Skip to content

G
gitlab-shell
Commit b34a88ce authored by Andrew Newdigate's avatar Andrew Newdigate
Browse files
Options

Vendor LabKit into GitLab-Shell 

Done using `govendor fetch -v gitlab.com/gitlab-org/labkit/...@master`
parent 39262b39
Expand all Hide whitespace changes
Inline Side-by-side
Showing with 16304 additions and 0 deletions
go/vendor/github.com/lightstep/lightstep-tracer-go/CHANGELOG.md 0 → 100644
View file @ b34a88ce
# Changelog
## [Pending Release](https://github.com/lightstep/lightstep-tracer-go/compare/v0.15.6...HEAD)
* Thrift transport is now removed.
* Requires go >= 1.7
* Imports `context` via the standard library instead of `golang.org/x/net/context`
* Fixes [#182](https://github.com/lightstep/lightstep-tracer-go/issues/182), so that `StartSpan` can now take `SpanReference`s to non-LightStep `SpanContext`s use
## [v0.15.6](https://github.com/lightstep/lightstep-tracer-go/compare/v0.15.5...v0.15.6)
* Minor update to `sendspan` to make it easier to pick which transport protocol to use.
* Add a new field to Options: DialOptions. These allow setting custom grpc dial options when using grpc.
* This is necessary to have customer balancers or interceptors.
* DialOptions shouldn't be set unless it is needed, it will default correctly.
* Added a new field to Endpoint: Scheme. Scheme can be used to override the default schemes (http/https) or set a custom scheme (for grpc).
* This is necessary for using custom grpc resolvers.
* If callers are using struct construction without field names (i.e. Endpoint{"host", port, ...}), they will need to add a new field (scheme = "").
* Scheme shouldn't be set unless it needs to overridden, it will default correctly.
## [v0.15.5](https://github.com/lightstep/lightstep-tracer-go/compare/v0.15.4...v0.15.5)
* Internal performance optimizations and a bug fix for issue [#161](https://github.com/lightstep/lightstep-tracer-go/issues/161)
## [v0.15.4](https://github.com/lightstep/lightstep-tracer-go/compare/v0.15.3...v0.15.4)
* This change affects LightStep's internal testing, not a functional change.
## [v0.15.3](https://github.com/lightstep/lightstep-tracer-go/compare/v0.15.2...v0.15.3)
* Adds compatibility for io.Writer and io.Reader in Inject/Extract, as required by Open Tracing.
## [v0.15.2](https://github.com/lightstep/lightstep-tracer-go/compare/v0.15.1...v0.15.2)
* Adds lightstep.GetLightStepReporterID.
## [v0.15.1](https://github.com/lightstep/lightstep-tracer-go/compare/v0.15.0...v0.15.1)
* Adds Gopkg.toml
## [v0.15.0](https://github.com/lightstep/lightstep-tracer-go/compare/v0.14.0...v0.15.0)
* We are replacing the internal diagnostic logging with a more flexible “Event” framework. This enables you to track and understand tracer problems with metrics and logging tools of your choice.
* We are also changed the types of the Close() and Flush() methods to take a context parameter to support cancellation. These changes are *not* backwards compatible and *you will need to update your instrumentation*. We are providing a NewTracerv0_14() method that is a drop-in replacement for the previous version.
## [v0.14.0](https://github.com/lightstep/lightstep-tracer-go/compare/v0.13.0...v0.14.0)
* Flush buffer syncronously on Close.
* Flush twice if a flush is already in flight.
* Remove gogo in favor of golang/protobuf.
* Requires grpc-go >= 1.4.0.
## [v0.13.0](https://github.com/lightstep/lightstep-tracer-go/compare/v0.12.0...v0.13.0)
* BasicTracer has been removed.
* Tracer now takes a SpanRecorder as an option.
* Tracer interface now includes Close and Flush.
* Tests redone with ginkgo/gomega.
## [v0.12.0](https://github.com/lightstep/lightstep-tracer-go/compare/v0.11.0...v0.12.0)
* Added CloseTracer function to flush and close a lightstep recorder.
## [v0.11.0](https://github.com/lightstep/lightstep-tracer-go/compare/v0.10.0...v0.11.0)
* Thrift transport is now deprecated, gRPC is the default.
go/vendor/github.com/lightstep/lightstep-tracer-go/Gopkg.lock 0 → 100644
View file @ b34a88ce
# This file is autogenerated, do not edit; changes may be undone by the next 'dep ensure'.
[[projects]]
branch = "master"
digest = "1:3b760d3b93f994df8eb1d9ebfad17d3e9e37edcb7f7efaa15b427c0d7a64f4e4"
name = "github.com/golang/protobuf"
packages = [
"proto",
"protoc-gen-go/descriptor",
"ptypes",
"ptypes/any",
"ptypes/duration",
"ptypes/timestamp",
]
pruneopts = ""
revision = "1e59b77b52bf8e4b449a57e6f79f21226d571845"
[[projects]]
digest = "1:32b27072cd55bd2fb7244de0425943d125da6a552ae2b6517cdd965a662baf18"
name = "github.com/onsi/ginkgo"
packages = [
".",
"config",
"internal/codelocation",
"internal/containernode",
"internal/failer",
"internal/leafnodes",
"internal/remote",
"internal/spec",
"internal/spec_iterator",
"internal/specrunner",
"internal/suite",
"internal/testingtproxy",
"internal/writer",
"reporters",
"reporters/stenographer",
"reporters/stenographer/support/go-colorable",
"reporters/stenographer/support/go-isatty",
"types",
]
pruneopts = ""
revision = "9eda700730cba42af70d53180f9dcce9266bc2bc"
version = "v1.4.0"
[[projects]]
digest = "1:a4e59d0b2821c983b58c317f141cd77df20570979632da8a7a352e5d12698de7"
name = "github.com/onsi/gomega"
packages = [
".",
"format",
"internal/assertion",
"internal/asyncassertion",
"internal/oraclematcher",
"internal/testingtsupport",
"matchers",
"matchers/support/goraph/bipartitegraph",
"matchers/support/goraph/edge",
"matchers/support/goraph/node",
"matchers/support/goraph/util",
"types",
]
pruneopts = ""
revision = "c893efa28eb45626cdaa76c9f653b62488858837"
version = "v1.2.0"
[[projects]]
digest = "1:78fb99d6011c2ae6c72f3293a83951311147b12b06a5ffa43abf750c4fab6ac5"
name = "github.com/opentracing/opentracing-go"
packages = [
".",
"log",
]
pruneopts = ""
revision = "1949ddbfd147afd4d964a9f00b24eb291e0e7c38"
version = "v1.0.2"
[[projects]]
branch = "master"
digest = "1:950b672f2ee80d0fc4c95a15a976ba9ee573a6fb8ede8a777770b2230776367e"
name = "golang.org/x/net"
packages = [
"context",
"html",
"html/atom",
"html/charset",
"http2",
"http2/hpack",
"idna",
"internal/timeseries",
"lex/httplex",
"trace",
]
pruneopts = ""
revision = "dc871a5d77e227f5bbf6545176ef3eeebf87e76e"
[[projects]]
branch = "master"
digest = "1:5da11ab130476e2736f32140f3c1aed2a3a96e9ba8963711a7d38db783d042bd"
name = "golang.org/x/sys"
packages = ["unix"]
pruneopts = ""
revision = "a0f4589a76f1f83070cb9e5613809e1d07b97c13"
[[projects]]
branch = "master"
digest = "1:1c70f7bb89783a026dc32920575a3feef48e065ef6e170ad227903e8194d7a36"
name = "golang.org/x/text"
packages = [
"collate",
"collate/build",
"encoding",
"encoding/charmap",
"encoding/htmlindex",
"encoding/internal",
"encoding/internal/identifier",
"encoding/japanese",
"encoding/korean",
"encoding/simplifiedchinese",
"encoding/traditionalchinese",
"encoding/unicode",
"internal/colltab",
"internal/gen",
"internal/tag",
"internal/triegen",
"internal/ucd",
"internal/utf8internal",
"language",
"runes",
"secure/bidirule",
"transform",
"unicode/bidi",
"unicode/cldr",
"unicode/norm",
"unicode/rangetable",
]
pruneopts = ""
revision = "be25de41fadfae372d6470bda81ca6beb55ef551"
[[projects]]
branch = "master"
digest = "1:6c15114fafeac4c833544476dec4207a3476799d50ab5165af79eb97806884cf"
name = "google.golang.org/genproto"
packages = [
"googleapis/api/annotations",
"googleapis/rpc/status",
]
pruneopts = ""
revision = "7f0da29060c682909f650ad8ed4e515bd74fa12a"
[[projects]]
digest = "1:6c00b4702c146631d30396090d40bfc0486f8b3af5c976d6f0daf2bc737cd7b2"
name = "google.golang.org/grpc"
packages = [
".",
"balancer",
"balancer/roundrobin",
"codes",
"connectivity",
"credentials",
"encoding",
"grpclb/grpc_lb_v1/messages",
"grpclog",
"internal",
"keepalive",
"metadata",
"naming",
"peer",
"resolver",
"resolver/dns",
"resolver/manual",
"resolver/passthrough",
"stats",
"status",
"tap",
"transport",
]
pruneopts = ""
revision = "be077907e29fdb945d351e4284eb5361e7f8924e"
version = "v1.8.1"
[[projects]]
branch = "v2"
digest = "1:f769ed60e075e4221612c2f4162fccc9d3795ef358fa463425e3b3d7a5debb27"
name = "gopkg.in/yaml.v2"
packages = ["."]
pruneopts = ""
revision = "287cf08546ab5e7e37d55a84f7ed3fd1db036de5"
[solve-meta]
analyzer-name = "dep"
analyzer-version = 1
input-imports = [
"github.com/golang/protobuf/proto",
"github.com/golang/protobuf/ptypes/timestamp",
"github.com/onsi/ginkgo",
"github.com/onsi/gomega",
"github.com/onsi/gomega/types",
"github.com/opentracing/opentracing-go",
"github.com/opentracing/opentracing-go/log",
"golang.org/x/net/context",
"google.golang.org/genproto/googleapis/api/annotations",
"google.golang.org/grpc",
"google.golang.org/grpc/credentials",
]
solver-name = "gps-cdcl"
solver-version = 1
go/vendor/github.com/lightstep/lightstep-tracer-go/Gopkg.toml 0 → 100644
View file @ b34a88ce
[[constraint]]
name = "github.com/onsi/ginkgo"
version = "1.4.0"
[[constraint]]
name = "github.com/onsi/gomega"
version = "1.2.0"
[[constraint]]
name = "github.com/opentracing/opentracing-go"
version = "1.0.2"
[[constraint]]
branch = "master"
name = "golang.org/x/net"
[[constraint]]
name = "google.golang.org/grpc"
version = "1.4.3"
go/vendor/github.com/lightstep/lightstep-tracer-go/LICENSE 0 → 100644
View file @ b34a88ce
The MIT License (MIT)
Copyright (c) 2016
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.
go/vendor/github.com/lightstep/lightstep-tracer-go/Makefile 0 → 100644
View file @ b34a88ce
# tools
GO=go
DOCKER_PRESENT = $(shell command -v docker 2> /dev/null)
LOCAL_GOPATH = $(PWD)/../../../../
default: build
.PHONY: default build test
# generate_fake: runs counterfeiter in docker container to generate fake classes
# $(1) output file path
# $(2) input file path
# $(3) class name
define generate_fake
docker run --rm -v $(LOCAL_GOPATH):/usergo \
lightstep/gobuild:latest /bin/bash -c "\
cd /usergo/src/github.com/lightstep/lightstep-tracer-go; \
counterfeiter -o $(1) $(2) $(3)"
endef
collectorpb/collectorpbfakes/fake_collector_service_client.go: collectorpb/collector.pb.go
$(call generate_fake,collectorpb/collectorpbfakes/fake_collector_service_client.go,collectorpb/collector.pb.go,CollectorServiceClient)
# gRPC
ifeq (,$(wildcard lightstep-tracer-common/collector.proto))
collectorpb/collector.pb.go:
else
collectorpb/collector.pb.go: lightstep-tracer-common/collector.proto
docker run --rm -v $(shell pwd)/lightstep-tracer-common:/input:ro -v $(shell pwd)/collectorpb:/output \
lightstep/grpc-gateway:latest \
protoc -I/root/go/src/tmp/vendor/github.com/grpc-ecosystem/grpc-gateway/third_party/googleapis --go_out=plugins=grpc:/output --proto_path=/input /input/collector.proto
endif
# gRPC
ifeq (,$(wildcard lightstep-tracer-common/collector.proto))
lightsteppb/lightstep.pb.go:
else
lightsteppb/lightstep.pb.go: lightstep-tracer-common/lightstep.proto
docker run --rm -v $(shell pwd)/lightstep-tracer-common:/input:ro -v $(shell pwd)/lightsteppb:/output \
lightstep/protoc:latest \
protoc --go_out=plugins=grpc:/output --proto_path=/input /input/lightstep.proto
endif
test: collectorpb/collector.pb.go lightsteppb/lightstep.pb.go \
collectorpb/collectorpbfakes/fake_collector_service_client.go \
lightstepfakes/fake_recorder.go
ifeq ($(DOCKER_PRESENT),)
$(error "docker not found. Please install from https://www.docker.com/")
endif
docker run --rm -v $(LOCAL_GOPATH):/usergo lightstep/gobuild:latest \
ginkgo -race -p /usergo/src/github.com/lightstep/lightstep-tracer-go
bash -c "! git grep -q '[g]ithub.com/golang/glog'"
build: collectorpb/collector.pb.go lightsteppb/lightstep.pb.go \
collectorpb/collectorpbfakes/fake_collector_service_client.go version.go \
lightstepfakes/fake_recorder.go
ifeq ($(DOCKER_PRESENT),)
$(error "docker not found. Please install from https://www.docker.com/")
endif
${GO} build github.com/lightstep/lightstep-tracer-go
# When releasing significant changes, make sure to update the semantic
# version number in `./VERSION`, merge changes, then run `make release_tag`.
version.go: VERSION
./tag_version.sh
release_tag:
git tag -a v`cat ./VERSION`
git push origin v`cat ./VERSION`
go/vendor/github.com/lightstep/lightstep-tracer-go/README.md 0 → 100644
View file @ b34a88ce
# lightstep-tracer-go
[![Circle CI](https://circleci.com/gh/lightstep/lightstep-tracer-go.svg?style=shield)](https://circleci.com/gh/lightstep/lightstep-tracer-go)
[![MIT license](http://img.shields.io/badge/license-MIT-blue.svg)](http://opensource.org/licenses/MIT)
[![GoDoc](https://godoc.org/github.com/lightstep/lightstep-tracer-go?status.svg)](https://godoc.org/github.com/lightstep/lightstep-tracer-go)
The LightStep distributed tracing library for Go.
## Installation
```
$ go get 'github.com/lightstep/lightstep-tracer-go'
```
## API Documentation
Godoc: https://godoc.org/github.com/lightstep/lightstep-tracer-go
## Initialization: Starting a new tracer
To initialize a tracer, configure it with a valid Access Token and optional tuning parameters. Register the tracer as the OpenTracing global tracer so that it will become available to your installed intstrumentations libraries.
```go
import (
"github.com/opentracing/opentracing-go"
"github.com/lightstep/lightstep-tracer-go"
)
func main() {
lightstepTracer := lightstep.NewTracer(lightstep.Options{
AccessToken: "YourAccessToken",
})
opentracing.SetGlobalTracer(lightstepTracer)
}
```
## Instrumenting Code: Using the OpenTracing API
All instrumentation should be done through the OpenTracing API, rather than using the lightstep tracer type directly. For API documentation and advice on instrumentation in general, see the opentracing godocs and the opentracing website.
- https://godoc.org/github.com/opentracing/opentracing-go
- http://opentracing.io
## Flushing and Closing: Managing the tracer lifecycle
As part of managaing your application lifecycle, the lightstep tracer extends the `opentracing.Tracer` interface with methods for manual flushing and closing. To access these methods, you can take the global tracer and typecast it to a `lightstep.Tracer`. As a convenience, the lightstep package provides static methods which perform the typecasting.
```go
import (
"context"
"github.com/opentracing/opentracing-go"
"github.com/lightstep/lightstep-tracer-go"
)
func shutdown(ctx context.Context) {
// access the running tracer
tracer := opentracing.GlobalTracer()
// typecast from opentracing.Tracer to lightstep.Tracer
lsTracer, ok := tracer.(lightstep.Tracer)
if (!ok) {
return
}
lsTracer.Close(ctx)
// or use static methods
lightstep.Close(ctx, tracer)
}
```
## Event Handling: Observing the LightStep tracer
In order to connect diagnostic information from the lightstep tracer into an application's logging and metrics systems, inject an event handler using the `OnEvent` static method. Events may be typecast to check for errors or specific events such as status reports.
```go
import (
"example/logger"
"example/metrics"
"github.com/lightstep/lightstep-tracer-go"
)
logAndMetricsHandler := func(event lightstep.Event){
switch event := event.(type) {
case EventStatusReport:
metrics.Count("tracer.dropped_spans", event.DroppedSpans())
case ErrorEvent:
logger.Error("LS Tracer error: %s", event)
default:
logger.Info("LS Tracer info: %s", event)
}
}
func main() {
// setup event handler first to catch startup errors
lightstep.SetGlobalEventHandler(logAndMetricsHandler)
lightstepTracer := lightstep.NewTracer(lightstep.Options{
AccessToken: "YourAccessToken",
})
opentracing.SetGlobalTracer(lightstepTracer)
}
```
Event handlers will receive events from any active tracers, as well as errors in static functions. It is suggested that you set up event handling before initializing your tracer to catch any errors on initialization.
## Advanced Configuration: Transport and Serialization Protocols
By default, the tracer will send information to LightStep using GRPC and Protocol Buffers which is the recommended configuration. If there are no specific transport protocol needs you have, there is no need to change this default.
There are three total options for transport protocols:
- [Protocol Buffers](https://developers.google.com/protocol-buffers/) over [GRPC](https://grpc.io/) - The recommended, default, and most performant solution.
- \[ EXPERIMENTAL \] [Protocol Buffers](https://developers.google.com/protocol-buffers/) over HTTP - New transport protocol supported for use cases where GRPC isn't an option. In order to enable HTTP you will need to configure the LightStep collectors receiving the data to accept HTTP traffic. Reach out to LightStep for support in this.
You can configure which transport protocol the tracer uses using the `UseGRPC`, and `UseHttp` flags in the options.
\ No newline at end of file
go/vendor/github.com/lightstep/lightstep-tracer-go/collector_client.go 0 → 100644
View file @ b34a88ce
package lightstep
import (
"context"
"io"
"net/http"
cpb "github.com/lightstep/lightstep-tracer-go/collectorpb"
)
// Connection describes a closable connection. Exposed for testing.
type Connection interface {
io.Closer
}
// ConnectorFactory is for testing purposes.
type ConnectorFactory func() (interface{}, Connection, error)
// collectorResponse encapsulates internal grpc/http responses.
type collectorResponse interface {
GetErrors() []string
Disable() bool
DevMode() bool
}
type reportRequest struct {
protoRequest *cpb.ReportRequest
httpRequest *http.Request
}
// collectorClient encapsulates internal grpc/http transports.
type collectorClient interface {
Report(context.Context, reportRequest) (collectorResponse, error)
Translate(context.Context, *reportBuffer) (reportRequest, error)
ConnectClient() (Connection, error)
ShouldReconnect() bool
}
func newCollectorClient(opts Options, reporterID uint64, attributes map[string]string) (collectorClient, error) {
if opts.UseHttp {
return newHTTPCollectorClient(opts, reporterID, attributes)
}
if opts.UseGRPC {
return newGrpcCollectorClient(opts, reporterID, attributes), nil
}
// No transport specified, defaulting to GRPC
return newGrpcCollectorClient(opts, reporterID, attributes), nil
}
go/vendor/github.com/lightstep/lightstep-tracer-go/collector_client_grpc.go 0 → 100644
View file @ b34a88ce
package lightstep
import (
"context"
"fmt"
"reflect"
"time"
"google.golang.org/grpc"
"google.golang.org/grpc/credentials"
// N.B.(jmacd): Do not use google.golang.org/glog in this package.
cpb "github.com/lightstep/lightstep-tracer-go/collectorpb"
)
const (
spansDropped = "spans.dropped"
logEncoderErrors = "log_encoder.errors"
)
var (
intType = reflect.TypeOf(int64(0))
)
// grpcCollectorClient specifies how to send reports back to a LightStep
// collector via grpc.
type grpcCollectorClient struct {
// auth and runtime information
attributes map[string]string
reporterID uint64
// accessToken is the access token used for explicit trace collection requests.
accessToken string
maxReportingPeriod time.Duration // set by GrpcOptions.MaxReportingPeriod
reconnectPeriod time.Duration // set by GrpcOptions.ReconnectPeriod
reportingTimeout time.Duration // set by GrpcOptions.ReportTimeout
// Remote service that will receive reports.
address string
grpcClient cpb.CollectorServiceClient
connTimestamp time.Time
dialOptions []grpc.DialOption
// converters
converter *protoConverter
// For testing purposes only
grpcConnectorFactory ConnectorFactory
}
func newGrpcCollectorClient(opts Options, reporterID uint64, attributes map[string]string) *grpcCollectorClient {
rec := &grpcCollectorClient{
attributes: attributes,
reporterID: reporterID,
accessToken: opts.AccessToken,
maxReportingPeriod: opts.ReportingPeriod,
reconnectPeriod: opts.ReconnectPeriod,
reportingTimeout: opts.ReportTimeout,
dialOptions: opts.DialOptions,
converter: newProtoConverter(opts),
grpcConnectorFactory: opts.ConnFactory,
}
if len(opts.Collector.Scheme) > 0 {
rec.address = opts.Collector.urlWithoutPath()
} else {
rec.address = opts.Collector.SocketAddress()
}
rec.dialOptions = append(rec.dialOptions, grpc.WithDefaultCallOptions(grpc.MaxCallSendMsgSize(opts.GRPCMaxCallSendMsgSizeBytes)))
if opts.Collector.Plaintext {
rec.dialOptions = append(rec.dialOptions, grpc.WithInsecure())
} else {
rec.dialOptions = append(rec.dialOptions, grpc.WithTransportCredentials(credentials.NewClientTLSFromCert(nil, "")))
}
return rec
}
func (client *grpcCollectorClient) ConnectClient() (Connection, error) {
now := time.Now()
var conn Connection
if client.grpcConnectorFactory != nil {
uncheckedClient, transport, err := client.grpcConnectorFactory()
if err != nil {
return nil, err
}
grpcClient, ok := uncheckedClient.(cpb.CollectorServiceClient)
if !ok {
return nil, fmt.Errorf("gRPC connector factory did not provide valid client")
}
conn = transport
client.grpcClient = grpcClient
} else {
transport, err := grpc.Dial(client.address, client.dialOptions...)
if err != nil {
return nil, err
}
conn = transport
client.grpcClient = cpb.NewCollectorServiceClient(transport)
}
client.connTimestamp = now
return conn, nil
}
func (client *grpcCollectorClient) ShouldReconnect() bool {
return time.Since(client.connTimestamp) > client.reconnectPeriod
}
func (client *grpcCollectorClient) Report(ctx context.Context, req reportRequest) (collectorResponse, error) {
if req.protoRequest == nil {
return nil, fmt.Errorf("protoRequest cannot be null")
}
resp, err := client.grpcClient.Report(ctx, req.protoRequest)
if err != nil {
return nil, err
}
return resp, nil
}
func (client *grpcCollectorClient) Translate(ctx context.Context, buffer *reportBuffer) (reportRequest, error) {
req := client.converter.toReportRequest(
client.reporterID,
client.attributes,
client.accessToken,
buffer,
)
return reportRequest{
protoRequest: req,
}, nil
}
go/vendor/github.com/lightstep/lightstep-tracer-go/collector_client_http.go 0 → 100644
View file @ b34a88ce
package lightstep
import (
"bytes"
"context"
"fmt"
"io/ioutil"
"net"
"net/http"
"net/url"
"time"
"github.com/golang/protobuf/proto"
"github.com/lightstep/lightstep-tracer-go/collectorpb"
)
var (
acceptHeader = http.CanonicalHeaderKey("Accept")
contentTypeHeader = http.CanonicalHeaderKey("Content-Type")
)
const (
collectorHTTPMethod = "POST"
collectorHTTPPath = "/api/v2/reports"
protoContentType = "application/octet-stream"
)
// grpcCollectorClient specifies how to send reports back to a LightStep
// collector via grpc.
type httpCollectorClient struct {
// auth and runtime information
reporterID uint64
accessToken string // accessToken is the access token used for explicit trace collection requests.
attributes map[string]string
reportTimeout time.Duration
reportingPeriod time.Duration
// Remote service that will receive reports.
url *url.URL
client *http.Client
// converters
converter *protoConverter
}
type transportCloser struct {
*http.Transport
}
func (closer transportCloser) Close() error {
closer.CloseIdleConnections()
return nil
}
func newHTTPCollectorClient(
opts Options,
reporterID uint64,
attributes map[string]string,
) (*httpCollectorClient, error) {
url, err := url.Parse(opts.Collector.URL())
if err != nil {
fmt.Println("collector config does not produce valid url", err)
return nil, err
}
url.Path = collectorHTTPPath
return &httpCollectorClient{
reporterID: reporterID,
accessToken: opts.AccessToken,
attributes: attributes,
reportTimeout: opts.ReportTimeout,
reportingPeriod: opts.ReportingPeriod,
url: url,
converter: newProtoConverter(opts),
}, nil
}
func (client *httpCollectorClient) ConnectClient() (Connection, error) {
// Use a transport independent from http.DefaultTransport to provide sane
// defaults that make sense in the context of the lightstep client. The
// differences are mostly on setting timeouts based on the report timeout
// and period.
transport := &http.Transport{
Proxy: http.ProxyFromEnvironment,
DialContext: (&net.Dialer{
Timeout: client.reportTimeout / 2,
DualStack: true,
}).DialContext,
// The collector responses are very small, there is no point asking for
// a compressed payload, explicitly disabling it.
DisableCompression: true,
IdleConnTimeout: 2 * client.reportingPeriod,
TLSHandshakeTimeout: client.reportTimeout / 2,
ResponseHeaderTimeout: client.reportTimeout,
ExpectContinueTimeout: client.reportTimeout,
MaxResponseHeaderBytes: 64 * 1024, // 64 KB, just a safeguard
}
client.client = &http.Client{
Transport: transport,
Timeout: client.reportTimeout,
}
return transportCloser{transport}, nil
}
func (client *httpCollectorClient) ShouldReconnect() bool {
// http.Transport will handle connection reuse under the hood
return false
}
func (client *httpCollectorClient) Report(context context.Context, req reportRequest) (collectorResponse, error) {
if req.httpRequest == nil {
return nil, fmt.Errorf("httpRequest cannot be null")
}
httpResponse, err := client.client.Do(req.httpRequest.WithContext(context))
if err != nil {
return nil, err
}
defer httpResponse.Body.Close()
response, err := client.toResponse(httpResponse)
if err != nil {
return nil, err
}
return response, nil
}
func (client *httpCollectorClient) Translate(ctx context.Context, buffer *reportBuffer) (reportRequest, error) {
httpRequest, err := client.toRequest(ctx, buffer)
if err != nil {
return reportRequest{}, err
}
return reportRequest{
httpRequest: httpRequest,
}, nil
}
func (client *httpCollectorClient) toRequest(
context context.Context,
buffer *reportBuffer,
) (*http.Request, error) {
protoRequest := client.converter.toReportRequest(
client.reporterID,
client.attributes,
client.accessToken,
buffer,
)
buf, err := proto.Marshal(protoRequest)
if err != nil {
return nil, err
}
requestBody := bytes.NewReader(buf)
request, err := http.NewRequest(collectorHTTPMethod, client.url.String(), requestBody)
if err != nil {
return nil, err
}
request = request.WithContext(context)
request.Header.Set(contentTypeHeader, protoContentType)
request.Header.Set(acceptHeader, protoContentType)
return request, nil
}
func (client *httpCollectorClient) toResponse(response *http.Response) (collectorResponse, error) {
if response.StatusCode != http.StatusOK {
return nil, fmt.Errorf("status code (%d) is not ok", response.StatusCode)
}
body, err := ioutil.ReadAll(response.Body)
if err != nil {
return nil, err
}
protoResponse := &collectorpb.ReportResponse{}
if err := proto.Unmarshal(body, protoResponse); err != nil {
return nil, err
}
return protoResponse, nil
}
go/vendor/github.com/lightstep/lightstep-tracer-go/event_handlers.go 0 → 100644
View file @ b34a88ce
package lightstep
import (
"log"
"sync"
"sync/atomic"
)
func init() {
SetGlobalEventHandler(NewEventLogOneError())
}
var eventHandler atomic.Value
// An EventHandler can be registered with SetGlobalEventHandler to
type EventHandler func(Event)
// emitEvent is a thread-safe function for emiting tracer events.
func emitEvent(event Event) {
handler := eventHandler.Load().(EventHandler)
handler(event)
}
// SetGlobalEventHandler sets a global handler to receive tracer events as they occur. Events
// may be emitted by the tracer, or by calls to static functions in this package.
// It is suggested that you set your EventHandler before starting your tracer,
// but it is safe to set a new handler at any time. Setting a new handler removes
// the previous one.
//
// The EventHandler is called synchronously – do not block in your handler as it
// may interfere with the tracer. If no EventHandler handler is set, a
// LogOnceOnError handler is set by default.
//
// NOTE: Event handling is for reporting purposes only. It is not intended as a
// mechanism for controling or restarting the tracer. Connection issues, retry
// logic, and other transient errors are handled internally by the tracer.
func SetGlobalEventHandler(handler EventHandler) {
eventHandler.Store(handler)
}
/*
SetGlobalEventHandler Handlers
*/
// NewEventLogger logs events using the standard go logger.
func NewEventLogger() EventHandler {
return logOnEvent
}
func logOnEvent(event Event) {
switch event := event.(type) {
case ErrorEvent:
log.Println("LS Tracer error: ", event)
default:
log.Println("LS Tracer event: ", event)
}
}
// NewEventLogOneError logs the first error event that occurs.
func NewEventLogOneError() EventHandler {
logger := logOneError{}
return logger.OnEvent
}
type logOneError struct {
sync.Once
}
func (l *logOneError) OnEvent(event Event) {
switch event := event.(type) {
case ErrorEvent:
l.Once.Do(func() {
log.Printf("LS Tracer error: (%s). NOTE: Set the SetGlobalEventHandler handler to log events.\n", event.Error())
})
}
}
// NewEventChannel returns an SetGlobalEventHandler callback handler, and a channel that
// produces the errors. When the channel buffer is full, subsequent errors will
// be dropped. A buffer size of less than one is incorrect, and will be adjusted
// to a buffer size of one.
func NewEventChannel(buffer int) (EventHandler, <-chan Event) {
if buffer < 1 {
buffer = 1
}
eventChan := make(chan Event, buffer)
handler := func(event Event) {
select {
case eventChan <- event:
default:
}
}
return handler, eventChan
}
go/vendor/github.com/lightstep/lightstep-tracer-go/events.go 0 → 100644
View file @ b34a88ce
package lightstep
import (
"errors"
"fmt"
"reflect"
"time"
opentracing "github.com/opentracing/opentracing-go"
)
// An Event is emitted by the LightStep tracer as a reporting mechanism. They are
// handled by registering an EventHandler callback via SetGlobalEventHandler. The
// emitted events may be cast to specific event types in order access additional
// information.
//
// NOTE: To ensure that events can be accurately identified, each event type contains
// a sentinel method matching the name of the type. This method is a no-op, it is
// only used for type coersion.
type Event interface {
Event()
String() string
}
// The ErrorEvent type can be used to filter events. The `Err` method
// retuns the underlying error.
type ErrorEvent interface {
Event
error
Err() error
}
// EventStartError occurs if the Options passed to NewTracer are invalid, and
// the Tracer has failed to start.
type EventStartError interface {
ErrorEvent
EventStartError()
}
type eventStartError struct {
err error
}
func newEventStartError(err error) *eventStartError {
return &eventStartError{err: err}
}
func (*eventStartError) Event() {}
func (*eventStartError) EventStartError() {}
func (e *eventStartError) String() string {
return e.err.Error()
}
func (e *eventStartError) Error() string {
return e.err.Error()
}
func (e *eventStartError) Err() error {
return e.err
}
// EventFlushErrorState lists the possible causes for a flush to fail.
type EventFlushErrorState string
// Constant strings corresponding to flush errors
const (
FlushErrorTracerClosed EventFlushErrorState = "flush failed, the tracer is closed."
FlushErrorTracerDisabled EventFlushErrorState = "flush failed, the tracer is disabled."
FlushErrorTransport EventFlushErrorState = "flush failed, could not send report to Collector"
FlushErrorReport EventFlushErrorState = "flush failed, report contained errors"
FlushErrorTranslate EventFlushErrorState = "flush failed, could not translate report"
)
var (
errFlushFailedTracerClosed = errors.New(string(FlushErrorTracerClosed))
)
// EventFlushError occurs when a flush fails to send. Call the `State` method to
// determine the type of error.
type EventFlushError interface {
ErrorEvent
EventFlushError()
State() EventFlushErrorState
}
type eventFlushError struct {
err error
state EventFlushErrorState
}
func newEventFlushError(err error, state EventFlushErrorState) *eventFlushError {
return &eventFlushError{err: err, state: state}
}
func (*eventFlushError) Event() {}
func (*eventFlushError) EventFlushError() {}
func (e *eventFlushError) State() EventFlushErrorState {
return e.state
}
func (e *eventFlushError) String() string {
return e.err.Error()
}
func (e *eventFlushError) Error() string {
return e.err.Error()
}
func (e *eventFlushError) Err() error {
return e.err
}
// EventConnectionError occurs when the tracer fails to maintain it's connection
// with the Collector.
type EventConnectionError interface {
ErrorEvent
EventConnectionError()
}
type eventConnectionError struct {
err error
}
func newEventConnectionError(err error) *eventConnectionError {
return &eventConnectionError{err: err}
}
func (*eventConnectionError) Event() {}
func (*eventConnectionError) EventConnectionError() {}
func (e *eventConnectionError) String() string {
return e.err.Error()
}
func (e *eventConnectionError) Error() string {
return e.err.Error()
}
func (e *eventConnectionError) Err() error {
return e.err
}
// EventStatusReport occurs on every successful flush. It contains all metrics
// collected since the previous succesful flush.
type EventStatusReport interface {
Event
EventStatusReport()
StartTime() time.Time
FinishTime() time.Time
Duration() time.Duration
SentSpans() int
DroppedSpans() int
EncodingErrors() int
}
type eventStatusReport struct {
startTime time.Time
finishTime time.Time
sentSpans int
droppedSpans int
encodingErrors int
}
func newEventStatusReport(
startTime, finishTime time.Time,
sentSpans, droppedSpans, encodingErrors int,
) *eventStatusReport {
return &eventStatusReport{
startTime: startTime,
finishTime: finishTime,
sentSpans: sentSpans,
droppedSpans: droppedSpans,
encodingErrors: encodingErrors,
}
}
func (*eventStatusReport) Event() {}
func (*eventStatusReport) EventStatusReport() {}
func (s *eventStatusReport) SetSentSpans(sent int) {
s.sentSpans = sent
}
func (s *eventStatusReport) StartTime() time.Time {
return s.startTime
}
func (s *eventStatusReport) FinishTime() time.Time {
return s.finishTime
}
func (s *eventStatusReport) Duration() time.Duration {
return s.finishTime.Sub(s.startTime)
}
func (s *eventStatusReport) SentSpans() int {
return s.sentSpans
}
func (s *eventStatusReport) DroppedSpans() int {
return s.droppedSpans
}
func (s *eventStatusReport) EncodingErrors() int {
return s.encodingErrors
}
func (s *eventStatusReport) String() string {
return fmt.Sprint(
"STATUS REPORT start: ", s.startTime,
", end: ", s.finishTime,
", dropped spans: ", s.droppedSpans,
", encoding errors: ", s.encodingErrors,
)
}
// EventUnsupportedTracer occurs when a tracer being passed to a helper function
// fails to typecast as a LightStep tracer.
type EventUnsupportedTracer interface {
ErrorEvent
EventUnsupportedTracer()
Tracer() opentracing.Tracer
}
type eventUnsupportedTracer struct {
tracer opentracing.Tracer
err error
}
func newEventUnsupportedTracer(tracer opentracing.Tracer) EventUnsupportedTracer {
return &eventUnsupportedTracer{
tracer: tracer,
err: fmt.Errorf("unsupported tracer type: %v", reflect.TypeOf(tracer)),
}
}
func (e *eventUnsupportedTracer) Event() {}
func (e *eventUnsupportedTracer) EventUnsupportedTracer() {}
func (e *eventUnsupportedTracer) Tracer() opentracing.Tracer {
return e.tracer
}
func (e *eventUnsupportedTracer) String() string {
return e.err.Error()
}
func (e *eventUnsupportedTracer) Error() string {
return e.err.Error()
}
func (e *eventUnsupportedTracer) Err() error {
return e.err
}
// EventUnsupportedValue occurs when a tracer encounters an unserializable tag
// or log field.
type EventUnsupportedValue interface {
ErrorEvent
EventUnsupportedValue()
Key() string
Value() interface{}
}
type eventUnsupportedValue struct {
key string
value interface{}
err error
}
func newEventUnsupportedValue(key string, value interface{}, err error) EventUnsupportedValue {
if err == nil {
err = fmt.Errorf(
"value `%v` of type `%T` for key `%s` is an unsupported type",
value, value, key,
)
}
return &eventUnsupportedValue{
key: key,
value: value,
err: err,
}
}
func (e *eventUnsupportedValue) Event() {}
func (e *eventUnsupportedValue) EventUnsupportedValue() {}
func (e *eventUnsupportedValue) Key() string {
return e.key
}
func (e *eventUnsupportedValue) Value() interface{} {
return e.value
}
func (e *eventUnsupportedValue) String() string {
return e.err.Error()
}
func (e *eventUnsupportedValue) Error() string {
return e.err.Error()
}
func (e *eventUnsupportedValue) Err() error {
return e.err
}
const tracerDisabled = "the tracer has been disabled"
// EventTracerDisabled occurs when a tracer is disabled by either the user or
// the collector.
type EventTracerDisabled interface {
Event
EventTracerDisabled()
}
type eventTracerDisabled struct{}
func newEventTracerDisabled() EventTracerDisabled {
return eventTracerDisabled{}
}
func (eventTracerDisabled) Event() {}
func (eventTracerDisabled) EventTracerDisabled() {}
func (eventTracerDisabled) String() string {
return tracerDisabled
}
go/vendor/github.com/opentracing/opentracing-go/CHANGELOG.md 0 → 100644
View file @ b34a88ce
Changes by Version
==================
1.1.0 (unreleased)
-------------------
- Deprecate InitGlobalTracer() in favor of SetGlobalTracer()
1.0.0 (2016-09-26)
-------------------
- This release implements OpenTracing Specification 1.0 (https://opentracing.io/spec)
go/vendor/github.com/opentracing/opentracing-go/LICENSE 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/opentracing/opentracing-go/Makefile 0 → 100644
View file @ b34a88ce
.DEFAULT_GOAL := test-and-lint
.PHONY: test-and-lint
test-and-lint: test lint
.PHONY: test
test:
go test -v -cover -race ./...
.PHONY: cover
cover:
go test -v -coverprofile=coverage.txt -covermode=atomic -race ./...
.PHONY: lint
lint:
go fmt ./...
golint ./...
@# Run again with magic to exit non-zero if golint outputs anything.
@! (golint ./... | read dummy)
go vet ./...
go/vendor/github.com/opentracing/opentracing-go/README.md 0 → 100644
View file @ b34a88ce
[![Gitter chat](http://img.shields.io/badge/gitter-join%20chat%20%E2%86%92-brightgreen.svg)](https://gitter.im/opentracing/public) [![Build Status](https://travis-ci.org/opentracing/opentracing-go.svg?branch=master)](https://travis-ci.org/opentracing/opentracing-go) [![GoDoc](https://godoc.org/github.com/opentracing/opentracing-go?status.svg)](http://godoc.org/github.com/opentracing/opentracing-go)
[![Sourcegraph Badge](https://sourcegraph.com/github.com/opentracing/opentracing-go/-/badge.svg)](https://sourcegraph.com/github.com/opentracing/opentracing-go?badge)
# OpenTracing API for Go
This package is a Go platform API for OpenTracing.
## Required Reading
In order to understand the Go platform API, one must first be familiar with the
[OpenTracing project](https://opentracing.io) and
[terminology](https://opentracing.io/specification/) more specifically.
## API overview for those adding instrumentation
Everyday consumers of this `opentracing` package really only need to worry
about a couple of key abstractions: the `StartSpan` function, the `Span`
interface, and binding a `Tracer` at `main()`-time. Here are code snippets
demonstrating some important use cases.
#### Singleton initialization
The simplest starting point is `./default_tracer.go`. As early as possible, call
```go
import "github.com/opentracing/opentracing-go"
import ".../some_tracing_impl"
func main() {
opentracing.SetGlobalTracer(
// tracing impl specific:
some_tracing_impl.New(...),
)
...
}
```
#### Non-Singleton initialization
If you prefer direct control to singletons, manage ownership of the
`opentracing.Tracer` implementation explicitly.
#### Creating a Span given an existing Go `context.Context`
If you use `context.Context` in your application, OpenTracing's Go library will
happily rely on it for `Span` propagation. To start a new (blocking child)
`Span`, you can use `StartSpanFromContext`.
```go
func xyz(ctx context.Context, ...) {
...
span, ctx := opentracing.StartSpanFromContext(ctx, "operation_name")
defer span.Finish()
span.LogFields(
log.String("event", "soft error"),
log.String("type", "cache timeout"),
log.Int("waited.millis", 1500))
...
}
```
#### Starting an empty trace by creating a "root span"
It's always possible to create a "root" `Span` with no parent or other causal
reference.
```go
func xyz() {
...
sp := opentracing.StartSpan("operation_name")
defer sp.Finish()
...
}
```
#### Creating a (child) Span given an existing (parent) Span
```go
func xyz(parentSpan opentracing.Span, ...) {
...
sp := opentracing.StartSpan(
"operation_name",
opentracing.ChildOf(parentSpan.Context()))
defer sp.Finish()
...
}
```
#### Serializing to the wire
```go
func makeSomeRequest(ctx context.Context) ... {
if span := opentracing.SpanFromContext(ctx); span != nil {
httpClient := &http.Client{}
httpReq, _ := http.NewRequest("GET", "http://myservice/", nil)
// Transmit the span's TraceContext as HTTP headers on our
// outbound request.
opentracing.GlobalTracer().Inject(
span.Context(),
opentracing.HTTPHeaders,
opentracing.HTTPHeadersCarrier(httpReq.Header))
resp, err := httpClient.Do(httpReq)
...
}
...
}
```
#### Deserializing from the wire
```go
http.HandleFunc("/", func(w http.ResponseWriter, req *http.Request) {
var serverSpan opentracing.Span
appSpecificOperationName := ...
wireContext, err := opentracing.GlobalTracer().Extract(
opentracing.HTTPHeaders,
opentracing.HTTPHeadersCarrier(req.Header))
if err != nil {
// Optionally record something about err here
}
// Create the span referring to the RPC client if available.
// If wireContext == nil, a root span will be created.
serverSpan = opentracing.StartSpan(
appSpecificOperationName,
ext.RPCServerOption(wireContext))
defer serverSpan.Finish()
ctx := opentracing.ContextWithSpan(context.Background(), serverSpan)
...
}
```
#### Conditionally capture a field using `log.Noop`
In some situations, you may want to dynamically decide whether or not
to log a field. For example, you may want to capture additional data,
such as a customer ID, in non-production environments:
```go
func Customer(order *Order) log.Field {
if os.Getenv("ENVIRONMENT") == "dev" {
return log.String("customer", order.Customer.ID)
}
return log.Noop()
}
```
#### Goroutine-safety
The entire public API is goroutine-safe and does not require external
synchronization.
## API pointers for those implementing a tracing system
Tracing system implementors may be able to reuse or copy-paste-modify the `basictracer` package, found [here](https://github.com/opentracing/basictracer-go). In particular, see `basictracer.New(...)`.
## API compatibility
For the time being, "mild" backwards-incompatible changes may be made without changing the major version number. As OpenTracing and `opentracing-go` mature, backwards compatibility will become more of a priority.
## Tracer test suite
A test suite is available in the [harness](https://godoc.org/github.com/opentracing/opentracing-go/harness) package that can assist Tracer implementors to assert that their Tracer is working correctly.
## Licensing
[Apache 2.0 License](./LICENSE).
go/vendor/github.com/opentracing/opentracing-go/ext/tags.go 0 → 100644
View file @ b34a88ce
package ext
import "github.com/opentracing/opentracing-go"
// These constants define common tag names recommended for better portability across
// tracing systems and languages/platforms.
//
// The tag names are defined as typed strings, so that in addition to the usual use
//
// span.setTag(TagName, value)
//
// they also support value type validation via this additional syntax:
//
// TagName.Set(span, value)
//
var (
//////////////////////////////////////////////////////////////////////
// SpanKind (client/server or producer/consumer)
//////////////////////////////////////////////////////////////////////
// SpanKind hints at relationship between spans, e.g. client/server
SpanKind = spanKindTagName("span.kind")
// SpanKindRPCClient marks a span representing the client-side of an RPC
// or other remote call
SpanKindRPCClientEnum = SpanKindEnum("client")
SpanKindRPCClient = opentracing.Tag{Key: string(SpanKind), Value: SpanKindRPCClientEnum}
// SpanKindRPCServer marks a span representing the server-side of an RPC
// or other remote call
SpanKindRPCServerEnum = SpanKindEnum("server")
SpanKindRPCServer = opentracing.Tag{Key: string(SpanKind), Value: SpanKindRPCServerEnum}
// SpanKindProducer marks a span representing the producer-side of a
// message bus
SpanKindProducerEnum = SpanKindEnum("producer")
SpanKindProducer = opentracing.Tag{Key: string(SpanKind), Value: SpanKindProducerEnum}
// SpanKindConsumer marks a span representing the consumer-side of a
// message bus
SpanKindConsumerEnum = SpanKindEnum("consumer")
SpanKindConsumer = opentracing.Tag{Key: string(SpanKind), Value: SpanKindConsumerEnum}
//////////////////////////////////////////////////////////////////////
// Component name
//////////////////////////////////////////////////////////////////////
// Component is a low-cardinality identifier of the module, library,
// or package that is generating a span.
Component = stringTagName("component")
//////////////////////////////////////////////////////////////////////
// Sampling hint
//////////////////////////////////////////////////////////////////////
// SamplingPriority determines the priority of sampling this Span.
SamplingPriority = uint16TagName("sampling.priority")
//////////////////////////////////////////////////////////////////////
// Peer tags. These tags can be emitted by either client-side of
// server-side to describe the other side/service in a peer-to-peer
// communications, like an RPC call.
//////////////////////////////////////////////////////////////////////
// PeerService records the service name of the peer.
PeerService = stringTagName("peer.service")
// PeerAddress records the address name of the peer. This may be a "ip:port",
// a bare "hostname", a FQDN or even a database DSN substring
// like "mysql://username@127.0.0.1:3306/dbname"
PeerAddress = stringTagName("peer.address")
// PeerHostname records the host name of the peer
PeerHostname = stringTagName("peer.hostname")
// PeerHostIPv4 records IP v4 host address of the peer
PeerHostIPv4 = ipv4Tag("peer.ipv4")
// PeerHostIPv6 records IP v6 host address of the peer
PeerHostIPv6 = stringTagName("peer.ipv6")
// PeerPort records port number of the peer
PeerPort = uint16TagName("peer.port")
//////////////////////////////////////////////////////////////////////
// HTTP Tags
//////////////////////////////////////////////////////////////////////
// HTTPUrl should be the URL of the request being handled in this segment
// of the trace, in standard URI format. The protocol is optional.
HTTPUrl = stringTagName("http.url")
// HTTPMethod is the HTTP method of the request, and is case-insensitive.
HTTPMethod = stringTagName("http.method")
// HTTPStatusCode is the numeric HTTP status code (200, 404, etc) of the
// HTTP response.
HTTPStatusCode = uint16TagName("http.status_code")
//////////////////////////////////////////////////////////////////////
// DB Tags
//////////////////////////////////////////////////////////////////////
// DBInstance is database instance name.
DBInstance = stringTagName("db.instance")
// DBStatement is a database statement for the given database type.
// It can be a query or a prepared statement (i.e., before substitution).
DBStatement = stringTagName("db.statement")
// DBType is a database type. For any SQL database, "sql".
// For others, the lower-case database category, e.g. "redis"
DBType = stringTagName("db.type")
// DBUser is a username for accessing database.
DBUser = stringTagName("db.user")
//////////////////////////////////////////////////////////////////////
// Message Bus Tag
//////////////////////////////////////////////////////////////////////
// MessageBusDestination is an address at which messages can be exchanged
MessageBusDestination = stringTagName("message_bus.destination")
//////////////////////////////////////////////////////////////////////
// Error Tag
//////////////////////////////////////////////////////////////////////
// Error indicates that operation represented by the span resulted in an error.
Error = boolTagName("error")
)
// ---
// SpanKindEnum represents common span types
type SpanKindEnum string
type spanKindTagName string
// Set adds a string tag to the `span`
func (tag spanKindTagName) Set(span opentracing.Span, value SpanKindEnum) {
span.SetTag(string(tag), value)
}
type rpcServerOption struct {
clientContext opentracing.SpanContext
}
func (r rpcServerOption) Apply(o *opentracing.StartSpanOptions) {
if r.clientContext != nil {
opentracing.ChildOf(r.clientContext).Apply(o)
}
SpanKindRPCServer.Apply(o)
}
// RPCServerOption returns a StartSpanOption appropriate for an RPC server span
// with `client` representing the metadata for the remote peer Span if available.
// In case client == nil, due to the client not being instrumented, this RPC
// server span will be a root span.
func RPCServerOption(client opentracing.SpanContext) opentracing.StartSpanOption {
return rpcServerOption{client}
}
// ---
type stringTagName string
// Set adds a string tag to the `span`
func (tag stringTagName) Set(span opentracing.Span, value string) {
span.SetTag(string(tag), value)
}
// ---
type uint32TagName string
// Set adds a uint32 tag to the `span`
func (tag uint32TagName) Set(span opentracing.Span, value uint32) {
span.SetTag(string(tag), value)
}
// ---
type uint16TagName string
// Set adds a uint16 tag to the `span`
func (tag uint16TagName) Set(span opentracing.Span, value uint16) {
span.SetTag(string(tag), value)
}
// ---
type boolTagName string
// Add adds a bool tag to the `span`
func (tag boolTagName) Set(span opentracing.Span, value bool) {
span.SetTag(string(tag), value)
}
type ipv4Tag string
// Set adds IP v4 host address of the peer as an uint32 value to the `span`, keep this for backward and zipkin compatibility
func (tag ipv4Tag) Set(span opentracing.Span, value uint32) {
span.SetTag(string(tag), value)
}
// SetString records IP v4 host address of the peer as a .-separated tuple to the `span`. E.g., "127.0.0.1"
func (tag ipv4Tag) SetString(span opentracing.Span, value string) {
span.SetTag(string(tag), value)
}
go/vendor/github.com/opentracing/opentracing-go/globaltracer.go 0 → 100644
View file @ b34a88ce
package opentracing
type registeredTracer struct {
tracer Tracer
isRegistered bool
}
var (
globalTracer = registeredTracer{NoopTracer{}, false}
)
// SetGlobalTracer sets the [singleton] opentracing.Tracer returned by
// GlobalTracer(). Those who use GlobalTracer (rather than directly manage an
// opentracing.Tracer instance) should call SetGlobalTracer as early as
// possible in main(), prior to calling the `StartSpan` global func below.
// Prior to calling `SetGlobalTracer`, any Spans started via the `StartSpan`
// (etc) globals are noops.
func SetGlobalTracer(tracer Tracer) {
globalTracer = registeredTracer{tracer, true}
}
// GlobalTracer returns the global singleton `Tracer` implementation.
// Before `SetGlobalTracer()` is called, the `GlobalTracer()` is a noop
// implementation that drops all data handed to it.
func GlobalTracer() Tracer {
return globalTracer.tracer
}
// StartSpan defers to `Tracer.StartSpan`. See `GlobalTracer()`.
func StartSpan(operationName string, opts ...StartSpanOption) Span {
return globalTracer.tracer.StartSpan(operationName, opts...)
}
// InitGlobalTracer is deprecated. Please use SetGlobalTracer.
func InitGlobalTracer(tracer Tracer) {
SetGlobalTracer(tracer)
}
// IsGlobalTracerRegistered returns a `bool` to indicate if a tracer has been globally registered
func IsGlobalTracerRegistered() bool {
return globalTracer.isRegistered
}
go/vendor/github.com/opentracing/opentracing-go/gocontext.go 0 → 100644
View file @ b34a88ce
package opentracing
import "context"
type contextKey struct{}
var activeSpanKey = contextKey{}
// ContextWithSpan returns a new `context.Context` that holds a reference to
// `span`'s SpanContext.
func ContextWithSpan(ctx context.Context, span Span) context.Context {
return context.WithValue(ctx, activeSpanKey, span)
}
// SpanFromContext returns the `Span` previously associated with `ctx`, or
// `nil` if no such `Span` could be found.
//
// NOTE: context.Context != SpanContext: the former is Go's intra-process
// context propagation mechanism, and the latter houses OpenTracing's per-Span
// identity and baggage information.
func SpanFromContext(ctx context.Context) Span {
val := ctx.Value(activeSpanKey)
if sp, ok := val.(Span); ok {
return sp
}
return nil
}
// StartSpanFromContext starts and returns a Span with `operationName`, using
// any Span found within `ctx` as a ChildOfRef. If no such parent could be
// found, StartSpanFromContext creates a root (parentless) Span.
//
// The second return value is a context.Context object built around the
// returned Span.
//
// Example usage:
//
// SomeFunction(ctx context.Context, ...) {
// sp, ctx := opentracing.StartSpanFromContext(ctx, "SomeFunction")
// defer sp.Finish()
// ...
// }
func StartSpanFromContext(ctx context.Context, operationName string, opts ...StartSpanOption) (Span, context.Context) {
return startSpanFromContextWithTracer(ctx, GlobalTracer(), operationName, opts...)
}
// startSpanFromContextWithTracer is factored out for testing purposes.
func startSpanFromContextWithTracer(ctx context.Context, tracer Tracer, operationName string, opts ...StartSpanOption) (Span, context.Context) {
if parentSpan := SpanFromContext(ctx); parentSpan != nil {
opts = append(opts, ChildOf(parentSpan.Context()))
}
span := tracer.StartSpan(operationName, opts...)
return span, ContextWithSpan(ctx, span)
}
go/vendor/github.com/opentracing/opentracing-go/log/field.go 0 → 100644
View file @ b34a88ce
package log
import (
"fmt"
"math"
)
type fieldType int
const (
stringType fieldType = iota
boolType
intType
int32Type
uint32Type
int64Type
uint64Type
float32Type
float64Type
errorType
objectType
lazyLoggerType
noopType
)
// Field instances are constructed via LogBool, LogString, and so on.
// Tracing implementations may then handle them via the Field.Marshal
// method.
//
// "heavily influenced by" (i.e., partially stolen from)
// https://github.com/uber-go/zap
type Field struct {
key string
fieldType fieldType
numericVal int64
stringVal string
interfaceVal interface{}
}
// String adds a string-valued key:value pair to a Span.LogFields() record
func String(key, val string) Field {
return Field{
key: key,
fieldType: stringType,
stringVal: val,
}
}
// Bool adds a bool-valued key:value pair to a Span.LogFields() record
func Bool(key string, val bool) Field {
var numericVal int64
if val {
numericVal = 1
}
return Field{
key: key,
fieldType: boolType,
numericVal: numericVal,
}
}
// Int adds an int-valued key:value pair to a Span.LogFields() record
func Int(key string, val int) Field {
return Field{
key: key,
fieldType: intType,
numericVal: int64(val),
}
}
// Int32 adds an int32-valued key:value pair to a Span.LogFields() record
func Int32(key string, val int32) Field {
return Field{
key: key,
fieldType: int32Type,
numericVal: int64(val),
}
}
// Int64 adds an int64-valued key:value pair to a Span.LogFields() record
func Int64(key string, val int64) Field {
return Field{
key: key,
fieldType: int64Type,
numericVal: val,
}
}
// Uint32 adds a uint32-valued key:value pair to a Span.LogFields() record
func Uint32(key string, val uint32) Field {
return Field{
key: key,
fieldType: uint32Type,
numericVal: int64(val),
}
}
// Uint64 adds a uint64-valued key:value pair to a Span.LogFields() record
func Uint64(key string, val uint64) Field {
return Field{
key: key,
fieldType: uint64Type,
numericVal: int64(val),
}
}
// Float32 adds a float32-valued key:value pair to a Span.LogFields() record
func Float32(key string, val float32) Field {
return Field{
key: key,
fieldType: float32Type,
numericVal: int64(math.Float32bits(val)),
}
}
// Float64 adds a float64-valued key:value pair to a Span.LogFields() record
func Float64(key string, val float64) Field {
return Field{
key: key,
fieldType: float64Type,
numericVal: int64(math.Float64bits(val)),
}
}
// Error adds an error with the key "error" to a Span.LogFields() record
func Error(err error) Field {
return Field{
key: "error",
fieldType: errorType,
interfaceVal: err,
}
}
// Object adds an object-valued key:value pair to a Span.LogFields() record
func Object(key string, obj interface{}) Field {
return Field{
key: key,
fieldType: objectType,
interfaceVal: obj,
}
}
// LazyLogger allows for user-defined, late-bound logging of arbitrary data
type LazyLogger func(fv Encoder)
// Lazy adds a LazyLogger to a Span.LogFields() record; the tracing
// implementation will call the LazyLogger function at an indefinite time in
// the future (after Lazy() returns).
func Lazy(ll LazyLogger) Field {
return Field{
fieldType: lazyLoggerType,
interfaceVal: ll,
}
}
// Noop creates a no-op log field that should be ignored by the tracer.
// It can be used to capture optional fields, for example those that should
// only be logged in non-production environment:
//
// func customerField(order *Order) log.Field {
// if os.Getenv("ENVIRONMENT") == "dev" {
// return log.String("customer", order.Customer.ID)
// }
// return log.Noop()
// }
//
// span.LogFields(log.String("event", "purchase"), customerField(order))
//
func Noop() Field {
return Field{
fieldType: noopType,
}
}
// Encoder allows access to the contents of a Field (via a call to
// Field.Marshal).
//
// Tracer implementations typically provide an implementation of Encoder;
// OpenTracing callers typically do not need to concern themselves with it.
type Encoder interface {
EmitString(key, value string)
EmitBool(key string, value bool)
EmitInt(key string, value int)
EmitInt32(key string, value int32)
EmitInt64(key string, value int64)
EmitUint32(key string, value uint32)
EmitUint64(key string, value uint64)
EmitFloat32(key string, value float32)
EmitFloat64(key string, value float64)
EmitObject(key string, value interface{})
EmitLazyLogger(value LazyLogger)
}
// Marshal passes a Field instance through to the appropriate
// field-type-specific method of an Encoder.
func (lf Field) Marshal(visitor Encoder) {
switch lf.fieldType {
case stringType:
visitor.EmitString(lf.key, lf.stringVal)
case boolType:
visitor.EmitBool(lf.key, lf.numericVal != 0)
case intType:
visitor.EmitInt(lf.key, int(lf.numericVal))
case int32Type:
visitor.EmitInt32(lf.key, int32(lf.numericVal))
case int64Type:
visitor.EmitInt64(lf.key, int64(lf.numericVal))
case uint32Type:
visitor.EmitUint32(lf.key, uint32(lf.numericVal))
case uint64Type:
visitor.EmitUint64(lf.key, uint64(lf.numericVal))
case float32Type:
visitor.EmitFloat32(lf.key, math.Float32frombits(uint32(lf.numericVal)))
case float64Type:
visitor.EmitFloat64(lf.key, math.Float64frombits(uint64(lf.numericVal)))
case errorType:
if err, ok := lf.interfaceVal.(error); ok {
visitor.EmitString(lf.key, err.Error())
} else {
visitor.EmitString(lf.key, "<nil>")
}
case objectType:
visitor.EmitObject(lf.key, lf.interfaceVal)
case lazyLoggerType:
visitor.EmitLazyLogger(lf.interfaceVal.(LazyLogger))
case noopType:
// intentionally left blank
}
}
// Key returns the field's key.
func (lf Field) Key() string {
return lf.key
}
// Value returns the field's value as interface{}.
func (lf Field) Value() interface{} {
switch lf.fieldType {
case stringType:
return lf.stringVal
case boolType:
return lf.numericVal != 0
case intType:
return int(lf.numericVal)
case int32Type:
return int32(lf.numericVal)
case int64Type:
return int64(lf.numericVal)
case uint32Type:
return uint32(lf.numericVal)
case uint64Type:
return uint64(lf.numericVal)
case float32Type:
return math.Float32frombits(uint32(lf.numericVal))
case float64Type:
return math.Float64frombits(uint64(lf.numericVal))
case errorType, objectType, lazyLoggerType:
return lf.interfaceVal
case noopType:
return nil
default:
return nil
}
}
// String returns a string representation of the key and value.
func (lf Field) String() string {
return fmt.Sprint(lf.key, ":", lf.Value())
}
go/vendor/github.com/opentracing/opentracing-go/log/util.go 0 → 100644
View file @ b34a88ce
package log
import "fmt"
// InterleavedKVToFields converts keyValues a la Span.LogKV() to a Field slice
// a la Span.LogFields().
func InterleavedKVToFields(keyValues ...interface{}) ([]Field, error) {
if len(keyValues)%2 != 0 {
return nil, fmt.Errorf("non-even keyValues len: %d", len(keyValues))
}
fields := make([]Field, len(keyValues)/2)
for i := 0; i*2 < len(keyValues); i++ {
key, ok := keyValues[i*2].(string)
if !ok {
return nil, fmt.Errorf(
"non-string key (pair #%d): %T",
i, keyValues[i*2])
}
switch typedVal := keyValues[i*2+1].(type) {
case bool:
fields[i] = Bool(key, typedVal)
case string:
fields[i] = String(key, typedVal)
case int:
fields[i] = Int(key, typedVal)
case int8:
fields[i] = Int32(key, int32(typedVal))
case int16:
fields[i] = Int32(key, int32(typedVal))
case int32:
fields[i] = Int32(key, typedVal)
case int64:
fields[i] = Int64(key, typedVal)
case uint:
fields[i] = Uint64(key, uint64(typedVal))
case uint64:
fields[i] = Uint64(key, typedVal)
case uint8:
fields[i] = Uint32(key, uint32(typedVal))
case uint16:
fields[i] = Uint32(key, uint32(typedVal))
case uint32:
fields[i] = Uint32(key, typedVal)
case float32:
fields[i] = Float32(key, typedVal)
case float64:
fields[i] = Float64(key, typedVal)
default:
// When in doubt, coerce to a string
fields[i] = String(key, fmt.Sprint(typedVal))
}
}
return fields, nil
}
go/vendor/github.com/opentracing/opentracing-go/noop.go 0 → 100644
View file @ b34a88ce
package opentracing
import "github.com/opentracing/opentracing-go/log"
// A NoopTracer is a trivial, minimum overhead implementation of Tracer
// for which all operations are no-ops.
//
// The primary use of this implementation is in libraries, such as RPC
// frameworks, that make tracing an optional feature controlled by the
// end user. A no-op implementation allows said libraries to use it
// as the default Tracer and to write instrumentation that does
// not need to keep checking if the tracer instance is nil.
//
// For the same reason, the NoopTracer is the default "global" tracer
// (see GlobalTracer and SetGlobalTracer functions).
//
// WARNING: NoopTracer does not support baggage propagation.
type NoopTracer struct{}
type noopSpan struct{}
type noopSpanContext struct{}
var (
defaultNoopSpanContext = noopSpanContext{}
defaultNoopSpan = noopSpan{}
defaultNoopTracer = NoopTracer{}
)
const (
emptyString = ""
)
// noopSpanContext:
func (n noopSpanContext) ForeachBaggageItem(handler func(k, v string) bool) {}
// noopSpan:
func (n noopSpan) Context() SpanContext { return defaultNoopSpanContext }
func (n noopSpan) SetBaggageItem(key, val string) Span { return defaultNoopSpan }
func (n noopSpan) BaggageItem(key string) string { return emptyString }
func (n noopSpan) SetTag(key string, value interface{}) Span { return n }
func (n noopSpan) LogFields(fields ...log.Field) {}
func (n noopSpan) LogKV(keyVals ...interface{}) {}
func (n noopSpan) Finish() {}
func (n noopSpan) FinishWithOptions(opts FinishOptions) {}
func (n noopSpan) SetOperationName(operationName string) Span { return n }
func (n noopSpan) Tracer() Tracer { return defaultNoopTracer }
func (n noopSpan) LogEvent(event string) {}
func (n noopSpan) LogEventWithPayload(event string, payload interface{}) {}
func (n noopSpan) Log(data LogData) {}
// StartSpan belongs to the Tracer interface.
func (n NoopTracer) StartSpan(operationName string, opts ...StartSpanOption) Span {
return defaultNoopSpan
}
// Inject belongs to the Tracer interface.
func (n NoopTracer) Inject(sp SpanContext, format interface{}, carrier interface{}) error {
return nil
}
// Extract belongs to the Tracer interface.
func (n NoopTracer) Extract(format interface{}, carrier interface{}) (SpanContext, error) {
return nil, ErrSpanContextNotFound
}
go/vendor/github.com/opentracing/opentracing-go/propagation.go 0 → 100644
View file @ b34a88ce
package opentracing
import (
"errors"
"net/http"
)
///////////////////////////////////////////////////////////////////////////////
// CORE PROPAGATION INTERFACES:
///////////////////////////////////////////////////////////////////////////////
var (
// ErrUnsupportedFormat occurs when the `format` passed to Tracer.Inject() or
// Tracer.Extract() is not recognized by the Tracer implementation.
ErrUnsupportedFormat = errors.New("opentracing: Unknown or unsupported Inject/Extract format")
// ErrSpanContextNotFound occurs when the `carrier` passed to
// Tracer.Extract() is valid and uncorrupted but has insufficient
// information to extract a SpanContext.
ErrSpanContextNotFound = errors.New("opentracing: SpanContext not found in Extract carrier")
// ErrInvalidSpanContext errors occur when Tracer.Inject() is asked to
// operate on a SpanContext which it is not prepared to handle (for
// example, since it was created by a different tracer implementation).
ErrInvalidSpanContext = errors.New("opentracing: SpanContext type incompatible with tracer")
// ErrInvalidCarrier errors occur when Tracer.Inject() or Tracer.Extract()
// implementations expect a different type of `carrier` than they are
// given.
ErrInvalidCarrier = errors.New("opentracing: Invalid Inject/Extract carrier")
// ErrSpanContextCorrupted occurs when the `carrier` passed to
// Tracer.Extract() is of the expected type but is corrupted.
ErrSpanContextCorrupted = errors.New("opentracing: SpanContext data corrupted in Extract carrier")
)
///////////////////////////////////////////////////////////////////////////////
// BUILTIN PROPAGATION FORMATS:
///////////////////////////////////////////////////////////////////////////////
// BuiltinFormat is used to demarcate the values within package `opentracing`
// that are intended for use with the Tracer.Inject() and Tracer.Extract()
// methods.
type BuiltinFormat byte
const (
// Binary represents SpanContexts as opaque binary data.
//
// For Tracer.Inject(): the carrier must be an `io.Writer`.
//
// For Tracer.Extract(): the carrier must be an `io.Reader`.
Binary BuiltinFormat = iota
// TextMap represents SpanContexts as key:value string pairs.
//
// Unlike HTTPHeaders, the TextMap format does not restrict the key or
// value character sets in any way.
//
// For Tracer.Inject(): the carrier must be a `TextMapWriter`.
//
// For Tracer.Extract(): the carrier must be a `TextMapReader`.
TextMap
// HTTPHeaders represents SpanContexts as HTTP header string pairs.
//
// Unlike TextMap, the HTTPHeaders format requires that the keys and values
// be valid as HTTP headers as-is (i.e., character casing may be unstable
// and special characters are disallowed in keys, values should be
// URL-escaped, etc).
//
// For Tracer.Inject(): the carrier must be a `TextMapWriter`.
//
// For Tracer.Extract(): the carrier must be a `TextMapReader`.
//
// See HTTPHeadersCarrier for an implementation of both TextMapWriter
// and TextMapReader that defers to an http.Header instance for storage.
// For example, Inject():
//
// carrier := opentracing.HTTPHeadersCarrier(httpReq.Header)
// err := span.Tracer().Inject(
// span.Context(), opentracing.HTTPHeaders, carrier)
//
// Or Extract():
//
// carrier := opentracing.HTTPHeadersCarrier(httpReq.Header)
// clientContext, err := tracer.Extract(
// opentracing.HTTPHeaders, carrier)
//
HTTPHeaders
)
// TextMapWriter is the Inject() carrier for the TextMap builtin format. With
// it, the caller can encode a SpanContext for propagation as entries in a map
// of unicode strings.
type TextMapWriter interface {
// Set a key:value pair to the carrier. Multiple calls to Set() for the
// same key leads to undefined behavior.
//
// NOTE: The backing store for the TextMapWriter may contain data unrelated
// to SpanContext. As such, Inject() and Extract() implementations that
// call the TextMapWriter and TextMapReader interfaces must agree on a
// prefix or other convention to distinguish their own key:value pairs.
Set(key, val string)
}
// TextMapReader is the Extract() carrier for the TextMap builtin format. With it,
// the caller can decode a propagated SpanContext as entries in a map of
// unicode strings.
type TextMapReader interface {
// ForeachKey returns TextMap contents via repeated calls to the `handler`
// function. If any call to `handler` returns a non-nil error, ForeachKey
// terminates and returns that error.
//
// NOTE: The backing store for the TextMapReader may contain data unrelated
// to SpanContext. As such, Inject() and Extract() implementations that
// call the TextMapWriter and TextMapReader interfaces must agree on a
// prefix or other convention to distinguish their own key:value pairs.
//
// The "foreach" callback pattern reduces unnecessary copying in some cases
// and also allows implementations to hold locks while the map is read.
ForeachKey(handler func(key, val string) error) error
}
// TextMapCarrier allows the use of regular map[string]string
// as both TextMapWriter and TextMapReader.
type TextMapCarrier map[string]string
// ForeachKey conforms to the TextMapReader interface.
func (c TextMapCarrier) ForeachKey(handler func(key, val string) error) error {
for k, v := range c {
if err := handler(k, v); err != nil {
return err
}
}
return nil
}
// Set implements Set() of opentracing.TextMapWriter
func (c TextMapCarrier) Set(key, val string) {
c[key] = val
}
// HTTPHeadersCarrier satisfies both TextMapWriter and TextMapReader.
//
// Example usage for server side:
//
// carrier := opentracing.HTTPHeadersCarrier(httpReq.Header)
// clientContext, err := tracer.Extract(opentracing.HTTPHeaders, carrier)
//
// Example usage for client side:
//
// carrier := opentracing.HTTPHeadersCarrier(httpReq.Header)
// err := tracer.Inject(
// span.Context(),
// opentracing.HTTPHeaders,
// carrier)
//
type HTTPHeadersCarrier http.Header
// Set conforms to the TextMapWriter interface.
func (c HTTPHeadersCarrier) Set(key, val string) {
h := http.Header(c)
h.Set(key, val)
}
// ForeachKey conforms to the TextMapReader interface.
func (c HTTPHeadersCarrier) ForeachKey(handler func(key, val string) error) error {
for k, vals := range c {
for _, v := range vals {
if err := handler(k, v); err != nil {
return err
}
}
}
return nil
}
go/vendor/github.com/opentracing/opentracing-go/span.go 0 → 100644
View file @ b34a88ce
package opentracing
import (
"time"
"github.com/opentracing/opentracing-go/log"
)
// SpanContext represents Span state that must propagate to descendant Spans and across process
// boundaries (e.g., a <trace_id, span_id, sampled> tuple).
type SpanContext interface {
// ForeachBaggageItem grants access to all baggage items stored in the
// SpanContext.
// The handler function will be called for each baggage key/value pair.
// The ordering of items is not guaranteed.
//
// The bool return value indicates if the handler wants to continue iterating
// through the rest of the baggage items; for example if the handler is trying to
// find some baggage item by pattern matching the name, it can return false
// as soon as the item is found to stop further iterations.
ForeachBaggageItem(handler func(k, v string) bool)
}
// Span represents an active, un-finished span in the OpenTracing system.
//
// Spans are created by the Tracer interface.
type Span interface {
// Sets the end timestamp and finalizes Span state.
//
// With the exception of calls to Context() (which are always allowed),
// Finish() must be the last call made to any span instance, and to do
// otherwise leads to undefined behavior.
Finish()
// FinishWithOptions is like Finish() but with explicit control over
// timestamps and log data.
FinishWithOptions(opts FinishOptions)
// Context() yields the SpanContext for this Span. Note that the return
// value of Context() is still valid after a call to Span.Finish(), as is
// a call to Span.Context() after a call to Span.Finish().
Context() SpanContext
// Sets or changes the operation name.
//
// Returns a reference to this Span for chaining.
SetOperationName(operationName string) Span
// Adds a tag to the span.
//
// If there is a pre-existing tag set for `key`, it is overwritten.
//
// Tag values can be numeric types, strings, or bools. The behavior of
// other tag value types is undefined at the OpenTracing level. If a
// tracing system does not know how to handle a particular value type, it
// may ignore the tag, but shall not panic.
//
// Returns a reference to this Span for chaining.
SetTag(key string, value interface{}) Span
// LogFields is an efficient and type-checked way to record key:value
// logging data about a Span, though the programming interface is a little
// more verbose than LogKV(). Here's an example:
//
// span.LogFields(
// log.String("event", "soft error"),
// log.String("type", "cache timeout"),
// log.Int("waited.millis", 1500))
//
// Also see Span.FinishWithOptions() and FinishOptions.BulkLogData.
LogFields(fields ...log.Field)
// LogKV is a concise, readable way to record key:value logging data about
// a Span, though unfortunately this also makes it less efficient and less
// type-safe than LogFields(). Here's an example:
//
// span.LogKV(
// "event", "soft error",
// "type", "cache timeout",
// "waited.millis", 1500)
//
// For LogKV (as opposed to LogFields()), the parameters must appear as
// key-value pairs, like
//
// span.LogKV(key1, val1, key2, val2, key3, val3, ...)
//
// The keys must all be strings. The values may be strings, numeric types,
// bools, Go error instances, or arbitrary structs.
//
// (Note to implementors: consider the log.InterleavedKVToFields() helper)
LogKV(alternatingKeyValues ...interface{})
// SetBaggageItem sets a key:value pair on this Span and its SpanContext
// that also propagates to descendants of this Span.
//
// SetBaggageItem() enables powerful functionality given a full-stack
// opentracing integration (e.g., arbitrary application data from a mobile
// app can make it, transparently, all the way into the depths of a storage
// system), and with it some powerful costs: use this feature with care.
//
// IMPORTANT NOTE #1: SetBaggageItem() will only propagate baggage items to
// *future* causal descendants of the associated Span.
//
// IMPORTANT NOTE #2: Use this thoughtfully and with care. Every key and
// value is copied into every local *and remote* child of the associated
// Span, and that can add up to a lot of network and cpu overhead.
//
// Returns a reference to this Span for chaining.
SetBaggageItem(restrictedKey, value string) Span
// Gets the value for a baggage item given its key. Returns the empty string
// if the value isn't found in this Span.
BaggageItem(restrictedKey string) string
// Provides access to the Tracer that created this Span.
Tracer() Tracer
// Deprecated: use LogFields or LogKV
LogEvent(event string)
// Deprecated: use LogFields or LogKV
LogEventWithPayload(event string, payload interface{})
// Deprecated: use LogFields or LogKV
Log(data LogData)
}
// LogRecord is data associated with a single Span log. Every LogRecord
// instance must specify at least one Field.
type LogRecord struct {
Timestamp time.Time
Fields []log.Field
}
// FinishOptions allows Span.FinishWithOptions callers to override the finish
// timestamp and provide log data via a bulk interface.
type FinishOptions struct {
// FinishTime overrides the Span's finish time, or implicitly becomes
// time.Now() if FinishTime.IsZero().
//
// FinishTime must resolve to a timestamp that's >= the Span's StartTime
// (per StartSpanOptions).
FinishTime time.Time
// LogRecords allows the caller to specify the contents of many LogFields()
// calls with a single slice. May be nil.
//
// None of the LogRecord.Timestamp values may be .IsZero() (i.e., they must
// be set explicitly). Also, they must be >= the Span's start timestamp and
// <= the FinishTime (or time.Now() if FinishTime.IsZero()). Otherwise the
// behavior of FinishWithOptions() is undefined.
//
// If specified, the caller hands off ownership of LogRecords at
// FinishWithOptions() invocation time.
//
// If specified, the (deprecated) BulkLogData must be nil or empty.
LogRecords []LogRecord
// BulkLogData is DEPRECATED.
BulkLogData []LogData
}
// LogData is DEPRECATED
type LogData struct {
Timestamp time.Time
Event string
Payload interface{}
}
// ToLogRecord converts a deprecated LogData to a non-deprecated LogRecord
func (ld *LogData) ToLogRecord() LogRecord {
var literalTimestamp time.Time
if ld.Timestamp.IsZero() {
literalTimestamp = time.Now()
} else {
literalTimestamp = ld.Timestamp
}
rval := LogRecord{
Timestamp: literalTimestamp,
}
if ld.Payload == nil {
rval.Fields = []log.Field{
log.String("event", ld.Event),
}
} else {
rval.Fields = []log.Field{
log.String("event", ld.Event),
log.Object("payload", ld.Payload),
}
}
return rval
}
go/vendor/github.com/opentracing/opentracing-go/tracer.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/philhofer/fwd/LICENSE.md 0 → 100644
View file @ b34a88ce
Copyright (c) 2014-2015, Philip Hofer
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.
\ No newline at end of file
go/vendor/github.com/philhofer/fwd/README.md 0 → 100644
View file @ b34a88ce
# fwd
import "github.com/philhofer/fwd"
The `fwd` package provides a buffered reader
and writer. Each has methods that help improve
the encoding/decoding performance of some binary
protocols.
The `fwd.Writer` and `fwd.Reader` type provide similar
functionality to their counterparts in `bufio`, plus
a few extra utility methods that simplify read-ahead
and write-ahead. I wrote this package to improve serialization
performance for <a href="http://github.com/tinylib/msgp">http://github.com/tinylib/msgp</a>,
where it provided about a 2x speedup over `bufio` for certain
workloads. However, care must be taken to understand the semantics of the
extra methods provided by this package, as they allow
the user to access and manipulate the buffer memory
directly.
The extra methods for `fwd.Reader` are `Peek`, `Skip`
and `Next`. `(*fwd.Reader).Peek`, unlike `(*bufio.Reader).Peek`,
will re-allocate the read buffer in order to accommodate arbitrarily
large read-ahead. `(*fwd.Reader).Skip` skips the next `n` bytes
in the stream, and uses the `io.Seeker` interface if the underlying
stream implements it. `(*fwd.Reader).Next` returns a slice pointing
to the next `n` bytes in the read buffer (like `Peek`), but also
increments the read position. This allows users to process streams
in arbitrary block sizes without having to manage appropriately-sized
slices. Additionally, obviating the need to copy the data from the
buffer to another location in memory can improve performance dramatically
in CPU-bound applications.
`fwd.Writer` only has one extra method, which is `(*fwd.Writer).Next`, which
returns a slice pointing to the next `n` bytes of the writer, and increments
the write position by the length of the returned slice. This allows users
to write directly to the end of the buffer.
## Constants
``` go
const (
// DefaultReaderSize is the default size of the read buffer
DefaultReaderSize = 2048
)
```
``` go
const (
// DefaultWriterSize is the
// default write buffer size.
DefaultWriterSize = 2048
)
```
## type Reader
``` go
type Reader struct {
// contains filtered or unexported fields
}
```
Reader is a buffered look-ahead reader
### func NewReader
``` go
func NewReader(r io.Reader) *Reader
```
NewReader returns a new *Reader that reads from 'r'
### func NewReaderSize
``` go
func NewReaderSize(r io.Reader, n int) *Reader
```
NewReaderSize returns a new *Reader that
reads from 'r' and has a buffer size 'n'
### func (\*Reader) BufferSize
``` go
func (r *Reader) BufferSize() int
```
BufferSize returns the total size of the buffer
### func (\*Reader) Buffered
``` go
func (r *Reader) Buffered() int
```
Buffered returns the number of bytes currently in the buffer
### func (\*Reader) Next
``` go
func (r *Reader) Next(n int) ([]byte, error)
```
Next returns the next 'n' bytes in the stream.
Unlike Peek, Next advances the reader position.
The returned bytes point to the same
data as the buffer, so the slice is
only valid until the next reader method call.
An EOF is considered an unexpected error.
If an the returned slice is less than the
length asked for, an error will be returned,
and the reader position will not be incremented.
### func (\*Reader) Peek
``` go
func (r *Reader) Peek(n int) ([]byte, error)
```
Peek returns the next 'n' buffered bytes,
reading from the underlying reader if necessary.
It will only return a slice shorter than 'n' bytes
if it also returns an error. Peek does not advance
the reader. EOF errors are *not* returned as
io.ErrUnexpectedEOF.
### func (\*Reader) Read
``` go
func (r *Reader) Read(b []byte) (int, error)
```
Read implements `io.Reader`
### func (\*Reader) ReadByte
``` go
func (r *Reader) ReadByte() (byte, error)
```
ReadByte implements `io.ByteReader`
### func (\*Reader) ReadFull
``` go
func (r *Reader) ReadFull(b []byte) (int, error)
```
ReadFull attempts to read len(b) bytes into
'b'. It returns the number of bytes read into
'b', and an error if it does not return len(b).
EOF is considered an unexpected error.
### func (\*Reader) Reset
``` go
func (r *Reader) Reset(rd io.Reader)
```
Reset resets the underlying reader
and the read buffer.
### func (\*Reader) Skip
``` go
func (r *Reader) Skip(n int) (int, error)
```
Skip moves the reader forward 'n' bytes.
Returns the number of bytes skipped and any
errors encountered. It is analogous to Seek(n, 1).
If the underlying reader implements io.Seeker, then
that method will be used to skip forward.
If the reader encounters
an EOF before skipping 'n' bytes, it
returns io.ErrUnexpectedEOF. If the
underlying reader implements io.Seeker, then
those rules apply instead. (Many implementations
will not return `io.EOF` until the next call
to Read.)
### func (\*Reader) WriteTo
``` go
func (r *Reader) WriteTo(w io.Writer) (int64, error)
```
WriteTo implements `io.WriterTo`
## type Writer
``` go
type Writer struct {
// contains filtered or unexported fields
}
```
Writer is a buffered writer
### func NewWriter
``` go
func NewWriter(w io.Writer) *Writer
```
NewWriter returns a new writer
that writes to 'w' and has a buffer
that is `DefaultWriterSize` bytes.
### func NewWriterSize
``` go
func NewWriterSize(w io.Writer, size int) *Writer
```
NewWriterSize returns a new writer
that writes to 'w' and has a buffer
that is 'size' bytes.
### func (\*Writer) BufferSize
``` go
func (w *Writer) BufferSize() int
```
BufferSize returns the maximum size of the buffer.
### func (\*Writer) Buffered
``` go
func (w *Writer) Buffered() int
```
Buffered returns the number of buffered bytes
in the reader.
### func (\*Writer) Flush
``` go
func (w *Writer) Flush() error
```
Flush flushes any buffered bytes
to the underlying writer.
### func (\*Writer) Next
``` go
func (w *Writer) Next(n int) ([]byte, error)
```
Next returns the next 'n' free bytes
in the write buffer, flushing the writer
as necessary. Next will return `io.ErrShortBuffer`
if 'n' is greater than the size of the write buffer.
Calls to 'next' increment the write position by
the size of the returned buffer.
### func (\*Writer) ReadFrom
``` go
func (w *Writer) ReadFrom(r io.Reader) (int64, error)
```
ReadFrom implements `io.ReaderFrom`
### func (\*Writer) Write
``` go
func (w *Writer) Write(p []byte) (int, error)
```
Write implements `io.Writer`
### func (\*Writer) WriteByte
``` go
func (w *Writer) WriteByte(b byte) error
```
WriteByte implements `io.ByteWriter`
### func (\*Writer) WriteString
``` go
func (w *Writer) WriteString(s string) (int, error)
```
WriteString is analogous to Write, but it takes a string.
- - -
Generated by [godoc2md](http://godoc.org/github.com/davecheney/godoc2md)
\ No newline at end of file
go/vendor/github.com/philhofer/fwd/reader.go 0 → 100644
View file @ b34a88ce
// The `fwd` package provides a buffered reader
// and writer. Each has methods that help improve
// the encoding/decoding performance of some binary
// protocols.
//
// The `fwd.Writer` and `fwd.Reader` type provide similar
// functionality to their counterparts in `bufio`, plus
// a few extra utility methods that simplify read-ahead
// and write-ahead. I wrote this package to improve serialization
// performance for http://github.com/tinylib/msgp,
// where it provided about a 2x speedup over `bufio` for certain
// workloads. However, care must be taken to understand the semantics of the
// extra methods provided by this package, as they allow
// the user to access and manipulate the buffer memory
// directly.
//
// The extra methods for `fwd.Reader` are `Peek`, `Skip`
// and `Next`. `(*fwd.Reader).Peek`, unlike `(*bufio.Reader).Peek`,
// will re-allocate the read buffer in order to accommodate arbitrarily
// large read-ahead. `(*fwd.Reader).Skip` skips the next `n` bytes
// in the stream, and uses the `io.Seeker` interface if the underlying
// stream implements it. `(*fwd.Reader).Next` returns a slice pointing
// to the next `n` bytes in the read buffer (like `Peek`), but also
// increments the read position. This allows users to process streams
// in arbitrary block sizes without having to manage appropriately-sized
// slices. Additionally, obviating the need to copy the data from the
// buffer to another location in memory can improve performance dramatically
// in CPU-bound applications.
//
// `fwd.Writer` only has one extra method, which is `(*fwd.Writer).Next`, which
// returns a slice pointing to the next `n` bytes of the writer, and increments
// the write position by the length of the returned slice. This allows users
// to write directly to the end of the buffer.
//
package fwd
import "io"
const (
// DefaultReaderSize is the default size of the read buffer
DefaultReaderSize = 2048
// minimum read buffer; straight from bufio
minReaderSize = 16
)
// NewReader returns a new *Reader that reads from 'r'
func NewReader(r io.Reader) *Reader {
return NewReaderSize(r, DefaultReaderSize)
}
// NewReaderSize returns a new *Reader that
// reads from 'r' and has a buffer size 'n'
func NewReaderSize(r io.Reader, n int) *Reader {
rd := &Reader{
r: r,
data: make([]byte, 0, max(minReaderSize, n)),
}
if s, ok := r.(io.Seeker); ok {
rd.rs = s
}
return rd
}
// Reader is a buffered look-ahead reader
type Reader struct {
r io.Reader // underlying reader
// data[n:len(data)] is buffered data; data[len(data):cap(data)] is free buffer space
data []byte // data
n int // read offset
state error // last read error
// if the reader past to NewReader was
// also an io.Seeker, this is non-nil
rs io.Seeker
}
// Reset resets the underlying reader
// and the read buffer.
func (r *Reader) Reset(rd io.Reader) {
r.r = rd
r.data = r.data[0:0]
r.n = 0
r.state = nil
if s, ok := rd.(io.Seeker); ok {
r.rs = s
} else {
r.rs = nil
}
}
// more() does one read on the underlying reader
func (r *Reader) more() {
// move data backwards so that
// the read offset is 0; this way
// we can supply the maximum number of
// bytes to the reader
if r.n != 0 {
if r.n < len(r.data) {
r.data = r.data[:copy(r.data[0:], r.data[r.n:])]
} else {
r.data = r.data[:0]
}
r.n = 0
}
var a int
a, r.state = r.r.Read(r.data[len(r.data):cap(r.data)])
if a == 0 && r.state == nil {
r.state = io.ErrNoProgress
return
} else if a > 0 && r.state == io.EOF {
// discard the io.EOF if we read more than 0 bytes.
// the next call to Read should return io.EOF again.
r.state = nil
}
r.data = r.data[:len(r.data)+a]
}
// pop error
func (r *Reader) err() (e error) {
e, r.state = r.state, nil
return
}
// pop error; EOF -> io.ErrUnexpectedEOF
func (r *Reader) noEOF() (e error) {
e, r.state = r.state, nil
if e == io.EOF {
e = io.ErrUnexpectedEOF
}
return
}
// buffered bytes
func (r *Reader) buffered() int { return len(r.data) - r.n }
// Buffered returns the number of bytes currently in the buffer
func (r *Reader) Buffered() int { return len(r.data) - r.n }
// BufferSize returns the total size of the buffer
func (r *Reader) BufferSize() int { return cap(r.data) }
// Peek returns the next 'n' buffered bytes,
// reading from the underlying reader if necessary.
// It will only return a slice shorter than 'n' bytes
// if it also returns an error. Peek does not advance
// the reader. EOF errors are *not* returned as
// io.ErrUnexpectedEOF.
func (r *Reader) Peek(n int) ([]byte, error) {
// in the degenerate case,
// we may need to realloc
// (the caller asked for more
// bytes than the size of the buffer)
if cap(r.data) < n {
old := r.data[r.n:]
r.data = make([]byte, n+r.buffered())
r.data = r.data[:copy(r.data, old)]
r.n = 0
}
// keep filling until
// we hit an error or
// read enough bytes
for r.buffered() < n && r.state == nil {
r.more()
}
// we must have hit an error
if r.buffered() < n {
return r.data[r.n:], r.err()
}
return r.data[r.n : r.n+n], nil
}
// Skip moves the reader forward 'n' bytes.
// Returns the number of bytes skipped and any
// errors encountered. It is analogous to Seek(n, 1).
// If the underlying reader implements io.Seeker, then
// that method will be used to skip forward.
//
// If the reader encounters
// an EOF before skipping 'n' bytes, it
// returns io.ErrUnexpectedEOF. If the
// underlying reader implements io.Seeker, then
// those rules apply instead. (Many implementations
// will not return `io.EOF` until the next call
// to Read.)
func (r *Reader) Skip(n int) (int, error) {
// fast path
if r.buffered() >= n {
r.n += n
return n, nil
}
// use seeker implementation
// if we can
if r.rs != nil {
return r.skipSeek(n)
}
// loop on filling
// and then erasing
o := n
for r.buffered() < n && r.state == nil {
r.more()
// we can skip forward
// up to r.buffered() bytes
step := min(r.buffered(), n)
r.n += step
n -= step
}
// at this point, n should be
// 0 if everything went smoothly
return o - n, r.noEOF()
}
// Next returns the next 'n' bytes in the stream.
// Unlike Peek, Next advances the reader position.
// The returned bytes point to the same
// data as the buffer, so the slice is
// only valid until the next reader method call.
// An EOF is considered an unexpected error.
// If an the returned slice is less than the
// length asked for, an error will be returned,
// and the reader position will not be incremented.
func (r *Reader) Next(n int) ([]byte, error) {
// in case the buffer is too small
if cap(r.data) < n {
old := r.data[r.n:]
r.data = make([]byte, n+r.buffered())
r.data = r.data[:copy(r.data, old)]
r.n = 0
}
// fill at least 'n' bytes
for r.buffered() < n && r.state == nil {
r.more()
}
if r.buffered() < n {
return r.data[r.n:], r.noEOF()
}
out := r.data[r.n : r.n+n]
r.n += n
return out, nil
}
// skipSeek uses the io.Seeker to seek forward.
// only call this function when n > r.buffered()
func (r *Reader) skipSeek(n int) (int, error) {
o := r.buffered()
// first, clear buffer
n -= o
r.n = 0
r.data = r.data[:0]
// then seek forward remaning bytes
i, err := r.rs.Seek(int64(n), 1)
return int(i) + o, err
}
// Read implements `io.Reader`
func (r *Reader) Read(b []byte) (int, error) {
// if we have data in the buffer, just
// return that.
if r.buffered() != 0 {
x := copy(b, r.data[r.n:])
r.n += x
return x, nil
}
var n int
// we have no buffered data; determine
// whether or not to buffer or call
// the underlying reader directly
if len(b) >= cap(r.data) {
n, r.state = r.r.Read(b)
} else {
r.more()
n = copy(b, r.data)
r.n = n
}
if n == 0 {
return 0, r.err()
}
return n, nil
}
// ReadFull attempts to read len(b) bytes into
// 'b'. It returns the number of bytes read into
// 'b', and an error if it does not return len(b).
// EOF is considered an unexpected error.
func (r *Reader) ReadFull(b []byte) (int, error) {
var n int // read into b
var nn int // scratch
l := len(b)
// either read buffered data,
// or read directly for the underlying
// buffer, or fetch more buffered data.
for n < l && r.state == nil {
if r.buffered() != 0 {
nn = copy(b[n:], r.data[r.n:])
n += nn
r.n += nn
} else if l-n > cap(r.data) {
nn, r.state = r.r.Read(b[n:])
n += nn
} else {
r.more()
}
}
if n < l {
return n, r.noEOF()
}
return n, nil
}
// ReadByte implements `io.ByteReader`
func (r *Reader) ReadByte() (byte, error) {
for r.buffered() < 1 && r.state == nil {
r.more()
}
if r.buffered() < 1 {
return 0, r.err()
}
b := r.data[r.n]
r.n++
return b, nil
}
// WriteTo implements `io.WriterTo`
func (r *Reader) WriteTo(w io.Writer) (int64, error) {
var (
i int64
ii int
err error
)
// first, clear buffer
if r.buffered() > 0 {
ii, err = w.Write(r.data[r.n:])
i += int64(ii)
if err != nil {
return i, err
}
r.data = r.data[0:0]
r.n = 0
}
for r.state == nil {
// here we just do
// 1:1 reads and writes
r.more()
if r.buffered() > 0 {
ii, err = w.Write(r.data)
i += int64(ii)
if err != nil {
return i, err
}
r.data = r.data[0:0]
r.n = 0
}
}
if r.state != io.EOF {
return i, r.err()
}
return i, nil
}
func min(a int, b int) int {
if a < b {
return a
}
return b
}
func max(a int, b int) int {
if a < b {
return b
}
return a
}
go/vendor/github.com/philhofer/fwd/writer.go 0 → 100644
View file @ b34a88ce
package fwd
import "io"
const (
// DefaultWriterSize is the
// default write buffer size.
DefaultWriterSize = 2048
minWriterSize = minReaderSize
)
// Writer is a buffered writer
type Writer struct {
w io.Writer // writer
buf []byte // 0:len(buf) is bufered data
}
// NewWriter returns a new writer
// that writes to 'w' and has a buffer
// that is `DefaultWriterSize` bytes.
func NewWriter(w io.Writer) *Writer {
if wr, ok := w.(*Writer); ok {
return wr
}
return &Writer{
w: w,
buf: make([]byte, 0, DefaultWriterSize),
}
}
// NewWriterSize returns a new writer
// that writes to 'w' and has a buffer
// that is 'size' bytes.
func NewWriterSize(w io.Writer, size int) *Writer {
if wr, ok := w.(*Writer); ok && cap(wr.buf) >= size {
return wr
}
return &Writer{
w: w,
buf: make([]byte, 0, max(size, minWriterSize)),
}
}
// Buffered returns the number of buffered bytes
// in the reader.
func (w *Writer) Buffered() int { return len(w.buf) }
// BufferSize returns the maximum size of the buffer.
func (w *Writer) BufferSize() int { return cap(w.buf) }
// Flush flushes any buffered bytes
// to the underlying writer.
func (w *Writer) Flush() error {
l := len(w.buf)
if l > 0 {
n, err := w.w.Write(w.buf)
// if we didn't write the whole
// thing, copy the unwritten
// bytes to the beginnning of the
// buffer.
if n < l && n > 0 {
w.pushback(n)
if err == nil {
err = io.ErrShortWrite
}
}
if err != nil {
return err
}
w.buf = w.buf[:0]
return nil
}
return nil
}
// Write implements `io.Writer`
func (w *Writer) Write(p []byte) (int, error) {
c, l, ln := cap(w.buf), len(w.buf), len(p)
avail := c - l
// requires flush
if avail < ln {
if err := w.Flush(); err != nil {
return 0, err
}
l = len(w.buf)
}
// too big to fit in buffer;
// write directly to w.w
if c < ln {
return w.w.Write(p)
}
// grow buf slice; copy; return
w.buf = w.buf[:l+ln]
return copy(w.buf[l:], p), nil
}
// WriteString is analogous to Write, but it takes a string.
func (w *Writer) WriteString(s string) (int, error) {
c, l, ln := cap(w.buf), len(w.buf), len(s)
avail := c - l
// requires flush
if avail < ln {
if err := w.Flush(); err != nil {
return 0, err
}
l = len(w.buf)
}
// too big to fit in buffer;
// write directly to w.w
//
// yes, this is unsafe. *but*
// io.Writer is not allowed
// to mutate its input or
// maintain a reference to it,
// per the spec in package io.
//
// plus, if the string is really
// too big to fit in the buffer, then
// creating a copy to write it is
// expensive (and, strictly speaking,
// unnecessary)
if c < ln {
return w.w.Write(unsafestr(s))
}
// grow buf slice; copy; return
w.buf = w.buf[:l+ln]
return copy(w.buf[l:], s), nil
}
// WriteByte implements `io.ByteWriter`
func (w *Writer) WriteByte(b byte) error {
if len(w.buf) == cap(w.buf) {
if err := w.Flush(); err != nil {
return err
}
}
w.buf = append(w.buf, b)
return nil
}
// Next returns the next 'n' free bytes
// in the write buffer, flushing the writer
// as necessary. Next will return `io.ErrShortBuffer`
// if 'n' is greater than the size of the write buffer.
// Calls to 'next' increment the write position by
// the size of the returned buffer.
func (w *Writer) Next(n int) ([]byte, error) {
c, l := cap(w.buf), len(w.buf)
if n > c {
return nil, io.ErrShortBuffer
}
avail := c - l
if avail < n {
if err := w.Flush(); err != nil {
return nil, err
}
l = len(w.buf)
}
w.buf = w.buf[:l+n]
return w.buf[l:], nil
}
// take the bytes from w.buf[n:len(w.buf)]
// and put them at the beginning of w.buf,
// and resize to the length of the copied segment.
func (w *Writer) pushback(n int) {
w.buf = w.buf[:copy(w.buf, w.buf[n:])]
}
// ReadFrom implements `io.ReaderFrom`
func (w *Writer) ReadFrom(r io.Reader) (int64, error) {
// anticipatory flush
if err := w.Flush(); err != nil {
return 0, err
}
w.buf = w.buf[0:cap(w.buf)] // expand buffer
var nn int64 // written
var err error // error
var x int // read
// 1:1 reads and writes
for err == nil {
x, err = r.Read(w.buf)
if x > 0 {
n, werr := w.w.Write(w.buf[:x])
nn += int64(n)
if err != nil {
if n < x && n > 0 {
w.pushback(n - x)
}
return nn, werr
}
if n < x {
w.pushback(n - x)
return nn, io.ErrShortWrite
}
} else if err == nil {
err = io.ErrNoProgress
break
}
}
if err != io.EOF {
return nn, err
}
// we only clear here
// because we are sure
// the writes have
// succeeded. otherwise,
// we retain the data in case
// future writes succeed.
w.buf = w.buf[0:0]
return nn, nil
}
go/vendor/github.com/philhofer/fwd/writer_appengine.go 0 → 100644
View file @ b34a88ce
// +build appengine
package fwd
func unsafestr(s string) []byte { return []byte(s) }
go/vendor/github.com/philhofer/fwd/writer_unsafe.go 0 → 100644
View file @ b34a88ce
// +build !appengine
package fwd
import (
"reflect"
"unsafe"
)
// unsafe cast string as []byte
func unsafestr(b string) []byte {
l := len(b)
return *(*[]byte)(unsafe.Pointer(&reflect.SliceHeader{
Len: l,
Cap: l,
Data: (*reflect.StringHeader)(unsafe.Pointer(&b)).Data,
}))
}
go/vendor/github.com/tinylib/msgp/LICENSE 0 → 100644
View file @ b34a88ce
Copyright (c) 2014 Philip Hofer
Portions Copyright (c) 2009 The Go Authors (license at http://golang.org) where indicated
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.
\ No newline at end of file
go/vendor/github.com/tinylib/msgp/msgp/advise_linux.go 0 → 100644
View file @ b34a88ce
// +build linux,!appengine
package msgp
import (
"os"
"syscall"
)
func adviseRead(mem []byte) {
syscall.Madvise(mem, syscall.MADV_SEQUENTIAL|syscall.MADV_WILLNEED)
}
func adviseWrite(mem []byte) {
syscall.Madvise(mem, syscall.MADV_SEQUENTIAL)
}
func fallocate(f *os.File, sz int64) error {
err := syscall.Fallocate(int(f.Fd()), 0, 0, sz)
if err == syscall.ENOTSUP {
return f.Truncate(sz)
}
return err
}
go/vendor/github.com/tinylib/msgp/msgp/advise_other.go 0 → 100644
View file @ b34a88ce
// +build !linux appengine
package msgp
import (
"os"
)
// TODO: darwin, BSD support
func adviseRead(mem []byte) {}
func adviseWrite(mem []byte) {}
func fallocate(f *os.File, sz int64) error {
return f.Truncate(sz)
}
go/vendor/github.com/tinylib/msgp/msgp/circular.go 0 → 100644
View file @ b34a88ce
package msgp
type timer interface {
StartTimer()
StopTimer()
}
// EndlessReader is an io.Reader
// that loops over the same data
// endlessly. It is used for benchmarking.
type EndlessReader struct {
tb timer
data []byte
offset int
}
// NewEndlessReader returns a new endless reader
func NewEndlessReader(b []byte, tb timer) *EndlessReader {
return &EndlessReader{tb: tb, data: b, offset: 0}
}
// Read implements io.Reader. In practice, it
// always returns (len(p), nil), although it
// fills the supplied slice while the benchmark
// timer is stopped.
func (c *EndlessReader) Read(p []byte) (int, error) {
c.tb.StopTimer()
var n int
l := len(p)
m := len(c.data)
for n < l {
nn := copy(p[n:], c.data[c.offset:])
n += nn
c.offset += nn
c.offset %= m
}
c.tb.StartTimer()
return n, nil
}
go/vendor/github.com/tinylib/msgp/msgp/defs.go 0 → 100644
View file @ b34a88ce
// This package is the support library for the msgp code generator (http://github.com/tinylib/msgp).
//
// This package defines the utilites used by the msgp code generator for encoding and decoding MessagePack
// from []byte and io.Reader/io.Writer types. Much of this package is devoted to helping the msgp code
// generator implement the Marshaler/Unmarshaler and Encodable/Decodable interfaces.
//
// This package defines four "families" of functions:
// - AppendXxxx() appends an object to a []byte in MessagePack encoding.
// - ReadXxxxBytes() reads an object from a []byte and returns the remaining bytes.
// - (*Writer).WriteXxxx() writes an object to the buffered *Writer type.
// - (*Reader).ReadXxxx() reads an object from a buffered *Reader type.
//
// Once a type has satisfied the `Encodable` and `Decodable` interfaces,
// it can be written and read from arbitrary `io.Writer`s and `io.Reader`s using
// msgp.Encode(io.Writer, msgp.Encodable)
// and
// msgp.Decode(io.Reader, msgp.Decodable)
//
// There are also methods for converting MessagePack to JSON without
// an explicit de-serialization step.
//
// For additional tips, tricks, and gotchas, please visit
// the wiki at http://github.com/tinylib/msgp
package msgp
const last4 = 0x0f
const first4 = 0xf0
const last5 = 0x1f
const first3 = 0xe0
const last7 = 0x7f
func isfixint(b byte) bool {
return b>>7 == 0
}
func isnfixint(b byte) bool {
return b&first3 == mnfixint
}
func isfixmap(b byte) bool {
return b&first4 == mfixmap
}
func isfixarray(b byte) bool {
return b&first4 == mfixarray
}
func isfixstr(b byte) bool {
return b&first3 == mfixstr
}
func wfixint(u uint8) byte {
return u & last7
}
func rfixint(b byte) uint8 {
return b
}
func wnfixint(i int8) byte {
return byte(i) | mnfixint
}
func rnfixint(b byte) int8 {
return int8(b)
}
func rfixmap(b byte) uint8 {
return b & last4
}
func wfixmap(u uint8) byte {
return mfixmap | (u & last4)
}
func rfixstr(b byte) uint8 {
return b & last5
}
func wfixstr(u uint8) byte {
return (u & last5) | mfixstr
}
func rfixarray(b byte) uint8 {
return (b & last4)
}
func wfixarray(u uint8) byte {
return (u & last4) | mfixarray
}
// These are all the byte
// prefixes defined by the
// msgpack standard
const (
// 0XXXXXXX
mfixint uint8 = 0x00
// 111XXXXX
mnfixint uint8 = 0xe0
// 1000XXXX
mfixmap uint8 = 0x80
// 1001XXXX
mfixarray uint8 = 0x90
// 101XXXXX
mfixstr uint8 = 0xa0
mnil uint8 = 0xc0
mfalse uint8 = 0xc2
mtrue uint8 = 0xc3
mbin8 uint8 = 0xc4
mbin16 uint8 = 0xc5
mbin32 uint8 = 0xc6
mext8 uint8 = 0xc7
mext16 uint8 = 0xc8
mext32 uint8 = 0xc9
mfloat32 uint8 = 0xca
mfloat64 uint8 = 0xcb
muint8 uint8 = 0xcc
muint16 uint8 = 0xcd
muint32 uint8 = 0xce
muint64 uint8 = 0xcf
mint8 uint8 = 0xd0
mint16 uint8 = 0xd1
mint32 uint8 = 0xd2
mint64 uint8 = 0xd3
mfixext1 uint8 = 0xd4
mfixext2 uint8 = 0xd5
mfixext4 uint8 = 0xd6
mfixext8 uint8 = 0xd7
mfixext16 uint8 = 0xd8
mstr8 uint8 = 0xd9
mstr16 uint8 = 0xda
mstr32 uint8 = 0xdb
marray16 uint8 = 0xdc
marray32 uint8 = 0xdd
mmap16 uint8 = 0xde
mmap32 uint8 = 0xdf
)
go/vendor/github.com/tinylib/msgp/msgp/edit.go 0 → 100644
View file @ b34a88ce
package msgp
import (
"math"
)
// Locate returns a []byte pointing to the field
// in a messagepack map with the provided key. (The returned []byte
// points to a sub-slice of 'raw'; Locate does no allocations.) If the
// key doesn't exist in the map, a zero-length []byte will be returned.
func Locate(key string, raw []byte) []byte {
s, n := locate(raw, key)
return raw[s:n]
}
// Replace takes a key ("key") in a messagepack map ("raw")
// and replaces its value with the one provided and returns
// the new []byte. The returned []byte may point to the same
// memory as "raw". Replace makes no effort to evaluate the validity
// of the contents of 'val'. It may use up to the full capacity of 'raw.'
// Replace returns 'nil' if the field doesn't exist or if the object in 'raw'
// is not a map.
func Replace(key string, raw []byte, val []byte) []byte {
start, end := locate(raw, key)
if start == end {
return nil
}
return replace(raw, start, end, val, true)
}
// CopyReplace works similarly to Replace except that the returned
// byte slice does not point to the same memory as 'raw'. CopyReplace
// returns 'nil' if the field doesn't exist or 'raw' isn't a map.
func CopyReplace(key string, raw []byte, val []byte) []byte {
start, end := locate(raw, key)
if start == end {
return nil
}
return replace(raw, start, end, val, false)
}
// Remove removes a key-value pair from 'raw'. It returns
// 'raw' unchanged if the key didn't exist.
func Remove(key string, raw []byte) []byte {
start, end := locateKV(raw, key)
if start == end {
return raw
}
raw = raw[:start+copy(raw[start:], raw[end:])]
return resizeMap(raw, -1)
}
// HasKey returns whether the map in 'raw' has
// a field with key 'key'
func HasKey(key string, raw []byte) bool {
sz, bts, err := ReadMapHeaderBytes(raw)
if err != nil {
return false
}
var field []byte
for i := uint32(0); i < sz; i++ {
field, bts, err = ReadStringZC(bts)
if err != nil {
return false
}
if UnsafeString(field) == key {
return true
}
}
return false
}
func replace(raw []byte, start int, end int, val []byte, inplace bool) []byte {
ll := end - start // length of segment to replace
lv := len(val)
if inplace {
extra := lv - ll
// fastest case: we're doing
// a 1:1 replacement
if extra == 0 {
copy(raw[start:], val)
return raw
} else if extra < 0 {
// 'val' smaller than replaced value
// copy in place and shift back
x := copy(raw[start:], val)
y := copy(raw[start+x:], raw[end:])
return raw[:start+x+y]
} else if extra < cap(raw)-len(raw) {
// 'val' less than (cap-len) extra bytes
// copy in place and shift forward
raw = raw[0 : len(raw)+extra]
// shift end forward
copy(raw[end+extra:], raw[end:])
copy(raw[start:], val)
return raw
}
}
// we have to allocate new space
out := make([]byte, len(raw)+len(val)-ll)
x := copy(out, raw[:start])
y := copy(out[x:], val)
copy(out[x+y:], raw[end:])
return out
}
// locate does a naive O(n) search for the map key; returns start, end
// (returns 0,0 on error)
func locate(raw []byte, key string) (start int, end int) {
var (
sz uint32
bts []byte
field []byte
err error
)
sz, bts, err = ReadMapHeaderBytes(raw)
if err != nil {
return
}
// loop and locate field
for i := uint32(0); i < sz; i++ {
field, bts, err = ReadStringZC(bts)
if err != nil {
return 0, 0
}
if UnsafeString(field) == key {
// start location
l := len(raw)
start = l - len(bts)
bts, err = Skip(bts)
if err != nil {
return 0, 0
}
end = l - len(bts)
return
}
bts, err = Skip(bts)
if err != nil {
return 0, 0
}
}
return 0, 0
}
// locate key AND value
func locateKV(raw []byte, key string) (start int, end int) {
var (
sz uint32
bts []byte
field []byte
err error
)
sz, bts, err = ReadMapHeaderBytes(raw)
if err != nil {
return 0, 0
}
for i := uint32(0); i < sz; i++ {
tmp := len(bts)
field, bts, err = ReadStringZC(bts)
if err != nil {
return 0, 0
}
if UnsafeString(field) == key {
start = len(raw) - tmp
bts, err = Skip(bts)
if err != nil {
return 0, 0
}
end = len(raw) - len(bts)
return
}
bts, err = Skip(bts)
if err != nil {
return 0, 0
}
}
return 0, 0
}
// delta is delta on map size
func resizeMap(raw []byte, delta int64) []byte {
var sz int64
switch raw[0] {
case mmap16:
sz = int64(big.Uint16(raw[1:]))
if sz+delta <= math.MaxUint16 {
big.PutUint16(raw[1:], uint16(sz+delta))
return raw
}
if cap(raw)-len(raw) >= 2 {
raw = raw[0 : len(raw)+2]
copy(raw[5:], raw[3:])
raw[0] = mmap32
big.PutUint32(raw[1:], uint32(sz+delta))
return raw
}
n := make([]byte, 0, len(raw)+5)
n = AppendMapHeader(n, uint32(sz+delta))
return append(n, raw[3:]...)
case mmap32:
sz = int64(big.Uint32(raw[1:]))
big.PutUint32(raw[1:], uint32(sz+delta))
return raw
default:
sz = int64(rfixmap(raw[0]))
if sz+delta < 16 {
raw[0] = wfixmap(uint8(sz + delta))
return raw
} else if sz+delta <= math.MaxUint16 {
if cap(raw)-len(raw) >= 2 {
raw = raw[0 : len(raw)+2]
copy(raw[3:], raw[1:])
raw[0] = mmap16
big.PutUint16(raw[1:], uint16(sz+delta))
return raw
}
n := make([]byte, 0, len(raw)+5)
n = AppendMapHeader(n, uint32(sz+delta))
return append(n, raw[1:]...)
}
if cap(raw)-len(raw) >= 4 {
raw = raw[0 : len(raw)+4]
copy(raw[5:], raw[1:])
raw[0] = mmap32
big.PutUint32(raw[1:], uint32(sz+delta))
return raw
}
n := make([]byte, 0, len(raw)+5)
n = AppendMapHeader(n, uint32(sz+delta))
return append(n, raw[1:]...)
}
}
go/vendor/github.com/tinylib/msgp/msgp/elsize.go 0 → 100644
View file @ b34a88ce
package msgp
// size of every object on the wire,
// plus type information. gives us
// constant-time type information
// for traversing composite objects.
//
var sizes = [256]bytespec{
mnil: {size: 1, extra: constsize, typ: NilType},
mfalse: {size: 1, extra: constsize, typ: BoolType},
mtrue: {size: 1, extra: constsize, typ: BoolType},
mbin8: {size: 2, extra: extra8, typ: BinType},
mbin16: {size: 3, extra: extra16, typ: BinType},
mbin32: {size: 5, extra: extra32, typ: BinType},
mext8: {size: 3, extra: extra8, typ: ExtensionType},
mext16: {size: 4, extra: extra16, typ: ExtensionType},
mext32: {size: 6, extra: extra32, typ: ExtensionType},
mfloat32: {size: 5, extra: constsize, typ: Float32Type},
mfloat64: {size: 9, extra: constsize, typ: Float64Type},
muint8: {size: 2, extra: constsize, typ: UintType},
muint16: {size: 3, extra: constsize, typ: UintType},
muint32: {size: 5, extra: constsize, typ: UintType},
muint64: {size: 9, extra: constsize, typ: UintType},
mint8: {size: 2, extra: constsize, typ: IntType},
mint16: {size: 3, extra: constsize, typ: IntType},
mint32: {size: 5, extra: constsize, typ: IntType},
mint64: {size: 9, extra: constsize, typ: IntType},
mfixext1: {size: 3, extra: constsize, typ: ExtensionType},
mfixext2: {size: 4, extra: constsize, typ: ExtensionType},
mfixext4: {size: 6, extra: constsize, typ: ExtensionType},
mfixext8: {size: 10, extra: constsize, typ: ExtensionType},
mfixext16: {size: 18, extra: constsize, typ: ExtensionType},
mstr8: {size: 2, extra: extra8, typ: StrType},
mstr16: {size: 3, extra: extra16, typ: StrType},
mstr32: {size: 5, extra: extra32, typ: StrType},
marray16: {size: 3, extra: array16v, typ: ArrayType},
marray32: {size: 5, extra: array32v, typ: ArrayType},
mmap16: {size: 3, extra: map16v, typ: MapType},
mmap32: {size: 5, extra: map32v, typ: MapType},
}
func init() {
// set up fixed fields
// fixint
for i := mfixint; i < 0x80; i++ {
sizes[i] = bytespec{size: 1, extra: constsize, typ: IntType}
}
// nfixint
for i := uint16(mnfixint); i < 0x100; i++ {
sizes[uint8(i)] = bytespec{size: 1, extra: constsize, typ: IntType}
}
// fixstr gets constsize,
// since the prefix yields the size
for i := mfixstr; i < 0xc0; i++ {
sizes[i] = bytespec{size: 1 + rfixstr(i), extra: constsize, typ: StrType}
}
// fixmap
for i := mfixmap; i < 0x90; i++ {
sizes[i] = bytespec{size: 1, extra: varmode(2 * rfixmap(i)), typ: MapType}
}
// fixarray
for i := mfixarray; i < 0xa0; i++ {
sizes[i] = bytespec{size: 1, extra: varmode(rfixarray(i)), typ: ArrayType}
}
}
// a valid bytespsec has
// non-zero 'size' and
// non-zero 'typ'
type bytespec struct {
size uint8 // prefix size information
extra varmode // extra size information
typ Type // type
_ byte // makes bytespec 4 bytes (yes, this matters)
}
// size mode
// if positive, # elements for composites
type varmode int8
const (
constsize varmode = 0 // constant size (size bytes + uint8(varmode) objects)
extra8 = -1 // has uint8(p[1]) extra bytes
extra16 = -2 // has be16(p[1:]) extra bytes
extra32 = -3 // has be32(p[1:]) extra bytes
map16v = -4 // use map16
map32v = -5 // use map32
array16v = -6 // use array16
array32v = -7 // use array32
)
func getType(v byte) Type {
return sizes[v].typ
}
go/vendor/github.com/tinylib/msgp/msgp/errors.go 0 → 100644
View file @ b34a88ce
package msgp
import (
"fmt"
"reflect"
)
const resumableDefault = false
var (
// ErrShortBytes is returned when the
// slice being decoded is too short to
// contain the contents of the message
ErrShortBytes error = errShort{}
// this error is only returned
// if we reach code that should
// be unreachable
fatal error = errFatal{}
)
// Error is the interface satisfied
// by all of the errors that originate
// from this package.
type Error interface {
error
// Resumable returns whether
// or not the error means that
// the stream of data is malformed
// and the information is unrecoverable.
Resumable() bool
}
// contextError allows msgp Error instances to be enhanced with additional
// context about their origin.
type contextError interface {
Error
// withContext must not modify the error instance - it must clone and
// return a new error with the context added.
withContext(ctx string) error
}
// Cause returns the underlying cause of an error that has been wrapped
// with additional context.
func Cause(e error) error {
out := e
if e, ok := e.(errWrapped); ok && e.cause != nil {
out = e.cause
}
return out
}
// Resumable returns whether or not the error means that the stream of data is
// malformed and the information is unrecoverable.
func Resumable(e error) bool {
if e, ok := e.(Error); ok {
return e.Resumable()
}
return resumableDefault
}
// WrapError wraps an error with additional context that allows the part of the
// serialized type that caused the problem to be identified. Underlying errors
// can be retrieved using Cause()
//
// The input error is not modified - a new error should be returned.
//
// ErrShortBytes is not wrapped with any context due to backward compatibility
// issues with the public API.
//
func WrapError(err error, ctx ...interface{}) error {
switch e := err.(type) {
case errShort:
return e
case contextError:
return e.withContext(ctxString(ctx))
default:
return errWrapped{cause: err, ctx: ctxString(ctx)}
}
}
// ctxString converts the incoming interface{} slice into a single string.
func ctxString(ctx []interface{}) string {
out := ""
for idx, cv := range ctx {
if idx > 0 {
out += "/"
}
out += fmt.Sprintf("%v", cv)
}
return out
}
func addCtx(ctx, add string) string {
if ctx != "" {
return add + "/" + ctx
} else {
return add
}
}
// errWrapped allows arbitrary errors passed to WrapError to be enhanced with
// context and unwrapped with Cause()
type errWrapped struct {
cause error
ctx string
}
func (e errWrapped) Error() string {
if e.ctx != "" {
return fmt.Sprintf("%s at %s", e.cause, e.ctx)
} else {
return e.cause.Error()
}
}
func (e errWrapped) Resumable() bool {
if e, ok := e.cause.(Error); ok {
return e.Resumable()
}
return resumableDefault
}
type errShort struct{}
func (e errShort) Error() string { return "msgp: too few bytes left to read object" }
func (e errShort) Resumable() bool { return false }
type errFatal struct {
ctx string
}
func (f errFatal) Error() string {
out := "msgp: fatal decoding error (unreachable code)"
if f.ctx != "" {
out += " at " + f.ctx
}
return out
}
func (f errFatal) Resumable() bool { return false }
func (f errFatal) withContext(ctx string) error { f.ctx = addCtx(f.ctx, ctx); return f }
// ArrayError is an error returned
// when decoding a fix-sized array
// of the wrong size
type ArrayError struct {
Wanted uint32
Got uint32
ctx string
}
// Error implements the error interface
func (a ArrayError) Error() string {
out := fmt.Sprintf("msgp: wanted array of size %d; got %d", a.Wanted, a.Got)
if a.ctx != "" {
out += " at " + a.ctx
}
return out
}
// Resumable is always 'true' for ArrayErrors
func (a ArrayError) Resumable() bool { return true }
func (a ArrayError) withContext(ctx string) error { a.ctx = addCtx(a.ctx, ctx); return a }
// IntOverflow is returned when a call
// would downcast an integer to a type
// with too few bits to hold its value.
type IntOverflow struct {
Value int64 // the value of the integer
FailedBitsize int // the bit size that the int64 could not fit into
ctx string
}
// Error implements the error interface
func (i IntOverflow) Error() string {
str := fmt.Sprintf("msgp: %d overflows int%d", i.Value, i.FailedBitsize)
if i.ctx != "" {
str += " at " + i.ctx
}
return str
}
// Resumable is always 'true' for overflows
func (i IntOverflow) Resumable() bool { return true }
func (i IntOverflow) withContext(ctx string) error { i.ctx = addCtx(i.ctx, ctx); return i }
// UintOverflow is returned when a call
// would downcast an unsigned integer to a type
// with too few bits to hold its value
type UintOverflow struct {
Value uint64 // value of the uint
FailedBitsize int // the bit size that couldn't fit the value
ctx string
}
// Error implements the error interface
func (u UintOverflow) Error() string {
str := fmt.Sprintf("msgp: %d overflows uint%d", u.Value, u.FailedBitsize)
if u.ctx != "" {
str += " at " + u.ctx
}
return str
}
// Resumable is always 'true' for overflows
func (u UintOverflow) Resumable() bool { return true }
func (u UintOverflow) withContext(ctx string) error { u.ctx = addCtx(u.ctx, ctx); return u }
// UintBelowZero is returned when a call
// would cast a signed integer below zero
// to an unsigned integer.
type UintBelowZero struct {
Value int64 // value of the incoming int
ctx string
}
// Error implements the error interface
func (u UintBelowZero) Error() string {
str := fmt.Sprintf("msgp: attempted to cast int %d to unsigned", u.Value)
if u.ctx != "" {
str += " at " + u.ctx
}
return str
}
// Resumable is always 'true' for overflows
func (u UintBelowZero) Resumable() bool { return true }
func (u UintBelowZero) withContext(ctx string) error {
u.ctx = ctx
return u
}
// A TypeError is returned when a particular
// decoding method is unsuitable for decoding
// a particular MessagePack value.
type TypeError struct {
Method Type // Type expected by method
Encoded Type // Type actually encoded
ctx string
}
// Error implements the error interface
func (t TypeError) Error() string {
out := fmt.Sprintf("msgp: attempted to decode type %q with method for %q", t.Encoded, t.Method)
if t.ctx != "" {
out += " at " + t.ctx
}
return out
}
// Resumable returns 'true' for TypeErrors
func (t TypeError) Resumable() bool { return true }
func (t TypeError) withContext(ctx string) error { t.ctx = addCtx(t.ctx, ctx); return t }
// returns either InvalidPrefixError or
// TypeError depending on whether or not
// the prefix is recognized
func badPrefix(want Type, lead byte) error {
t := sizes[lead].typ
if t == InvalidType {
return InvalidPrefixError(lead)
}
return TypeError{Method: want, Encoded: t}
}
// InvalidPrefixError is returned when a bad encoding
// uses a prefix that is not recognized in the MessagePack standard.
// This kind of error is unrecoverable.
type InvalidPrefixError byte
// Error implements the error interface
func (i InvalidPrefixError) Error() string {
return fmt.Sprintf("msgp: unrecognized type prefix 0x%x", byte(i))
}
// Resumable returns 'false' for InvalidPrefixErrors
func (i InvalidPrefixError) Resumable() bool { return false }
// ErrUnsupportedType is returned
// when a bad argument is supplied
// to a function that takes `interface{}`.
type ErrUnsupportedType struct {
T reflect.Type
ctx string
}
// Error implements error
func (e *ErrUnsupportedType) Error() string {
out := fmt.Sprintf("msgp: type %q not supported", e.T)
if e.ctx != "" {
out += " at " + e.ctx
}
return out
}
// Resumable returns 'true' for ErrUnsupportedType
func (e *ErrUnsupportedType) Resumable() bool { return true }
func (e *ErrUnsupportedType) withContext(ctx string) error {
o := *e
o.ctx = addCtx(o.ctx, ctx)
return &o
}
go/vendor/github.com/tinylib/msgp/msgp/extension.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/tinylib/msgp/msgp/file.go 0 → 100644
View file @ b34a88ce
// +build linux darwin dragonfly freebsd netbsd openbsd
// +build !appengine
package msgp
import (
"os"
"syscall"
)
// ReadFile reads a file into 'dst' using
// a read-only memory mapping. Consequently,
// the file must be mmap-able, and the
// Unmarshaler should never write to
// the source memory. (Methods generated
// by the msgp tool obey that constraint, but
// user-defined implementations may not.)
//
// Reading and writing through file mappings
// is only efficient for large files; small
// files are best read and written using
// the ordinary streaming interfaces.
//
func ReadFile(dst Unmarshaler, file *os.File) error {
stat, err := file.Stat()
if err != nil {
return err
}
data, err := syscall.Mmap(int(file.Fd()), 0, int(stat.Size()), syscall.PROT_READ, syscall.MAP_SHARED)
if err != nil {
return err
}
adviseRead(data)
_, err = dst.UnmarshalMsg(data)
uerr := syscall.Munmap(data)
if err == nil {
err = uerr
}
return err
}
// MarshalSizer is the combination
// of the Marshaler and Sizer
// interfaces.
type MarshalSizer interface {
Marshaler
Sizer
}
// WriteFile writes a file from 'src' using
// memory mapping. It overwrites the entire
// contents of the previous file.
// The mapping size is calculated
// using the `Msgsize()` method
// of 'src', so it must produce a result
// equal to or greater than the actual encoded
// size of the object. Otherwise,
// a fault (SIGBUS) will occur.
//
// Reading and writing through file mappings
// is only efficient for large files; small
// files are best read and written using
// the ordinary streaming interfaces.
//
// NOTE: The performance of this call
// is highly OS- and filesystem-dependent.
// Users should take care to test that this
// performs as expected in a production environment.
// (Linux users should run a kernel and filesystem
// that support fallocate(2) for the best results.)
func WriteFile(src MarshalSizer, file *os.File) error {
sz := src.Msgsize()
err := fallocate(file, int64(sz))
if err != nil {
return err
}
data, err := syscall.Mmap(int(file.Fd()), 0, sz, syscall.PROT_READ|syscall.PROT_WRITE, syscall.MAP_SHARED)
if err != nil {
return err
}
adviseWrite(data)
chunk := data[:0]
chunk, err = src.MarshalMsg(chunk)
if err != nil {
return err
}
uerr := syscall.Munmap(data)
if uerr != nil {
return uerr
}
return file.Truncate(int64(len(chunk)))
}
go/vendor/github.com/tinylib/msgp/msgp/file_port.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/tinylib/msgp/msgp/integers.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/tinylib/msgp/msgp/json.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/tinylib/msgp/msgp/json_bytes.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/tinylib/msgp/msgp/number.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/tinylib/msgp/msgp/purego.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/tinylib/msgp/msgp/read.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/tinylib/msgp/msgp/read_bytes.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/tinylib/msgp/msgp/size.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/tinylib/msgp/msgp/unsafe.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/tinylib/msgp/msgp/write.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/tinylib/msgp/msgp/write_bytes.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/uber/jaeger-client-go/LICENSE 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/uber/jaeger-client-go/config/config.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/uber/jaeger-client-go/config/options.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/github.com/uber/jaeger-client-go/log/logger.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gitlab.com/gitlab-org/labkit/LICENSE 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gitlab.com/gitlab-org/labkit/correlation/base62.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gitlab.com/gitlab-org/labkit/correlation/context.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gitlab.com/gitlab-org/labkit/correlation/doc.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gitlab.com/gitlab-org/labkit/tracing/doc.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gitlab.com/gitlab-org/labkit/tracing/env_injector.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gitlab.com/gitlab-org/labkit/tracing/errors.go 0 → 100644
View file @ b34a88ce
package tracing
import (
"fmt"
)
// ErrConfiguration is returned when the tracer is not properly configured.
var ErrConfiguration = fmt.Errorf("Tracing is not properly configured")
go/vendor/gitlab.com/gitlab-org/labkit/tracing/impl/constants.go 0 → 100644
View file @ b34a88ce
package impl
const keyStrictConnectionParsing = "strict_parsing"
go/vendor/gitlab.com/gitlab-org/labkit/tracing/inbound_http.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gopkg.in/DataDog/dd-trace-go.v1/LICENSE 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gopkg.in/DataDog/dd-trace-go.v1/LICENSE-3rdparty.csv 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gopkg.in/DataDog/dd-trace-go.v1/ddtrace/Gopkg.toml 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gopkg.in/DataDog/dd-trace-go.v1/ddtrace/ddtrace.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gopkg.in/DataDog/dd-trace-go.v1/ddtrace/ext/db.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gopkg.in/DataDog/dd-trace-go.v1/ddtrace/ext/peer.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/gopkg.in/DataDog/dd-trace-go.v1/ddtrace/ext/tags.go 0 → 100644
View file @ b34a88ce
This diff is collapsed.
go/vendor/vendor.json
View file @ b34a88ce
This diff is collapsed.
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7