkplumbingwg/network-attachment-definition-client v1.7.1
github.com/onsi/ginkgo/v2 v2.13.2
github.com/onsi/gomega v1.30.0
github.com/pkg/errors v0.9.1 // indirect
github.com/k8snetworkplumbingwg/network-attachment-definition-client v1.7.6
github.com/onsi/ginkgo/v2 v2.20.1
github.com/onsi/gomega v1.34.1
github.com/prometheus/client_golang v1.16.0
github.com/spf13/pflag v1.0.5
github.com/vishvananda/netlink v1.1.1-0.20210330154013-f5de75959ad5
golang.org/x/net v0.23.0
golang.org/x/sys v0.18.0
golang.org/x/net v0.28.0
golang.org/x/sys v0.23.0
google.golang.org/grpc v1.58.3
gopkg.in/natefinch/lumberjack.v2 v2.0.0
k8s.io/api v0.29.0
k8s.io/apimachinery v0.29.0
k8s.io/client-go v1.5.2
k8s.io/client-go v0.29.0
k8s.io/klog v1.0.0
k8s.io/klog/v2 v2.110.1
k8s.io/kube-openapi v0.0.0-20231010175941-2dd684a91f00 // indirect
k8s.io/kubelet v0.27.5
sigs.k8s.io/yaml v1.3.0 // indirect
)
require (
github.com/prometheus/client_golang v1.16.0
github.com/spf13/pflag v1.0.5
)
require (
github.com/Masterminds/semver/v3 v3.2.1 // indirect
github.com/BurntSushi/toml v0.3.1 // indirect
github.com/beorn7/perks v1.0.1 // indirect
github.com/cespare/xxhash/v2 v2.2.0 // indirect
github.com/davecgh/go-spew v1.1.1 // indirect
github.com/emicklei/go-restful/v3 v3.11.0 // indirect
github.com/evanphx/json-patch v4.12.0+incompatible // indirect
github.com/go-logr/logr v1.4.2 // indirect
github.com/go-openapi/jsonpointer v0.19.6 // indirect
github.com/go-openapi/jsonreference v0.20.2 // indirect
github.com/go-openapi/swag v0.22.3 // indirect
github.com/go-task/slim-sprig v0.0.0-20230315185526-52ccab3ef572 // indirect
github.com/go-task/slim-sprig/v3 v3.0.0 // indirect
github.com/gogo/protobuf v1.3.2 // indirect
github.com/golang/groupcache v0.0.0-20210331224755-41bb18bfe9da // indirect
github.com/golang/protobuf v1.5.3 // indirect
github.com/google/gnostic v0.7.0 // indirect
github.com/google/gnostic-models v0.6.9-0.20230804172637-c7be7c783f49 // indirect
github.com/google/go-cmp v0.6.0 // indirect
github.com/google/gofuzz v1.2.0 // indirect
github.com/google/pprof v0.0.0-20210720184732-4bb14d4b1be1 // indirect
github.com/google/pprof v0.0.0-20240727154555-813a5fbdbec8 // indirect
github.com/google/uuid v1.3.0 // indirect
github.com/imdario/mergo v0.3.11 // indirect
github.com/josharian/intern v1.0.0 // indirect
@@ -60,53 +53,26 @@ require (
github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd // indirect
github.com/modern-go/reflect2 v1.0.2 // indirect
github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822 // indirect
github.com/pkg/errors v0.9.1 // indirect
github.com/prometheus/client_model v0.4.0 // indirect
github.com/prometheus/common v0.44.0 // indirect
github.com/prometheus/procfs v0.10.1 // indirect
github.com/vishvananda/netns v0.0.4 // indirect
golang.org/x/exp v0.0.0-20240719175910-8a7402abbf56 // indirect
golang.org/x/oauth2 v0.10.0 // indirect
golang.org/x/term v0.18.0 // indirect
golang.org/x/text v0.14.0 // indirect
golang.org/x/term v0.23.0 // indirect
golang.org/x/text v0.17.0 // indirect
golang.org/x/time v0.3.0 // indirect
golang.org/x/tools v0.14.0 // indirect
golang.org/x/tools v0.24.0 // indirect
google.golang.org/appengine v1.6.7 // indirect
google.golang.org/genproto/googleapis/rpc v0.0.0-20230822172742-b8732ec3820d // indirect
google.golang.org/protobuf v1.33.0 // indirect
google.golang.org/protobuf v1.34.1 // indirect
gopkg.in/inf.v0 v0.9.1 // indirect
gopkg.in/yaml.v2 v2.4.0 // indirect
gopkg.in/yaml.v3 v3.0.1 // indirect
k8s.io/kube-openapi v0.0.0-20231010175941-2dd684a91f00 // indirect
k8s.io/utils v0.0.0-20230726121419-3b25d923346b // indirect
sigs.k8s.io/json v0.0.0-20221116044647-bc3834ca7abd // indirect
sigs.k8s.io/structured-merge-diff/v4 v4.4.1 // indirect
)
replace (
github.com/gogo/protobuf => github.com/gogo/protobuf v1.3.2
k8s.io/api => k8s.io/api v0.29.0
k8s.io/apiextensions-apiserver => k8s.io/apiextensions-apiserver v0.29.0
k8s.io/apimachinery => k8s.io/apimachinery v0.29.0
k8s.io/apiserver => k8s.io/apiserver v0.29.0
k8s.io/cli-runtime => k8s.io/cli-runtime v0.29.0
k8s.io/client-go => k8s.io/client-go v0.29.0
k8s.io/cloud-provider => k8s.io/cloud-provider v0.29.0
k8s.io/cluster-bootstrap => k8s.io/cluster-bootstrap v0.29.0
k8s.io/code-generator => k8s.io/code-generator v0.29.0
k8s.io/component-base => k8s.io/component-base v0.29.0
k8s.io/component-helpers => k8s.io/component-helpers v0.29.0
k8s.io/controller-manager => k8s.io/controller-manager v0.29.0
k8s.io/cri-api => k8s.io/cri-api v0.29.0
k8s.io/csi-translation-lib => k8s.io/csi-translation-lib v0.29.0
k8s.io/kube-aggregator => k8s.io/kube-aggregator v0.29.0
k8s.io/kube-controller-manager => k8s.io/kube-controller-manager v0.29.0
k8s.io/kube-openapi => k8s.io/kube-openapi v0.0.0-20230501164219-8b0f38b5fd1f
k8s.io/kube-proxy => k8s.io/kube-proxy v0.29.0
k8s.io/kube-scheduler => k8s.io/kube-scheduler v0.29.0
k8s.io/kubectl => k8s.io/kubectl v0.29.0
k8s.io/kubelet => k8s.io/kubelet v0.29.0
k8s.io/kubernetes => k8s.io/kubernetes v1.29.0
k8s.io/legacy-cloud-providers => k8s.io/legacy-cloud-providers v0.29.0
k8s.io/metrics => k8s.io/metrics v0.29.0
k8s.io/mount-utils => k8s.io/mount-utils v0.29.0
k8s.io/pod-security-admission => k8s.io/pod-security-admission v0.29.0
k8s.io/sample-apiserver => k8s.io/sample-apiserver v0.29.0
sigs.k8s.io/yaml v1.3.0 // indirect
)

1512
go.sum

File diff suppressed because it is too large Load Diff

View File

@@ -80,3 +80,5 @@ echo "Building kubeconfig_generator"
go build -o "${DEST_DIR}"/kubeconfig_generator ${BUILD_ARGS} -ldflags "${LDFLAGS}" ./cmd/kubeconfig_generator
echo "Building cert-approver"
go build -o "${DEST_DIR}"/cert-approver ${BUILD_ARGS} -ldflags "${LDFLAGS}" ./cmd/cert-approver
echo "Building passthru CNI"
go build -o "${DEST_DIR}"/passthru ${BUILD_ARGS} -ldflags "${LDFLAGS}" ./cmd/passthru-cni

View File

@@ -1,5 +1,5 @@
# This Dockerfile is used to build the image available on DockerHub
FROM --platform=$BUILDPLATFORM golang:1.21 as build
FROM --platform=$BUILDPLATFORM golang:1.23 as build
# Add everything
ADD . /usr/src/multus-cni
@@ -8,7 +8,7 @@ ARG TARGETPLATFORM
RUN cd /usr/src/multus-cni && \
./hack/build-go.sh
FROM gcr.io/distroless/base-debian11:latest
FROM gcr.io/distroless/base-debian12:latest
LABEL org.opencontainers.image.source https://github.com/k8snetworkplumbingwg/multus-cni
COPY --from=build /usr/src/multus-cni/bin /usr/src/multus-cni/bin
COPY --from=build /usr/src/multus-cni/LICENSE /usr/src/multus-cni/LICENSE
@@ -18,4 +18,5 @@ COPY --from=build /usr/src/multus-cni/bin/install_multus /
COPY --from=build /usr/src/multus-cni/bin/thin_entrypoint /
COPY --from=build /usr/src/multus-cni/bin/kubeconfig_generator /
COPY --from=build /usr/src/multus-cni/bin/cert-approver /
CMD ["/thin_entrypoint"]
ENTRYPOINT ["/thin_entrypoint"]

View File

@@ -1,5 +1,5 @@
# This Dockerfile is used to build the image available on DockerHub
FROM --platform=$BUILDPLATFORM golang:1.21 as build
FROM --platform=$BUILDPLATFORM golang:1.23 as build
# Add everything
ADD . /usr/src/multus-cni
@@ -8,7 +8,7 @@ ARG TARGETPLATFORM
RUN cd /usr/src/multus-cni && \
./hack/build-go.sh
FROM gcr.io/distroless/base-debian11:debug
FROM gcr.io/distroless/base-debian12:debug
LABEL org.opencontainers.image.source https://github.com/k8snetworkplumbingwg/multus-cni
COPY --from=build /usr/src/multus-cni/bin /usr/src/multus-cni/bin
COPY --from=build /usr/src/multus-cni/LICENSE /usr/src/multus-cni/LICENSE
@@ -18,4 +18,5 @@ COPY --from=build /usr/src/multus-cni/bin/install_multus /
COPY --from=build /usr/src/multus-cni/bin/thin_entrypoint /
COPY --from=build /usr/src/multus-cni/bin/kubeconfig_generator /
COPY --from=build /usr/src/multus-cni/bin/cert-approver /
CMD ["/thin_entrypoint"]
ENTRYPOINT ["/thin_entrypoint"]

View File

@@ -1,14 +1,15 @@
# This Dockerfile is used to build the image available on DockerHub
FROM golang:1.21 as build
FROM --platform=$BUILDPLATFORM golang:1.23 as build
# Add everything
ADD . /usr/src/multus-cni
ARG TARGETPLATFORM
RUN cd /usr/src/multus-cni && \
./hack/build-go.sh
FROM debian:stable-slim
LABEL org.opencontainers.image.source https://github.com/k8snetworkplumbingwg/multus-cni
LABEL org.opencontainers.image.source=https://github.com/k8snetworkplumbingwg/multus-cni
COPY --from=build /usr/src/multus-cni/bin /usr/src/multus-cni/bin
COPY --from=build /usr/src/multus-cni/LICENSE /usr/src/multus-cni/LICENSE
COPY --from=build /usr/src/multus-cni/bin/cert-approver /

View File

@@ -18,9 +18,11 @@ package k8sclient
import (
"context"
"encoding/json"
"errors"
"fmt"
"net"
"os"
"path/filepath"
"regexp"
"strings"
"syscall"
@@ -82,6 +84,20 @@ func (c *ClientInfo) GetPod(namespace, name string) (*v1.Pod, error) {
return c.Client.CoreV1().Pods(namespace).Get(context.TODO(), name, metav1.GetOptions{})
}
// GetPodContext gets pod from kubernetes with context
func (c *ClientInfo) GetPodContext(ctx context.Context, namespace, name string) (*v1.Pod, error) {
if c.PodInformer != nil {
logging.Debugf("GetPod for [%s/%s] will use informer cache", namespace, name)
return listers.NewPodLister(c.PodInformer.GetIndexer()).Pods(namespace).Get(name)
}
return c.Client.CoreV1().Pods(namespace).Get(ctx, name, metav1.GetOptions{})
}
// GetPodAPILiveQuery does a live API query for the pod, instead of using informers, for cases when a failure occurred, as to prevent a cache miss.
func (c *ClientInfo) GetPodAPILiveQuery(ctx context.Context, namespace, name string) (*v1.Pod, error) {
return c.Client.CoreV1().Pods(namespace).Get(ctx, name, metav1.GetOptions{})
}
// DeletePod deletes a pod from kubernetes
func (c *ClientInfo) DeletePod(namespace, name string) error {
return c.Client.CoreV1().Pods(namespace).Delete(context.TODO(), name, metav1.DeleteOptions{})
@@ -288,7 +304,7 @@ func getKubernetesDelegate(client *ClientInfo, net *types.NetworkSelectionElemen
// Get resourceName annotation from NetworkAttachmentDefinition
deviceID := ""
resourceName, ok := customResource.GetAnnotations()[resourceNameAnnot]
if ok && pod.Name != "" && pod.Namespace != "" {
if ok && pod != nil && pod.Name != "" && pod.Namespace != "" {
// ResourceName annotation is found; try to get device info from resourceMap
logging.Debugf("getKubernetesDelegate: found resourceName annotation : %s", resourceName)
@@ -524,31 +540,119 @@ func getNetDelegate(client *ClientInfo, pod *v1.Pod, netname, confdir, namespace
} else {
// option4) if file path (absolute), then load it directly
if strings.HasSuffix(netname, ".conflist") {
confList, err := libcni.ConfListFromFile(netname)
confList, err := LoadChainedPluginsFromFile(netname)
if err != nil {
return nil, resourceMap, logging.Errorf("error loading CNI conflist file %s: %v", netname, err)
}
configBytes = confList.Bytes
} else {
conf, err := libcni.ConfFromFile(netname)
delegate, err := types.LoadDelegateNetConfFromConfList(confList, nil, "", "")
if err != nil {
return nil, resourceMap, logging.Errorf("error loading CNI config file %s: %v", netname, err)
return nil, resourceMap, err
}
if conf.Network.Type == "" {
return nil, resourceMap, logging.Errorf("error loading CNI config file %s: no 'type'; perhaps this is a .conflist?", netname)
}
configBytes = conf.Bytes
return delegate, resourceMap, nil
}
delegate, err := types.LoadDelegateNetConf(configBytes, nil, "", "")
// Or it's not a conflist...
// after libcni v1.2.3 there's no support support this old-school method with non-conflists.
// this method doesn't check if there's a 0 length plugins field, that is.
conf, err := libcni.ConfFromFile(netname)
if err != nil {
return nil, resourceMap, logging.Errorf("error loading CNI config file %s: %v", netname, err)
}
if conf.Network.Type == "" {
return nil, resourceMap, logging.Errorf("error loading CNI config file %s: no 'type'; perhaps this is supposed to be a .conflist?", netname)
}
delegate, err := types.LoadDelegateNetConf(conf.Bytes, nil, "", "")
if err != nil {
return nil, resourceMap, err
}
return delegate, resourceMap, nil
}
}
return nil, resourceMap, logging.Errorf("getNetDelegate: cannot find network: %v", netname)
}
func loadSubdirectoryChain(bytes []byte, cniconfdir string) (*libcni.NetworkConfigList, error) {
// Load the network configuration from the byte array
conf, err := libcni.NetworkConfFromBytes(bytes)
if err != nil {
return nil, fmt.Errorf("error loading network config from bytes: %v", err)
}
// Check if plugins need to be loaded from files
if !conf.LoadOnlyInlinedPlugins && cniconfdir != "" {
// Let's validate that conf.Name
// From the CNI spec:
// > Must start with an alphanumeric character, optionally followed by any combination of one or more alphanumeric characters,
// > underscore, dot (.) or hyphen (-). Must not contain characters disallowed in file paths.
if !regexp.MustCompile(`^[a-zA-Z0-9][a-zA-Z0-9_.-]*$`).MatchString(conf.Name) {
return nil, fmt.Errorf("invalid network config name: %s", conf.Name)
}
plugins, err := libcni.NetworkPluginConfsFromFiles(cniconfdir, conf.Name)
if err != nil {
return nil, fmt.Errorf("error loading plugin configs: %v", err)
}
conf.Plugins = append(conf.Plugins, plugins...)
}
if len(conf.Plugins) == 0 {
return nil, fmt.Errorf("no plugin configs found")
}
return conf, nil
}
// LoadChainedDelegatesFromBytes loads a CNI configuration byte array and returns a DelegateNetConf with the chain added.
func LoadChainedDelegatesFromBytes(bytes []byte, cniconfdir string) *types.DelegateNetConf {
conf, err := loadSubdirectoryChain(bytes, cniconfdir)
if err != nil {
logging.Errorf("LoadChainedDelegatesFromBytes: %v", err)
return nil
}
// Create and return a DelegateNetConf from the configuration list
delegate, err := types.LoadDelegateNetConfFromConfList(conf, nil, "", "")
if err != nil {
logging.Errorf("LoadChainedDelegatesFromBytes: error loading delegate network config: %v", err)
return nil
}
return delegate
}
// LoadChainedPluginsFromFile loads a CNI configuration file and returns the NetworkConfigList
func LoadChainedPluginsFromFile(filename string) (*libcni.NetworkConfigList, error) {
cleanPath := filepath.Clean(filename)
// stat the file to make sure it's a normal file.
info, err := os.Stat(cleanPath)
if err != nil {
return nil, err
}
if !info.Mode().IsRegular() {
return nil, errors.New("CNI configuration path is not a regular file")
}
bytes, err := os.ReadFile(cleanPath)
if err != nil {
return nil, fmt.Errorf("error reading %s: %w", filename, err)
}
logging.Debugf("LoadChainedPluginsFromFile: %s", filename)
conf, err := loadSubdirectoryChain(bytes, filepath.Dir(filename))
if err != nil {
return nil, err
}
logging.Debugf("Loaded SubdirectoryChain: %+v", conf)
return conf, nil
}
// GetDefaultNetworks parses 'defaultNetwork' config, gets network json and put it into netconf.Delegates.
func GetDefaultNetworks(pod *v1.Pod, conf *types.NetConf, kubeClient *ClientInfo, resourceMap map[string]*types.ResourceInfo) (map[string]*types.ResourceInfo, error) {
logging.Debugf("GetDefaultNetworks: %v, %v, %v, %v", pod, conf, kubeClient, resourceMap)
@@ -575,7 +679,7 @@ func GetDefaultNetworks(pod *v1.Pod, conf *types.NetConf, kubeClient *ClientInfo
delegates = append(delegates, delegate)
// Pod in kube-system namespace does not have default network for now.
if !types.CheckSystemNamespaces(pod.ObjectMeta.Namespace, conf.SystemNamespaces) {
if pod != nil && !types.CheckSystemNamespaces(pod.ObjectMeta.Namespace, conf.SystemNamespaces) {
for _, netname := range conf.DefaultNetworks {
delegate, resourceMap, err := getNetDelegate(kubeClient, pod, netname, conf.ConfDir, conf.MultusNamespace, resourceMap)
if err != nil {

View File

@@ -52,11 +52,12 @@ const (
)
var (
version = "master@git"
commit = "unknown commit"
date = "unknown date"
gitTreeState = ""
releaseStatus = ""
version = "master@git"
commit = "unknown commit"
date = "unknown date"
gitTreeState = ""
releaseStatus = ""
errPodNotFound = fmt.Errorf("pod not found during Multus GetPod")
)
// PrintVersionString ...
@@ -131,15 +132,49 @@ func saveDelegates(containerID, dataDir string, delegates []*types.DelegateNetCo
return err
}
func deleteDelegates(containerID, dataDir string) error {
logging.Debugf("deleteDelegates: %s, %s", containerID, dataDir)
path := filepath.Join(dataDir, containerID)
if err := os.Remove(path); err != nil {
return logging.Errorf("deleteDelegates: error in deleting the delegates : %v", err)
func getValidAttachmentFromCache(b []byte) (string, string, error) {
type simpleCacheV1 struct {
Kind string `json:"kind"`
ContainerID string `json:"containerId"`
IfName string `json:"ifName"`
}
return nil
cache := &simpleCacheV1{}
if err := json.Unmarshal(b, cache); err != nil {
return "", "", fmt.Errorf("getValidAttachmentFromCache: invalid json: %v", err)
}
if cache.ContainerID == "" || cache.IfName == "" {
return "", "", fmt.Errorf("invalid cache: containerID:%q, ifName:%q", cache.ContainerID, cache.IfName)
}
return cache.ContainerID, cache.IfName, nil
}
func gatherValidAttachmentsFromCache(cniDir string) ([]cnitypes.GCAttachment, error) {
cacheDir := filepath.Join(cniDir, "results")
dirEntries, err := os.ReadDir(cacheDir)
if err != nil {
return nil, err
}
allAttachments := []cnitypes.GCAttachment{}
for _, dirEnt := range dirEntries {
path := filepath.Join(cacheDir, dirEnt.Name())
delegatesBytes, err := os.ReadFile(path)
// if delegates cannot read that, skipped for now (because cannot recover).
if err != nil {
logging.Errorf("gatherSavedDelegates: cannot read %q, skipped to add", path)
continue
}
containerID, ifName, err := getValidAttachmentFromCache(delegatesBytes)
if err != nil {
logging.Errorf("gatherSavedDelegates: cannot read cache, skipped to add: %v", err)
continue
}
allAttachments = append(allAttachments, cnitypes.GCAttachment{ContainerID: containerID, IfName: ifName})
}
return allAttachments, nil
}
func validateIfName(nsname string, ifname string) error {
@@ -223,16 +258,25 @@ func confDel(rt *libcni.RuntimeConf, rawNetconf []byte, multusNetconf *types.Net
return err
}
func conflistAdd(rt *libcni.RuntimeConf, rawnetconflist []byte, multusNetconf *types.NetConf, exec invoke.Exec) (cnitypes.Result, error) {
func conflistAdd(rt *libcni.RuntimeConf, rawnetconflist []byte, cniConfList *libcni.NetworkConfigList, multusNetconf *types.NetConf, exec invoke.Exec) (cnitypes.Result, error) {
logging.Debugf("conflistAdd: %v, %s", rt, string(rawnetconflist))
// In part, adapted from K8s pkg/kubelet/dockershim/network/cni/cni.go
binDirs := filepath.SplitList(os.Getenv("CNI_PATH"))
binDirs = append([]string{multusNetconf.BinDir}, binDirs...)
cniNet := libcni.NewCNIConfigWithCacheDir(binDirs, multusNetconf.CNIDir, exec)
confList, err := libcni.ConfListFromBytes(rawnetconflist)
if err != nil {
return nil, logging.Errorf("conflistAdd: error converting the raw bytes into a conflist: %v", err)
var confList *libcni.NetworkConfigList
var err error
// This may wind up being set during parsing the default network config.
// In this case -- we'll use it as passed. Otherwise, we'll recalculate it.
if len(cniConfList.Plugins) > 0 {
confList = cniConfList
} else {
confList, err = libcni.NetworkConfFromBytes(rawnetconflist)
if err != nil {
return nil, logging.Errorf("conflistAdd: error converting the raw bytes into a conflist: %v", err)
}
}
result, err := cniNet.AddNetworkList(context.Background(), confList, rt)
@@ -326,7 +370,8 @@ func DelegateAdd(exec invoke.Exec, kubeClient *k8s.ClientInfo, pod *v1.Pod, dele
var result cnitypes.Result
var err error
if delegate.ConfListPlugin {
result, err = conflistAdd(rt, delegate.Bytes, multusNetconf, exec)
// TODO: why are we passing bytes here? don't we have a better representation of it?
result, err = conflistAdd(rt, delegate.Bytes, &delegate.CNINetworkConfigList, multusNetconf, exec)
if err != nil {
return nil, err
}
@@ -542,21 +587,33 @@ func GetPod(kubeClient *k8s.ClientInfo, k8sArgs *types.K8sArgs, isDel bool) (*v1
var pod *v1.Pod
if err := wait.PollImmediate(pollDuration, shortPollTimeout, func() (bool, error) {
var getErr error
pod, getErr = kubeClient.GetPod(podNamespace, podName)
// Use context with a short timeout so the call to API server doesn't take too long.
ctx, cancel := context.WithTimeout(context.TODO(), pollDuration)
defer cancel()
pod, getErr = kubeClient.GetPodContext(ctx, podNamespace, podName)
if isCriticalRequestRetriable(getErr) || retryOnNotFound(getErr) {
return false, nil
}
return pod != nil, getErr
}); err != nil {
if isDel && errors.IsNotFound(err) {
// On DEL pod may already be gone from apiserver/informer
return nil, nil
if errors.IsNotFound(err) {
// When pods are not found, this is "OK", it's a known condition for rapidly deleted pods, we'll just warn on it.
if !isDel {
logging.Verbosef("Warning: GetPod for [%s/%s] resulted in pod not found during CNI ADD (pod may have already been deleted): %v", podNamespace, podName, err)
}
return nil, errPodNotFound
}
// Try one more time to get the pod directly from the apiserver;
// TODO: figure out why static pods don't show up via the informer
// and always hit this case.
pod, err = kubeClient.GetPod(podNamespace, podName)
ctx, cancel := context.WithTimeout(context.TODO(), pollDuration)
defer cancel()
pod, err = kubeClient.GetPodAPILiveQuery(ctx, podNamespace, podName)
if err != nil {
if errors.IsNotFound(err) {
logging.Verbosef("Warning: On live query retry, [%s/%s] pod not found during CNI ADD (pod may have already been deleted): %v", podNamespace, podName, err)
return nil, errPodNotFound
}
return nil, cmdErr(k8sArgs, "error waiting for pod: %v", err)
}
}
@@ -602,6 +659,11 @@ func CmdAdd(args *skel.CmdArgs, exec invoke.Exec, kubeClient *k8s.ClientInfo) (c
pod, err := GetPod(kubeClient, k8sArgs, false)
if err != nil {
if err == errPodNotFound {
emptyresult := emptyCNIResult(args, "1.0.0")
logging.Verbosef("CmdAdd: Warning: pod [%s/%s] not found, exiting with empty CNI result: %v", k8sArgs.K8S_POD_NAMESPACE, k8sArgs.K8S_POD_NAME, emptyresult)
return emptyresult, nil
}
return nil, err
}
@@ -623,6 +685,36 @@ func CmdAdd(args *skel.CmdArgs, exec invoke.Exec, kubeClient *k8s.ClientInfo) (c
return nil, cmdErr(k8sArgs, "error loading k8s delegates k8s args: %v", err)
}
// we add to the auxiliary CNI chain here.
if n.AuxiliaryCNIChainName != "" {
logging.Debugf("Using AuxiliaryCNIChainName: %v", n.AuxiliaryCNIChainName)
// create an passthru cni conflist configuration with our aux chain cni chain name.
jsonString := fmt.Sprintf(`{"cniVersion":"%s","name":"%s","plugins":[{"type":"passthru","name":"passthru-cni"}]}`, n.CNIVersion, n.AuxiliaryCNIChainName)
// Convert the JSON string to a byte array
byteArray := []byte(jsonString)
// Let's try to get the cni path from the ClusterNetwork
if !strings.Contains(n.ClusterNetwork, "/") {
return nil, cmdErr(k8sArgs, "auxiliary chain used but ClusterNetwork must be a path, and it is not a path: %v", n.ClusterNetwork)
}
// Get the directory part of the ClusterNetwork path
// TODO: This could probably be improved.
cniPath := filepath.Dir(n.ClusterNetwork)
// Load chained delegates
delegate := k8s.LoadChainedDelegatesFromBytes(byteArray, cniPath)
if delegate != nil {
// Only if additional plugins were listed do we add this aux chain delegate.
if len(delegate.ConfList.Plugins) > 1 {
// Add the resulting delegate to n.Delegates
n.Delegates = append(n.Delegates, delegate)
}
}
}
// cache the multus config
if err := saveDelegates(args.ContainerID, n.CNIDir, n.Delegates); err != nil {
return nil, cmdErr(k8sArgs, "error saving the delegates: %v", err)
@@ -757,15 +849,17 @@ func CmdAdd(args *skel.CmdArgs, exec invoke.Exec, kubeClient *k8s.ClientInfo) (c
}
}
// set the network status annotation in apiserver, only in case Multus as kubeconfig
// set the network status annotation in apiserver, only in case Multus has kubeconfig
if kubeClient != nil && kc != nil {
if !types.CheckSystemNamespaces(string(k8sArgs.K8S_POD_NAME), n.SystemNamespaces) {
err = k8s.SetNetworkStatus(kubeClient, k8sArgs, netStatus, n)
if err != nil {
if strings.Contains(err.Error(), "failed to query the pod") {
return nil, cmdErr(k8sArgs, "error setting the networks status, pod was already deleted: %v", err)
if strings.Contains(err.Error(), `pod "`) && strings.Contains(err.Error(), `" not found`) {
// Tolerate issues with writing the status due to pod deletion, and log them.
logging.Verbosef("warning: tolerated failure writing network status (pod not found): %v", err)
} else {
return nil, cmdErr(k8sArgs, "error setting the networks status: %v", err)
}
return nil, cmdErr(k8sArgs, "error setting the networks status: %v", err)
}
}
}
@@ -925,3 +1019,125 @@ func CmdDel(args *skel.CmdArgs, exec invoke.Exec, kubeClient *k8s.ClientInfo) er
return e
}
// CmdStatus ...
func CmdStatus(args *skel.CmdArgs, exec invoke.Exec, kubeClient *k8s.ClientInfo) error {
n, err := types.LoadNetConf(args.StdinData)
logging.Debugf("CmdStatus: %v, %v, %v", args, exec, kubeClient)
if err != nil {
return cmdErr(nil, "error loading netconf: %v", err)
}
kubeClient, err = k8s.GetK8sClient(n.Kubeconfig, kubeClient)
if err != nil {
return cmdErr(nil, "error getting k8s client: %v", err)
}
if n.ReadinessIndicatorFile != "" {
if err := types.GetReadinessIndicatorFile(n.ReadinessIndicatorFile); err != nil {
return cmdErr(nil, "have you checked that your default network is ready? still waiting for readinessindicatorfile @ %v. pollimmediate error: %v", n.ReadinessIndicatorFile, err)
}
}
if n.ClusterNetwork != "" {
_, err = k8s.GetDefaultNetworks(nil, n, kubeClient, nil)
if err != nil {
return cmdErr(nil, "failed to get clusterNetwork: %v", err)
}
// First delegate is always the master plugin
n.Delegates[0].MasterPlugin = true
}
// invoke delegate's STATUS command
// we only need to check cluster network status
binDirs := filepath.SplitList(os.Getenv("CNI_PATH"))
binDirs = append([]string{n.BinDir}, binDirs...)
cniNet := libcni.NewCNIConfigWithCacheDir(binDirs, n.CNIDir, exec)
conf, err := libcni.ConfListFromBytes(n.Delegates[0].Bytes)
if err != nil {
return logging.Errorf("error in converting the raw bytes to conf: %v", err)
}
err = cniNet.GetStatusNetworkList(context.TODO(), conf)
if err != nil {
return logging.Errorf("error in STATUS command: %v", err)
}
return nil
}
// CmdGC ...
func CmdGC(args *skel.CmdArgs, exec invoke.Exec, kubeClient *k8s.ClientInfo) error {
n, err := types.LoadNetConf(args.StdinData)
logging.Debugf("CmdStatus: %v, %v, %v", args, exec, kubeClient)
if err != nil {
return cmdErr(nil, "error loading netconf: %v", err)
}
kubeClient, err = k8s.GetK8sClient(n.Kubeconfig, kubeClient)
if err != nil {
return cmdErr(nil, "error getting k8s client: %v", err)
}
if n.ReadinessIndicatorFile != "" {
if err := types.GetReadinessIndicatorFile(n.ReadinessIndicatorFile); err != nil {
return cmdErr(nil, "have you checked that your default network is ready? still waiting for readinessindicatorfile @ %v. pollimmediate error: %v", n.ReadinessIndicatorFile, err)
}
}
if n.ClusterNetwork != "" {
_, err = k8s.GetDefaultNetworks(nil, n, kubeClient, nil)
if err != nil {
return cmdErr(nil, "failed to get clusterNetwork: %v", err)
}
// First delegate is always the master plugin
n.Delegates[0].MasterPlugin = true
}
// invoke delegate's GC command
// we only need to check cluster network status
binDirs := filepath.SplitList(os.Getenv("CNI_PATH"))
binDirs = append([]string{n.BinDir}, binDirs...)
cniNet := libcni.NewCNIConfigWithCacheDir(binDirs, n.CNIDir, exec)
conf, err := libcni.ConfListFromBytes(n.Delegates[0].Bytes)
if err != nil {
return logging.Errorf("error in converting the raw bytes to conf: %v", err)
}
validAttachments, err := gatherValidAttachmentsFromCache(n.CNIDir)
if err != nil {
return logging.Errorf("error in gather valid attachments: %v", err)
}
err = cniNet.GCNetworkList(context.TODO(), conf, &libcni.GCArgs{
ValidAttachments: validAttachments,
})
if err != nil {
return logging.Errorf("error in GC command: %v", err)
}
return nil
}
func emptyCNIResult(args *skel.CmdArgs, cniVersion string) *cni100.Result {
return &cni100.Result{
CNIVersion: cniVersion,
Interfaces: []*cni100.Interface{
{
Name: args.IfName,
Sandbox: args.Netns,
},
},
IPs: []*cni100.IPConfig{
{
Address: net.IPNet{
IP: net.ParseIP("0.0.0.0"),
Mask: net.CIDRMask(0, 32),
},
Gateway: net.ParseIP("0.0.0.0"),
},
},
}
}

View File

@@ -26,10 +26,8 @@ import (
types020 "github.com/containernetworking/cni/pkg/types/020"
"github.com/containernetworking/plugins/pkg/ns"
"github.com/containernetworking/plugins/pkg/testutils"
"gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/k8sclient"
"gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/logging"
testhelpers "gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/testing"
"gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/types"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/client-go/tools/record"
@@ -43,20 +41,6 @@ var _ = Describe("multus operations", func() {
err := saveScratchNetConf("123456789", "", meme)
Expect(err).To(HaveOccurred())
})
It("fails to delete delegates with bad filepath", func() {
err := deleteDelegates("123456789", "bad!file!~?Path$^")
Expect(err).To(HaveOccurred())
})
It("delete delegates given good filepath", func() {
os.MkdirAll("/opt/cni/bin", 0755)
d1 := []byte("blah")
os.WriteFile("/opt/cni/bin/123456789", d1, 0644)
err := deleteDelegates("123456789", "/opt/cni/bin")
Expect(err).NotTo(HaveOccurred())
})
})
var _ = Describe("multus operations cniVersion 0.2.0 config", func() {
@@ -769,66 +753,4 @@ var _ = Describe("multus operations cniVersion 0.2.0 config", func() {
Expect(fExec.delIndex).To(Equal(len(fExec.plugins)))
})
It("fails to execute confListDel given no 'plugins' key", func() {
args := &skel.CmdArgs{
ContainerID: "123456789",
Netns: testNS.Path(),
IfName: "eth0",
StdinData: []byte(`{
"name": "node-cni-network",
"type": "multus",
"readinessindicatorfile": "/tmp/foo.multus.conf",
"defaultnetworkwaitseconds": 3,
"delegates": [{
"name": "weave1",
"cniVersion": "0.2.0",
"type": "weave-net"
},{
"name": "other1",
"cniVersion": "0.2.0",
"type": "other-plugin"
}]
}`),
}
fExec := newFakeExec()
expectedResult1 := &types020.Result{
CNIVersion: "0.2.0",
IP4: &types020.IPConfig{
IP: *testhelpers.EnsureCIDR("1.1.1.2/24"),
},
}
expectedConf1 := `{
"name": "weave1",
"cniVersion": "0.2.0",
"type": "weave-net"
}`
fExec.addPlugin020(nil, "eth0", expectedConf1, expectedResult1, nil)
expectedResult2 := &types020.Result{
CNIVersion: "0.2.0",
IP4: &types020.IPConfig{
IP: *testhelpers.EnsureCIDR("1.1.1.5/24"),
},
}
expectedConf2 := `{
"name": "other1",
"cniVersion": "0.2.0",
"type": "other-plugin"
}`
fExec.addPlugin020(nil, "net1", expectedConf2, expectedResult2, nil)
fakeMultusNetConf := types.NetConf{
BinDir: "/opt/cni/bin",
}
// use fExec for the exec param
rawnetconflist := []byte(`{"cniVersion":"0.2.0","name":"weave1","type":"weave-net"}`)
k8sargs, err := k8sclient.GetK8sArgs(args)
n, err := types.LoadNetConf(args.StdinData)
rt, _ := types.CreateCNIRuntimeConf(args, k8sargs, args.IfName, n.RuntimeConfig, nil)
err = conflistDel(rt, rawnetconflist, &fakeMultusNetConf, fExec)
Expect(err).To(HaveOccurred())
})
})

View File

@@ -26,10 +26,8 @@ import (
cni040 "github.com/containernetworking/cni/pkg/types/040"
"github.com/containernetworking/plugins/pkg/ns"
"github.com/containernetworking/plugins/pkg/testutils"
"gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/k8sclient"
"gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/logging"
testhelpers "gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/testing"
"gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/types"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
. "github.com/onsi/ginkgo/v2"
@@ -1502,67 +1500,4 @@ var _ = Describe("multus operations cniVersion 0.4.0 config", func() {
Expect(fExec.delIndex).To(Equal(len(fExec.plugins)))
})
It("fails to execute confListDel given no 'plugins' key", func() {
args := &skel.CmdArgs{
ContainerID: "123456789",
Netns: testNS.Path(),
IfName: "eth0",
StdinData: []byte(`{
"name": "node-cni-network",
"type": "multus",
"readinessindicatorfile": "/tmp/foo.multus.conf",
"defaultnetworkwaitseconds": 3,
"delegates": [{
"name": "weave1",
"cniVersion": "0.4.0",
"type": "weave-net"
},{
"name": "other1",
"cniVersion": "0.4.0",
"type": "other-plugin"
}]
}`),
}
fExec := newFakeExec()
expectedResult1 := &cni040.Result{
CNIVersion: "0.4.0",
IPs: []*cni040.IPConfig{{
Address: *testhelpers.EnsureCIDR("1.1.1.2/24"),
},
},
}
expectedConf1 := `{
"name": "weave1",
"cniVersion": "0.4.0",
"type": "weave-net"
}`
fExec.addPlugin040(nil, "eth0", expectedConf1, expectedResult1, nil)
expectedResult2 := &cni040.Result{
CNIVersion: "0.4.0",
IPs: []*cni040.IPConfig{{
Address: *testhelpers.EnsureCIDR("1.1.1.5/24"),
},
},
}
expectedConf2 := `{
"name": "other1",
"cniVersion": "0.4.0",
"type": "other-plugin"
}`
fExec.addPlugin040(nil, "net1", expectedConf2, expectedResult2, nil)
fakeMultusNetConf := types.NetConf{
BinDir: "/opt/cni/bin",
}
// use fExec for the exec param
rawnetconflist := []byte(`{"cniVersion":"0.4.0","name":"weave1","type":"weave-net"}`)
k8sargs, err := k8sclient.GetK8sArgs(args)
n, err := types.LoadNetConf(args.StdinData)
rt, _ := types.CreateCNIRuntimeConf(args, k8sargs, args.IfName, n.RuntimeConfig, nil)
err = conflistDel(rt, rawnetconflist, &fakeMultusNetConf, fExec)
Expect(err).To(HaveOccurred())
})
})

View File

@@ -20,6 +20,7 @@ import (
"context"
"fmt"
"os"
"path/filepath"
"reflect"
"sync"
"time"
@@ -28,10 +29,8 @@ import (
cni100 "github.com/containernetworking/cni/pkg/types/100"
"github.com/containernetworking/plugins/pkg/ns"
"github.com/containernetworking/plugins/pkg/testutils"
"gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/k8sclient"
"gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/logging"
testhelpers "gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/testing"
"gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/types"
. "github.com/onsi/ginkgo/v2"
. "github.com/onsi/gomega"
@@ -1163,7 +1162,47 @@ var _ = Describe("multus operations cniVersion 1.0.0 config", func() {
Expect(fExec.delIndex).To(Equal(len(fExec.plugins)))
})
It("fails to execute confListDel given no 'plugins' key", func() {
})
var _ = Describe("multus operations cniVersion 1.1.0 config", func() {
var testNS ns.NetNS
var tmpDir string
configPath := "/tmp/foo.multus.conf"
var cancel context.CancelFunc
BeforeEach(func() {
// Create a new NetNS so we don't modify the host
var err error
testNS, err = testutils.NewNS()
Expect(err).NotTo(HaveOccurred())
os.Setenv("CNI_NETNS", testNS.Path())
os.Setenv("CNI_PATH", "/some/path")
tmpDir, err = os.MkdirTemp("", "multus_tmp")
Expect(err).NotTo(HaveOccurred())
// Touch the default network file.
os.OpenFile(configPath, os.O_RDONLY|os.O_CREATE, 0755)
_, cancel = context.WithCancel(context.TODO())
})
AfterEach(func() {
cancel()
// Cleanup default network file.
if _, errStat := os.Stat(configPath); errStat == nil {
errRemove := os.Remove(configPath)
Expect(errRemove).NotTo(HaveOccurred())
}
Expect(testNS.Close()).To(Succeed())
os.Unsetenv("CNI_PATH")
os.Unsetenv("CNI_ARGS")
err := os.RemoveAll(tmpDir)
Expect(err).NotTo(HaveOccurred())
})
It("executes delegates with CNI Check", func() {
args := &skel.CmdArgs{
ContainerID: "123456789",
Netns: testNS.Path(),
@@ -1171,59 +1210,116 @@ var _ = Describe("multus operations cniVersion 1.0.0 config", func() {
StdinData: []byte(`{
"name": "node-cni-network",
"type": "multus",
"readinessindicatorfile": "/tmp/foo.multus.conf",
"defaultnetworkfile": "/tmp/foo.multus.conf",
"defaultnetworkwaitseconds": 3,
"delegates": [{
"name": "weave1",
"cniVersion": "1.0.0",
"type": "weave-net"
"cniVersion": "1.1.0",
"plugins": [{
"type": "weave-net"
}]
},{
"name": "other1",
"cniVersion": "1.0.0",
"type": "other-plugin"
"cniVersion": "1.1.0",
"plugins": [{
"type": "other-plugin"
}]
}]
}`),
}
logging.SetLogLevel("verbose")
fExec := newFakeExec()
expectedResult1 := &cni100.Result{
CNIVersion: "1.0.0",
IPs: []*cni100.IPConfig{{
Address: *testhelpers.EnsureCIDR("1.1.1.2/24"),
},
},
}
expectedConf1 := `{
"name": "weave1",
"cniVersion": "1.0.0",
"cniVersion": "1.1.0",
"type": "weave-net"
}`
fExec.addPlugin100(nil, "eth0", expectedConf1, expectedResult1, nil)
fExec.addPlugin100(nil, "", expectedConf1, nil, nil)
expectedResult2 := &cni100.Result{
CNIVersion: "1.0.0",
IPs: []*cni100.IPConfig{{
Address: *testhelpers.EnsureCIDR("1.1.1.5/24"),
},
},
err := CmdStatus(args, fExec, nil)
Expect(err).NotTo(HaveOccurred())
// we only execute once for cluster network, not additional one
Expect(fExec.statusIndex).To(Equal(1))
})
It("executes delegates with CNI GC", func() {
tmpCNIDir := tmpDir + "/cniData"
err := os.Mkdir(tmpCNIDir, 0777)
Expect(err).NotTo(HaveOccurred())
cniCacheDir := filepath.Join(tmpCNIDir, "/results")
err = os.Mkdir(cniCacheDir, 0777)
Expect(err).NotTo(HaveOccurred())
//create fake cniResult file
err = os.WriteFile(filepath.Join(cniCacheDir, "cbr0-3f6940ab5ab43bc522569d15b23f8c1bbde1d7678b080398506924fc01d72755-eth0"), []byte(`{"kind":"cniCacheV1","containerId":"3f6940ab5ab43bc522569d15b23f8c1bbde1d7678b080398506924fc01d72755","config":"eyJjbmlWZXJzaW9uIjoiMC4zLjEiLCJuYW1lIjoiY2JyMCIsInBsdWdpbnMiOlt7ImNhcGFiaWxpdGllcyI6eyJpby5rdWJlcm5ldGVzLmNyaS5wb2QtYW5ub3RhdGlvbnMiOnRydWV9LCJkZWxlZ2F0ZSI6eyJoYWlycGluTW9kZSI6dHJ1ZSwiaXNEZWZhdWx0R2F0ZXdheSI6dHJ1ZX0sInR5cGUiOiJmbGFubmVsIn0seyJjYXBhYmlsaXRpZXMiOnsicG9ydE1hcHBpbmdzIjp0cnVlfSwidHlwZSI6InBvcnRtYXAifV19","ifName":"eth0","networkName":"cbr0","netns":"/var/run/netns/8b8677c8-8929-4746-8206-514069760f6e","cniArgs":[["IgnoreUnknown","true"],["K8S_POD_NAMESPACE","default"],["K8S_POD_NAME","macvlan"],["K8S_POD_INFRA_CONTAINER_ID","3f6940ab5ab43bc522569d15b23f8c1bbde1d7678b080398506924fc01d72755"],["K8S_POD_UID","f0bfbd5b-096d-48ef-998c-da26743dd0cb"],["IgnoreUnknown","1"],["K8S_POD_NAMESPACE","default"],["K8S_POD_NAME","macvlan"],["K8S_POD_INFRA_CONTAINER_ID","3f6940ab5ab43bc522569d15b23f8c1bbde1d7678b080398506924fc01d72755"],["K8S_POD_UID","f0bfbd5b-096d-48ef-998c-da26743dd0cb"]],"result":{"cniVersion":"0.3.1","dns":{},"interfaces":[{"mac":"ea:19:25:a2:a1:93","name":"cni0"},{"mac":"ba:76:61:2f:8b:ca","name":"vethc42d3d18"},{"mac":"7e:57:6a:9b:6b:b5","name":"eth0","sandbox":"/var/run/netns/8b8677c8-8929-4746-8206-514069760f6e"}],"ips":[{"address":"10.244.1.4/24","gateway":"10.244.1.1","interface":2,"version":"4"}],"routes":[{"dst":"10.244.0.0/16"},{"dst":"0.0.0.0/0","gw":"10.244.1.1"}]}}`), 0666)
Expect(err).NotTo(HaveOccurred())
err = os.WriteFile(filepath.Join(cniCacheDir, "macvlan-conf-1-3f6940ab5ab43bc522569d15b23f8c1bbde1d7678b080398506924fc01d72755-net1"), []byte(`{"kind":"cniCacheV1","containerId":"3f6940ab5ab43bc522569d15b23f8c1bbde1d7678b080398506924fc01d72755","config":"eyJjbmlWZXJzaW9uIjoiMC4zLjEiLCJpcGFtIjp7ImFkZHJlc3NlcyI6W3siYWRkcmVzcyI6IjEwLjEuMS4xMDEvMjQifV0sInR5cGUiOiJzdGF0aWMifSwibWFzdGVyIjoiZXRoMSIsIm1vZGUiOiJicmlkZ2UiLCJuYW1lIjoibWFjdmxhbi1jb25mLTEiLCJ0eXBlIjoibWFjdmxhbiJ9","ifName":"net1","networkName":"macvlan-conf-1","netns":"/var/run/netns/8b8677c8-8929-4746-8206-514069760f6e","cniArgs":[["IgnoreUnknown","true"],["K8S_POD_NAMESPACE","default"],["K8S_POD_NAME","macvlan"],["K8S_POD_INFRA_CONTAINER_ID","3f6940ab5ab43bc522569d15b23f8c1bbde1d7678b080398506924fc01d72755"],["K8S_POD_UID","f0bfbd5b-096d-48ef-998c-da26743dd0cb"],["IgnoreUnknown","1"],["K8S_POD_NAMESPACE","default"],["K8S_POD_NAME","macvlan"],["K8S_POD_INFRA_CONTAINER_ID","3f6940ab5ab43bc522569d15b23f8c1bbde1d7678b080398506924fc01d72755"],["K8S_POD_UID","f0bfbd5b-096d-48ef-998c-da26743dd0cb"]],"result":{"cniVersion":"0.3.1","dns":{},"interfaces":[{"mac":"36:b3:c5:29:ad:b8","name":"net1","sandbox":"/var/run/netns/8b8677c8-8929-4746-8206-514069760f6e"}],"ips":[{"address":"10.1.1.101/24","interface":0,"version":"4"}]}}`), 0666)
Expect(err).NotTo(HaveOccurred())
args := &skel.CmdArgs{
ContainerID: "123456789",
Netns: testNS.Path(),
IfName: "eth0",
StdinData: []byte(fmt.Sprintf(`{
"name": "node-cni-network",
"type": "multus",
"defaultnetworkfile": "/tmp/foo.multus.conf",
"defaultnetworkwaitseconds": 3,
"cniDir": "%s",
"delegates": [{
"name": "weave1",
"cniVersion": "1.1.0",
"plugins": [{
"type": "weave-net"
}]
},{
"name": "other1",
"cniVersion": "1.1.0",
"plugins": [{
"type": "other-plugin"
}]
}]
}`, tmpCNIDir)),
}
expectedConf2 := `{
"name": "other1",
"cniVersion": "1.0.0",
"type": "other-plugin"
}`
fExec.addPlugin100(nil, "net1", expectedConf2, expectedResult2, nil)
fakeMultusNetConf := types.NetConf{
BinDir: "/opt/cni/bin",
}
// use fExec for the exec param
rawnetconflist := []byte(`{"cniVersion":"1.0.0","name":"weave1","type":"weave-net"}`)
k8sargs, err := k8sclient.GetK8sArgs(args)
n, err := types.LoadNetConf(args.StdinData)
rt, _ := types.CreateCNIRuntimeConf(args, k8sargs, args.IfName, n.RuntimeConfig, nil)
logging.SetLogLevel("verbose")
err = conflistDel(rt, rawnetconflist, &fakeMultusNetConf, fExec)
Expect(err).To(HaveOccurred())
fExec := newFakeExec()
expectedConf1 := `{
"cni.dev/attachments": [
{
"containerID": "3f6940ab5ab43bc522569d15b23f8c1bbde1d7678b080398506924fc01d72755",
"ifname": "eth0"
},
{
"containerID": "3f6940ab5ab43bc522569d15b23f8c1bbde1d7678b080398506924fc01d72755",
"ifname": "net1"
}
],
"cni.dev/valid-attachments": [
{
"containerID": "3f6940ab5ab43bc522569d15b23f8c1bbde1d7678b080398506924fc01d72755",
"ifname": "eth0"
},
{
"containerID": "3f6940ab5ab43bc522569d15b23f8c1bbde1d7678b080398506924fc01d72755",
"ifname": "net1"
}
],
"cniVersion": "1.1.0",
"name": "weave1",
"type": "weave-net"
}`
fExec.addPlugin100(nil, "", expectedConf1, nil, nil)
err = CmdGC(args, fExec, nil)
Expect(err).NotTo(HaveOccurred())
// we only execute once for cluster network, not additional one
Expect(fExec.gcIndex).To(Equal(1))
err = os.RemoveAll(tmpCNIDir)
Expect(err).NotTo(HaveOccurred())
})
})

View File

@@ -58,6 +58,8 @@ type fakeExec struct {
addIndex int
delIndex int
chkIndex int
statusIndex int
gcIndex int
expectedDelSkip int
plugins map[string]*fakePlugin
}
@@ -168,6 +170,14 @@ func (f *fakeExec) ExecPlugin(_ context.Context, pluginPath string, stdinData []
Expect(len(f.plugins)).To(BeNumerically(">", f.delIndex))
index = len(f.plugins) - f.expectedDelSkip - f.delIndex - 1
f.delIndex++
case "GC":
Expect(len(f.plugins)).To(BeNumerically(">", f.statusIndex))
index = f.gcIndex
f.gcIndex++
case "STATUS":
Expect(len(f.plugins)).To(BeNumerically(">", f.statusIndex))
index = f.statusIndex
f.statusIndex++
default:
// Should never be reached
Expect(false).To(BeTrue())

View File

@@ -74,6 +74,24 @@ func CmdDel(args *skel.CmdArgs) error {
return nil
}
// CmdGC implements the CNI spec GC command handler
func CmdGC(args *skel.CmdArgs) error {
_, _, err := postRequest(args, WaitUntilAPIReady)
if err != nil {
return logging.Errorf("CmdGC (shim): %v", err)
}
return nil
}
// CmdStatus implements the CNI spec STATUS command handler
func CmdStatus(args *skel.CmdArgs) error {
_, _, err := postRequest(args, WaitUntilAPIReady)
if err != nil {
return logging.Errorf("CmdStatus (shim): %v", err)
}
return nil
}
func postRequest(args *skel.CmdArgs, readinessCheck readyCheckFunc) (*Response, string, error) {
multusShimConfig, err := shimConfig(args.StdinData)
if err != nil {

View File

@@ -55,6 +55,7 @@ type MultusConf struct {
Type string `json:"type"`
CniDir string `json:"cniDir,omitempty"`
CniConfigDir string `json:"cniConfigDir,omitempty"`
AuxiliaryCNIChainName string `json:"auxiliaryCNIChainName,omitempty"`
DaemonSocketDir string `json:"daemonSocketDir,omitempty"`
MultusConfigFile string `json:"multusConfigFile,omitempty"`
MultusMasterCni string `json:"multusMasterCNI,omitempty"`

View File

@@ -95,6 +95,10 @@ func (s *Server) HandleCNIRequest(cmd string, k8sArgs *types.K8sArgs, cniCmdArgs
err = s.cmdDel(cniCmdArgs, k8sArgs)
case "CHECK":
err = s.cmdCheck(cniCmdArgs, k8sArgs)
case "GC":
err = s.cmdGC(cniCmdArgs, k8sArgs)
case "STATUS":
err = s.cmdStatus(cniCmdArgs, k8sArgs)
default:
return []byte(""), fmt.Errorf("unknown cmd type: %s", cmd)
}
@@ -614,6 +618,28 @@ func (s *Server) cmdCheck(cmdArgs *skel.CmdArgs, k8sArgs *types.K8sArgs) error {
return multus.CmdCheck(cmdArgs, s.exec, s.kubeclient)
}
func (s *Server) cmdGC(cmdArgs *skel.CmdArgs, k8sArgs *types.K8sArgs) error {
namespace := string(k8sArgs.K8S_POD_NAMESPACE)
podName := string(k8sArgs.K8S_POD_NAME)
if namespace == "" || podName == "" {
return fmt.Errorf("required CNI variable missing. pod name: %s; pod namespace: %s", podName, namespace)
}
logging.Debugf("CmdGC for [%s/%s]. CNI conf: %+v", namespace, podName, *cmdArgs)
return multus.CmdGC(cmdArgs, s.exec, s.kubeclient)
}
func (s *Server) cmdStatus(cmdArgs *skel.CmdArgs, k8sArgs *types.K8sArgs) error {
namespace := string(k8sArgs.K8S_POD_NAMESPACE)
podName := string(k8sArgs.K8S_POD_NAME)
if namespace == "" || podName == "" {
return fmt.Errorf("required CNI variable missing. pod name: %s; pod namespace: %s", podName, namespace)
}
logging.Debugf("CmdStatus for [%s/%s]. CNI conf: %+v", namespace, podName, *cmdArgs)
return multus.CmdStatus(cmdArgs, s.exec, s.kubeclient)
}
func serializeResult(result cnitypes.Result) ([]byte, error) {
// cni result is converted to latest here and decoded to specific cni version at multus-shim
realResult, err := cni100.NewResultFromResult(result)

View File

@@ -26,6 +26,7 @@ import (
"github.com/containernetworking/cni/libcni"
"github.com/containernetworking/cni/pkg/skel"
"github.com/containernetworking/cni/pkg/types"
cni100 "github.com/containernetworking/cni/pkg/types/100"
"github.com/containernetworking/cni/pkg/version"
nadutils "github.com/k8snetworkplumbingwg/network-attachment-definition-client/pkg/utils"
@@ -62,6 +63,112 @@ func LoadDelegateNetConfList(bytes []byte, delegateConf *DelegateNetConf) error
return nil
}
// ConvertNetworkConfigListToNetConfList converts a libcni.NetworkConfigList to a NetConfList
func ConvertNetworkConfigListToNetConfList(ncList *libcni.NetworkConfigList) (*types.NetConfList, error) {
// Convert Plugins from []*libcni.PluginConfig to []*types.PluginConf
var plugins []*types.PluginConf
for _, plugin := range ncList.Plugins {
plugins = append(plugins, plugin.Network)
}
// Create NetConfList
netConfList := &types.NetConfList{
CNIVersion: ncList.CNIVersion,
Name: ncList.Name,
DisableCheck: ncList.DisableCheck,
DisableGC: ncList.DisableGC,
Plugins: plugins,
}
return netConfList, nil
}
// LoadDelegateNetConfFromConfList converts a libcni.NetworkConfigList into a DelegateNetConf structure
func LoadDelegateNetConfFromConfList(confList *libcni.NetworkConfigList, netElement *NetworkSelectionElement, deviceID string, resourceName string) (*DelegateNetConf, error) {
var err error
logging.Debugf("LoadDelegateNetConfFromConfList: %v, %v, %s", confList, netElement, deviceID)
// Convert libcni.NetworkConfigList to NetConfList
netConfList, err := ConvertNetworkConfigListToNetConfList(confList)
if err != nil {
return nil, err
}
delegateConf := &DelegateNetConf{
Name: netConfList.Name,
ConfList: *netConfList,
CNINetworkConfigList: *confList,
ConfListPlugin: true,
}
// Convert the plugins back to bytes for consistency
pluginsBytes, err := json.Marshal(netConfList)
if err != nil {
return nil, logging.Errorf("LoadDelegateNetConfFromConfList: error marshaling netConfList: %v", err)
}
delegateConf.Bytes = pluginsBytes
if deviceID != "" {
pluginsBytes, err = addDeviceIDInConfList(pluginsBytes, deviceID)
if err != nil {
return nil, logging.Errorf("LoadDelegateNetConfFromConfList: failed to add deviceID in NetConfList bytes: %v", err)
}
delegateConf.ResourceName = resourceName
delegateConf.DeviceID = deviceID
}
if netElement != nil && netElement.CNIArgs != nil {
pluginsBytes, err = addCNIArgsInConfList(pluginsBytes, netElement.CNIArgs)
if err != nil {
return nil, logging.Errorf("LoadDelegateNetConfFromConfList: failed to add cni-args in NetConfList bytes: %v", err)
}
delegateConf.Bytes = pluginsBytes
}
if netElement != nil {
if netElement.Name != "" {
// Overwrite CNI config name with net-attach-def name
delegateConf.Name = fmt.Sprintf("%s/%s", netElement.Namespace, netElement.Name)
}
if netElement.InterfaceRequest != "" {
delegateConf.IfnameRequest = netElement.InterfaceRequest
}
if netElement.MacRequest != "" {
delegateConf.MacRequest = netElement.MacRequest
}
if netElement.IPRequest != nil {
delegateConf.IPRequest = netElement.IPRequest
}
if netElement.BandwidthRequest != nil {
delegateConf.BandwidthRequest = netElement.BandwidthRequest
}
if netElement.PortMappingsRequest != nil {
delegateConf.PortMappingsRequest = netElement.PortMappingsRequest
}
if netElement.GatewayRequest != nil {
var list []net.IP
if delegateConf.GatewayRequest != nil {
list = append(*delegateConf.GatewayRequest, *netElement.GatewayRequest...)
} else {
list = *netElement.GatewayRequest
}
delegateConf.GatewayRequest = &list
}
if netElement.InfinibandGUIDRequest != "" {
delegateConf.InfinibandGUIDRequest = netElement.InfinibandGUIDRequest
}
if netElement.DeviceID != "" {
if deviceID != "" {
logging.Debugf("Warning: Both RuntimeConfig and ResourceMap provide deviceID. Ignoring RuntimeConfig")
} else {
delegateConf.DeviceID = netElement.DeviceID
}
}
}
return delegateConf, nil
}
// LoadDelegateNetConf converts raw CNI JSON into a DelegateNetConf structure
func LoadDelegateNetConf(bytes []byte, netElement *NetworkSelectionElement, deviceID string, resourceName string) (*DelegateNetConf, error) {
var err error

View File

@@ -18,10 +18,10 @@ package types
import (
"net"
"gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/logging"
"github.com/containernetworking/cni/libcni"
"github.com/containernetworking/cni/pkg/types"
cni100 "github.com/containernetworking/cni/pkg/types/100"
"gopkg.in/k8snetworkplumbingwg/multus-cni.v4/pkg/logging"
v1 "k8s.io/api/core/v1"
)
@@ -57,6 +57,7 @@ type NetConf struct {
NamespaceIsolation bool `json:"namespaceIsolation"`
RawNonIsolatedNamespaces string `json:"globalNamespaces"`
NonIsolatedNamespaces []string `json:"-"`
AuxiliaryCNIChainName string `json:"auxiliaryCNIChainName,omitempty"`
// Option to set system namespaces (to avoid to add defaultNetworks)
SystemNamespaces []string `json:"systemNamespaces"`
@@ -99,6 +100,7 @@ type BandwidthEntry struct {
type DelegateNetConf struct {
Conf types.NetConf
ConfList types.NetConfList
CNINetworkConfigList libcni.NetworkConfigList
Name string
IfnameRequest string `json:"ifnameRequest,omitempty"`
MacRequest string `json:"macRequest,omitempty"`

View File

@@ -1 +0,0 @@
_fuzz/

View File

@@ -1,27 +0,0 @@
run:
deadline: 2m
linters:
disable-all: true
enable:
- misspell
- govet
- staticcheck
- errcheck
- unparam
- ineffassign
- nakedret
- gocyclo
- dupl
- goimports
- revive
- gosec
- gosimple
- typecheck
- unused
linters-settings:
gofmt:
simplify: true
dupl:
threshold: 600

View File

@@ -1,214 +0,0 @@
# Changelog
## 3.2.0 (2022-11-28)
### Added
- #190: Added text marshaling and unmarshaling
- #167: Added JSON marshalling for constraints (thanks @SimonTheLeg)
- #173: Implement encoding.TextMarshaler and encoding.TextUnmarshaler on Version (thanks @MarkRosemaker)
- #179: Added New() version constructor (thanks @kazhuravlev)
### Changed
- #182/#183: Updated CI testing setup
### Fixed
- #186: Fixing issue where validation of constraint section gave false positives
- #176: Fix constraints check with *-0 (thanks @mtt0)
- #181: Fixed Caret operator (^) gives unexpected results when the minor version in constraint is 0 (thanks @arshchimni)
- #161: Fixed godoc (thanks @afirth)
## 3.1.1 (2020-11-23)
### Fixed
- #158: Fixed issue with generated regex operation order that could cause problem
## 3.1.0 (2020-04-15)
### Added
- #131: Add support for serializing/deserializing SQL (thanks @ryancurrah)
### Changed
- #148: More accurate validation messages on constraints
## 3.0.3 (2019-12-13)
### Fixed
- #141: Fixed issue with <= comparison
## 3.0.2 (2019-11-14)
### Fixed
- #134: Fixed broken constraint checking with ^0.0 (thanks @krmichelos)
## 3.0.1 (2019-09-13)
### Fixed
- #125: Fixes issue with module path for v3
## 3.0.0 (2019-09-12)
This is a major release of the semver package which includes API changes. The Go
API is compatible with ^1. The Go API was not changed because many people are using
`go get` without Go modules for their applications and API breaking changes cause
errors which we have or would need to support.
The changes in this release are the handling based on the data passed into the
functions. These are described in the added and changed sections below.
### Added
- StrictNewVersion function. This is similar to NewVersion but will return an
error if the version passed in is not a strict semantic version. For example,
1.2.3 would pass but v1.2.3 or 1.2 would fail because they are not strictly
speaking semantic versions. This function is faster, performs fewer operations,
and uses fewer allocations than NewVersion.
- Fuzzing has been performed on NewVersion, StrictNewVersion, and NewConstraint.
The Makefile contains the operations used. For more information on you can start
on Wikipedia at https://en.wikipedia.org/wiki/Fuzzing
- Now using Go modules
### Changed
- NewVersion has proper prerelease and metadata validation with error messages
to signal an issue with either of them
- ^ now operates using a similar set of rules to npm/js and Rust/Cargo. If the
version is >=1 the ^ ranges works the same as v1. For major versions of 0 the
rules have changed. The minor version is treated as the stable version unless
a patch is specified and then it is equivalent to =. One difference from npm/js
is that prereleases there are only to a specific version (e.g. 1.2.3).
Prereleases here look over multiple versions and follow semantic version
ordering rules. This pattern now follows along with the expected and requested
handling of this packaged by numerous users.
## 1.5.0 (2019-09-11)
### Added
- #103: Add basic fuzzing for `NewVersion()` (thanks @jesse-c)
### Changed
- #82: Clarify wildcard meaning in range constraints and update tests for it (thanks @greysteil)
- #83: Clarify caret operator range for pre-1.0.0 dependencies (thanks @greysteil)
- #72: Adding docs comment pointing to vert for a cli
- #71: Update the docs on pre-release comparator handling
- #89: Test with new go versions (thanks @thedevsaddam)
- #87: Added $ to ValidPrerelease for better validation (thanks @jeremycarroll)
### Fixed
- #78: Fix unchecked error in example code (thanks @ravron)
- #70: Fix the handling of pre-releases and the 0.0.0 release edge case
- #97: Fixed copyright file for proper display on GitHub
- #107: Fix handling prerelease when sorting alphanum and num
- #109: Fixed where Validate sometimes returns wrong message on error
## 1.4.2 (2018-04-10)
### Changed
- #72: Updated the docs to point to vert for a console appliaction
- #71: Update the docs on pre-release comparator handling
### Fixed
- #70: Fix the handling of pre-releases and the 0.0.0 release edge case
## 1.4.1 (2018-04-02)
### Fixed
- Fixed #64: Fix pre-release precedence issue (thanks @uudashr)
## 1.4.0 (2017-10-04)
### Changed
- #61: Update NewVersion to parse ints with a 64bit int size (thanks @zknill)
## 1.3.1 (2017-07-10)
### Fixed
- Fixed #57: number comparisons in prerelease sometimes inaccurate
## 1.3.0 (2017-05-02)
### Added
- #45: Added json (un)marshaling support (thanks @mh-cbon)
- Stability marker. See https://masterminds.github.io/stability/
### Fixed
- #51: Fix handling of single digit tilde constraint (thanks @dgodd)
### Changed
- #55: The godoc icon moved from png to svg
## 1.2.3 (2017-04-03)
### Fixed
- #46: Fixed 0.x.x and 0.0.x in constraints being treated as *
## Release 1.2.2 (2016-12-13)
### Fixed
- #34: Fixed issue where hyphen range was not working with pre-release parsing.
## Release 1.2.1 (2016-11-28)
### Fixed
- #24: Fixed edge case issue where constraint "> 0" does not handle "0.0.1-alpha"
properly.
## Release 1.2.0 (2016-11-04)
### Added
- #20: Added MustParse function for versions (thanks @adamreese)
- #15: Added increment methods on versions (thanks @mh-cbon)
### Fixed
- Issue #21: Per the SemVer spec (section 9) a pre-release is unstable and
might not satisfy the intended compatibility. The change here ignores pre-releases
on constraint checks (e.g., ~ or ^) when a pre-release is not part of the
constraint. For example, `^1.2.3` will ignore pre-releases while
`^1.2.3-alpha` will include them.
## Release 1.1.1 (2016-06-30)
### Changed
- Issue #9: Speed up version comparison performance (thanks @sdboyer)
- Issue #8: Added benchmarks (thanks @sdboyer)
- Updated Go Report Card URL to new location
- Updated Readme to add code snippet formatting (thanks @mh-cbon)
- Updating tagging to v[SemVer] structure for compatibility with other tools.
## Release 1.1.0 (2016-03-11)
- Issue #2: Implemented validation to provide reasons a versions failed a
constraint.
## Release 1.0.1 (2015-12-31)
- Fixed #1: * constraint failing on valid versions.
## Release 1.0.0 (2015-10-20)
- Initial release

View File

@@ -1,19 +0,0 @@
Copyright (C) 2014-2019, Matt Butcher and Matt Farina
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE.

View File

@@ -1,30 +0,0 @@
GOPATH=$(shell go env GOPATH)
GOLANGCI_LINT=$(GOPATH)/bin/golangci-lint
.PHONY: lint
lint: $(GOLANGCI_LINT)
@echo "==> Linting codebase"
@$(GOLANGCI_LINT) run
.PHONY: test
test:
@echo "==> Running tests"
GO111MODULE=on go test -v
.PHONY: test-cover
test-cover:
@echo "==> Running Tests with coverage"
GO111MODULE=on go test -cover .
.PHONY: fuzz
fuzz:
@echo "==> Running Fuzz Tests"
go test -fuzz=FuzzNewVersion -fuzztime=15s .
go test -fuzz=FuzzStrictNewVersion -fuzztime=15s .
go test -fuzz=FuzzNewConstraint -fuzztime=15s .
$(GOLANGCI_LINT):
# Install golangci-lint. The configuration for it is in the .golangci.yml
# file in the root of the repository
echo ${GOPATH}
curl -sfL https://install.goreleaser.com/github.com/golangci/golangci-lint.sh | sh -s -- -b $(GOPATH)/bin v1.17.1

View File

@@ -1,258 +0,0 @@
# SemVer
The `semver` package provides the ability to work with [Semantic Versions](http://semver.org) in Go. Specifically it provides the ability to:
* Parse semantic versions
* Sort semantic versions
* Check if a semantic version fits within a set of constraints
* Optionally work with a `v` prefix
[![Stability:
Active](https://masterminds.github.io/stability/active.svg)](https://masterminds.github.io/stability/active.html)
[![](https://github.com/Masterminds/semver/workflows/Tests/badge.svg)](https://github.com/Masterminds/semver/actions)
[![GoDoc](https://img.shields.io/static/v1?label=godoc&message=reference&color=blue)](https://pkg.go.dev/github.com/Masterminds/semver/v3)
[![Go Report Card](https://goreportcard.com/badge/github.com/Masterminds/semver)](https://goreportcard.com/report/github.com/Masterminds/semver)
If you are looking for a command line tool for version comparisons please see
[vert](https://github.com/Masterminds/vert) which uses this library.
## Package Versions
Note, import `github.com/github.com/Masterminds/semver/v3` to use the latest version.
There are three major versions fo the `semver` package.
* 3.x.x is the stable and active version. This version is focused on constraint
compatibility for range handling in other tools from other languages. It has
a similar API to the v1 releases. The development of this version is on the master
branch. The documentation for this version is below.
* 2.x was developed primarily for [dep](https://github.com/golang/dep). There are
no tagged releases and the development was performed by [@sdboyer](https://github.com/sdboyer).
There are API breaking changes from v1. This version lives on the [2.x branch](https://github.com/Masterminds/semver/tree/2.x).
* 1.x.x is the original release. It is no longer maintained. You should use the
v3 release instead. You can read the documentation for the 1.x.x release
[here](https://github.com/Masterminds/semver/blob/release-1/README.md).
## Parsing Semantic Versions
There are two functions that can parse semantic versions. The `StrictNewVersion`
function only parses valid version 2 semantic versions as outlined in the
specification. The `NewVersion` function attempts to coerce a version into a
semantic version and parse it. For example, if there is a leading v or a version
listed without all 3 parts (e.g. `v1.2`) it will attempt to coerce it into a valid
semantic version (e.g., 1.2.0). In both cases a `Version` object is returned
that can be sorted, compared, and used in constraints.
When parsing a version an error is returned if there is an issue parsing the
version. For example,
v, err := semver.NewVersion("1.2.3-beta.1+build345")
The version object has methods to get the parts of the version, compare it to
other versions, convert the version back into a string, and get the original
string. Getting the original string is useful if the semantic version was coerced
into a valid form.
## Sorting Semantic Versions
A set of versions can be sorted using the `sort` package from the standard library.
For example,
```go
raw := []string{"1.2.3", "1.0", "1.3", "2", "0.4.2",}
vs := make([]*semver.Version, len(raw))
for i, r := range raw {
v, err := semver.NewVersion(r)
if err != nil {
t.Errorf("Error parsing version: %s", err)
}
vs[i] = v
}
sort.Sort(semver.Collection(vs))
```
## Checking Version Constraints
There are two methods for comparing versions. One uses comparison methods on
`Version` instances and the other uses `Constraints`. There are some important
differences to notes between these two methods of comparison.
1. When two versions are compared using functions such as `Compare`, `LessThan`,
and others it will follow the specification and always include prereleases
within the comparison. It will provide an answer that is valid with the
comparison section of the spec at https://semver.org/#spec-item-11
2. When constraint checking is used for checks or validation it will follow a
different set of rules that are common for ranges with tools like npm/js
and Rust/Cargo. This includes considering prereleases to be invalid if the
ranges does not include one. If you want to have it include pre-releases a
simple solution is to include `-0` in your range.
3. Constraint ranges can have some complex rules including the shorthand use of
~ and ^. For more details on those see the options below.
There are differences between the two methods or checking versions because the
comparison methods on `Version` follow the specification while comparison ranges
are not part of the specification. Different packages and tools have taken it
upon themselves to come up with range rules. This has resulted in differences.
For example, npm/js and Cargo/Rust follow similar patterns while PHP has a
different pattern for ^. The comparison features in this package follow the
npm/js and Cargo/Rust lead because applications using it have followed similar
patters with their versions.
Checking a version against version constraints is one of the most featureful
parts of the package.
```go
c, err := semver.NewConstraint(">= 1.2.3")
if err != nil {
// Handle constraint not being parsable.
}
v, err := semver.NewVersion("1.3")
if err != nil {
// Handle version not being parsable.
}
// Check if the version meets the constraints. The a variable will be true.
a := c.Check(v)
```
### Basic Comparisons
There are two elements to the comparisons. First, a comparison string is a list
of space or comma separated AND comparisons. These are then separated by || (OR)
comparisons. For example, `">= 1.2 < 3.0.0 || >= 4.2.3"` is looking for a
comparison that's greater than or equal to 1.2 and less than 3.0.0 or is
greater than or equal to 4.2.3.
The basic comparisons are:
* `=`: equal (aliased to no operator)
* `!=`: not equal
* `>`: greater than
* `<`: less than
* `>=`: greater than or equal to
* `<=`: less than or equal to
### Working With Prerelease Versions
Pre-releases, for those not familiar with them, are used for software releases
prior to stable or generally available releases. Examples of prereleases include
development, alpha, beta, and release candidate releases. A prerelease may be
a version such as `1.2.3-beta.1` while the stable release would be `1.2.3`. In the
order of precedence, prereleases come before their associated releases. In this
example `1.2.3-beta.1 < 1.2.3`.
According to the Semantic Version specification prereleases may not be
API compliant with their release counterpart. It says,
> A pre-release version indicates that the version is unstable and might not satisfy the intended compatibility requirements as denoted by its associated normal version.
SemVer comparisons using constraints without a prerelease comparator will skip
prerelease versions. For example, `>=1.2.3` will skip prereleases when looking
at a list of releases while `>=1.2.3-0` will evaluate and find prereleases.
The reason for the `0` as a pre-release version in the example comparison is
because pre-releases can only contain ASCII alphanumerics and hyphens (along with
`.` separators), per the spec. Sorting happens in ASCII sort order, again per the
spec. The lowest character is a `0` in ASCII sort order
(see an [ASCII Table](http://www.asciitable.com/))
Understanding ASCII sort ordering is important because A-Z comes before a-z. That
means `>=1.2.3-BETA` will return `1.2.3-alpha`. What you might expect from case
sensitivity doesn't apply here. This is due to ASCII sort ordering which is what
the spec specifies.
### Hyphen Range Comparisons
There are multiple methods to handle ranges and the first is hyphens ranges.
These look like:
* `1.2 - 1.4.5` which is equivalent to `>= 1.2 <= 1.4.5`
* `2.3.4 - 4.5` which is equivalent to `>= 2.3.4 <= 4.5`
### Wildcards In Comparisons
The `x`, `X`, and `*` characters can be used as a wildcard character. This works
for all comparison operators. When used on the `=` operator it falls
back to the patch level comparison (see tilde below). For example,
* `1.2.x` is equivalent to `>= 1.2.0, < 1.3.0`
* `>= 1.2.x` is equivalent to `>= 1.2.0`
* `<= 2.x` is equivalent to `< 3`
* `*` is equivalent to `>= 0.0.0`
### Tilde Range Comparisons (Patch)
The tilde (`~`) comparison operator is for patch level ranges when a minor
version is specified and major level changes when the minor number is missing.
For example,
* `~1.2.3` is equivalent to `>= 1.2.3, < 1.3.0`
* `~1` is equivalent to `>= 1, < 2`
* `~2.3` is equivalent to `>= 2.3, < 2.4`
* `~1.2.x` is equivalent to `>= 1.2.0, < 1.3.0`
* `~1.x` is equivalent to `>= 1, < 2`
### Caret Range Comparisons (Major)
The caret (`^`) comparison operator is for major level changes once a stable
(1.0.0) release has occurred. Prior to a 1.0.0 release the minor versions acts
as the API stability level. This is useful when comparisons of API versions as a
major change is API breaking. For example,
* `^1.2.3` is equivalent to `>= 1.2.3, < 2.0.0`
* `^1.2.x` is equivalent to `>= 1.2.0, < 2.0.0`
* `^2.3` is equivalent to `>= 2.3, < 3`
* `^2.x` is equivalent to `>= 2.0.0, < 3`
* `^0.2.3` is equivalent to `>=0.2.3 <0.3.0`
* `^0.2` is equivalent to `>=0.2.0 <0.3.0`
* `^0.0.3` is equivalent to `>=0.0.3 <0.0.4`
* `^0.0` is equivalent to `>=0.0.0 <0.1.0`
* `^0` is equivalent to `>=0.0.0 <1.0.0`
## Validation
In addition to testing a version against a constraint, a version can be validated
against a constraint. When validation fails a slice of errors containing why a
version didn't meet the constraint is returned. For example,
```go
c, err := semver.NewConstraint("<= 1.2.3, >= 1.4")
if err != nil {
// Handle constraint not being parseable.
}
v, err := semver.NewVersion("1.3")
if err != nil {
// Handle version not being parseable.
}
// Validate a version against a constraint.
a, msgs := c.Validate(v)
// a is false
for _, m := range msgs {
fmt.Println(m)
// Loops over the errors which would read
// "1.3 is greater than 1.2.3"
// "1.3 is less than 1.4"
}
```
## Contribute
If you find an issue or want to contribute please file an [issue](https://github.com/Masterminds/semver/issues)
or [create a pull request](https://github.com/Masterminds/semver/pulls).
## Security
Security is an important consideration for this project. The project currently
uses the following tools to help discover security issues:
* [CodeQL](https://github.com/Masterminds/semver)
* [gosec](https://github.com/securego/gosec)
* Daily Fuzz testing
If you believe you have found a security vulnerability you can privately disclose
it through the [GitHub security page](https://github.com/Masterminds/semver/security).

View File

@@ -1,19 +0,0 @@
# Security Policy
## Supported Versions
The following versions of semver are currently supported:
| Version | Supported |
| ------- | ------------------ |
| 3.x | :white_check_mark: |
| 2.x | :x: |
| 1.x | :x: |
Fixes are only released for the latest minor version in the form of a patch release.
## Reporting a Vulnerability
You can privately disclose a vulnerability through GitHubs
[private vulnerability reporting](https://github.com/Masterminds/semver/security/advisories)
mechanism.

View File

@@ -1,24 +0,0 @@
package semver
// Collection is a collection of Version instances and implements the sort
// interface. See the sort package for more details.
// https://golang.org/pkg/sort/
type Collection []*Version
// Len returns the length of a collection. The number of Version instances
// on the slice.
func (c Collection) Len() int {
return len(c)
}
// Less is needed for the sort interface to compare two Version objects on the
// slice. If checks if one is less than the other.
func (c Collection) Less(i, j int) bool {
return c[i].LessThan(c[j])
}
// Swap is needed for the sort interface to replace the Version objects
// at two different positions in the slice.
func (c Collection) Swap(i, j int) {
c[i], c[j] = c[j], c[i]
}

View File

@@ -1,594 +0,0 @@
package semver
import (
"bytes"
"errors"
"fmt"
"regexp"
"strings"
)
// Constraints is one or more constraint that a semantic version can be
// checked against.
type Constraints struct {
constraints [][]*constraint
}
// NewConstraint returns a Constraints instance that a Version instance can
// be checked against. If there is a parse error it will be returned.
func NewConstraint(c string) (*Constraints, error) {
// Rewrite - ranges into a comparison operation.
c = rewriteRange(c)
ors := strings.Split(c, "||")
or := make([][]*constraint, len(ors))
for k, v := range ors {
// TODO: Find a way to validate and fetch all the constraints in a simpler form
// Validate the segment
if !validConstraintRegex.MatchString(v) {
return nil, fmt.Errorf("improper constraint: %s", v)
}
cs := findConstraintRegex.FindAllString(v, -1)
if cs == nil {
cs = append(cs, v)
}
result := make([]*constraint, len(cs))
for i, s := range cs {
pc, err := parseConstraint(s)
if err != nil {
return nil, err
}
result[i] = pc
}
or[k] = result
}
o := &Constraints{constraints: or}
return o, nil
}
// Check tests if a version satisfies the constraints.
func (cs Constraints) Check(v *Version) bool {
// TODO(mattfarina): For v4 of this library consolidate the Check and Validate
// functions as the underlying functions make that possible now.
// loop over the ORs and check the inner ANDs
for _, o := range cs.constraints {
joy := true
for _, c := range o {
if check, _ := c.check(v); !check {
joy = false
break
}
}
if joy {
return true
}
}
return false
}
// Validate checks if a version satisfies a constraint. If not a slice of
// reasons for the failure are returned in addition to a bool.
func (cs Constraints) Validate(v *Version) (bool, []error) {
// loop over the ORs and check the inner ANDs
var e []error
// Capture the prerelease message only once. When it happens the first time
// this var is marked
var prerelesase bool
for _, o := range cs.constraints {
joy := true
for _, c := range o {
// Before running the check handle the case there the version is
// a prerelease and the check is not searching for prereleases.
if c.con.pre == "" && v.pre != "" {
if !prerelesase {
em := fmt.Errorf("%s is a prerelease version and the constraint is only looking for release versions", v)
e = append(e, em)
prerelesase = true
}
joy = false
} else {
if _, err := c.check(v); err != nil {
e = append(e, err)
joy = false
}
}
}
if joy {
return true, []error{}
}
}
return false, e
}
func (cs Constraints) String() string {
buf := make([]string, len(cs.constraints))
var tmp bytes.Buffer
for k, v := range cs.constraints {
tmp.Reset()
vlen := len(v)
for kk, c := range v {
tmp.WriteString(c.string())
// Space separate the AND conditions
if vlen > 1 && kk < vlen-1 {
tmp.WriteString(" ")
}
}
buf[k] = tmp.String()
}
return strings.Join(buf, " || ")
}
// UnmarshalText implements the encoding.TextUnmarshaler interface.
func (cs *Constraints) UnmarshalText(text []byte) error {
temp, err := NewConstraint(string(text))
if err != nil {
return err
}
*cs = *temp
return nil
}
// MarshalText implements the encoding.TextMarshaler interface.
func (cs Constraints) MarshalText() ([]byte, error) {
return []byte(cs.String()), nil
}
var constraintOps map[string]cfunc
var constraintRegex *regexp.Regexp
var constraintRangeRegex *regexp.Regexp
// Used to find individual constraints within a multi-constraint string
var findConstraintRegex *regexp.Regexp
// Used to validate an segment of ANDs is valid
var validConstraintRegex *regexp.Regexp
const cvRegex string = `v?([0-9|x|X|\*]+)(\.[0-9|x|X|\*]+)?(\.[0-9|x|X|\*]+)?` +
`(-([0-9A-Za-z\-]+(\.[0-9A-Za-z\-]+)*))?` +
`(\+([0-9A-Za-z\-]+(\.[0-9A-Za-z\-]+)*))?`
func init() {
constraintOps = map[string]cfunc{
"": constraintTildeOrEqual,
"=": constraintTildeOrEqual,
"!=": constraintNotEqual,
">": constraintGreaterThan,
"<": constraintLessThan,
">=": constraintGreaterThanEqual,
"=>": constraintGreaterThanEqual,
"<=": constraintLessThanEqual,
"=<": constraintLessThanEqual,
"~": constraintTilde,
"~>": constraintTilde,
"^": constraintCaret,
}
ops := `=||!=|>|<|>=|=>|<=|=<|~|~>|\^`
constraintRegex = regexp.MustCompile(fmt.Sprintf(
`^\s*(%s)\s*(%s)\s*$`,
ops,
cvRegex))
constraintRangeRegex = regexp.MustCompile(fmt.Sprintf(
`\s*(%s)\s+-\s+(%s)\s*`,
cvRegex, cvRegex))
findConstraintRegex = regexp.MustCompile(fmt.Sprintf(
`(%s)\s*(%s)`,
ops,
cvRegex))
// The first time a constraint shows up will look slightly different from
// future times it shows up due to a leading space or comma in a given
// string.
validConstraintRegex = regexp.MustCompile(fmt.Sprintf(
`^(\s*(%s)\s*(%s)\s*)((?:\s+|,\s*)(%s)\s*(%s)\s*)*$`,
ops,
cvRegex,
ops,
cvRegex))
}
// An individual constraint
type constraint struct {
// The version used in the constraint check. For example, if a constraint
// is '<= 2.0.0' the con a version instance representing 2.0.0.
con *Version
// The original parsed version (e.g., 4.x from != 4.x)
orig string
// The original operator for the constraint
origfunc string
// When an x is used as part of the version (e.g., 1.x)
minorDirty bool
dirty bool
patchDirty bool
}
// Check if a version meets the constraint
func (c *constraint) check(v *Version) (bool, error) {
return constraintOps[c.origfunc](v, c)
}
// String prints an individual constraint into a string
func (c *constraint) string() string {
return c.origfunc + c.orig
}
type cfunc func(v *Version, c *constraint) (bool, error)
func parseConstraint(c string) (*constraint, error) {
if len(c) > 0 {
m := constraintRegex.FindStringSubmatch(c)
if m == nil {
return nil, fmt.Errorf("improper constraint: %s", c)
}
cs := &constraint{
orig: m[2],
origfunc: m[1],
}
ver := m[2]
minorDirty := false
patchDirty := false
dirty := false
if isX(m[3]) || m[3] == "" {
ver = fmt.Sprintf("0.0.0%s", m[6])
dirty = true
} else if isX(strings.TrimPrefix(m[4], ".")) || m[4] == "" {
minorDirty = true
dirty = true
ver = fmt.Sprintf("%s.0.0%s", m[3], m[6])
} else if isX(strings.TrimPrefix(m[5], ".")) || m[5] == "" {
dirty = true
patchDirty = true
ver = fmt.Sprintf("%s%s.0%s", m[3], m[4], m[6])
}
con, err := NewVersion(ver)
if err != nil {
// The constraintRegex should catch any regex parsing errors. So,
// we should never get here.
return nil, errors.New("constraint Parser Error")
}
cs.con = con
cs.minorDirty = minorDirty
cs.patchDirty = patchDirty
cs.dirty = dirty
return cs, nil
}
// The rest is the special case where an empty string was passed in which
// is equivalent to * or >=0.0.0
con, err := StrictNewVersion("0.0.0")
if err != nil {
// The constraintRegex should catch any regex parsing errors. So,
// we should never get here.
return nil, errors.New("constraint Parser Error")
}
cs := &constraint{
con: con,
orig: c,
origfunc: "",
minorDirty: false,
patchDirty: false,
dirty: true,
}
return cs, nil
}
// Constraint functions
func constraintNotEqual(v *Version, c *constraint) (bool, error) {
if c.dirty {
// If there is a pre-release on the version but the constraint isn't looking
// for them assume that pre-releases are not compatible. See issue 21 for
// more details.
if v.Prerelease() != "" && c.con.Prerelease() == "" {
return false, fmt.Errorf("%s is a prerelease version and the constraint is only looking for release versions", v)
}
if c.con.Major() != v.Major() {
return true, nil
}
if c.con.Minor() != v.Minor() && !c.minorDirty {
return true, nil
} else if c.minorDirty {
return false, fmt.Errorf("%s is equal to %s", v, c.orig)
} else if c.con.Patch() != v.Patch() && !c.patchDirty {
return true, nil
} else if c.patchDirty {
// Need to handle prereleases if present
if v.Prerelease() != "" || c.con.Prerelease() != "" {
eq := comparePrerelease(v.Prerelease(), c.con.Prerelease()) != 0
if eq {
return true, nil
}
return false, fmt.Errorf("%s is equal to %s", v, c.orig)
}
return false, fmt.Errorf("%s is equal to %s", v, c.orig)
}
}
eq := v.Equal(c.con)
if eq {
return false, fmt.Errorf("%s is equal to %s", v, c.orig)
}
return true, nil
}
func constraintGreaterThan(v *Version, c *constraint) (bool, error) {
// If there is a pre-release on the version but the constraint isn't looking
// for them assume that pre-releases are not compatible. See issue 21 for
// more details.
if v.Prerelease() != "" && c.con.Prerelease() == "" {
return false, fmt.Errorf("%s is a prerelease version and the constraint is only looking for release versions", v)
}
var eq bool
if !c.dirty {
eq = v.Compare(c.con) == 1
if eq {
return true, nil
}
return false, fmt.Errorf("%s is less than or equal to %s", v, c.orig)
}
if v.Major() > c.con.Major() {
return true, nil
} else if v.Major() < c.con.Major() {
return false, fmt.Errorf("%s is less than or equal to %s", v, c.orig)
} else if c.minorDirty {
// This is a range case such as >11. When the version is something like
// 11.1.0 is it not > 11. For that we would need 12 or higher
return false, fmt.Errorf("%s is less than or equal to %s", v, c.orig)
} else if c.patchDirty {
// This is for ranges such as >11.1. A version of 11.1.1 is not greater
// which one of 11.2.1 is greater
eq = v.Minor() > c.con.Minor()
if eq {
return true, nil
}
return false, fmt.Errorf("%s is less than or equal to %s", v, c.orig)
}
// If we have gotten here we are not comparing pre-preleases and can use the
// Compare function to accomplish that.
eq = v.Compare(c.con) == 1
if eq {
return true, nil
}
return false, fmt.Errorf("%s is less than or equal to %s", v, c.orig)
}
func constraintLessThan(v *Version, c *constraint) (bool, error) {
// If there is a pre-release on the version but the constraint isn't looking
// for them assume that pre-releases are not compatible. See issue 21 for
// more details.
if v.Prerelease() != "" && c.con.Prerelease() == "" {
return false, fmt.Errorf("%s is a prerelease version and the constraint is only looking for release versions", v)
}
eq := v.Compare(c.con) < 0
if eq {
return true, nil
}
return false, fmt.Errorf("%s is greater than or equal to %s", v, c.orig)
}
func constraintGreaterThanEqual(v *Version, c *constraint) (bool, error) {
// If there is a pre-release on the version but the constraint isn't looking
// for them assume that pre-releases are not compatible. See issue 21 for
// more details.
if v.Prerelease() != "" && c.con.Prerelease() == "" {
return false, fmt.Errorf("%s is a prerelease version and the constraint is only looking for release versions", v)
}
eq := v.Compare(c.con) >= 0
if eq {
return true, nil
}
return false, fmt.Errorf("%s is less than %s", v, c.orig)
}
func constraintLessThanEqual(v *Version, c *constraint) (bool, error) {
// If there is a pre-release on the version but the constraint isn't looking
// for them assume that pre-releases are not compatible. See issue 21 for
// more details.
if v.Prerelease() != "" && c.con.Prerelease() == "" {
return false, fmt.Errorf("%s is a prerelease version and the constraint is only looking for release versions", v)
}
var eq bool
if !c.dirty {
eq = v.Compare(c.con) <= 0
if eq {
return true, nil
}
return false, fmt.Errorf("%s is greater than %s", v, c.orig)
}
if v.Major() > c.con.Major() {
return false, fmt.Errorf("%s is greater than %s", v, c.orig)
} else if v.Major() == c.con.Major() && v.Minor() > c.con.Minor() && !c.minorDirty {
return false, fmt.Errorf("%s is greater than %s", v, c.orig)
}
return true, nil
}
// ~*, ~>* --> >= 0.0.0 (any)
// ~2, ~2.x, ~2.x.x, ~>2, ~>2.x ~>2.x.x --> >=2.0.0, <3.0.0
// ~2.0, ~2.0.x, ~>2.0, ~>2.0.x --> >=2.0.0, <2.1.0
// ~1.2, ~1.2.x, ~>1.2, ~>1.2.x --> >=1.2.0, <1.3.0
// ~1.2.3, ~>1.2.3 --> >=1.2.3, <1.3.0
// ~1.2.0, ~>1.2.0 --> >=1.2.0, <1.3.0
func constraintTilde(v *Version, c *constraint) (bool, error) {
// If there is a pre-release on the version but the constraint isn't looking
// for them assume that pre-releases are not compatible. See issue 21 for
// more details.
if v.Prerelease() != "" && c.con.Prerelease() == "" {
return false, fmt.Errorf("%s is a prerelease version and the constraint is only looking for release versions", v)
}
if v.LessThan(c.con) {
return false, fmt.Errorf("%s is less than %s", v, c.orig)
}
// ~0.0.0 is a special case where all constraints are accepted. It's
// equivalent to >= 0.0.0.
if c.con.Major() == 0 && c.con.Minor() == 0 && c.con.Patch() == 0 &&
!c.minorDirty && !c.patchDirty {
return true, nil
}
if v.Major() != c.con.Major() {
return false, fmt.Errorf("%s does not have same major version as %s", v, c.orig)
}
if v.Minor() != c.con.Minor() && !c.minorDirty {
return false, fmt.Errorf("%s does not have same major and minor version as %s", v, c.orig)
}
return true, nil
}
// When there is a .x (dirty) status it automatically opts in to ~. Otherwise
// it's a straight =
func constraintTildeOrEqual(v *Version, c *constraint) (bool, error) {
// If there is a pre-release on the version but the constraint isn't looking
// for them assume that pre-releases are not compatible. See issue 21 for
// more details.
if v.Prerelease() != "" && c.con.Prerelease() == "" {
return false, fmt.Errorf("%s is a prerelease version and the constraint is only looking for release versions", v)
}
if c.dirty {
return constraintTilde(v, c)
}
eq := v.Equal(c.con)
if eq {
return true, nil
}
return false, fmt.Errorf("%s is not equal to %s", v, c.orig)
}
// ^* --> (any)
// ^1.2.3 --> >=1.2.3 <2.0.0
// ^1.2 --> >=1.2.0 <2.0.0
// ^1 --> >=1.0.0 <2.0.0
// ^0.2.3 --> >=0.2.3 <0.3.0
// ^0.2 --> >=0.2.0 <0.3.0
// ^0.0.3 --> >=0.0.3 <0.0.4
// ^0.0 --> >=0.0.0 <0.1.0
// ^0 --> >=0.0.0 <1.0.0
func constraintCaret(v *Version, c *constraint) (bool, error) {
// If there is a pre-release on the version but the constraint isn't looking
// for them assume that pre-releases are not compatible. See issue 21 for
// more details.
if v.Prerelease() != "" && c.con.Prerelease() == "" {
return false, fmt.Errorf("%s is a prerelease version and the constraint is only looking for release versions", v)
}
// This less than handles prereleases
if v.LessThan(c.con) {
return false, fmt.Errorf("%s is less than %s", v, c.orig)
}
var eq bool
// ^ when the major > 0 is >=x.y.z < x+1
if c.con.Major() > 0 || c.minorDirty {
// ^ has to be within a major range for > 0. Everything less than was
// filtered out with the LessThan call above. This filters out those
// that greater but not within the same major range.
eq = v.Major() == c.con.Major()
if eq {
return true, nil
}
return false, fmt.Errorf("%s does not have same major version as %s", v, c.orig)
}
// ^ when the major is 0 and minor > 0 is >=0.y.z < 0.y+1
if c.con.Major() == 0 && v.Major() > 0 {
return false, fmt.Errorf("%s does not have same major version as %s", v, c.orig)
}
// If the con Minor is > 0 it is not dirty
if c.con.Minor() > 0 || c.patchDirty {
eq = v.Minor() == c.con.Minor()
if eq {
return true, nil
}
return false, fmt.Errorf("%s does not have same minor version as %s. Expected minor versions to match when constraint major version is 0", v, c.orig)
}
// ^ when the minor is 0 and minor > 0 is =0.0.z
if c.con.Minor() == 0 && v.Minor() > 0 {
return false, fmt.Errorf("%s does not have same minor version as %s", v, c.orig)
}
// At this point the major is 0 and the minor is 0 and not dirty. The patch
// is not dirty so we need to check if they are equal. If they are not equal
eq = c.con.Patch() == v.Patch()
if eq {
return true, nil
}
return false, fmt.Errorf("%s does not equal %s. Expect version and constraint to equal when major and minor versions are 0", v, c.orig)
}
func isX(x string) bool {
switch x {
case "x", "*", "X":
return true
default:
return false
}
}
func rewriteRange(i string) string {
m := constraintRangeRegex.FindAllStringSubmatch(i, -1)
if m == nil {
return i
}
o := i
for _, v := range m {
t := fmt.Sprintf(">= %s, <= %s ", v[1], v[11])
o = strings.Replace(o, v[0], t, 1)
}
return o
}

View File

@@ -1,184 +0,0 @@
/*
Package semver provides the ability to work with Semantic Versions (http://semver.org) in Go.
Specifically it provides the ability to:
- Parse semantic versions
- Sort semantic versions
- Check if a semantic version fits within a set of constraints
- Optionally work with a `v` prefix
# Parsing Semantic Versions
There are two functions that can parse semantic versions. The `StrictNewVersion`
function only parses valid version 2 semantic versions as outlined in the
specification. The `NewVersion` function attempts to coerce a version into a
semantic version and parse it. For example, if there is a leading v or a version
listed without all 3 parts (e.g. 1.2) it will attempt to coerce it into a valid
semantic version (e.g., 1.2.0). In both cases a `Version` object is returned
that can be sorted, compared, and used in constraints.
When parsing a version an optional error can be returned if there is an issue
parsing the version. For example,
v, err := semver.NewVersion("1.2.3-beta.1+b345")
The version object has methods to get the parts of the version, compare it to
other versions, convert the version back into a string, and get the original
string. For more details please see the documentation
at https://godoc.org/github.com/Masterminds/semver.
# Sorting Semantic Versions
A set of versions can be sorted using the `sort` package from the standard library.
For example,
raw := []string{"1.2.3", "1.0", "1.3", "2", "0.4.2",}
vs := make([]*semver.Version, len(raw))
for i, r := range raw {
v, err := semver.NewVersion(r)
if err != nil {
t.Errorf("Error parsing version: %s", err)
}
vs[i] = v
}
sort.Sort(semver.Collection(vs))
# Checking Version Constraints and Comparing Versions
There are two methods for comparing versions. One uses comparison methods on
`Version` instances and the other is using Constraints. There are some important
differences to notes between these two methods of comparison.
1. When two versions are compared using functions such as `Compare`, `LessThan`,
and others it will follow the specification and always include prereleases
within the comparison. It will provide an answer valid with the comparison
spec section at https://semver.org/#spec-item-11
2. When constraint checking is used for checks or validation it will follow a
different set of rules that are common for ranges with tools like npm/js
and Rust/Cargo. This includes considering prereleases to be invalid if the
ranges does not include on. If you want to have it include pre-releases a
simple solution is to include `-0` in your range.
3. Constraint ranges can have some complex rules including the shorthard use of
~ and ^. For more details on those see the options below.
There are differences between the two methods or checking versions because the
comparison methods on `Version` follow the specification while comparison ranges
are not part of the specification. Different packages and tools have taken it
upon themselves to come up with range rules. This has resulted in differences.
For example, npm/js and Cargo/Rust follow similar patterns which PHP has a
different pattern for ^. The comparison features in this package follow the
npm/js and Cargo/Rust lead because applications using it have followed similar
patters with their versions.
Checking a version against version constraints is one of the most featureful
parts of the package.
c, err := semver.NewConstraint(">= 1.2.3")
if err != nil {
// Handle constraint not being parsable.
}
v, err := semver.NewVersion("1.3")
if err != nil {
// Handle version not being parsable.
}
// Check if the version meets the constraints. The a variable will be true.
a := c.Check(v)
# Basic Comparisons
There are two elements to the comparisons. First, a comparison string is a list
of comma or space separated AND comparisons. These are then separated by || (OR)
comparisons. For example, `">= 1.2 < 3.0.0 || >= 4.2.3"` is looking for a
comparison that's greater than or equal to 1.2 and less than 3.0.0 or is
greater than or equal to 4.2.3. This can also be written as
`">= 1.2, < 3.0.0 || >= 4.2.3"`
The basic comparisons are:
- `=`: equal (aliased to no operator)
- `!=`: not equal
- `>`: greater than
- `<`: less than
- `>=`: greater than or equal to
- `<=`: less than or equal to
# Hyphen Range Comparisons
There are multiple methods to handle ranges and the first is hyphens ranges.
These look like:
- `1.2 - 1.4.5` which is equivalent to `>= 1.2, <= 1.4.5`
- `2.3.4 - 4.5` which is equivalent to `>= 2.3.4 <= 4.5`
# Wildcards In Comparisons
The `x`, `X`, and `*` characters can be used as a wildcard character. This works
for all comparison operators. When used on the `=` operator it falls
back to the tilde operation. For example,
- `1.2.x` is equivalent to `>= 1.2.0 < 1.3.0`
- `>= 1.2.x` is equivalent to `>= 1.2.0`
- `<= 2.x` is equivalent to `<= 3`
- `*` is equivalent to `>= 0.0.0`
Tilde Range Comparisons (Patch)
The tilde (`~`) comparison operator is for patch level ranges when a minor
version is specified and major level changes when the minor number is missing.
For example,
- `~1.2.3` is equivalent to `>= 1.2.3 < 1.3.0`
- `~1` is equivalent to `>= 1, < 2`
- `~2.3` is equivalent to `>= 2.3 < 2.4`
- `~1.2.x` is equivalent to `>= 1.2.0 < 1.3.0`
- `~1.x` is equivalent to `>= 1 < 2`
Caret Range Comparisons (Major)
The caret (`^`) comparison operator is for major level changes once a stable
(1.0.0) release has occurred. Prior to a 1.0.0 release the minor versions acts
as the API stability level. This is useful when comparisons of API versions as a
major change is API breaking. For example,
- `^1.2.3` is equivalent to `>= 1.2.3, < 2.0.0`
- `^1.2.x` is equivalent to `>= 1.2.0, < 2.0.0`
- `^2.3` is equivalent to `>= 2.3, < 3`
- `^2.x` is equivalent to `>= 2.0.0, < 3`
- `^0.2.3` is equivalent to `>=0.2.3 <0.3.0`
- `^0.2` is equivalent to `>=0.2.0 <0.3.0`
- `^0.0.3` is equivalent to `>=0.0.3 <0.0.4`
- `^0.0` is equivalent to `>=0.0.0 <0.1.0`
- `^0` is equivalent to `>=0.0.0 <1.0.0`
# Validation
In addition to testing a version against a constraint, a version can be validated
against a constraint. When validation fails a slice of errors containing why a
version didn't meet the constraint is returned. For example,
c, err := semver.NewConstraint("<= 1.2.3, >= 1.4")
if err != nil {
// Handle constraint not being parseable.
}
v, _ := semver.NewVersion("1.3")
if err != nil {
// Handle version not being parseable.
}
// Validate a version against a constraint.
a, msgs := c.Validate(v)
// a is false
for _, m := range msgs {
fmt.Println(m)
// Loops over the errors which would read
// "1.3 is greater than 1.2.3"
// "1.3 is less than 1.4"
}
*/
package semver

View File

@@ -1,639 +0,0 @@
package semver
import (
"bytes"
"database/sql/driver"
"encoding/json"
"errors"
"fmt"
"regexp"
"strconv"
"strings"
)
// The compiled version of the regex created at init() is cached here so it
// only needs to be created once.
var versionRegex *regexp.Regexp
var (
// ErrInvalidSemVer is returned a version is found to be invalid when
// being parsed.
ErrInvalidSemVer = errors.New("Invalid Semantic Version")
// ErrEmptyString is returned when an empty string is passed in for parsing.
ErrEmptyString = errors.New("Version string empty")
// ErrInvalidCharacters is returned when invalid characters are found as
// part of a version
ErrInvalidCharacters = errors.New("Invalid characters in version")
// ErrSegmentStartsZero is returned when a version segment starts with 0.
// This is invalid in SemVer.
ErrSegmentStartsZero = errors.New("Version segment starts with 0")
// ErrInvalidMetadata is returned when the metadata is an invalid format
ErrInvalidMetadata = errors.New("Invalid Metadata string")
// ErrInvalidPrerelease is returned when the pre-release is an invalid format
ErrInvalidPrerelease = errors.New("Invalid Prerelease string")
)
// semVerRegex is the regular expression used to parse a semantic version.
const semVerRegex string = `v?([0-9]+)(\.[0-9]+)?(\.[0-9]+)?` +
`(-([0-9A-Za-z\-]+(\.[0-9A-Za-z\-]+)*))?` +
`(\+([0-9A-Za-z\-]+(\.[0-9A-Za-z\-]+)*))?`
// Version represents a single semantic version.
type Version struct {
major, minor, patch uint64
pre string
metadata string
original string
}
func init() {
versionRegex = regexp.MustCompile("^" + semVerRegex + "$")
}
const (
num string = "0123456789"
allowed string = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ-" + num
)
// StrictNewVersion parses a given version and returns an instance of Version or
// an error if unable to parse the version. Only parses valid semantic versions.
// Performs checking that can find errors within the version.
// If you want to coerce a version such as 1 or 1.2 and parse it as the 1.x
// releases of semver did, use the NewVersion() function.
func StrictNewVersion(v string) (*Version, error) {
// Parsing here does not use RegEx in order to increase performance and reduce
// allocations.
if len(v) == 0 {
return nil, ErrEmptyString
}
// Split the parts into [0]major, [1]minor, and [2]patch,prerelease,build
parts := strings.SplitN(v, ".", 3)
if len(parts) != 3 {
return nil, ErrInvalidSemVer
}
sv := &Version{
original: v,
}
// check for prerelease or build metadata
var extra []string
if strings.ContainsAny(parts[2], "-+") {
// Start with the build metadata first as it needs to be on the right
extra = strings.SplitN(parts[2], "+", 2)
if len(extra) > 1 {
// build metadata found
sv.metadata = extra[1]
parts[2] = extra[0]
}
extra = strings.SplitN(parts[2], "-", 2)
if len(extra) > 1 {
// prerelease found
sv.pre = extra[1]
parts[2] = extra[0]
}
}
// Validate the number segments are valid. This includes only having positive
// numbers and no leading 0's.
for _, p := range parts {
if !containsOnly(p, num) {
return nil, ErrInvalidCharacters
}
if len(p) > 1 && p[0] == '0' {
return nil, ErrSegmentStartsZero
}
}
// Extract the major, minor, and patch elements onto the returned Version
var err error
sv.major, err = strconv.ParseUint(parts[0], 10, 64)
if err != nil {
return nil, err
}
sv.minor, err = strconv.ParseUint(parts[1], 10, 64)
if err != nil {
return nil, err
}
sv.patch, err = strconv.ParseUint(parts[2], 10, 64)
if err != nil {
return nil, err
}
// No prerelease or build metadata found so returning now as a fastpath.
if sv.pre == "" && sv.metadata == "" {
return sv, nil
}
if sv.pre != "" {
if err = validatePrerelease(sv.pre); err != nil {
return nil, err
}
}
if sv.metadata != "" {
if err = validateMetadata(sv.metadata); err != nil {
return nil, err
}
}
return sv, nil
}
// NewVersion parses a given version and returns an instance of Version or
// an error if unable to parse the version. If the version is SemVer-ish it
// attempts to convert it to SemVer. If you want to validate it was a strict
// semantic version at parse time see StrictNewVersion().
func NewVersion(v string) (*Version, error) {
m := versionRegex.FindStringSubmatch(v)
if m == nil {
return nil, ErrInvalidSemVer
}
sv := &Version{
metadata: m[8],
pre: m[5],
original: v,
}
var err error
sv.major, err = strconv.ParseUint(m[1], 10, 64)
if err != nil {
return nil, fmt.Errorf("Error parsing version segment: %s", err)
}
if m[2] != "" {
sv.minor, err = strconv.ParseUint(strings.TrimPrefix(m[2], "."), 10, 64)
if err != nil {
return nil, fmt.Errorf("Error parsing version segment: %s", err)
}
} else {
sv.minor = 0
}
if m[3] != "" {
sv.patch, err = strconv.ParseUint(strings.TrimPrefix(m[3], "."), 10, 64)
if err != nil {
return nil, fmt.Errorf("Error parsing version segment: %s", err)
}
} else {
sv.patch = 0
}
// Perform some basic due diligence on the extra parts to ensure they are
// valid.
if sv.pre != "" {
if err = validatePrerelease(sv.pre); err != nil {
return nil, err
}
}
if sv.metadata != "" {
if err = validateMetadata(sv.metadata); err != nil {
return nil, err
}
}
return sv, nil
}
// New creates a new instance of Version with each of the parts passed in as
// arguments instead of parsing a version string.
func New(major, minor, patch uint64, pre, metadata string) *Version {
v := Version{
major: major,
minor: minor,
patch: patch,
pre: pre,
metadata: metadata,
original: "",
}
v.original = v.String()
return &v
}
// MustParse parses a given version and panics on error.
func MustParse(v string) *Version {
sv, err := NewVersion(v)
if err != nil {
panic(err)
}
return sv
}
// String converts a Version object to a string.
// Note, if the original version contained a leading v this version will not.
// See the Original() method to retrieve the original value. Semantic Versions
// don't contain a leading v per the spec. Instead it's optional on
// implementation.
func (v Version) String() string {
var buf bytes.Buffer
fmt.Fprintf(&buf, "%d.%d.%d", v.major, v.minor, v.patch)
if v.pre != "" {
fmt.Fprintf(&buf, "-%s", v.pre)
}
if v.metadata != "" {
fmt.Fprintf(&buf, "+%s", v.metadata)
}
return buf.String()
}
// Original returns the original value passed in to be parsed.
func (v *Version) Original() string {
return v.original
}
// Major returns the major version.
func (v Version) Major() uint64 {
return v.major
}
// Minor returns the minor version.
func (v Version) Minor() uint64 {
return v.minor
}
// Patch returns the patch version.
func (v Version) Patch() uint64 {
return v.patch
}
// Prerelease returns the pre-release version.
func (v Version) Prerelease() string {
return v.pre
}
// Metadata returns the metadata on the version.
func (v Version) Metadata() string {
return v.metadata
}
// originalVPrefix returns the original 'v' prefix if any.
func (v Version) originalVPrefix() string {
// Note, only lowercase v is supported as a prefix by the parser.
if v.original != "" && v.original[:1] == "v" {
return v.original[:1]
}
return ""
}
// IncPatch produces the next patch version.
// If the current version does not have prerelease/metadata information,
// it unsets metadata and prerelease values, increments patch number.
// If the current version has any of prerelease or metadata information,
// it unsets both values and keeps current patch value
func (v Version) IncPatch() Version {
vNext := v
// according to http://semver.org/#spec-item-9
// Pre-release versions have a lower precedence than the associated normal version.
// according to http://semver.org/#spec-item-10
// Build metadata SHOULD be ignored when determining version precedence.
if v.pre != "" {
vNext.metadata = ""
vNext.pre = ""
} else {
vNext.metadata = ""
vNext.pre = ""
vNext.patch = v.patch + 1
}
vNext.original = v.originalVPrefix() + "" + vNext.String()
return vNext
}
// IncMinor produces the next minor version.
// Sets patch to 0.
// Increments minor number.
// Unsets metadata.
// Unsets prerelease status.
func (v Version) IncMinor() Version {
vNext := v
vNext.metadata = ""
vNext.pre = ""
vNext.patch = 0
vNext.minor = v.minor + 1
vNext.original = v.originalVPrefix() + "" + vNext.String()
return vNext
}
// IncMajor produces the next major version.
// Sets patch to 0.
// Sets minor to 0.
// Increments major number.
// Unsets metadata.
// Unsets prerelease status.
func (v Version) IncMajor() Version {
vNext := v
vNext.metadata = ""
vNext.pre = ""
vNext.patch = 0
vNext.minor = 0
vNext.major = v.major + 1
vNext.original = v.originalVPrefix() + "" + vNext.String()
return vNext
}
// SetPrerelease defines the prerelease value.
// Value must not include the required 'hyphen' prefix.
func (v Version) SetPrerelease(prerelease string) (Version, error) {
vNext := v
if len(prerelease) > 0 {
if err := validatePrerelease(prerelease); err != nil {
return vNext, err
}
}
vNext.pre = prerelease
vNext.original = v.originalVPrefix() + "" + vNext.String()
return vNext, nil
}
// SetMetadata defines metadata value.
// Value must not include the required 'plus' prefix.
func (v Version) SetMetadata(metadata string) (Version, error) {
vNext := v
if len(metadata) > 0 {
if err := validateMetadata(metadata); err != nil {
return vNext, err
}
}
vNext.metadata = metadata
vNext.original = v.originalVPrefix() + "" + vNext.String()
return vNext, nil
}
// LessThan tests if one version is less than another one.
func (v *Version) LessThan(o *Version) bool {
return v.Compare(o) < 0
}
// GreaterThan tests if one version is greater than another one.
func (v *Version) GreaterThan(o *Version) bool {
return v.Compare(o) > 0
}
// Equal tests if two versions are equal to each other.
// Note, versions can be equal with different metadata since metadata
// is not considered part of the comparable version.
func (v *Version) Equal(o *Version) bool {
return v.Compare(o) == 0
}
// Compare compares this version to another one. It returns -1, 0, or 1 if
// the version smaller, equal, or larger than the other version.
//
// Versions are compared by X.Y.Z. Build metadata is ignored. Prerelease is
// lower than the version without a prerelease. Compare always takes into account
// prereleases. If you want to work with ranges using typical range syntaxes that
// skip prereleases if the range is not looking for them use constraints.
func (v *Version) Compare(o *Version) int {
// Compare the major, minor, and patch version for differences. If a
// difference is found return the comparison.
if d := compareSegment(v.Major(), o.Major()); d != 0 {
return d
}
if d := compareSegment(v.Minor(), o.Minor()); d != 0 {
return d
}
if d := compareSegment(v.Patch(), o.Patch()); d != 0 {
return d
}
// At this point the major, minor, and patch versions are the same.
ps := v.pre
po := o.Prerelease()
if ps == "" && po == "" {
return 0
}
if ps == "" {
return 1
}
if po == "" {
return -1
}
return comparePrerelease(ps, po)
}
// UnmarshalJSON implements JSON.Unmarshaler interface.
func (v *Version) UnmarshalJSON(b []byte) error {
var s string
if err := json.Unmarshal(b, &s); err != nil {
return err
}
temp, err := NewVersion(s)
if err != nil {
return err
}
v.major = temp.major
v.minor = temp.minor
v.patch = temp.patch
v.pre = temp.pre
v.metadata = temp.metadata
v.original = temp.original
return nil
}
// MarshalJSON implements JSON.Marshaler interface.
func (v Version) MarshalJSON() ([]byte, error) {
return json.Marshal(v.String())
}
// UnmarshalText implements the encoding.TextUnmarshaler interface.
func (v *Version) UnmarshalText(text []byte) error {
temp, err := NewVersion(string(text))
if err != nil {
return err
}
*v = *temp
return nil
}
// MarshalText implements the encoding.TextMarshaler interface.
func (v Version) MarshalText() ([]byte, error) {
return []byte(v.String()), nil
}
// Scan implements the SQL.Scanner interface.
func (v *Version) Scan(value interface{}) error {
var s string
s, _ = value.(string)
temp, err := NewVersion(s)
if err != nil {
return err
}
v.major = temp.major
v.minor = temp.minor
v.patch = temp.patch
v.pre = temp.pre
v.metadata = temp.metadata
v.original = temp.original
return nil
}
// Value implements the Driver.Valuer interface.
func (v Version) Value() (driver.Value, error) {
return v.String(), nil
}
func compareSegment(v, o uint64) int {
if v < o {
return -1
}
if v > o {
return 1
}
return 0
}
func comparePrerelease(v, o string) int {
// split the prelease versions by their part. The separator, per the spec,
// is a .
sparts := strings.Split(v, ".")
oparts := strings.Split(o, ".")
// Find the longer length of the parts to know how many loop iterations to
// go through.
slen := len(sparts)
olen := len(oparts)
l := slen
if olen > slen {
l = olen
}
// Iterate over each part of the prereleases to compare the differences.
for i := 0; i < l; i++ {
// Since the lentgh of the parts can be different we need to create
// a placeholder. This is to avoid out of bounds issues.
stemp := ""
if i < slen {
stemp = sparts[i]
}
otemp := ""
if i < olen {
otemp = oparts[i]
}
d := comparePrePart(stemp, otemp)
if d != 0 {
return d
}
}
// Reaching here means two versions are of equal value but have different
// metadata (the part following a +). They are not identical in string form
// but the version comparison finds them to be equal.
return 0
}
func comparePrePart(s, o string) int {
// Fastpath if they are equal
if s == o {
return 0
}
// When s or o are empty we can use the other in an attempt to determine
// the response.
if s == "" {
if o != "" {
return -1
}
return 1
}
if o == "" {
if s != "" {
return 1
}
return -1
}
// When comparing strings "99" is greater than "103". To handle
// cases like this we need to detect numbers and compare them. According
// to the semver spec, numbers are always positive. If there is a - at the
// start like -99 this is to be evaluated as an alphanum. numbers always
// have precedence over alphanum. Parsing as Uints because negative numbers
// are ignored.
oi, n1 := strconv.ParseUint(o, 10, 64)
si, n2 := strconv.ParseUint(s, 10, 64)
// The case where both are strings compare the strings
if n1 != nil && n2 != nil {
if s > o {
return 1
}
return -1
} else if n1 != nil {
// o is a string and s is a number
return -1
} else if n2 != nil {
// s is a string and o is a number
return 1
}
// Both are numbers
if si > oi {
return 1
}
return -1
}
// Like strings.ContainsAny but does an only instead of any.
func containsOnly(s string, comp string) bool {
return strings.IndexFunc(s, func(r rune) bool {
return !strings.ContainsRune(comp, r)
}) == -1
}
// From the spec, "Identifiers MUST comprise only
// ASCII alphanumerics and hyphen [0-9A-Za-z-]. Identifiers MUST NOT be empty.
// Numeric identifiers MUST NOT include leading zeroes.". These segments can
// be dot separated.
func validatePrerelease(p string) error {
eparts := strings.Split(p, ".")
for _, p := range eparts {
if containsOnly(p, num) {
if len(p) > 1 && p[0] == '0' {
return ErrSegmentStartsZero
}
} else if !containsOnly(p, allowed) {
return ErrInvalidPrerelease
}
}
return nil
}
// From the spec, "Build metadata MAY be denoted by
// appending a plus sign and a series of dot separated identifiers immediately
// following the patch or pre-release version. Identifiers MUST comprise only
// ASCII alphanumerics and hyphen [0-9A-Za-z-]. Identifiers MUST NOT be empty."
func validateMetadata(m string) error {
eparts := strings.Split(m, ".")
for _, p := range eparts {
if !containsOnly(p, allowed) {
return ErrInvalidMetadata
}
}
return nil
}

View File

@@ -23,6 +23,7 @@ package libcni
import (
"context"
"encoding/json"
"errors"
"fmt"
"os"
"path/filepath"
@@ -66,17 +67,23 @@ type RuntimeConf struct {
CacheDir string
}
type NetworkConfig struct {
Network *types.NetConf
// Use PluginConfig instead of NetworkConfig, the NetworkConfig
// backwards-compat alias will be removed in a future release.
type NetworkConfig = PluginConfig
type PluginConfig struct {
Network *types.PluginConf
Bytes []byte
}
type NetworkConfigList struct {
Name string
CNIVersion string
DisableCheck bool
Plugins []*NetworkConfig
Bytes []byte
Name string
CNIVersion string
DisableCheck bool
DisableGC bool
LoadOnlyInlinedPlugins bool
Plugins []*PluginConfig
Bytes []byte
}
type NetworkAttachment struct {
@@ -100,19 +107,21 @@ type CNI interface {
GetNetworkListCachedResult(net *NetworkConfigList, rt *RuntimeConf) (types.Result, error)
GetNetworkListCachedConfig(net *NetworkConfigList, rt *RuntimeConf) ([]byte, *RuntimeConf, error)
AddNetwork(ctx context.Context, net *NetworkConfig, rt *RuntimeConf) (types.Result, error)
CheckNetwork(ctx context.Context, net *NetworkConfig, rt *RuntimeConf) error
DelNetwork(ctx context.Context, net *NetworkConfig, rt *RuntimeConf) error
GetNetworkCachedResult(net *NetworkConfig, rt *RuntimeConf) (types.Result, error)
GetNetworkCachedConfig(net *NetworkConfig, rt *RuntimeConf) ([]byte, *RuntimeConf, error)
AddNetwork(ctx context.Context, net *PluginConfig, rt *RuntimeConf) (types.Result, error)
CheckNetwork(ctx context.Context, net *PluginConfig, rt *RuntimeConf) error
DelNetwork(ctx context.Context, net *PluginConfig, rt *RuntimeConf) error
GetNetworkCachedResult(net *PluginConfig, rt *RuntimeConf) (types.Result, error)
GetNetworkCachedConfig(net *PluginConfig, rt *RuntimeConf) ([]byte, *RuntimeConf, error)
ValidateNetworkList(ctx context.Context, net *NetworkConfigList) ([]string, error)
ValidateNetwork(ctx context.Context, net *NetworkConfig) ([]string, error)
ValidateNetwork(ctx context.Context, net *PluginConfig) ([]string, error)
GCNetworkList(ctx context.Context, net *NetworkConfigList, args *GCArgs) error
GetStatusNetworkList(ctx context.Context, net *NetworkConfigList) error
GetCachedAttachments(containerID string) ([]*NetworkAttachment, error)
GetVersionInfo(ctx context.Context, pluginType string) (version.PluginInfo, error)
}
type CNIConfig struct {
@@ -143,7 +152,7 @@ func NewCNIConfigWithCacheDir(path []string, cacheDir string, exec invoke.Exec)
}
}
func buildOneConfig(name, cniVersion string, orig *NetworkConfig, prevResult types.Result, rt *RuntimeConf) (*NetworkConfig, error) {
func buildOneConfig(name, cniVersion string, orig *PluginConfig, prevResult types.Result, rt *RuntimeConf) (*PluginConfig, error) {
var err error
inject := map[string]interface{}{
@@ -179,7 +188,7 @@ func buildOneConfig(name, cniVersion string, orig *NetworkConfig, prevResult typ
// capabilities include "portMappings", and the CapabilityArgs map includes a
// "portMappings" key, that key and its value are added to the "runtimeConfig"
// dictionary to be passed to the plugin's stdin.
func injectRuntimeConfig(orig *NetworkConfig, rt *RuntimeConf) (*NetworkConfig, error) {
func injectRuntimeConfig(orig *PluginConfig, rt *RuntimeConf) (*PluginConfig, error) {
var err error
rc := make(map[string]interface{})
@@ -400,7 +409,7 @@ func (c *CNIConfig) GetNetworkListCachedResult(list *NetworkConfigList, rt *Runt
// GetNetworkCachedResult returns the cached Result of the previous
// AddNetwork() operation for a network, or an error.
func (c *CNIConfig) GetNetworkCachedResult(net *NetworkConfig, rt *RuntimeConf) (types.Result, error) {
func (c *CNIConfig) GetNetworkCachedResult(net *PluginConfig, rt *RuntimeConf) (types.Result, error) {
return c.getCachedResult(net.Network.Name, net.Network.CNIVersion, rt)
}
@@ -412,7 +421,7 @@ func (c *CNIConfig) GetNetworkListCachedConfig(list *NetworkConfigList, rt *Runt
// GetNetworkCachedConfig copies the input RuntimeConf to output
// RuntimeConf with fields updated with info from the cached Config.
func (c *CNIConfig) GetNetworkCachedConfig(net *NetworkConfig, rt *RuntimeConf) ([]byte, *RuntimeConf, error) {
func (c *CNIConfig) GetNetworkCachedConfig(net *PluginConfig, rt *RuntimeConf) ([]byte, *RuntimeConf, error) {
return c.getCachedConfig(net.Network.Name, rt)
}
@@ -422,6 +431,9 @@ func (c *CNIConfig) GetCachedAttachments(containerID string) ([]*NetworkAttachme
dirPath := filepath.Join(c.getCacheDir(&RuntimeConf{}), "results")
entries, err := os.ReadDir(dirPath)
if err != nil {
if os.IsNotExist(err) {
return nil, nil
}
return nil, err
}
@@ -475,7 +487,7 @@ func (c *CNIConfig) GetCachedAttachments(containerID string) ([]*NetworkAttachme
return attachments, nil
}
func (c *CNIConfig) addNetwork(ctx context.Context, name, cniVersion string, net *NetworkConfig, prevResult types.Result, rt *RuntimeConf) (types.Result, error) {
func (c *CNIConfig) addNetwork(ctx context.Context, name, cniVersion string, net *PluginConfig, prevResult types.Result, rt *RuntimeConf) (types.Result, error) {
c.ensureExec()
pluginPath, err := c.exec.FindInPath(net.Network.Type, c.Path)
if err != nil {
@@ -517,7 +529,7 @@ func (c *CNIConfig) AddNetworkList(ctx context.Context, list *NetworkConfigList,
return result, nil
}
func (c *CNIConfig) checkNetwork(ctx context.Context, name, cniVersion string, net *NetworkConfig, prevResult types.Result, rt *RuntimeConf) error {
func (c *CNIConfig) checkNetwork(ctx context.Context, name, cniVersion string, net *PluginConfig, prevResult types.Result, rt *RuntimeConf) error {
c.ensureExec()
pluginPath, err := c.exec.FindInPath(net.Network.Type, c.Path)
if err != nil {
@@ -559,7 +571,7 @@ func (c *CNIConfig) CheckNetworkList(ctx context.Context, list *NetworkConfigLis
return nil
}
func (c *CNIConfig) delNetwork(ctx context.Context, name, cniVersion string, net *NetworkConfig, prevResult types.Result, rt *RuntimeConf) error {
func (c *CNIConfig) delNetwork(ctx context.Context, name, cniVersion string, net *PluginConfig, prevResult types.Result, rt *RuntimeConf) error {
c.ensureExec()
pluginPath, err := c.exec.FindInPath(net.Network.Type, c.Path)
if err != nil {
@@ -595,14 +607,12 @@ func (c *CNIConfig) DelNetworkList(ctx context.Context, list *NetworkConfigList,
}
}
if cachedResult != nil {
_ = c.cacheDel(list.Name, rt)
}
_ = c.cacheDel(list.Name, rt)
return nil
}
func pluginDescription(net *types.NetConf) string {
func pluginDescription(net *types.PluginConf) string {
if net == nil {
return "<missing>"
}
@@ -616,7 +626,7 @@ func pluginDescription(net *types.NetConf) string {
}
// AddNetwork executes the plugin with the ADD command
func (c *CNIConfig) AddNetwork(ctx context.Context, net *NetworkConfig, rt *RuntimeConf) (types.Result, error) {
func (c *CNIConfig) AddNetwork(ctx context.Context, net *PluginConfig, rt *RuntimeConf) (types.Result, error) {
result, err := c.addNetwork(ctx, net.Network.Name, net.Network.CNIVersion, net, nil, rt)
if err != nil {
return nil, err
@@ -630,7 +640,7 @@ func (c *CNIConfig) AddNetwork(ctx context.Context, net *NetworkConfig, rt *Runt
}
// CheckNetwork executes the plugin with the CHECK command
func (c *CNIConfig) CheckNetwork(ctx context.Context, net *NetworkConfig, rt *RuntimeConf) error {
func (c *CNIConfig) CheckNetwork(ctx context.Context, net *PluginConfig, rt *RuntimeConf) error {
// CHECK was added in CNI spec version 0.4.0 and higher
if gtet, err := version.GreaterThanOrEqualTo(net.Network.CNIVersion, "0.4.0"); err != nil {
return err
@@ -646,7 +656,7 @@ func (c *CNIConfig) CheckNetwork(ctx context.Context, net *NetworkConfig, rt *Ru
}
// DelNetwork executes the plugin with the DEL command
func (c *CNIConfig) DelNetwork(ctx context.Context, net *NetworkConfig, rt *RuntimeConf) error {
func (c *CNIConfig) DelNetwork(ctx context.Context, net *PluginConfig, rt *RuntimeConf) error {
var cachedResult types.Result
// Cached result on DEL was added in CNI spec version 0.4.0 and higher
@@ -706,7 +716,7 @@ func (c *CNIConfig) ValidateNetworkList(ctx context.Context, list *NetworkConfig
// ValidateNetwork checks that a configuration is reasonably valid.
// It uses the same logic as ValidateNetworkList)
// Returns a list of capabilities
func (c *CNIConfig) ValidateNetwork(ctx context.Context, net *NetworkConfig) ([]string, error) {
func (c *CNIConfig) ValidateNetwork(ctx context.Context, net *PluginConfig) ([]string, error) {
caps := []string{}
for c, ok := range net.Network.Capabilities {
if ok {
@@ -758,15 +768,23 @@ func (c *CNIConfig) GetVersionInfo(ctx context.Context, pluginType string) (vers
// - dump the list of cached attachments, and issue deletes as necessary
// - issue a GC to the underlying plugins (if the version is high enough)
func (c *CNIConfig) GCNetworkList(ctx context.Context, list *NetworkConfigList, args *GCArgs) error {
// If DisableGC is set, then don't bother GCing at all.
if list.DisableGC {
return nil
}
// First, get the list of cached attachments
cachedAttachments, err := c.GetCachedAttachments("")
if err != nil {
return nil
}
validAttachments := make(map[types.GCAttachment]interface{}, len(args.ValidAttachments))
for _, a := range args.ValidAttachments {
validAttachments[a] = nil
var validAttachments map[types.GCAttachment]interface{}
if args != nil {
validAttachments = make(map[types.GCAttachment]interface{}, len(args.ValidAttachments))
for _, a := range args.ValidAttachments {
validAttachments[a] = nil
}
}
var errs []error
@@ -799,10 +817,15 @@ func (c *CNIConfig) GCNetworkList(ctx context.Context, list *NetworkConfigList,
// now, if the version supports it, issue a GC
if gt, _ := version.GreaterThanOrEqualTo(list.CNIVersion, "1.1.0"); gt {
inject := map[string]interface{}{
"name": list.Name,
"cniVersion": list.CNIVersion,
"cni.dev/valid-attachments": args.ValidAttachments,
"name": list.Name,
"cniVersion": list.CNIVersion,
}
if args != nil {
inject["cni.dev/valid-attachments"] = args.ValidAttachments
// #1101: spec used incorrect variable name
inject["cni.dev/attachments"] = args.ValidAttachments
}
for _, plugin := range list.Plugins {
// build config here
pluginConfig, err := InjectConf(plugin, inject)
@@ -815,10 +838,10 @@ func (c *CNIConfig) GCNetworkList(ctx context.Context, list *NetworkConfigList,
}
}
return joinErrors(errs...)
return errors.Join(errs...)
}
func (c *CNIConfig) gcNetwork(ctx context.Context, net *NetworkConfig) error {
func (c *CNIConfig) gcNetwork(ctx context.Context, net *PluginConfig) error {
c.ensureExec()
pluginPath, err := c.exec.FindInPath(net.Network.Type, c.Path)
if err != nil {
@@ -853,7 +876,7 @@ func (c *CNIConfig) GetStatusNetworkList(ctx context.Context, list *NetworkConfi
return nil
}
func (c *CNIConfig) getStatusNetwork(ctx context.Context, net *NetworkConfig) error {
func (c *CNIConfig) getStatusNetwork(ctx context.Context, net *PluginConfig) error {
c.ensureExec()
pluginPath, err := c.exec.FindInPath(net.Network.Type, c.Path)
if err != nil {

View File

@@ -20,11 +20,10 @@ import (
"fmt"
"os"
"path/filepath"
"slices"
"sort"
"strings"
"github.com/Masterminds/semver/v3"
"github.com/containernetworking/cni/pkg/types"
"github.com/containernetworking/cni/pkg/version"
)
@@ -46,9 +45,16 @@ func (e NoConfigsFoundError) Error() string {
return fmt.Sprintf(`no net configurations found in %s`, e.Dir)
}
func ConfFromBytes(bytes []byte) (*NetworkConfig, error) {
conf := &NetworkConfig{Bytes: bytes, Network: &types.NetConf{}}
if err := json.Unmarshal(bytes, conf.Network); err != nil {
// This will not validate that the plugins actually belong to the netconfig by ensuring
// that they are loaded from a directory named after the networkName, relative to the network config.
//
// Since here we are just accepting raw bytes, the caller is responsible for ensuring that the plugin
// config provided here actually "belongs" to the networkconfig in question.
func NetworkPluginConfFromBytes(pluginConfBytes []byte) (*PluginConfig, error) {
// TODO why are we creating a struct that holds both the byte representation and the deserialized
// representation, and returning that, instead of just returning the deserialized representation?
conf := &PluginConfig{Bytes: pluginConfBytes, Network: &types.PluginConf{}}
if err := json.Unmarshal(pluginConfBytes, conf.Network); err != nil {
return nil, fmt.Errorf("error parsing configuration: %w", err)
}
if conf.Network.Type == "" {
@@ -57,17 +63,35 @@ func ConfFromBytes(bytes []byte) (*NetworkConfig, error) {
return conf, nil
}
func ConfFromFile(filename string) (*NetworkConfig, error) {
bytes, err := os.ReadFile(filename)
// Given a path to a directory containing a network configuration, and the name of a network,
// loads all plugin definitions found at path `networkConfPath/networkName/*.conf`
func NetworkPluginConfsFromFiles(networkConfPath, networkName string) ([]*PluginConfig, error) {
var pConfs []*PluginConfig
pluginConfPath := filepath.Join(networkConfPath, networkName)
pluginConfFiles, err := ConfFiles(pluginConfPath, []string{".conf"})
if err != nil {
return nil, fmt.Errorf("error reading %s: %w", filename, err)
return nil, fmt.Errorf("failed to read plugin config files in %s: %w", pluginConfPath, err)
}
return ConfFromBytes(bytes)
for _, pluginConfFile := range pluginConfFiles {
pluginConfBytes, err := os.ReadFile(pluginConfFile)
if err != nil {
return nil, fmt.Errorf("error reading %s: %w", pluginConfFile, err)
}
pluginConf, err := NetworkPluginConfFromBytes(pluginConfBytes)
if err != nil {
return nil, err
}
pConfs = append(pConfs, pluginConf)
}
return pConfs, nil
}
func ConfListFromBytes(bytes []byte) (*NetworkConfigList, error) {
func NetworkConfFromBytes(confBytes []byte) (*NetworkConfigList, error) {
rawList := make(map[string]interface{})
if err := json.Unmarshal(bytes, &rawList); err != nil {
if err := json.Unmarshal(confBytes, &rawList); err != nil {
return nil, fmt.Errorf("error parsing configuration list: %w", err)
}
@@ -92,24 +116,20 @@ func ConfListFromBytes(bytes []byte) (*NetworkConfigList, error) {
rawVersions, ok := rawList["cniVersions"]
if ok {
// Parse the current package CNI version
currentVersion, err := semver.NewVersion(version.Current())
if err != nil {
panic("CNI version is invalid semver!")
}
rvs, ok := rawVersions.([]interface{})
if !ok {
return nil, fmt.Errorf("error parsing configuration list: invalid type for cniVersions: %T", rvs)
}
vs := make([]*semver.Version, 0, len(rvs))
vs := make([]string, 0, len(rvs))
for i, rv := range rvs {
v, ok := rv.(string)
if !ok {
return nil, fmt.Errorf("error parsing configuration list: invalid type for cniVersions index %d: %T", i, rv)
}
if v, err := semver.NewVersion(v); err != nil {
gt, err := version.GreaterThan(v, version.Current())
if err != nil {
return nil, fmt.Errorf("error parsing configuration list: invalid cniVersions entry %s at index %d: %w", v, i, err)
} else if !v.GreaterThan(currentVersion) {
} else if !gt {
// Skip versions "greater" than this implementation of the spec
vs = append(vs, v)
}
@@ -117,50 +137,91 @@ func ConfListFromBytes(bytes []byte) (*NetworkConfigList, error) {
// if cniVersion was already set, append it to the list for sorting.
if cniVersion != "" {
if v, err := semver.NewVersion(cniVersion); err != nil {
gt, err := version.GreaterThan(cniVersion, version.Current())
if err != nil {
return nil, fmt.Errorf("error parsing configuration list: invalid cniVersion %s: %w", cniVersion, err)
} else if !v.GreaterThan(currentVersion) {
} else if !gt {
// ignore any versions higher than the current implemented spec version
vs = append(vs, v)
vs = append(vs, cniVersion)
}
}
sort.Sort(semver.Collection(vs))
slices.SortFunc[[]string](vs, func(v1, v2 string) int {
if v1 == v2 {
return 0
}
if gt, _ := version.GreaterThan(v1, v2); gt {
return 1
}
return -1
})
if len(vs) > 0 {
cniVersion = vs[len(vs)-1].String()
cniVersion = vs[len(vs)-1]
}
}
disableCheck := false
if rawDisableCheck, ok := rawList["disableCheck"]; ok {
disableCheck, ok = rawDisableCheck.(bool)
readBool := func(key string) (bool, error) {
rawVal, ok := rawList[key]
if !ok {
disableCheckStr, ok := rawDisableCheck.(string)
if !ok {
return nil, fmt.Errorf("error parsing configuration list: invalid disableCheck type %T", rawDisableCheck)
}
switch {
case strings.ToLower(disableCheckStr) == "false":
disableCheck = false
case strings.ToLower(disableCheckStr) == "true":
disableCheck = true
default:
return nil, fmt.Errorf("error parsing configuration list: invalid disableCheck value %q", disableCheckStr)
}
return false, nil
}
if b, ok := rawVal.(bool); ok {
return b, nil
}
s, ok := rawVal.(string)
if !ok {
return false, fmt.Errorf("error parsing configuration list: invalid type %T for %s", rawVal, key)
}
s = strings.ToLower(s)
switch s {
case "false":
return false, nil
case "true":
return true, nil
}
return false, fmt.Errorf("error parsing configuration list: invalid value %q for %s", s, key)
}
disableCheck, err := readBool("disableCheck")
if err != nil {
return nil, err
}
disableGC, err := readBool("disableGC")
if err != nil {
return nil, err
}
loadOnlyInlinedPlugins, err := readBool("loadOnlyInlinedPlugins")
if err != nil {
return nil, err
}
list := &NetworkConfigList{
Name: name,
DisableCheck: disableCheck,
CNIVersion: cniVersion,
Bytes: bytes,
Name: name,
DisableCheck: disableCheck,
DisableGC: disableGC,
LoadOnlyInlinedPlugins: loadOnlyInlinedPlugins,
CNIVersion: cniVersion,
Bytes: confBytes,
}
var plugins []interface{}
plug, ok := rawList["plugins"]
if !ok {
return nil, fmt.Errorf("error parsing configuration list: no 'plugins' key")
// We can have a `plugins` list key in the main conf,
// We can also have `loadOnlyInlinedPlugins == true`
//
// If `plugins` is there, then `loadOnlyInlinedPlugins` can be true
//
// If plugins is NOT there, then `loadOnlyInlinedPlugins` cannot be true
//
// We have to have at least some plugins.
if !ok && loadOnlyInlinedPlugins {
return nil, fmt.Errorf("error parsing configuration list: `loadOnlyInlinedPlugins` is true, and no 'plugins' key")
} else if !ok && !loadOnlyInlinedPlugins {
return list, nil
}
plugins, ok = plug.([]interface{})
if !ok {
return nil, fmt.Errorf("error parsing configuration list: invalid 'plugins' type %T", plug)
@@ -180,24 +241,68 @@ func ConfListFromBytes(bytes []byte) (*NetworkConfigList, error) {
}
list.Plugins = append(list.Plugins, netConf)
}
return list, nil
}
func ConfListFromFile(filename string) (*NetworkConfigList, error) {
func NetworkConfFromFile(filename string) (*NetworkConfigList, error) {
bytes, err := os.ReadFile(filename)
if err != nil {
return nil, fmt.Errorf("error reading %s: %w", filename, err)
}
return ConfListFromBytes(bytes)
conf, err := NetworkConfFromBytes(bytes)
if err != nil {
return nil, err
}
if !conf.LoadOnlyInlinedPlugins {
plugins, err := NetworkPluginConfsFromFiles(filepath.Dir(filename), conf.Name)
if err != nil {
return nil, err
}
conf.Plugins = append(conf.Plugins, plugins...)
}
if len(conf.Plugins) == 0 {
// Having 0 plugins for a given network is not necessarily a problem,
// but return as error for caller to decide, since they tried to load
return nil, fmt.Errorf("no plugin configs found")
}
return conf, nil
}
// Deprecated: This file format is no longer supported, use NetworkConfXXX and NetworkPluginXXX functions
func ConfFromBytes(bytes []byte) (*NetworkConfig, error) {
return NetworkPluginConfFromBytes(bytes)
}
// Deprecated: This file format is no longer supported, use NetworkConfXXX and NetworkPluginXXX functions
func ConfFromFile(filename string) (*NetworkConfig, error) {
bytes, err := os.ReadFile(filename)
if err != nil {
return nil, fmt.Errorf("error reading %s: %w", filename, err)
}
return ConfFromBytes(bytes)
}
func ConfListFromBytes(bytes []byte) (*NetworkConfigList, error) {
return NetworkConfFromBytes(bytes)
}
func ConfListFromFile(filename string) (*NetworkConfigList, error) {
return NetworkConfFromFile(filename)
}
// ConfFiles simply returns a slice of all files in the provided directory
// with extensions matching the provided set.
func ConfFiles(dir string, extensions []string) ([]string, error) {
// In part, adapted from rkt/networking/podenv.go#listFiles
files, err := os.ReadDir(dir)
switch {
case err == nil: // break
case os.IsNotExist(err):
// If folder not there, return no error - only return an
// error if we cannot read contents or there are no contents.
return nil, nil
default:
return nil, err
@@ -218,6 +323,7 @@ func ConfFiles(dir string, extensions []string) ([]string, error) {
return confFiles, nil
}
// Deprecated: This file format is no longer supported, use NetworkConfXXX and NetworkPluginXXX functions
func LoadConf(dir, name string) (*NetworkConfig, error) {
files, err := ConfFiles(dir, []string{".conf", ".json"})
switch {
@@ -241,6 +347,15 @@ func LoadConf(dir, name string) (*NetworkConfig, error) {
}
func LoadConfList(dir, name string) (*NetworkConfigList, error) {
return LoadNetworkConf(dir, name)
}
// LoadNetworkConf looks at all the network configs in a given dir,
// loads and parses them all, and returns the first one with an extension of `.conf`
// that matches the provided network name predicate.
func LoadNetworkConf(dir, name string) (*NetworkConfigList, error) {
// TODO this .conflist/.conf extension thing is confusing and inexact
// for implementors. We should pick one extension for everything and stick with it.
files, err := ConfFiles(dir, []string{".conflist"})
if err != nil {
return nil, err
@@ -248,7 +363,7 @@ func LoadConfList(dir, name string) (*NetworkConfigList, error) {
sort.Strings(files)
for _, confFile := range files {
conf, err := ConfListFromFile(confFile)
conf, err := NetworkConfFromFile(confFile)
if err != nil {
return nil, err
}
@@ -257,7 +372,7 @@ func LoadConfList(dir, name string) (*NetworkConfigList, error) {
}
}
// Try and load a network configuration file (instead of list)
// Deprecated: Try and load a network configuration file (instead of list)
// from the same name, then upconvert.
singleConf, err := LoadConf(dir, name)
if err != nil {
@@ -273,7 +388,8 @@ func LoadConfList(dir, name string) (*NetworkConfigList, error) {
return ConfListFromConf(singleConf)
}
func InjectConf(original *NetworkConfig, newValues map[string]interface{}) (*NetworkConfig, error) {
// InjectConf takes a PluginConfig and inserts additional values into it, ensuring the result is serializable.
func InjectConf(original *PluginConfig, newValues map[string]interface{}) (*PluginConfig, error) {
config := make(map[string]interface{})
err := json.Unmarshal(original.Bytes, &config)
if err != nil {
@@ -297,12 +413,14 @@ func InjectConf(original *NetworkConfig, newValues map[string]interface{}) (*Net
return nil, err
}
return ConfFromBytes(newBytes)
return NetworkPluginConfFromBytes(newBytes)
}
// ConfListFromConf "upconverts" a network config in to a NetworkConfigList,
// with the single network as the only entry in the list.
func ConfListFromConf(original *NetworkConfig) (*NetworkConfigList, error) {
//
// Deprecated: Non-conflist file formats are unsupported, use NetworkConfXXX and NetworkPluginXXX functions
func ConfListFromConf(original *PluginConfig) (*NetworkConfigList, error) {
// Re-deserialize the config's json, then make a raw map configlist.
// This may seem a bit strange, but it's to make the Bytes fields
// actually make sense. Otherwise, the generated json is littered with

View File

@@ -1,58 +0,0 @@
// Copyright 2022 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Copyright the CNI authors
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
//
// Adapted from errors/join.go from go 1.20
// This package can be removed once the toolchain is updated to 1.20
package libcni
func joinErrors(errs ...error) error {
n := 0
for _, err := range errs {
if err != nil {
n++
}
}
if n == 0 {
return nil
}
e := &multiError{
errs: make([]error, 0, n),
}
for _, err := range errs {
if err != nil {
e.errs = append(e.errs, err)
}
}
return e
}
type multiError struct {
errs []error
}
func (e *multiError) Error() string {
var b []byte
for i, err := range e.errs {
if i > 0 {
b = append(b, '\n')
}
b = append(b, err.Error()...)
}
return string(b)
}

View File

@@ -1,10 +1,10 @@
// Copyright 2017 Google LLC. All Rights Reserved.
// Copyright 2022 CNI authors
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
@@ -12,5 +12,10 @@
// See the License for the specific language governing permissions and
// limitations under the License.
// Package compiler provides support functions to generated compiler code.
package compiler
package ns
import "github.com/containernetworking/cni/pkg/types"
func CheckNetNS(nsPath string) (bool, *types.Error) {
return false, nil
}

View File

@@ -56,8 +56,12 @@ func (n *IPNet) UnmarshalJSON(data []byte) error {
return nil
}
// NetConf describes a network.
type NetConf struct {
// Use PluginConf instead of NetConf, the NetConf
// backwards-compat alias will be removed in a future release.
type NetConf = PluginConf
// PluginConf describes a plugin configuration for a specific network.
type PluginConf struct {
CNIVersion string `json:"cniVersion,omitempty"`
Name string `json:"name,omitempty"`
@@ -83,11 +87,8 @@ type GCAttachment struct {
// Note: DNS should be omit if DNS is empty but default Marshal function
// will output empty structure hence need to write a Marshal function
func (n *NetConf) MarshalJSON() ([]byte, error) {
// use type alias to escape recursion for json.Marshal() to MarshalJSON()
type fixObjType = NetConf
bytes, err := json.Marshal(fixObjType(*n)) //nolint:all
func (n *PluginConf) MarshalJSON() ([]byte, error) {
bytes, err := json.Marshal(*n)
if err != nil {
return nil, err
}
@@ -117,9 +118,10 @@ func (i *IPAM) IsEmpty() bool {
type NetConfList struct {
CNIVersion string `json:"cniVersion,omitempty"`
Name string `json:"name,omitempty"`
DisableCheck bool `json:"disableCheck,omitempty"`
Plugins []*NetConf `json:"plugins,omitempty"`
Name string `json:"name,omitempty"`
DisableCheck bool `json:"disableCheck,omitempty"`
DisableGC bool `json:"disableGC,omitempty"`
Plugins []*PluginConf `json:"plugins,omitempty"`
}
// Result is an interface that provides the result of plugin execution

View File

@@ -142,3 +142,27 @@ func GreaterThanOrEqualTo(version, otherVersion string) (bool, error) {
}
return false, nil
}
// GreaterThan returns true if the first version is greater than the second
func GreaterThan(version, otherVersion string) (bool, error) {
firstMajor, firstMinor, firstMicro, err := ParseVersion(version)
if err != nil {
return false, err
}
secondMajor, secondMinor, secondMicro, err := ParseVersion(otherVersion)
if err != nil {
return false, err
}
if firstMajor > secondMajor {
return true, nil
} else if firstMajor == secondMajor {
if firstMinor > secondMinor {
return true, nil
} else if firstMinor == secondMinor && firstMicro > secondMicro {
return true, nil
}
}
return false, nil
}

View File

@@ -63,7 +63,7 @@ func NewResult(version string, resultBytes []byte) (types.Result, error) {
// ParsePrevResult parses a prevResult in a NetConf structure and sets
// the NetConf's PrevResult member to the parsed Result object.
func ParsePrevResult(conf *types.NetConf) error {
func ParsePrevResult(conf *types.PluginConf) error {
if conf.RawPrevResult == nil {
return nil
}

View File

@@ -1,6 +1,7 @@
# A minimal logging API for Go
[![Go Reference](https://pkg.go.dev/badge/github.com/go-logr/logr.svg)](https://pkg.go.dev/github.com/go-logr/logr)
[![Go Report Card](https://goreportcard.com/badge/github.com/go-logr/logr)](https://goreportcard.com/report/github.com/go-logr/logr)
[![OpenSSF Scorecard](https://api.securityscorecards.dev/projects/github.com/go-logr/logr/badge)](https://securityscorecards.dev/viewer/?platform=github.com&org=go-logr&repo=logr)
logr offers an(other) opinion on how Go programs and libraries can do logging
@@ -91,11 +92,12 @@ logr design but also left out some parts and changed others:
| Adding a name to a logger | `WithName` | no API |
| Modify verbosity of log entries in a call chain | `V` | no API |
| Grouping of key/value pairs | not supported | `WithGroup`, `GroupValue` |
| Pass context for extracting additional values | no API | API variants like `InfoCtx` |
The high-level slog API is explicitly meant to be one of many different APIs
that can be layered on top of a shared `slog.Handler`. logr is one such
alternative API, with [interoperability](#slog-interoperability) provided by the [`slogr`](slogr)
package.
alternative API, with [interoperability](#slog-interoperability) provided by
some conversion functions.
### Inspiration
@@ -145,24 +147,24 @@ There are implementations for the following logging libraries:
## slog interoperability
Interoperability goes both ways, using the `logr.Logger` API with a `slog.Handler`
and using the `slog.Logger` API with a `logr.LogSink`. [slogr](./slogr) provides `NewLogr` and
`NewSlogHandler` API calls to convert between a `logr.Logger` and a `slog.Handler`.
and using the `slog.Logger` API with a `logr.LogSink`. `FromSlogHandler` and
`ToSlogHandler` convert between a `logr.Logger` and a `slog.Handler`.
As usual, `slog.New` can be used to wrap such a `slog.Handler` in the high-level
slog API. `slogr` itself leaves that to the caller.
slog API.
## Using a `logr.Sink` as backend for slog
### Using a `logr.LogSink` as backend for slog
Ideally, a logr sink implementation should support both logr and slog by
implementing both the normal logr interface(s) and `slogr.SlogSink`. Because
implementing both the normal logr interface(s) and `SlogSink`. Because
of a conflict in the parameters of the common `Enabled` method, it is [not
possible to implement both slog.Handler and logr.Sink in the same
type](https://github.com/golang/go/issues/59110).
If both are supported, log calls can go from the high-level APIs to the backend
without the need to convert parameters. `NewLogr` and `NewSlogHandler` can
without the need to convert parameters. `FromSlogHandler` and `ToSlogHandler` can
convert back and forth without adding additional wrappers, with one exception:
when `Logger.V` was used to adjust the verbosity for a `slog.Handler`, then
`NewSlogHandler` has to use a wrapper which adjusts the verbosity for future
`ToSlogHandler` has to use a wrapper which adjusts the verbosity for future
log calls.
Such an implementation should also support values that implement specific
@@ -187,13 +189,13 @@ Not supporting slog has several drawbacks:
These drawbacks are severe enough that applications using a mixture of slog and
logr should switch to a different backend.
## Using a `slog.Handler` as backend for logr
### Using a `slog.Handler` as backend for logr
Using a plain `slog.Handler` without support for logr works better than the
other direction:
- All logr verbosity levels can be mapped 1:1 to their corresponding slog level
by negating them.
- Stack unwinding is done by the `slogr.SlogSink` and the resulting program
- Stack unwinding is done by the `SlogSink` and the resulting program
counter is passed to the `slog.Handler`.
- Names added via `Logger.WithName` are gathered and recorded in an additional
attribute with `logger` as key and the names separated by slash as value.
@@ -205,27 +207,39 @@ ideally support both `logr.Marshaler` and `slog.Valuer`. If compatibility
with logr implementations without slog support is not important, then
`slog.Valuer` is sufficient.
## Context support for slog
### Context support for slog
Storing a logger in a `context.Context` is not supported by
slog. `logr.NewContext` and `logr.FromContext` can be used with slog like this
to fill this gap:
slog. `NewContextWithSlogLogger` and `FromContextAsSlogLogger` can be
used to fill this gap. They store and retrieve a `slog.Logger` pointer
under the same context key that is also used by `NewContext` and
`FromContext` for `logr.Logger` value.
func HandlerFromContext(ctx context.Context) slog.Handler {
logger, err := logr.FromContext(ctx)
if err == nil {
return slogr.NewSlogHandler(logger)
}
return slog.Default().Handler()
}
When `NewContextWithSlogLogger` is followed by `FromContext`, the latter will
automatically convert the `slog.Logger` to a
`logr.Logger`. `FromContextAsSlogLogger` does the same for the other direction.
func ContextWithHandler(ctx context.Context, handler slog.Handler) context.Context {
return logr.NewContext(ctx, slogr.NewLogr(handler))
}
With this approach, binaries which use either slog or logr are as efficient as
possible with no unnecessary allocations. This is also why the API stores a
`slog.Logger` pointer: when storing a `slog.Handler`, creating a `slog.Logger`
on retrieval would need to allocate one.
The downside is that storing and retrieving a `slog.Handler` needs more
allocations compared to using a `logr.Logger`. Therefore the recommendation is
to use the `logr.Logger` API in code which uses contextual logging.
The downside is that switching back and forth needs more allocations. Because
logr is the API that is already in use by different packages, in particular
Kubernetes, the recommendation is to use the `logr.Logger` API in code which
uses contextual logging.
An alternative to adding values to a logger and storing that logger in the
context is to store the values in the context and to configure a logging
backend to extract those values when emitting log entries. This only works when
log calls are passed the context, which is not supported by the logr API.
With the slog API, it is possible, but not
required. https://github.com/veqryn/slog-context is a package for slog which
provides additional support code for this approach. It also contains wrappers
for the context functions in logr, so developers who prefer to not use the logr
APIs directly can use those instead and the resulting code will still be
interoperable with logr.
## FAQ

33
vendor/github.com/go-logr/logr/context.go generated vendored Normal file
View File

@@ -0,0 +1,33 @@
/*
Copyright 2023 The logr Authors.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/
package logr
// contextKey is how we find Loggers in a context.Context. With Go < 1.21,
// the value is always a Logger value. With Go >= 1.21, the value can be a
// Logger value or a slog.Logger pointer.
type contextKey struct{}
// notFoundError exists to carry an IsNotFound method.
type notFoundError struct{}
func (notFoundError) Error() string {
return "no logr.Logger was present"
}
func (notFoundError) IsNotFound() bool {
return true
}

49
vendor/github.com/go-logr/logr/context_noslog.go generated vendored Normal file
View File

@@ -0,0 +1,49 @@
//go:build !go1.21
// +build !go1.21
/*
Copyright 2019 The logr Authors.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/
package logr
import (
"context"
)
// FromContext returns a Logger from ctx or an error if no Logger is found.
func FromContext(ctx context.Context) (Logger, error) {
if v, ok := ctx.Value(contextKey{}).(Logger); ok {
return v, nil
}
return Logger{}, notFoundError{}
}
// FromContextOrDiscard returns a Logger from ctx. If no Logger is found, this
// returns a Logger that discards all log messages.
func FromContextOrDiscard(ctx context.Context) Logger {
if v, ok := ctx.Value(contextKey{}).(Logger); ok {
return v
}
return Discard()
}
// NewContext returns a new Context, derived from ctx, which carries the
// provided Logger.
func NewContext(ctx context.Context, logger Logger) context.Context {
return context.WithValue(ctx, contextKey{}, logger)
}

83
vendor/github.com/go-logr/logr/context_slog.go generated vendored Normal file
View File

@@ -0,0 +1,83 @@
//go:build go1.21
// +build go1.21
/*
Copyright 2019 The logr Authors.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/
package logr
import (
"context"
"fmt"
"log/slog"
)
// FromContext returns a Logger from ctx or an error if no Logger is found.
func FromContext(ctx context.Context) (Logger, error) {
v := ctx.Value(contextKey{})
if v == nil {
return Logger{}, notFoundError{}
}
switch v := v.(type) {
case Logger:
return v, nil
case *slog.Logger:
return FromSlogHandler(v.Handler()), nil
default:
// Not reached.
panic(fmt.Sprintf("unexpected value type for logr context key: %T", v))
}
}
// FromContextAsSlogLogger returns a slog.Logger from ctx or nil if no such Logger is found.
func FromContextAsSlogLogger(ctx context.Context) *slog.Logger {
v := ctx.Value(contextKey{})
if v == nil {
return nil
}
switch v := v.(type) {
case Logger:
return slog.New(ToSlogHandler(v))
case *slog.Logger:
return v
default:
// Not reached.
panic(fmt.Sprintf("unexpected value type for logr context key: %T", v))
}
}
// FromContextOrDiscard returns a Logger from ctx. If no Logger is found, this
// returns a Logger that discards all log messages.
func FromContextOrDiscard(ctx context.Context) Logger {
if logger, err := FromContext(ctx); err == nil {
return logger
}
return Discard()
}
// NewContext returns a new Context, derived from ctx, which carries the
// provided Logger.
func NewContext(ctx context.Context, logger Logger) context.Context {
return context.WithValue(ctx, contextKey{}, logger)
}
// NewContextWithSlogLogger returns a new Context, derived from ctx, which carries the
// provided slog.Logger.
func NewContextWithSlogLogger(ctx context.Context, logger *slog.Logger) context.Context {
return context.WithValue(ctx, contextKey{}, logger)
}

View File

@@ -100,6 +100,11 @@ type Options struct {
// details, see docs for Go's time.Layout.
TimestampFormat string
// LogInfoLevel tells funcr what key to use to log the info level.
// If not specified, the info level will be logged as "level".
// If this is set to "", the info level will not be logged at all.
LogInfoLevel *string
// Verbosity tells funcr which V logs to produce. Higher values enable
// more logs. Info logs at or below this level will be written, while logs
// above this level will be discarded.
@@ -213,6 +218,10 @@ func newFormatter(opts Options, outfmt outputFormat) Formatter {
if opts.MaxLogDepth == 0 {
opts.MaxLogDepth = defaultMaxLogDepth
}
if opts.LogInfoLevel == nil {
opts.LogInfoLevel = new(string)
*opts.LogInfoLevel = "level"
}
f := Formatter{
outputFormat: outfmt,
prefix: "",
@@ -233,6 +242,8 @@ type Formatter struct {
valuesStr string
depth int
opts *Options
groupName string // for slog groups
groups []groupDef
}
// outputFormat indicates which outputFormat to use.
@@ -245,6 +256,13 @@ const (
outputJSON
)
// groupDef represents a saved group. The values may be empty, but we don't
// know if we need to render the group until the final record is rendered.
type groupDef struct {
name string
values string
}
// PseudoStruct is a list of key-value pairs that gets logged as a struct.
type PseudoStruct []any
@@ -252,63 +270,125 @@ type PseudoStruct []any
func (f Formatter) render(builtins, args []any) string {
// Empirically bytes.Buffer is faster than strings.Builder for this.
buf := bytes.NewBuffer(make([]byte, 0, 1024))
if f.outputFormat == outputJSON {
buf.WriteByte('{')
buf.WriteByte('{') // for the whole record
}
// Render builtins
vals := builtins
if hook := f.opts.RenderBuiltinsHook; hook != nil {
vals = hook(f.sanitize(vals))
}
f.flatten(buf, vals, false, false) // keys are ours, no need to escape
f.flatten(buf, vals, false) // keys are ours, no need to escape
continuing := len(builtins) > 0
if len(f.valuesStr) > 0 {
if continuing {
if f.outputFormat == outputJSON {
buf.WriteByte(',')
} else {
buf.WriteByte(' ')
}
// Turn the inner-most group into a string
argsStr := func() string {
buf := bytes.NewBuffer(make([]byte, 0, 1024))
vals = args
if hook := f.opts.RenderArgsHook; hook != nil {
vals = hook(f.sanitize(vals))
}
continuing = true
buf.WriteString(f.valuesStr)
f.flatten(buf, vals, true) // escape user-provided keys
return buf.String()
}()
// Render the stack of groups from the inside out.
bodyStr := f.renderGroup(f.groupName, f.valuesStr, argsStr)
for i := len(f.groups) - 1; i >= 0; i-- {
grp := &f.groups[i]
if grp.values == "" && bodyStr == "" {
// no contents, so we must elide the whole group
continue
}
bodyStr = f.renderGroup(grp.name, grp.values, bodyStr)
}
vals = args
if hook := f.opts.RenderArgsHook; hook != nil {
vals = hook(f.sanitize(vals))
if bodyStr != "" {
if continuing {
buf.WriteByte(f.comma())
}
buf.WriteString(bodyStr)
}
f.flatten(buf, vals, continuing, true) // escape user-provided keys
if f.outputFormat == outputJSON {
buf.WriteByte('}')
buf.WriteByte('}') // for the whole record
}
return buf.String()
}
// flatten renders a list of key-value pairs into a buffer. If continuing is
// true, it assumes that the buffer has previous values and will emit a
// separator (which depends on the output format) before the first pair it
// writes. If escapeKeys is true, the keys are assumed to have
// non-JSON-compatible characters in them and must be evaluated for escapes.
// renderGroup returns a string representation of the named group with rendered
// values and args. If the name is empty, this will return the values and args,
// joined. If the name is not empty, this will return a single key-value pair,
// where the value is a grouping of the values and args. If the values and
// args are both empty, this will return an empty string, even if the name was
// specified.
func (f Formatter) renderGroup(name string, values string, args string) string {
buf := bytes.NewBuffer(make([]byte, 0, 1024))
needClosingBrace := false
if name != "" && (values != "" || args != "") {
buf.WriteString(f.quoted(name, true)) // escape user-provided keys
buf.WriteByte(f.colon())
buf.WriteByte('{')
needClosingBrace = true
}
continuing := false
if values != "" {
buf.WriteString(values)
continuing = true
}
if args != "" {
if continuing {
buf.WriteByte(f.comma())
}
buf.WriteString(args)
}
if needClosingBrace {
buf.WriteByte('}')
}
return buf.String()
}
// flatten renders a list of key-value pairs into a buffer. If escapeKeys is
// true, the keys are assumed to have non-JSON-compatible characters in them
// and must be evaluated for escapes.
//
// This function returns a potentially modified version of kvList, which
// ensures that there is a value for every key (adding a value if needed) and
// that each key is a string (substituting a key if needed).
func (f Formatter) flatten(buf *bytes.Buffer, kvList []any, continuing bool, escapeKeys bool) []any {
func (f Formatter) flatten(buf *bytes.Buffer, kvList []any, escapeKeys bool) []any {
// This logic overlaps with sanitize() but saves one type-cast per key,
// which can be measurable.
if len(kvList)%2 != 0 {
kvList = append(kvList, noValue)
}
copied := false
for i := 0; i < len(kvList); i += 2 {
k, ok := kvList[i].(string)
if !ok {
if !copied {
newList := make([]any, len(kvList))
copy(newList, kvList)
kvList = newList
copied = true
}
k = f.nonStringKey(kvList[i])
kvList[i] = k
}
v := kvList[i+1]
if i > 0 || continuing {
if i > 0 {
if f.outputFormat == outputJSON {
buf.WriteByte(',')
buf.WriteByte(f.comma())
} else {
// In theory the format could be something we don't understand. In
// practice, we control it, so it won't be.
@@ -316,24 +396,35 @@ func (f Formatter) flatten(buf *bytes.Buffer, kvList []any, continuing bool, esc
}
}
if escapeKeys {
buf.WriteString(prettyString(k))
} else {
// this is faster
buf.WriteByte('"')
buf.WriteString(k)
buf.WriteByte('"')
}
if f.outputFormat == outputJSON {
buf.WriteByte(':')
} else {
buf.WriteByte('=')
}
buf.WriteString(f.quoted(k, escapeKeys))
buf.WriteByte(f.colon())
buf.WriteString(f.pretty(v))
}
return kvList
}
func (f Formatter) quoted(str string, escape bool) string {
if escape {
return prettyString(str)
}
// this is faster
return `"` + str + `"`
}
func (f Formatter) comma() byte {
if f.outputFormat == outputJSON {
return ','
}
return ' '
}
func (f Formatter) colon() byte {
if f.outputFormat == outputJSON {
return ':'
}
return '='
}
func (f Formatter) pretty(value any) string {
return f.prettyWithFlags(value, 0, 0)
}
@@ -407,12 +498,12 @@ func (f Formatter) prettyWithFlags(value any, flags uint32, depth int) string {
}
for i := 0; i < len(v); i += 2 {
if i > 0 {
buf.WriteByte(',')
buf.WriteByte(f.comma())
}
k, _ := v[i].(string) // sanitize() above means no need to check success
// arbitrary keys might need escaping
buf.WriteString(prettyString(k))
buf.WriteByte(':')
buf.WriteByte(f.colon())
buf.WriteString(f.prettyWithFlags(v[i+1], 0, depth+1))
}
if flags&flagRawStruct == 0 {
@@ -481,7 +572,7 @@ func (f Formatter) prettyWithFlags(value any, flags uint32, depth int) string {
continue
}
if printComma {
buf.WriteByte(',')
buf.WriteByte(f.comma())
}
printComma = true // if we got here, we are rendering a field
if fld.Anonymous && fld.Type.Kind() == reflect.Struct && name == "" {
@@ -492,10 +583,8 @@ func (f Formatter) prettyWithFlags(value any, flags uint32, depth int) string {
name = fld.Name
}
// field names can't contain characters which need escaping
buf.WriteByte('"')
buf.WriteString(name)
buf.WriteByte('"')
buf.WriteByte(':')
buf.WriteString(f.quoted(name, false))
buf.WriteByte(f.colon())
buf.WriteString(f.prettyWithFlags(v.Field(i).Interface(), 0, depth+1))
}
if flags&flagRawStruct == 0 {
@@ -520,7 +609,7 @@ func (f Formatter) prettyWithFlags(value any, flags uint32, depth int) string {
buf.WriteByte('[')
for i := 0; i < v.Len(); i++ {
if i > 0 {
buf.WriteByte(',')
buf.WriteByte(f.comma())
}
e := v.Index(i)
buf.WriteString(f.prettyWithFlags(e.Interface(), 0, depth+1))
@@ -534,7 +623,7 @@ func (f Formatter) prettyWithFlags(value any, flags uint32, depth int) string {
i := 0
for it.Next() {
if i > 0 {
buf.WriteByte(',')
buf.WriteByte(f.comma())
}
// If a map key supports TextMarshaler, use it.
keystr := ""
@@ -556,7 +645,7 @@ func (f Formatter) prettyWithFlags(value any, flags uint32, depth int) string {
}
}
buf.WriteString(keystr)
buf.WriteByte(':')
buf.WriteByte(f.colon())
buf.WriteString(f.prettyWithFlags(it.Value().Interface(), 0, depth+1))
i++
}
@@ -706,6 +795,24 @@ func (f Formatter) sanitize(kvList []any) []any {
return kvList
}
// startGroup opens a new group scope (basically a sub-struct), which locks all
// the current saved values and starts them anew. This is needed to satisfy
// slog.
func (f *Formatter) startGroup(name string) {
// Unnamed groups are just inlined.
if name == "" {
return
}
n := len(f.groups)
f.groups = append(f.groups[:n:n], groupDef{f.groupName, f.valuesStr})
// Start collecting new values.
f.groupName = name
f.valuesStr = ""
f.values = nil
}
// Init configures this Formatter from runtime info, such as the call depth
// imposed by logr itself.
// Note that this receiver is a pointer, so depth can be saved.
@@ -740,7 +847,10 @@ func (f Formatter) FormatInfo(level int, msg string, kvList []any) (prefix, args
if policy := f.opts.LogCaller; policy == All || policy == Info {
args = append(args, "caller", f.caller())
}
args = append(args, "level", level, "msg", msg)
if key := *f.opts.LogInfoLevel; key != "" {
args = append(args, key, level)
}
args = append(args, "msg", msg)
return prefix, f.render(args, kvList)
}
@@ -793,7 +903,7 @@ func (f *Formatter) AddValues(kvList []any) {
// Pre-render values, so we don't have to do it on each Info/Error call.
buf := bytes.NewBuffer(make([]byte, 0, 1024))
f.flatten(buf, vals, false, true) // escape user-provided keys
f.flatten(buf, vals, true) // escape user-provided keys
f.valuesStr = buf.String()
}

105
vendor/github.com/go-logr/logr/funcr/slogsink.go generated vendored Normal file
View File

@@ -0,0 +1,105 @@
//go:build go1.21
// +build go1.21
/*
Copyright 2023 The logr Authors.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/
package funcr
import (
"context"
"log/slog"
"github.com/go-logr/logr"
)
var _ logr.SlogSink = &fnlogger{}
const extraSlogSinkDepth = 3 // 2 for slog, 1 for SlogSink
func (l fnlogger) Handle(_ context.Context, record slog.Record) error {
kvList := make([]any, 0, 2*record.NumAttrs())
record.Attrs(func(attr slog.Attr) bool {
kvList = attrToKVs(attr, kvList)
return true
})
if record.Level >= slog.LevelError {
l.WithCallDepth(extraSlogSinkDepth).Error(nil, record.Message, kvList...)
} else {
level := l.levelFromSlog(record.Level)
l.WithCallDepth(extraSlogSinkDepth).Info(level, record.Message, kvList...)
}
return nil
}
func (l fnlogger) WithAttrs(attrs []slog.Attr) logr.SlogSink {
kvList := make([]any, 0, 2*len(attrs))
for _, attr := range attrs {
kvList = attrToKVs(attr, kvList)
}
l.AddValues(kvList)
return &l
}
func (l fnlogger) WithGroup(name string) logr.SlogSink {
l.startGroup(name)
return &l
}
// attrToKVs appends a slog.Attr to a logr-style kvList. It handle slog Groups
// and other details of slog.
func attrToKVs(attr slog.Attr, kvList []any) []any {
attrVal := attr.Value.Resolve()
if attrVal.Kind() == slog.KindGroup {
groupVal := attrVal.Group()
grpKVs := make([]any, 0, 2*len(groupVal))
for _, attr := range groupVal {
grpKVs = attrToKVs(attr, grpKVs)
}
if attr.Key == "" {
// slog says we have to inline these
kvList = append(kvList, grpKVs...)
} else {
kvList = append(kvList, attr.Key, PseudoStruct(grpKVs))
}
} else if attr.Key != "" {
kvList = append(kvList, attr.Key, attrVal.Any())
}
return kvList
}
// levelFromSlog adjusts the level by the logger's verbosity and negates it.
// It ensures that the result is >= 0. This is necessary because the result is
// passed to a LogSink and that API did not historically document whether
// levels could be negative or what that meant.
//
// Some example usage:
//
// logrV0 := getMyLogger()
// logrV2 := logrV0.V(2)
// slogV2 := slog.New(logr.ToSlogHandler(logrV2))
// slogV2.Debug("msg") // =~ logrV2.V(4) =~ logrV0.V(6)
// slogV2.Info("msg") // =~ logrV2.V(0) =~ logrV0.V(2)
// slogv2.Warn("msg") // =~ logrV2.V(-4) =~ logrV0.V(0)
func (l fnlogger) levelFromSlog(level slog.Level) int {
result := -level
if result < 0 {
result = 0 // because LogSink doesn't expect negative V levels
}
return int(result)
}

View File

@@ -207,10 +207,6 @@ limitations under the License.
// those.
package logr
import (
"context"
)
// New returns a new Logger instance. This is primarily used by libraries
// implementing LogSink, rather than end users. Passing a nil sink will create
// a Logger which discards all log lines.
@@ -410,45 +406,6 @@ func (l Logger) IsZero() bool {
return l.sink == nil
}
// contextKey is how we find Loggers in a context.Context.
type contextKey struct{}
// FromContext returns a Logger from ctx or an error if no Logger is found.
func FromContext(ctx context.Context) (Logger, error) {
if v, ok := ctx.Value(contextKey{}).(Logger); ok {
return v, nil
}
return Logger{}, notFoundError{}
}
// notFoundError exists to carry an IsNotFound method.
type notFoundError struct{}
func (notFoundError) Error() string {
return "no logr.Logger was present"
}
func (notFoundError) IsNotFound() bool {
return true
}
// FromContextOrDiscard returns a Logger from ctx. If no Logger is found, this
// returns a Logger that discards all log messages.
func FromContextOrDiscard(ctx context.Context) Logger {
if v, ok := ctx.Value(contextKey{}).(Logger); ok {
return v
}
return Discard()
}
// NewContext returns a new Context, derived from ctx, which carries the
// provided Logger.
func NewContext(ctx context.Context, logger Logger) context.Context {
return context.WithValue(ctx, contextKey{}, logger)
}
// RuntimeInfo holds information that the logr "core" library knows which
// LogSinks might want to know.
type RuntimeInfo struct {

View File

@@ -17,18 +17,16 @@ See the License for the specific language governing permissions and
limitations under the License.
*/
package slogr
package logr
import (
"context"
"log/slog"
"github.com/go-logr/logr"
)
type slogHandler struct {
// May be nil, in which case all logs get discarded.
sink logr.LogSink
sink LogSink
// Non-nil if sink is non-nil and implements SlogSink.
slogSink SlogSink
@@ -54,7 +52,7 @@ func (l *slogHandler) GetLevel() slog.Level {
return l.levelBias
}
func (l *slogHandler) Enabled(ctx context.Context, level slog.Level) bool {
func (l *slogHandler) Enabled(_ context.Context, level slog.Level) bool {
return l.sink != nil && (level >= slog.LevelError || l.sink.Enabled(l.levelFromSlog(level)))
}
@@ -72,9 +70,7 @@ func (l *slogHandler) Handle(ctx context.Context, record slog.Record) error {
kvList := make([]any, 0, 2*record.NumAttrs())
record.Attrs(func(attr slog.Attr) bool {
if attr.Key != "" {
kvList = append(kvList, l.addGroupPrefix(attr.Key), attr.Value.Resolve().Any())
}
kvList = attrToKVs(attr, l.groupPrefix, kvList)
return true
})
if record.Level >= slog.LevelError {
@@ -90,15 +86,15 @@ func (l *slogHandler) Handle(ctx context.Context, record slog.Record) error {
// are called by Handle, code in slog gets skipped.
//
// This offset currently (Go 1.21.0) works for calls through
// slog.New(NewSlogHandler(...)). There's no guarantee that the call
// slog.New(ToSlogHandler(...)). There's no guarantee that the call
// chain won't change. Wrapping the handler will also break unwinding. It's
// still better than not adjusting at all....
//
// This cannot be done when constructing the handler because NewLogr needs
// This cannot be done when constructing the handler because FromSlogHandler needs
// access to the original sink without this adjustment. A second copy would
// work, but then WithAttrs would have to be called for both of them.
func (l *slogHandler) sinkWithCallDepth() logr.LogSink {
if sink, ok := l.sink.(logr.CallDepthLogSink); ok {
func (l *slogHandler) sinkWithCallDepth() LogSink {
if sink, ok := l.sink.(CallDepthLogSink); ok {
return sink.WithCallDepth(2)
}
return l.sink
@@ -109,60 +105,88 @@ func (l *slogHandler) WithAttrs(attrs []slog.Attr) slog.Handler {
return l
}
copy := *l
clone := *l
if l.slogSink != nil {
copy.slogSink = l.slogSink.WithAttrs(attrs)
copy.sink = copy.slogSink
clone.slogSink = l.slogSink.WithAttrs(attrs)
clone.sink = clone.slogSink
} else {
kvList := make([]any, 0, 2*len(attrs))
for _, attr := range attrs {
if attr.Key != "" {
kvList = append(kvList, l.addGroupPrefix(attr.Key), attr.Value.Resolve().Any())
}
kvList = attrToKVs(attr, l.groupPrefix, kvList)
}
copy.sink = l.sink.WithValues(kvList...)
clone.sink = l.sink.WithValues(kvList...)
}
return &copy
return &clone
}
func (l *slogHandler) WithGroup(name string) slog.Handler {
if l.sink == nil {
return l
}
copy := *l
if l.slogSink != nil {
copy.slogSink = l.slogSink.WithGroup(name)
copy.sink = l.slogSink
} else {
copy.groupPrefix = copy.addGroupPrefix(name)
if name == "" {
// slog says to inline empty groups
return l
}
return &copy
clone := *l
if l.slogSink != nil {
clone.slogSink = l.slogSink.WithGroup(name)
clone.sink = clone.slogSink
} else {
clone.groupPrefix = addPrefix(clone.groupPrefix, name)
}
return &clone
}
func (l *slogHandler) addGroupPrefix(name string) string {
if l.groupPrefix == "" {
// attrToKVs appends a slog.Attr to a logr-style kvList. It handle slog Groups
// and other details of slog.
func attrToKVs(attr slog.Attr, groupPrefix string, kvList []any) []any {
attrVal := attr.Value.Resolve()
if attrVal.Kind() == slog.KindGroup {
groupVal := attrVal.Group()
grpKVs := make([]any, 0, 2*len(groupVal))
prefix := groupPrefix
if attr.Key != "" {
prefix = addPrefix(groupPrefix, attr.Key)
}
for _, attr := range groupVal {
grpKVs = attrToKVs(attr, prefix, grpKVs)
}
kvList = append(kvList, grpKVs...)
} else if attr.Key != "" {
kvList = append(kvList, addPrefix(groupPrefix, attr.Key), attrVal.Any())
}
return kvList
}
func addPrefix(prefix, name string) string {
if prefix == "" {
return name
}
return l.groupPrefix + groupSeparator + name
if name == "" {
return prefix
}
return prefix + groupSeparator + name
}
// levelFromSlog adjusts the level by the logger's verbosity and negates it.
// It ensures that the result is >= 0. This is necessary because the result is
// passed to a logr.LogSink and that API did not historically document whether
// passed to a LogSink and that API did not historically document whether
// levels could be negative or what that meant.
//
// Some example usage:
// logrV0 := getMyLogger()
// logrV2 := logrV0.V(2)
// slogV2 := slog.New(slogr.NewSlogHandler(logrV2))
// slogV2.Debug("msg") // =~ logrV2.V(4) =~ logrV0.V(6)
// slogV2.Info("msg") // =~ logrV2.V(0) =~ logrV0.V(2)
// slogv2.Warn("msg") // =~ logrV2.V(-4) =~ logrV0.V(0)
//
// logrV0 := getMyLogger()
// logrV2 := logrV0.V(2)
// slogV2 := slog.New(logr.ToSlogHandler(logrV2))
// slogV2.Debug("msg") // =~ logrV2.V(4) =~ logrV0.V(6)
// slogV2.Info("msg") // =~ logrV2.V(0) =~ logrV0.V(2)
// slogv2.Warn("msg") // =~ logrV2.V(-4) =~ logrV0.V(0)
func (l *slogHandler) levelFromSlog(level slog.Level) int {
result := -level
result += l.levelBias // in case the original logr.Logger had a V level
result += l.levelBias // in case the original Logger had a V level
if result < 0 {
result = 0 // because logr.LogSink doesn't expect negative V levels
result = 0 // because LogSink doesn't expect negative V levels
}
return int(result)
}

100
vendor/github.com/go-logr/logr/slogr.go generated vendored Normal file
View File

@@ -0,0 +1,100 @@
//go:build go1.21
// +build go1.21
/*
Copyright 2023 The logr Authors.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/
package logr
import (
"context"
"log/slog"
)
// FromSlogHandler returns a Logger which writes to the slog.Handler.
//
// The logr verbosity level is mapped to slog levels such that V(0) becomes
// slog.LevelInfo and V(4) becomes slog.LevelDebug.
func FromSlogHandler(handler slog.Handler) Logger {
if handler, ok := handler.(*slogHandler); ok {
if handler.sink == nil {
return Discard()
}
return New(handler.sink).V(int(handler.levelBias))
}
return New(&slogSink{handler: handler})
}
// ToSlogHandler returns a slog.Handler which writes to the same sink as the Logger.
//
// The returned logger writes all records with level >= slog.LevelError as
// error log entries with LogSink.Error, regardless of the verbosity level of
// the Logger:
//
// logger := <some Logger with 0 as verbosity level>
// slog.New(ToSlogHandler(logger.V(10))).Error(...) -> logSink.Error(...)
//
// The level of all other records gets reduced by the verbosity
// level of the Logger and the result is negated. If it happens
// to be negative, then it gets replaced by zero because a LogSink
// is not expected to handled negative levels:
//
// slog.New(ToSlogHandler(logger)).Debug(...) -> logger.GetSink().Info(level=4, ...)
// slog.New(ToSlogHandler(logger)).Warning(...) -> logger.GetSink().Info(level=0, ...)
// slog.New(ToSlogHandler(logger)).Info(...) -> logger.GetSink().Info(level=0, ...)
// slog.New(ToSlogHandler(logger.V(4))).Info(...) -> logger.GetSink().Info(level=4, ...)
func ToSlogHandler(logger Logger) slog.Handler {
if sink, ok := logger.GetSink().(*slogSink); ok && logger.GetV() == 0 {
return sink.handler
}
handler := &slogHandler{sink: logger.GetSink(), levelBias: slog.Level(logger.GetV())}
if slogSink, ok := handler.sink.(SlogSink); ok {
handler.slogSink = slogSink
}
return handler
}
// SlogSink is an optional interface that a LogSink can implement to support
// logging through the slog.Logger or slog.Handler APIs better. It then should
// also support special slog values like slog.Group. When used as a
// slog.Handler, the advantages are:
//
// - stack unwinding gets avoided in favor of logging the pre-recorded PC,
// as intended by slog
// - proper grouping of key/value pairs via WithGroup
// - verbosity levels > slog.LevelInfo can be recorded
// - less overhead
//
// Both APIs (Logger and slog.Logger/Handler) then are supported equally
// well. Developers can pick whatever API suits them better and/or mix
// packages which use either API in the same binary with a common logging
// implementation.
//
// This interface is necessary because the type implementing the LogSink
// interface cannot also implement the slog.Handler interface due to the
// different prototype of the common Enabled method.
//
// An implementation could support both interfaces in two different types, but then
// additional interfaces would be needed to convert between those types in FromSlogHandler
// and ToSlogHandler.
type SlogSink interface {
LogSink
Handle(ctx context.Context, record slog.Record) error
WithAttrs(attrs []slog.Attr) SlogSink
WithGroup(name string) SlogSink
}

View File

@@ -23,10 +23,11 @@ limitations under the License.
//
// See the README in the top-level [./logr] package for a discussion of
// interoperability.
//
// Deprecated: use the main logr package instead.
package slogr
import (
"context"
"log/slog"
"github.com/go-logr/logr"
@@ -34,75 +35,27 @@ import (
// NewLogr returns a logr.Logger which writes to the slog.Handler.
//
// The logr verbosity level is mapped to slog levels such that V(0) becomes
// slog.LevelInfo and V(4) becomes slog.LevelDebug.
// Deprecated: use [logr.FromSlogHandler] instead.
func NewLogr(handler slog.Handler) logr.Logger {
if handler, ok := handler.(*slogHandler); ok {
if handler.sink == nil {
return logr.Discard()
}
return logr.New(handler.sink).V(int(handler.levelBias))
}
return logr.New(&slogSink{handler: handler})
return logr.FromSlogHandler(handler)
}
// NewSlogHandler returns a slog.Handler which writes to the same sink as the logr.Logger.
//
// The returned logger writes all records with level >= slog.LevelError as
// error log entries with LogSink.Error, regardless of the verbosity level of
// the logr.Logger:
//
// logger := <some logr.Logger with 0 as verbosity level>
// slog.New(NewSlogHandler(logger.V(10))).Error(...) -> logSink.Error(...)
//
// The level of all other records gets reduced by the verbosity
// level of the logr.Logger and the result is negated. If it happens
// to be negative, then it gets replaced by zero because a LogSink
// is not expected to handled negative levels:
//
// slog.New(NewSlogHandler(logger)).Debug(...) -> logger.GetSink().Info(level=4, ...)
// slog.New(NewSlogHandler(logger)).Warning(...) -> logger.GetSink().Info(level=0, ...)
// slog.New(NewSlogHandler(logger)).Info(...) -> logger.GetSink().Info(level=0, ...)
// slog.New(NewSlogHandler(logger.V(4))).Info(...) -> logger.GetSink().Info(level=4, ...)
// Deprecated: use [logr.ToSlogHandler] instead.
func NewSlogHandler(logger logr.Logger) slog.Handler {
if sink, ok := logger.GetSink().(*slogSink); ok && logger.GetV() == 0 {
return sink.handler
}
return logr.ToSlogHandler(logger)
}
handler := &slogHandler{sink: logger.GetSink(), levelBias: slog.Level(logger.GetV())}
if slogSink, ok := handler.sink.(SlogSink); ok {
handler.slogSink = slogSink
}
return handler
// ToSlogHandler returns a slog.Handler which writes to the same sink as the logr.Logger.
//
// Deprecated: use [logr.ToSlogHandler] instead.
func ToSlogHandler(logger logr.Logger) slog.Handler {
return logr.ToSlogHandler(logger)
}
// SlogSink is an optional interface that a LogSink can implement to support
// logging through the slog.Logger or slog.Handler APIs better. It then should
// also support special slog values like slog.Group. When used as a
// slog.Handler, the advantages are:
// logging through the slog.Logger or slog.Handler APIs better.
//
// - stack unwinding gets avoided in favor of logging the pre-recorded PC,
// as intended by slog
// - proper grouping of key/value pairs via WithGroup
// - verbosity levels > slog.LevelInfo can be recorded
// - less overhead
//
// Both APIs (logr.Logger and slog.Logger/Handler) then are supported equally
// well. Developers can pick whatever API suits them better and/or mix
// packages which use either API in the same binary with a common logging
// implementation.
//
// This interface is necessary because the type implementing the LogSink
// interface cannot also implement the slog.Handler interface due to the
// different prototype of the common Enabled method.
//
// An implementation could support both interfaces in two different types, but then
// additional interfaces would be needed to convert between those types in NewLogr
// and NewSlogHandler.
type SlogSink interface {
logr.LogSink
Handle(ctx context.Context, record slog.Record) error
WithAttrs(attrs []slog.Attr) SlogSink
WithGroup(name string) SlogSink
}
// Deprecated: use [logr.SlogSink] instead.
type SlogSink = logr.SlogSink

View File

@@ -17,24 +17,22 @@ See the License for the specific language governing permissions and
limitations under the License.
*/
package slogr
package logr
import (
"context"
"log/slog"
"runtime"
"time"
"github.com/go-logr/logr"
)
var (
_ logr.LogSink = &slogSink{}
_ logr.CallDepthLogSink = &slogSink{}
_ Underlier = &slogSink{}
_ LogSink = &slogSink{}
_ CallDepthLogSink = &slogSink{}
_ Underlier = &slogSink{}
)
// Underlier is implemented by the LogSink returned by NewLogr.
// Underlier is implemented by the LogSink returned by NewFromLogHandler.
type Underlier interface {
// GetUnderlying returns the Handler used by the LogSink.
GetUnderlying() slog.Handler
@@ -54,7 +52,7 @@ type slogSink struct {
handler slog.Handler
}
func (l *slogSink) Init(info logr.RuntimeInfo) {
func (l *slogSink) Init(info RuntimeInfo) {
l.callDepth = info.CallDepth
}
@@ -62,7 +60,7 @@ func (l *slogSink) GetUnderlying() slog.Handler {
return l.handler
}
func (l *slogSink) WithCallDepth(depth int) logr.LogSink {
func (l *slogSink) WithCallDepth(depth int) LogSink {
newLogger := *l
newLogger.callDepth += depth
return &newLogger
@@ -93,18 +91,18 @@ func (l *slogSink) log(err error, msg string, level slog.Level, kvList ...interf
record.AddAttrs(slog.Any(errKey, err))
}
record.Add(kvList...)
l.handler.Handle(context.Background(), record)
_ = l.handler.Handle(context.Background(), record)
}
func (l slogSink) WithName(name string) logr.LogSink {
func (l slogSink) WithName(name string) LogSink {
if l.name != "" {
l.name = l.name + "/"
l.name += "/"
}
l.name += name
return &l
}
func (l slogSink) WithValues(kvList ...interface{}) logr.LogSink {
func (l slogSink) WithValues(kvList ...interface{}) LogSink {
l.handler = l.handler.WithAttrs(kvListToAttrs(kvList...))
return &l
}

View File

@@ -1,5 +1,24 @@
# Changelog
## Release 3.2.3 (2022-11-29)
### Changed
- Updated docs (thanks @book987 @aJetHorn @neelayu @pellizzetti @apricote @SaigyoujiYuyuko233 @AlekSi)
- #348: Updated huandu/xstrings which fixed a snake case bug (thanks @yxxhero)
- #353: Updated masterminds/semver which included bug fixes
- #354: Updated golang.org/x/crypto which included bug fixes
## Release 3.2.2 (2021-02-04)
This is a re-release of 3.2.1 to satisfy something with the Go module system.
## Release 3.2.1 (2021-02-04)
### Changed
- Upgraded `Masterminds/goutils` to `v1.1.1`. see the [Security Advisory](https://github.com/Masterminds/goutils/security/advisories/GHSA-xg2h-wx96-xgxr)
## Release 3.2.0 (2020-12-14)
### Added

View File

@@ -1,4 +1,4 @@
# Slim-Sprig: Template functions for Go templates [![GoDoc](https://godoc.org/github.com/go-task/slim-sprig?status.svg)](https://godoc.org/github.com/go-task/slim-sprig) [![Go Report Card](https://goreportcard.com/badge/github.com/go-task/slim-sprig)](https://goreportcard.com/report/github.com/go-task/slim-sprig)
# Slim-Sprig: Template functions for Go templates [![Go Reference](https://pkg.go.dev/badge/github.com/go-task/slim-sprig/v3.svg)](https://pkg.go.dev/github.com/go-task/slim-sprig/v3)
Slim-Sprig is a fork of [Sprig](https://github.com/Masterminds/sprig), but with
all functions that depend on external (non standard library) or crypto packages

View File

@@ -1,6 +1,6 @@
# https://taskfile.dev
version: '2'
version: '3'
tasks:
default:

View File

@@ -1,203 +0,0 @@
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf
of any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
APPENDIX: How to apply the Apache License to your work.
To apply the Apache License to your work, attach the following
boilerplate notice, with the fields enclosed by brackets "[]"
replaced with your own identifying information. (Don't include
the brackets!) The text should be enclosed in the appropriate
comment syntax for the file format. We also recommend that a
file or class name and description of purpose be included on the
same "printed page" as the copyright notice for easier
identification within third-party archives.
Copyright [yyyy] [name of copyright owner]
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

View File

@@ -1,4 +0,0 @@
# Compiler support code
This directory contains compiler support code used by Gnostic and Gnostic
extensions.

View File

@@ -1,28 +0,0 @@
// Copyright 2017 Google LLC. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package compiler
import (
"github.com/google/gnostic-models/compiler"
)
// Context contains state of the compiler as it traverses a document.
type Context = compiler.Context
// NewContextWithExtensions returns a new object representing the compiler state
var NewContextWithExtensions = compiler.NewContextWithExtensions
// NewContext returns a new object representing the compiler state
var NewContext = compiler.NewContext

View File

@@ -1,31 +0,0 @@
// Copyright 2017 Google LLC. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package compiler
import (
"github.com/google/gnostic-models/compiler"
)
// Error represents compiler errors and their location in the document.
type Error = compiler.Error
// NewError creates an Error.
var NewError = compiler.NewError
// ErrorGroup is a container for groups of Error values.
type ErrorGroup = compiler.ErrorGroup
// NewErrorGroupOrNil returns a new ErrorGroup for a slice of errors or nil if the slice is empty.
var NewErrorGroupOrNil = compiler.NewErrorGroupOrNil

View File

@@ -1,25 +0,0 @@
// Copyright 2017 Google LLC. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package compiler
import (
"github.com/google/gnostic-models/compiler"
)
// ExtensionHandler describes a binary that is called by the compiler to handle specification extensions.
type ExtensionHandler = compiler.ExtensionHandler
// CallExtension calls a binary extension handler.
var CallExtension = compiler.CallExtension

View File

@@ -1,105 +0,0 @@
// Copyright 2017 Google LLC. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package compiler
import (
"github.com/google/gnostic-models/compiler"
)
// compiler helper functions, usually called from generated code
// UnpackMap gets a *yaml.Node if possible.
var UnpackMap = compiler.UnpackMap
// SortedKeysForMap returns the sorted keys of a yamlv2.MapSlice.
var SortedKeysForMap = compiler.SortedKeysForMap
// MapHasKey returns true if a yamlv2.MapSlice contains a specified key.
var MapHasKey = compiler.MapHasKey
// MapValueForKey gets the value of a map value for a specified key.
var MapValueForKey = compiler.MapValueForKey
// ConvertInterfaceArrayToStringArray converts an array of interfaces to an array of strings, if possible.
var ConvertInterfaceArrayToStringArray = compiler.ConvertInterfaceArrayToStringArray
// SequenceNodeForNode returns a node if it is a SequenceNode.
var SequenceNodeForNode = compiler.SequenceNodeForNode
// BoolForScalarNode returns the bool value of a node.
var BoolForScalarNode = compiler.BoolForScalarNode
// IntForScalarNode returns the integer value of a node.
var IntForScalarNode = compiler.IntForScalarNode
// FloatForScalarNode returns the float value of a node.
var FloatForScalarNode = compiler.FloatForScalarNode
// StringForScalarNode returns the string value of a node.
var StringForScalarNode = compiler.StringForScalarNode
// StringArrayForSequenceNode converts a sequence node to an array of strings, if possible.
var StringArrayForSequenceNode = compiler.StringArrayForSequenceNode
// MissingKeysInMap identifies which keys from a list of required keys are not in a map.
var MissingKeysInMap = compiler.MissingKeysInMap
// InvalidKeysInMap returns keys in a map that don't match a list of allowed keys and patterns.
var InvalidKeysInMap = compiler.InvalidKeysInMap
// NewNullNode creates a new Null node.
var NewNullNode = compiler.NewNullNode
// NewMappingNode creates a new Mapping node.
var NewMappingNode = compiler.NewMappingNode
// NewSequenceNode creates a new Sequence node.
var NewSequenceNode = compiler.NewSequenceNode
// NewScalarNodeForString creates a new node to hold a string.
var NewScalarNodeForString = compiler.NewScalarNodeForString
// NewSequenceNodeForStringArray creates a new node to hold an array of strings.
var NewSequenceNodeForStringArray = compiler.NewSequenceNodeForStringArray
// NewScalarNodeForBool creates a new node to hold a bool.
var NewScalarNodeForBool = compiler.NewScalarNodeForBool
// NewScalarNodeForFloat creates a new node to hold a float.
var NewScalarNodeForFloat = compiler.NewScalarNodeForFloat
// NewScalarNodeForInt creates a new node to hold an integer.
var NewScalarNodeForInt = compiler.NewScalarNodeForInt
// PluralProperties returns the string "properties" pluralized.
var PluralProperties = compiler.PluralProperties
// StringArrayContainsValue returns true if a string array contains a specified value.
var StringArrayContainsValue = compiler.StringArrayContainsValue
// StringArrayContainsValues returns true if a string array contains all of a list of specified values.
var StringArrayContainsValues = compiler.StringArrayContainsValues
// StringValue returns the string value of an item.
var StringValue = compiler.StringValue
// Description returns a human-readable represention of an item.
var Description = compiler.Description
// Display returns a description of a node for use in error messages.
var Display = compiler.Display
// Marshal creates a yaml version of a structure in our preferred style
var Marshal = compiler.Marshal

View File

@@ -1,61 +0,0 @@
// Copyright 2017 Google LLC. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package compiler
import (
"github.com/google/gnostic-models/compiler"
)
// EnableFileCache turns on file caching.
var EnableFileCache = compiler.EnableFileCache
// EnableInfoCache turns on parsed info caching.
var EnableInfoCache = compiler.EnableInfoCache
// DisableFileCache turns off file caching.
var DisableFileCache = compiler.DisableFileCache
// DisableInfoCache turns off parsed info caching.
var DisableInfoCache = compiler.DisableInfoCache
// RemoveFromFileCache removes an entry from the file cache.
var RemoveFromFileCache = compiler.RemoveFromFileCache
// RemoveFromInfoCache removes an entry from the info cache.
var RemoveFromInfoCache = compiler.RemoveFromInfoCache
// GetInfoCache returns the info cache map.
var GetInfoCache = compiler.GetInfoCache
// ClearFileCache clears the file cache.
var ClearFileCache = compiler.ClearFileCache
// ClearInfoCache clears the info cache.
var ClearInfoCache = compiler.ClearInfoCache
// ClearCaches clears all caches.
var ClearCaches = compiler.ClearCaches
// FetchFile gets a specified file from the local filesystem or a remote location.
var FetchFile = compiler.FetchFile
// ReadBytesForFile reads the bytes of a file.
var ReadBytesForFile = compiler.ReadBytesForFile
// ReadInfoFromBytes unmarshals a file as a *yaml.Node.
var ReadInfoFromBytes = compiler.ReadInfoFromBytes
// ReadInfoForRef reads a file and return the fragment needed to resolve a $ref.
var ReadInfoForRef = compiler.ReadInfoForRef

File diff suppressed because it is too large Load Diff

Some files were not shown because too many files have changed in this diff Show More