|
1 .. -*- coding: utf-8 -*- |
|
2 |
|
3 RQL language (Relation Query Language) |
|
4 ====================================== |
|
5 |
|
6 XXX see also RQL documentation in source rql/doc. |
|
7 |
|
8 |
|
9 Introduction |
|
10 ------------ |
|
11 * RQL language focuses on browsing relations. |
|
12 * Attributes are considered as particular relations. |
|
13 * RQL is inspired from SQL but is a high level language. |
|
14 * A good knowledge of Erudi's schemas defining the application is required. |
|
15 |
|
16 |
|
17 Types of requests |
|
18 ----------------- |
|
19 |
|
20 Search (`Any`) |
|
21 query the repository to extract entities and/or attributes. |
|
22 |
|
23 Insertion (`INSERT`) |
|
24 insert new entities in the database. |
|
25 |
|
26 Updates of entities, creation of relations (`SET`) |
|
27 update existing entities in the database, or create relations between existing |
|
28 entities |
|
29 |
|
30 Deletion of entities or relations (`DELETE`) |
|
31 delete existing entities and relations from the database. |
|
32 |
|
33 |
|
34 Variables and typing |
|
35 -------------------- |
|
36 |
|
37 Entities and values to browse and/or select are set in the query through *variables* |
|
38 which should be written in capital letters. |
|
39 |
|
40 The possible types for each variable can be deducted from the schema depending on |
|
41 the conditions expressed in the query. |
|
42 |
|
43 You can force the possible types for a variable thanks to the special relation `is`. |
|
44 |
|
45 |
|
46 |
|
47 Built-in types |
|
48 -------------- |
|
49 * `String` (literal: between double or single quotes). |
|
50 * `Int`, `Float` (separator is '.'). |
|
51 * `Date`, `Datetime`, `Time` (literal: pattern YYYY/MM/DD[ hh:mm] or keywords |
|
52 `TODAY` and `NOW`). |
|
53 * `Boolean` (keywords `TRUE` et `FALSE`). |
|
54 * keyword `NULL`. |
|
55 |
|
56 Operators |
|
57 ---------- |
|
58 * Logical operators: `AND`, `OR`, `,`. |
|
59 * Mathematical operators: `+`, `-`, `*`, `/`. |
|
60 * Comparison operators: `=`, `<`, `<=`, `>=`, `>`, `~=`, `LIKE`, `IN`. |
|
61 |
|
62 * The operator `=` is the default operator. |
|
63 |
|
64 * The operator `LIKE` / `~=` allows the use of the character `%` in a string |
|
65 to indicate that the string should start/end with a prefix/suffix:: |
|
66 |
|
67 Any X WHERE X nom ~= 'Th%' |
|
68 Any X WHERE X nom LIKE '%lt' |
|
69 |
|
70 * The operator `IN` allows to provide a list of possible values:: |
|
71 |
|
72 Any X WHERE X nom IN ('chauvat', 'fayolle', 'di mascio', 'thenault') |
|
73 |
|
74 Search query |
|
75 ------------ |
|
76 |
|
77 [`DISTINCT`] <entity type> V1(, V2)\* |
|
78 [`GROUPBY` V1(, V2)\*] [`ORDERBY` <orderterms>] |
|
79 [`WHERE` <condition>] |
|
80 [`LIMIT` <value>] [`OFFSET` <value>] |
|
81 |
|
82 :entity type: |
|
83 Type of the selected variable |
|
84 Special type `Any` is equivalent to not specify a type |
|
85 :condition: |
|
86 list of relations to browse following the pattern |
|
87 `V1 relation V2|<static value>` |
|
88 :orderterms: |
|
89 Setting of the selection order : variable or column number followed by the |
|
90 sorting method (`ASC`, `DESC`), ASC being the default value. |
|
91 :note for grouped queries: |
|
92 For grouped queries (e.g. using function `GROUPBY`), all the selected |
|
93 variables must be grouped or aggregated. |
|
94 |
|
95 Examples - search |
|
96 ~~~~~~~~~~~~~~~~~ |
|
97 :: |
|
98 |
|
99 Any X WHERE X eid 53 |
|
100 Personne X |
|
101 Personne X WHERE X travaille_pour S, S nom "logilab" |
|
102 Any E,COUNT(X) GROUPBY E ORDERBY EN WHERE X is E, E name EN |
|
103 Any E,COUNT(X) GROUPBY E ORDERBY 2 WHERE X is E |
|
104 |
|
105 |
|
106 Advanced features |
|
107 ~~~~~~~~~~~~~~~~~ |
|
108 * Aggregate functions: `COUNT`, `MIN`, `MAX`, `SUM`. |
|
109 * String functions:`UPPER`, `LOWER`. |
|
110 * Optional relations: |
|
111 |
|
112 * They allow to select entities related to others or not. |
|
113 |
|
114 * You should use `?` behind the variable to specify the relation to itself is |
|
115 optional. |
|
116 |
|
117 - Project anomalies related to a version or not:: |
|
118 |
|
119 Any X,V WHERE X concerns P, P eid 42, X corrected_in V? |
|
120 |
|
121 - All the cards and the project they document otherwise :: |
|
122 |
|
123 Any C,P WHERE C is Card, P? documented_by C |
|
124 |
|
125 Negation |
|
126 ~~~~~~~~ |
|
127 * A query such as `Document X WHERE NOT X owned_by U` is equivalent to |
|
128 "the documents which do not have relation `owned_by`". |
|
129 * Whereas the query `Document X WHERE NOT X owned_by U, U login "syt"` |
|
130 is equivalent to "the documents which do not have relation `owned_by` |
|
131 with the user syt". They could have a relation with other users. |
|
132 |
|
133 |
|
134 Identity |
|
135 ~~~~~~~~ |
|
136 |
|
137 We could use the special relation `identity` in a query in order to add a |
|
138 condition of identity between two variables. This is equivalent to ``is`` |
|
139 in Python. |
|
140 |
|
141 Any A WHERE A comments B, A identity B |
|
142 |
|
143 returns the set of objects which comment themselves. The relation `identity` |
|
144 is very usefull while defining security rules with `RQLExpressions`. |
|
145 |
|
146 Insertion queries |
|
147 ----------------- |
|
148 `INSERT` <entity type> V1(, <entity type> V2)\* `:` <assignments> |
|
149 [`WHERE` <condition>] |
|
150 |
|
151 :assignments: |
|
152 list of relations to assign such as `V1 relation V2|<static value>` |
|
153 |
|
154 The condition allow to define the variables we would use in assignments. |
|
155 |
|
156 Be careful, if a condition is specified, the insertion is done *for each result |
|
157 returned by the condition*. |
|
158 |
|
159 Examples - insertion |
|
160 ~~~~~~~~~~~~~~~~~~~~~ |
|
161 * Insertion of a new person named 'bidule':: |
|
162 |
|
163 INSERT Person X: X name 'bidule' |
|
164 |
|
165 * Insertion of a new person named 'bidule', another named |
|
166 'chouette' and a relation 'friend' between them:: |
|
167 |
|
168 INSERT Person X, Person Y: X name 'bidule', Y name 'chouette', X friend Y |
|
169 |
|
170 * Insertion of a new person named 'bidule' and a relation 'friend'with an |
|
171 existing person 'chouette':: |
|
172 |
|
173 INSERT Person X: X name 'bidule', X friend Y WHERE Y name 'chouette' |
|
174 |
|
175 |
|
176 Update queries |
|
177 -------------- |
|
178 `SET` <assignments> |
|
179 [`WHERE` <condition>] |
|
180 |
|
181 Be careful, if a condition is specified, the update is done *for each result |
|
182 returned by the condition*. |
|
183 |
|
184 Examples - update |
|
185 ~~~~~~~~~~~~~~~~~ |
|
186 * Renaming of the person named 'bidule' to 'toto', with change on the first name:: |
|
187 |
|
188 SET X name 'toto', X firstname 'original' WHERE X is 'Person', X name 'bidule' |
|
189 |
|
190 * Insertion of a relation of type 'know' between two objects linked with the relation |
|
191 of type 'friend' :: |
|
192 |
|
193 SET X know Y WHERE X friend Y |
|
194 |
|
195 Deletion queries |
|
196 ---------------- |
|
197 `DELETE` (<entity type> V) | (V1 relation v2),... |
|
198 [`WHERE` <condition>] |
|
199 |
|
200 |
|
201 Be careful, if a condition is specified, the deletion is done *for each result |
|
202 returned by the condition*. |
|
203 |
|
204 |
|
205 Examples |
|
206 ~~~~~~~~ |
|
207 * Deletion of the person named 'toto':: |
|
208 |
|
209 DELETE Person X WHERE X name 'toto' |
|
210 |
|
211 * Deletion of all the relations of type 'friend' linked to the person named |
|
212 'toto':: |
|
213 |
|
214 DELETE X friend Y WHERE X is 'Person', X name 'toto' |
|
215 |
|
216 |