dataSourcePlan.go 6.0 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250
  1. // Copyright 2021-2022 EMQ Technologies Co., Ltd.
  2. //
  3. // Licensed under the Apache License, Version 2.0 (the "License");
  4. // you may not use this file except in compliance with the License.
  5. // You may obtain a copy of the License at
  6. //
  7. // http://www.apache.org/licenses/LICENSE-2.0
  8. //
  9. // Unless required by applicable law or agreed to in writing, software
  10. // distributed under the License is distributed on an "AS IS" BASIS,
  11. // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  12. // See the License for the specific language governing permissions and
  13. // limitations under the License.
  14. package planner
  15. import (
  16. "fmt"
  17. "sort"
  18. "strings"
  19. "github.com/lf-edge/ekuiper/pkg/ast"
  20. "github.com/lf-edge/ekuiper/pkg/message"
  21. )
  22. type DataSourcePlan struct {
  23. baseLogicalPlan
  24. name ast.StreamName
  25. // calculated properties
  26. // initialized with stream definition, pruned with rule
  27. metaFields []string
  28. // pass-on and converted state. For schemaless, the value is always nil
  29. streamFields map[string]*ast.JsonStreamField
  30. // pass-on properties
  31. isSchemaless bool
  32. streamStmt *ast.StreamStmt
  33. allMeta bool
  34. isBinary bool
  35. iet bool
  36. timestampFormat string
  37. timestampField string
  38. // intermediate status
  39. isWildCard bool
  40. fields map[string]*ast.JsonStreamField
  41. metaMap map[string]string
  42. }
  43. func (p DataSourcePlan) Init() *DataSourcePlan {
  44. p.baseLogicalPlan.self = &p
  45. p.baseLogicalPlan.setPlanType(DATASOURCE)
  46. return &p
  47. }
  48. func (p *DataSourcePlan) BuildExplainInfo(id int64) {
  49. info := ""
  50. if p.name != "" {
  51. info += "StreamName: " + string(p.name)
  52. }
  53. if p.fields != nil && len(p.fields) != 0 {
  54. info += ", Fields:[ "
  55. keys := make([]string, 0, len(p.fields))
  56. for k := range p.fields {
  57. keys = append(keys, k)
  58. }
  59. sort.Strings(keys)
  60. for i := 0; i < len(keys); i++ {
  61. info += keys[i]
  62. if i != len(keys)-1 {
  63. info += ", "
  64. }
  65. }
  66. info += " ]"
  67. }
  68. if p.streamFields != nil && len(p.streamFields) != 0 {
  69. info += ", StreamFields:[ "
  70. keys := make([]string, 0, len(p.streamFields))
  71. for k := range p.streamFields {
  72. keys = append(keys, k)
  73. }
  74. sort.Strings(keys)
  75. for i := 0; i < len(keys); i++ {
  76. info += keys[i]
  77. if i != len(keys)-1 {
  78. info += ", "
  79. }
  80. }
  81. info += " ]"
  82. }
  83. p.baseLogicalPlan.ExplainInfo.ID = id
  84. p.baseLogicalPlan.ExplainInfo.Info = info
  85. }
  86. // PushDownPredicate Presume no children for data source
  87. func (p *DataSourcePlan) PushDownPredicate(condition ast.Expr) (ast.Expr, LogicalPlan) {
  88. if p.streamStmt.StreamType == ast.TypeTable {
  89. return condition, p.self
  90. }
  91. owned, other := p.extract(condition)
  92. if owned != nil {
  93. // Add a filter plan for children
  94. f := FilterPlan{
  95. condition: owned,
  96. }.Init()
  97. f.SetChildren([]LogicalPlan{p})
  98. return other, f
  99. }
  100. return other, p
  101. }
  102. func (p *DataSourcePlan) extract(expr ast.Expr) (ast.Expr, ast.Expr) {
  103. s, hasDefault := getRefSources(expr)
  104. l := len(s)
  105. if hasDefault {
  106. l += 1
  107. }
  108. switch len(s) {
  109. case 0:
  110. return expr, nil
  111. case 1:
  112. if s[0] == p.name || s[0] == ast.DefaultStream {
  113. return expr, nil
  114. } else {
  115. return nil, expr
  116. }
  117. default:
  118. if be, ok := expr.(*ast.BinaryExpr); ok && be.OP == ast.AND {
  119. ul, pl := p.extract(be.LHS)
  120. ur, pr := p.extract(be.RHS)
  121. owned := combine(ul, ur)
  122. other := combine(pl, pr)
  123. return owned, other
  124. }
  125. return nil, expr
  126. }
  127. }
  128. func (p *DataSourcePlan) PruneColumns(fields []ast.Expr) error {
  129. // init values
  130. err := p.getProps()
  131. if err != nil {
  132. return err
  133. }
  134. p.fields = make(map[string]*ast.JsonStreamField)
  135. if !p.allMeta {
  136. p.metaMap = make(map[string]string)
  137. }
  138. if p.timestampField != "" {
  139. if !p.isSchemaless {
  140. tsf, ok := p.streamFields[p.timestampField]
  141. if !ok {
  142. return fmt.Errorf("timestamp field %s not found", p.timestampField)
  143. }
  144. p.fields[p.timestampField] = tsf
  145. } else {
  146. p.fields[p.timestampField] = nil
  147. }
  148. }
  149. for _, field := range fields {
  150. switch f := field.(type) {
  151. case *ast.Wildcard:
  152. p.isWildCard = true
  153. case *ast.FieldRef:
  154. if !p.isWildCard && (f.StreamName == ast.DefaultStream || f.StreamName == p.name) {
  155. if _, ok := p.fields[f.Name]; !ok {
  156. sf, err := p.getField(f.Name, f.StreamName == p.name)
  157. if err != nil {
  158. return err
  159. }
  160. if p.isSchemaless || sf != nil {
  161. p.fields[f.Name] = sf
  162. }
  163. }
  164. }
  165. case *ast.MetaRef:
  166. if p.allMeta {
  167. break
  168. }
  169. if f.StreamName == ast.DefaultStream || f.StreamName == p.name {
  170. if f.Name == "*" {
  171. p.allMeta = true
  172. p.metaMap = nil
  173. } else if !p.allMeta {
  174. p.metaMap[strings.ToLower(f.Name)] = f.Name
  175. }
  176. }
  177. case *ast.SortField:
  178. if !p.isWildCard {
  179. sf, err := p.getField(f.Name, f.StreamName == p.name)
  180. if err != nil {
  181. return err
  182. }
  183. if p.isSchemaless || sf != nil {
  184. p.fields[f.Name] = sf
  185. }
  186. }
  187. default:
  188. return fmt.Errorf("unsupported field %v", field)
  189. }
  190. }
  191. p.getAllFields()
  192. return nil
  193. }
  194. func (p *DataSourcePlan) getField(name string, strict bool) (*ast.JsonStreamField, error) {
  195. if !p.isSchemaless {
  196. r, ok := p.streamFields[name]
  197. if !ok {
  198. if strict {
  199. return nil, fmt.Errorf("field %s not found in stream %s", name, p.name)
  200. }
  201. } else {
  202. return r, nil
  203. }
  204. }
  205. // always return nil for schemaless
  206. return nil, nil
  207. }
  208. // Do not prune fields now for preprocessor
  209. // TODO provide field information to the source for it to prune
  210. func (p *DataSourcePlan) getAllFields() {
  211. if !p.isWildCard {
  212. p.streamFields = p.fields
  213. }
  214. p.metaFields = make([]string, 0, len(p.metaMap))
  215. for _, v := range p.metaMap {
  216. p.metaFields = append(p.metaFields, v)
  217. }
  218. // for consistency of results for testing
  219. sort.Strings(p.metaFields)
  220. p.fields = nil
  221. p.metaMap = nil
  222. }
  223. func (p *DataSourcePlan) getProps() error {
  224. if p.iet {
  225. if p.streamStmt.Options.TIMESTAMP != "" {
  226. p.timestampField = p.streamStmt.Options.TIMESTAMP
  227. } else {
  228. return fmt.Errorf("preprocessor is set to be event time but stream option TIMESTAMP not found")
  229. }
  230. }
  231. if p.streamStmt.Options.TIMESTAMP_FORMAT != "" {
  232. p.timestampFormat = p.streamStmt.Options.TIMESTAMP_FORMAT
  233. }
  234. if strings.EqualFold(p.streamStmt.Options.FORMAT, message.FormatBinary) {
  235. p.isBinary = true
  236. }
  237. return nil
  238. }