forked from influxdata/telegraf
-
Notifications
You must be signed in to change notification settings - Fork 0
/
parser.go
244 lines (204 loc) · 5.26 KB
/
parser.go
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
package influx
import (
"errors"
"fmt"
"io"
"strings"
"sync"
"time"
"github.com/influxdata/telegraf"
)
const (
maxErrorBufferSize = 1024
)
var (
ErrNoMetric = errors.New("no metric in line")
)
type TimeFunc func() time.Time
// ParseError indicates a error in the parsing of the text.
type ParseError struct {
Offset int
LineOffset int
LineNumber int
Column int
msg string
buf string
}
func (e *ParseError) Error() string {
buffer := e.buf[e.LineOffset:]
eol := strings.IndexAny(buffer, "\n")
if eol >= 0 {
buffer = strings.TrimSuffix(buffer[:eol], "\r")
}
if len(buffer) > maxErrorBufferSize {
startEllipsis := true
offset := e.Offset - e.LineOffset
start := offset - maxErrorBufferSize
if start < 0 {
startEllipsis = false
start = 0
}
// if we trimmed it the column won't line up. it'll always be the last character,
// because the parser doesn't continue past it, but point it out anyway so
// it's obvious where the issue is.
buffer = buffer[start:offset] + "<-- here"
if startEllipsis {
buffer = "..." + buffer
}
}
return fmt.Sprintf("metric parse error: %s at %d:%d: %q", e.msg, e.LineNumber, e.Column, buffer)
}
// Parser is an InfluxDB Line Protocol parser that implements the
// parsers.Parser interface.
type Parser struct {
DefaultTags map[string]string
sync.Mutex
*machine
handler *MetricHandler
}
// NewParser returns a Parser than accepts line protocol
func NewParser(handler *MetricHandler) *Parser {
return &Parser{
machine: NewMachine(handler),
handler: handler,
}
}
// NewSeriesParser returns a Parser than accepts a measurement and tagset
func NewSeriesParser(handler *MetricHandler) *Parser {
return &Parser{
machine: NewSeriesMachine(handler),
handler: handler,
}
}
func (p *Parser) SetTimeFunc(f TimeFunc) {
p.handler.SetTimeFunc(f)
}
func (p *Parser) Parse(input []byte) ([]telegraf.Metric, error) {
p.Lock()
defer p.Unlock()
metrics := make([]telegraf.Metric, 0)
p.machine.SetData(input)
for {
err := p.machine.Next()
if err == EOF {
break
}
if err != nil {
return nil, &ParseError{
Offset: p.machine.Position(),
LineOffset: p.machine.LineOffset(),
LineNumber: p.machine.LineNumber(),
Column: p.machine.Column(),
msg: err.Error(),
buf: string(input),
}
}
metric, err := p.handler.Metric()
if err != nil {
return nil, err
}
if metric == nil {
continue
}
metrics = append(metrics, metric)
}
p.applyDefaultTags(metrics)
return metrics, nil
}
func (p *Parser) ParseLine(line string) (telegraf.Metric, error) {
metrics, err := p.Parse([]byte(line))
if err != nil {
return nil, err
}
if len(metrics) < 1 {
return nil, ErrNoMetric
}
return metrics[0], nil
}
func (p *Parser) SetDefaultTags(tags map[string]string) {
p.DefaultTags = tags
}
func (p *Parser) applyDefaultTags(metrics []telegraf.Metric) {
if len(p.DefaultTags) == 0 {
return
}
for _, m := range metrics {
p.applyDefaultTagsSingle(m)
}
}
func (p *Parser) applyDefaultTagsSingle(metric telegraf.Metric) {
for k, v := range p.DefaultTags {
if !metric.HasTag(k) {
metric.AddTag(k, v)
}
}
}
// StreamParser is an InfluxDB Line Protocol parser. It is not safe for
// concurrent use in multiple goroutines.
type StreamParser struct {
machine *streamMachine
handler *MetricHandler
}
func NewStreamParser(r io.Reader) *StreamParser {
handler := NewMetricHandler()
return &StreamParser{
machine: NewStreamMachine(r, handler),
handler: handler,
}
}
// SetTimeFunc changes the function used to determine the time of metrics
// without a timestamp. The default TimeFunc is time.Now. Useful mostly for
// testing, or perhaps if you want all metrics to have the same timestamp.
func (sp *StreamParser) SetTimeFunc(f TimeFunc) {
sp.handler.SetTimeFunc(f)
}
func (sp *StreamParser) SetTimePrecision(u time.Duration) {
sp.handler.SetTimePrecision(u)
}
// Next parses the next item from the stream. You can repeat calls to this
// function if it returns ParseError to get the next metric or error.
func (sp *StreamParser) Next() (telegraf.Metric, error) {
err := sp.machine.Next()
if err == EOF {
return nil, err
}
if e, ok := err.(*readErr); ok {
return nil, e.Err
}
if err != nil {
return nil, &ParseError{
Offset: sp.machine.Position(),
LineOffset: sp.machine.LineOffset(),
LineNumber: sp.machine.LineNumber(),
Column: sp.machine.Column(),
msg: err.Error(),
buf: sp.machine.LineText(),
}
}
metric, err := sp.handler.Metric()
if err != nil {
return nil, err
}
return metric, nil
}
// Position returns the current byte offset into the data.
func (sp *StreamParser) Position() int {
return sp.machine.Position()
}
// LineOffset returns the byte offset of the current line.
func (sp *StreamParser) LineOffset() int {
return sp.machine.LineOffset()
}
// LineNumber returns the current line number. Lines are counted based on the
// regular expression `\r?\n`.
func (sp *StreamParser) LineNumber() int {
return sp.machine.LineNumber()
}
// Column returns the current column.
func (sp *StreamParser) Column() int {
return sp.machine.Column()
}
// LineText returns the text of the current line that has been parsed so far.
func (sp *StreamParser) LineText() string {
return sp.machine.LineText()
}