prometheus/rules/parser.y

280 lines
10 KiB
Plaintext
Raw Normal View History

// Copyright 2013 The Prometheus Authors
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
%{
package rules
import (
clientmodel "github.com/prometheus/client_golang/model"
"github.com/prometheus/prometheus/rules/ast"
"github.com/prometheus/prometheus/storage/metric"
)
%}
%union {
num clientmodel.SampleValue
str string
ruleNode ast.Node
ruleNodeSlice []ast.Node
boolean bool
labelNameSlice clientmodel.LabelNames
labelSet clientmodel.LabelSet
labelMatcher *metric.LabelMatcher
labelMatchers metric.LabelMatchers
vectorMatching *vectorMatching
}
/* We simulate multiple start symbols for closely-related grammars via dummy tokens. See
http://www.gnu.org/software/bison/manual/html_node/Multiple-start_002dsymbols.html
Reason: we want to be able to parse lists of named rules as well as single expressions.
*/
%token START_RULES START_EXPRESSION
%token <str> IDENTIFIER STRING DURATION METRICNAME
%token <num> NUMBER
%token PERMANENT GROUP_OP KEEPING_EXTRA OFFSET MATCH_OP
%token <str> AGGR_OP CMP_OP ADDITIVE_OP MULT_OP MATCH_MOD
%token ALERT IF FOR WITH SUMMARY DESCRIPTION
%type <ruleNodeSlice> func_arg_list
%type <labelNameSlice> label_list grouping_opts
%type <labelSet> label_assign label_assign_list rule_labels
%type <labelMatcher> label_match
%type <labelMatchers> label_match_list label_matches
%type <vectorMatching> vector_matching
%type <ruleNode> rule_expr func_arg
%type <boolean> qualifier extra_labels_opts
Implement offset operator. This allows changing the time offset for individual instant and range vectors in a query. For example, this returns the value of `foo` 5 minutes in the past relative to the current query evaluation time: foo offset 5m Note that the `offset` modifier always needs to follow the selector immediately. I.e. the following would be correct: sum(foo offset 5m) // GOOD. While the following would be *incorrect*: sum(foo) offset 5m // INVALID. The same works for range vectors. This returns the 5-minutes-rate that `foo` had a week ago: rate(foo[5m] offset 1w) This change touches the following components: * Lexer/parser: additions to correctly parse the new `offset`/`OFFSET` keyword. * AST: vector and matrix nodes now have an additional `offset` field. This is used during their evaluation to adjust query and result times appropriately. * Query analyzer: now works on separate sets of ranges and instants per offset. Isolating different offsets from each other completely in this way keeps the preloading code relatively simple. No storage engine changes were needed by this change. The rules tests have been changed to not probe the internal implementation details of the query analyzer anymore (how many instants and ranges have been preloaded). This would also become too cumbersome to test with the new model, and measuring the result of the query should be sufficient. This fixes https://github.com/prometheus/prometheus/issues/529 This fixed https://github.com/prometheus/promdash/issues/201
2015-02-18 01:30:41 +00:00
%type <str> for_duration metric_name label_match_type offset_opts
%right '='
%left CMP_OP
%left ADDITIVE_OP
%left MULT_OP
%start start
%%
start : START_RULES rules_stat_list
| START_EXPRESSION saved_rule_expr
;
rules_stat_list : /* empty */
| rules_stat_list rules_stat
;
saved_rule_expr : rule_expr
{ yylex.(*RulesLexer).parsedExpr = $1 }
;
rules_stat : qualifier metric_name rule_labels '=' rule_expr
{
rule, err := CreateRecordingRule($2, $3, $5, $1)
if err != nil { yylex.Error(err.Error()); return 1 }
yylex.(*RulesLexer).parsedRules = append(yylex.(*RulesLexer).parsedRules, rule)
}
| ALERT IDENTIFIER IF rule_expr for_duration WITH rule_labels SUMMARY STRING DESCRIPTION STRING
{
rule, err := CreateAlertingRule($2, $4, $5, $7, $9, $11)
if err != nil { yylex.Error(err.Error()); return 1 }
yylex.(*RulesLexer).parsedRules = append(yylex.(*RulesLexer).parsedRules, rule)
}
;
for_duration : /* empty */
{ $$ = "0s" }
| FOR DURATION
{ $$ = $2 }
;
qualifier : /* empty */
{ $$ = false }
| PERMANENT
{ $$ = true }
;
metric_name : METRICNAME
{ $$ = $1 }
| IDENTIFIER
{ $$ = $1 }
;
rule_labels : /* empty */
{ $$ = clientmodel.LabelSet{} }
| '{' label_assign_list '}'
{ $$ = $2 }
| '{' '}'
{ $$ = clientmodel.LabelSet{} }
label_assign_list : label_assign
{ $$ = $1 }
| label_assign_list ',' label_assign
{ for k, v := range $3 { $$[k] = v } }
;
label_assign : IDENTIFIER '=' STRING
{ $$ = clientmodel.LabelSet{ clientmodel.LabelName($1): clientmodel.LabelValue($3) } }
;
label_matches : /* empty */
{ $$ = metric.LabelMatchers{} }
| '{' '}'
{ $$ = metric.LabelMatchers{} }
| '{' label_match_list '}'
{ $$ = $2 }
;
label_match_list : label_match
{ $$ = metric.LabelMatchers{$1} }
| label_match_list ',' label_match
{ $$ = append($$, $3) }
;
label_match : IDENTIFIER label_match_type STRING
{
var err error
$$, err = newLabelMatcher($2, clientmodel.LabelName($1), clientmodel.LabelValue($3))
if err != nil { yylex.Error(err.Error()); return 1 }
}
;
label_match_type : '='
{ $$ = "=" }
| CMP_OP
{ $$ = $1 }
;
Implement offset operator. This allows changing the time offset for individual instant and range vectors in a query. For example, this returns the value of `foo` 5 minutes in the past relative to the current query evaluation time: foo offset 5m Note that the `offset` modifier always needs to follow the selector immediately. I.e. the following would be correct: sum(foo offset 5m) // GOOD. While the following would be *incorrect*: sum(foo) offset 5m // INVALID. The same works for range vectors. This returns the 5-minutes-rate that `foo` had a week ago: rate(foo[5m] offset 1w) This change touches the following components: * Lexer/parser: additions to correctly parse the new `offset`/`OFFSET` keyword. * AST: vector and matrix nodes now have an additional `offset` field. This is used during their evaluation to adjust query and result times appropriately. * Query analyzer: now works on separate sets of ranges and instants per offset. Isolating different offsets from each other completely in this way keeps the preloading code relatively simple. No storage engine changes were needed by this change. The rules tests have been changed to not probe the internal implementation details of the query analyzer anymore (how many instants and ranges have been preloaded). This would also become too cumbersome to test with the new model, and measuring the result of the query should be sufficient. This fixes https://github.com/prometheus/prometheus/issues/529 This fixed https://github.com/prometheus/promdash/issues/201
2015-02-18 01:30:41 +00:00
offset_opts : /* empty */
{ $$ = "0s" }
| OFFSET DURATION
{ $$ = $2 }
;
rule_expr : '(' rule_expr ')'
{ $$ = $2 }
Implement offset operator. This allows changing the time offset for individual instant and range vectors in a query. For example, this returns the value of `foo` 5 minutes in the past relative to the current query evaluation time: foo offset 5m Note that the `offset` modifier always needs to follow the selector immediately. I.e. the following would be correct: sum(foo offset 5m) // GOOD. While the following would be *incorrect*: sum(foo) offset 5m // INVALID. The same works for range vectors. This returns the 5-minutes-rate that `foo` had a week ago: rate(foo[5m] offset 1w) This change touches the following components: * Lexer/parser: additions to correctly parse the new `offset`/`OFFSET` keyword. * AST: vector and matrix nodes now have an additional `offset` field. This is used during their evaluation to adjust query and result times appropriately. * Query analyzer: now works on separate sets of ranges and instants per offset. Isolating different offsets from each other completely in this way keeps the preloading code relatively simple. No storage engine changes were needed by this change. The rules tests have been changed to not probe the internal implementation details of the query analyzer anymore (how many instants and ranges have been preloaded). This would also become too cumbersome to test with the new model, and measuring the result of the query should be sufficient. This fixes https://github.com/prometheus/prometheus/issues/529 This fixed https://github.com/prometheus/promdash/issues/201
2015-02-18 01:30:41 +00:00
| '{' label_match_list '}' offset_opts
{
var err error
$$, err = NewVectorSelector($2, $4)
if err != nil { yylex.Error(err.Error()); return 1 }
}
| metric_name label_matches offset_opts
{
var err error
m, err := metric.NewLabelMatcher(metric.Equal, clientmodel.MetricNameLabel, clientmodel.LabelValue($1))
if err != nil { yylex.Error(err.Error()); return 1 }
$2 = append($2, m)
Implement offset operator. This allows changing the time offset for individual instant and range vectors in a query. For example, this returns the value of `foo` 5 minutes in the past relative to the current query evaluation time: foo offset 5m Note that the `offset` modifier always needs to follow the selector immediately. I.e. the following would be correct: sum(foo offset 5m) // GOOD. While the following would be *incorrect*: sum(foo) offset 5m // INVALID. The same works for range vectors. This returns the 5-minutes-rate that `foo` had a week ago: rate(foo[5m] offset 1w) This change touches the following components: * Lexer/parser: additions to correctly parse the new `offset`/`OFFSET` keyword. * AST: vector and matrix nodes now have an additional `offset` field. This is used during their evaluation to adjust query and result times appropriately. * Query analyzer: now works on separate sets of ranges and instants per offset. Isolating different offsets from each other completely in this way keeps the preloading code relatively simple. No storage engine changes were needed by this change. The rules tests have been changed to not probe the internal implementation details of the query analyzer anymore (how many instants and ranges have been preloaded). This would also become too cumbersome to test with the new model, and measuring the result of the query should be sufficient. This fixes https://github.com/prometheus/prometheus/issues/529 This fixed https://github.com/prometheus/promdash/issues/201
2015-02-18 01:30:41 +00:00
$$, err = NewVectorSelector($2, $3)
if err != nil { yylex.Error(err.Error()); return 1 }
}
| IDENTIFIER '(' func_arg_list ')'
{
var err error
$$, err = NewFunctionCall($1, $3)
if err != nil { yylex.Error(err.Error()); return 1 }
}
| IDENTIFIER '(' ')'
{
var err error
$$, err = NewFunctionCall($1, []ast.Node{})
if err != nil { yylex.Error(err.Error()); return 1 }
}
Implement offset operator. This allows changing the time offset for individual instant and range vectors in a query. For example, this returns the value of `foo` 5 minutes in the past relative to the current query evaluation time: foo offset 5m Note that the `offset` modifier always needs to follow the selector immediately. I.e. the following would be correct: sum(foo offset 5m) // GOOD. While the following would be *incorrect*: sum(foo) offset 5m // INVALID. The same works for range vectors. This returns the 5-minutes-rate that `foo` had a week ago: rate(foo[5m] offset 1w) This change touches the following components: * Lexer/parser: additions to correctly parse the new `offset`/`OFFSET` keyword. * AST: vector and matrix nodes now have an additional `offset` field. This is used during their evaluation to adjust query and result times appropriately. * Query analyzer: now works on separate sets of ranges and instants per offset. Isolating different offsets from each other completely in this way keeps the preloading code relatively simple. No storage engine changes were needed by this change. The rules tests have been changed to not probe the internal implementation details of the query analyzer anymore (how many instants and ranges have been preloaded). This would also become too cumbersome to test with the new model, and measuring the result of the query should be sufficient. This fixes https://github.com/prometheus/prometheus/issues/529 This fixed https://github.com/prometheus/promdash/issues/201
2015-02-18 01:30:41 +00:00
| rule_expr '[' DURATION ']' offset_opts
{
var err error
Implement offset operator. This allows changing the time offset for individual instant and range vectors in a query. For example, this returns the value of `foo` 5 minutes in the past relative to the current query evaluation time: foo offset 5m Note that the `offset` modifier always needs to follow the selector immediately. I.e. the following would be correct: sum(foo offset 5m) // GOOD. While the following would be *incorrect*: sum(foo) offset 5m // INVALID. The same works for range vectors. This returns the 5-minutes-rate that `foo` had a week ago: rate(foo[5m] offset 1w) This change touches the following components: * Lexer/parser: additions to correctly parse the new `offset`/`OFFSET` keyword. * AST: vector and matrix nodes now have an additional `offset` field. This is used during their evaluation to adjust query and result times appropriately. * Query analyzer: now works on separate sets of ranges and instants per offset. Isolating different offsets from each other completely in this way keeps the preloading code relatively simple. No storage engine changes were needed by this change. The rules tests have been changed to not probe the internal implementation details of the query analyzer anymore (how many instants and ranges have been preloaded). This would also become too cumbersome to test with the new model, and measuring the result of the query should be sufficient. This fixes https://github.com/prometheus/prometheus/issues/529 This fixed https://github.com/prometheus/promdash/issues/201
2015-02-18 01:30:41 +00:00
$$, err = NewMatrixSelector($1, $3, $5)
if err != nil { yylex.Error(err.Error()); return 1 }
}
| AGGR_OP '(' rule_expr ')' grouping_opts extra_labels_opts
{
var err error
$$, err = NewVectorAggregation($1, $3, $5, $6)
if err != nil { yylex.Error(err.Error()); return 1 }
}
| AGGR_OP grouping_opts extra_labels_opts '(' rule_expr ')'
{
var err error
$$, err = NewVectorAggregation($1, $5, $2, $3)
if err != nil { yylex.Error(err.Error()); return 1 }
}
/* Yacc can only attach associativity to terminals, so we
* have to list all operators here. */
| rule_expr ADDITIVE_OP vector_matching rule_expr
{
var err error
$$, err = NewArithExpr($2, $1, $4, $3)
if err != nil { yylex.Error(err.Error()); return 1 }
}
| rule_expr MULT_OP vector_matching rule_expr
{
var err error
$$, err = NewArithExpr($2, $1, $4, $3)
if err != nil { yylex.Error(err.Error()); return 1 }
}
| rule_expr CMP_OP vector_matching rule_expr
{
var err error
$$, err = NewArithExpr($2, $1, $4, $3)
if err != nil { yylex.Error(err.Error()); return 1 }
}
| NUMBER
{ $$ = ast.NewScalarLiteral($1)}
;
extra_labels_opts : /* empty */
{ $$ = false }
| KEEPING_EXTRA
{ $$ = true }
;
vector_matching : /* empty */
{ $$ = nil }
| MATCH_OP '(' label_list ')'
{
var err error
$$, err = newVectorMatching("", $3, nil)
if err != nil { yylex.Error(err.Error()); return 1 }
}
| MATCH_OP '(' label_list ')' MATCH_MOD '(' label_list ')'
{
var err error
$$, err = newVectorMatching($5, $3, $7)
if err != nil { yylex.Error(err.Error()); return 1 }
}
;
grouping_opts :
{ $$ = clientmodel.LabelNames{} }
| GROUP_OP '(' label_list ')'
{ $$ = $3 }
;
label_list : IDENTIFIER
{ $$ = clientmodel.LabelNames{clientmodel.LabelName($1)} }
| label_list ',' IDENTIFIER
{ $$ = append($$, clientmodel.LabelName($3)) }
;
func_arg_list : func_arg
{ $$ = []ast.Node{$1} }
| func_arg_list ',' func_arg
{ $$ = append($$, $3) }
;
func_arg : rule_expr
{ $$ = $1 }
| STRING
{ $$ = ast.NewStringLiteral($1) }
;
%%