rustango 0.30.20

Django-shaped batteries-included web framework for Rust: ORM + migrations + auto-admin + multi-tenancy + audit log + auth (sessions, JWT, OAuth2/OIDC, HMAC) + APIs (ViewSet, OpenAPI auto-derive, JSON:API) + jobs (in-mem + Postgres) + email + media (S3 / R2 / B2 / MinIO + presigned uploads + collections + tags) + production middleware (CSRF, CSP, rate-limiting, compression, idempotency, etc.).
Documentation
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
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
//! Query layer for rustango.
//!
//! v0.1 ships a typed `QuerySet<T>` that builds an `AND`-joined `WHERE`
//! clause and compiles to the dialect-neutral `SelectQuery` IR in
//! `rustango-core`. `UpdateBuilder<T>` mirrors the same shape for `UPDATE`,
//! and `QuerySet<T>` itself is the input to bulk delete. The dynamic
//! resolver lands in week 5.

use std::marker::PhantomData;

use crate::core::{
    AggregateExpr, AggregateQuery, Assignment, DeleteQuery, Filter, Model, ModelSchema, Op,
    OrderClause, QueryError, SelectQuery, SqlValue, TypedAssignment, TypedExpr, UpdateQuery,
    WhereExpr,
};

/// A lazy builder for a `SELECT` over `T`.
///
/// Filters are accumulated in insertion order; nothing touches the schema
/// until `compile` is called, so the builder never panics on bad input.
///
/// Two filter shapes are accepted and may be mixed freely:
/// * [`Self::filter`] / [`Self::eq`] — string-keyed, validated at
///   `compile` time.
/// * [`Self::where_`] — typed (`User::id.gt(10)`); the column is already
///   resolved, so it bypasses the schema lookup at compile time.
pub struct QuerySet<T: Model> {
    pending: Vec<PendingFilter>,
    limit: Option<i64>,
    offset: Option<i64>,
    /// FK field names registered for [`Self::select_related`] — slice
    /// 9.0d. Each name resolves to a `Join` against the FK target at
    /// `compile()` time, so the SELECT pulls the parent rows along
    /// with the children in a single SQL round trip.
    select_related: Vec<String>,
    /// `(field_name, desc)` pairs registered via [`Self::order_by`].
    /// Slice 9.0b. Resolved against the schema at `compile()` time.
    order_by: Vec<(String, bool)>,
    _model: PhantomData<fn() -> T>,
}

/// Filter accumulator entry — keeps insertion order across string-keyed and
/// typed filter calls. Each entry contributes one node to the final
/// `WhereExpr::And` clause.
enum PendingFilter {
    /// String-keyed; resolved against the schema at `compile` time.
    Raw(RawFilter),
    /// Already resolved by a typed [`Column`](crate::core::Column).
    Resolved(Filter),
    /// Typed sub-expression (built via `.and()` / `.or()` on the
    /// typed-column API). Already validated; contributes a whole
    /// sub-tree to the WHERE clause.
    Expr(WhereExpr),
}

#[derive(Debug, Clone)]
struct RawFilter {
    field: String,
    op: Op,
    value: SqlValue,
}

#[derive(Debug, Clone)]
struct RawAssignment {
    field: String,
    value: SqlValue,
}

impl<T: Model> Default for QuerySet<T> {
    fn default() -> Self {
        Self::new()
    }
}

impl<T: Model> QuerySet<T> {
    #[must_use]
    pub fn new() -> Self {
        Self {
            pending: Vec::new(),
            limit: None,
            offset: None,
            select_related: Vec::new(),
            order_by: Vec::new(),
            _model: PhantomData,
        }
    }

    /// Append `ORDER BY` columns. Slice 9.0b.
    ///
    /// Each entry is a `(field_name, desc)` pair where `field_name`
    /// is a Rust-side field on the model — schema validation runs at
    /// `compile()` time. Multiple `.order_by(...)` calls compose;
    /// subsequent calls append after earlier ones (left-to-right
    /// precedence).
    ///
    /// ```ignore
    /// let posts = Post::objects()
    ///     .order_by(&[("published_at", true)])  // newest first
    ///     .fetch_on(conn).await?;
    /// ```
    ///
    /// To sort by multiple columns:
    ///
    /// ```ignore
    /// .order_by(&[("category", false), ("published_at", true)])
    /// ```
    #[must_use]
    pub fn order_by(mut self, items: &[(&str, bool)]) -> Self {
        for (field, desc) in items {
            self.order_by.push(((*field).to_owned(), *desc));
        }
        self
    }

    /// Eagerly load a `ForeignKey<Parent>` field via a `LEFT JOIN` —
    /// Django's `select_related`. Pass the field name on `T` (not the
    /// FK column or the parent table); subsequent `fetch_on` returns
    /// rows where each `ForeignKey<Parent>` is `Loaded` after a
    /// **single** SQL query, no N+1.
    ///
    /// ```ignore
    /// let posts: Vec<Post> = Post::objects()
    ///     .select_related("author")
    ///     .fetch_on(conn).await?;
    /// // post.author is ForeignKey::Loaded { pk, value }
    /// ```
    ///
    /// Multiple `.select_related()` calls compose: each adds another
    /// `LEFT JOIN` to the same SELECT. Schema validation (the field
    /// exists, is an FK, has a primary-key target) happens at
    /// `compile()` time.
    #[must_use]
    pub fn select_related(mut self, field: impl Into<String>) -> Self {
        self.select_related.push(field.into());
        self
    }

    /// Cap the number of returned rows. `None` removes any previously set limit.
    #[must_use]
    pub fn limit(mut self, n: i64) -> Self {
        self.limit = Some(n);
        self
    }

    /// Skip the first `n` rows. Pair with [`limit`](Self::limit) for paging.
    #[must_use]
    pub fn offset(mut self, n: i64) -> Self {
        self.offset = Some(n);
        self
    }

    /// Append a `WHERE field <op> value` predicate.
    ///
    /// `field` is the Rust-side field name; the column is looked up from the
    /// schema at compile time.
    #[must_use]
    pub fn filter(mut self, field: impl Into<String>, op: Op, value: impl Into<SqlValue>) -> Self {
        self.pending.push(PendingFilter::Raw(RawFilter {
            field: field.into(),
            op,
            value: value.into(),
        }));
        self
    }

    /// Sugar for `filter(field, Op::Eq, value)`.
    #[must_use]
    pub fn eq(self, field: impl Into<String>, value: impl Into<SqlValue>) -> Self {
        self.filter(field, Op::Eq, value)
    }

    /// Append a typed predicate or boolean expression built via the
    /// [`Column`](crate::core::Column) API. Accepts either a single
    /// [`TypedFilter`](crate::core::TypedFilter) (`User::id.gt(10)`)
    /// AND-join a raw [`WhereExpr`] into the accumulated WHERE clause.
    /// Useful when the model doesn't have typed columns derived (so
    /// [`Self::where_`] isn't available) and the caller needs to
    /// express OR / NOT / nested predicates that the string-keyed
    /// [`Self::filter`] can't reach.
    ///
    /// The expression is fully validated against `T::SCHEMA` at
    /// `compile()` time — passing an unknown column or a wrong-typed
    /// value returns [`QueryError::UnknownField`] /
    /// [`QueryError::TypeMismatch`] just like the typed paths.
    #[must_use]
    pub fn where_raw(mut self, expr: WhereExpr) -> Self {
        self.pending.push(PendingFilter::Expr(expr));
        self
    }

    /// or a composed [`TypedExpr`] (`User::id.eq(1).or(User::id.eq(2))`).
    /// Every `.where_()` call AND-joins its argument into the
    /// queryset's accumulated WHERE clause.
    #[must_use]
    pub fn where_<E: Into<TypedExpr<T>>>(mut self, predicate: E) -> Self {
        let expr = predicate.into().into_expr();
        // Hoist a bare predicate into the legacy `Resolved` slot so
        // the resulting WhereExpr stays a flat AND-of-predicates for
        // simple chains — preserves the v0.6 `as_flat_and()` shape.
        match expr {
            WhereExpr::Predicate(filter) => {
                self.pending.push(PendingFilter::Resolved(filter));
            }
            other => {
                self.pending.push(PendingFilter::Expr(other));
            }
        }
        self
    }

    /// Validate the accumulated filters against `T::SCHEMA` and lower to
    /// the dialect-neutral `SelectQuery` IR.
    ///
    /// # Errors
    /// Returns [`QueryError::UnknownField`] if a filter names a field not
    /// present on the model, and [`QueryError::TypeMismatch`] if the bound
    /// value's type does not match the field's declared type.
    pub fn compile(self) -> Result<SelectQuery, QueryError> {
        let model: &'static ModelSchema = T::SCHEMA;
        let where_clause = resolve_pending(model, self.pending)?;
        let joins = lower_select_related(model, &self.select_related)?;
        let order_by = lower_order_by(model, &self.order_by)?;
        Ok(SelectQuery {
            model,
            where_clause,
            search: None,
            joins,
            order_by,
            limit: self.limit,
            offset: self.offset,
        })
    }

    /// Lower this queryset to a `DeleteQuery` — same WHERE clause, no projection.
    ///
    /// # Errors
    /// As [`QuerySet::compile`].
    pub fn compile_delete(self) -> Result<DeleteQuery, QueryError> {
        let model: &'static ModelSchema = T::SCHEMA;
        let where_clause = resolve_pending(model, self.pending)?;
        Ok(DeleteQuery {
            model,
            where_clause,
        })
    }

    /// Start an `UpdateBuilder` carrying this queryset's filters as the WHERE clause.
    #[must_use]
    pub fn update(self) -> UpdateBuilder<T> {
        UpdateBuilder {
            qs: self,
            set: Vec::new(),
        }
    }

    /// Start an [`AggregateBuilder`] carrying this queryset's filters as the
    /// WHERE clause. Chain `.group_by`, `.annotate`, `.having`, `.order_by`,
    /// `.limit`, `.offset` then call `.compile()` to get an [`AggregateQuery`].
    #[must_use]
    pub fn aggregate(self) -> AggregateBuilder<T> {
        AggregateBuilder {
            qs: self,
            group_by: Vec::new(),
            aggregates: Vec::new(),
            having: None,
            order_by: Vec::new(),
            limit: None,
            offset: None,
        }
    }
}

/// Accumulates `SET column = value` assignments, then compiles to an `UpdateQuery`.
///
/// Constructed via [`QuerySet::update`]. The queryset's filters become the
/// WHERE clause; an empty queryset produces an unfiltered update affecting
/// every row.
pub struct UpdateBuilder<T: Model> {
    qs: QuerySet<T>,
    set: Vec<PendingAssignment>,
}

enum PendingAssignment {
    Raw(RawAssignment),
    Resolved(Assignment),
}

impl<T: Model> UpdateBuilder<T> {
    /// Append a `SET field = value` assignment. Last write wins for repeated fields.
    #[must_use]
    pub fn set(mut self, field: impl Into<String>, value: impl Into<SqlValue>) -> Self {
        self.set.push(PendingAssignment::Raw(RawAssignment {
            field: field.into(),
            value: value.into(),
        }));
        self
    }

    /// Append a typed `SET column = value` from a [`Column`](crate::core::Column).
    #[must_use]
    pub fn set_typed(mut self, assignment: TypedAssignment<T>) -> Self {
        self.set
            .push(PendingAssignment::Resolved(assignment.into_assignment()));
        self
    }

    /// Validate against `T::SCHEMA` and lower to an `UpdateQuery`.
    ///
    /// # Errors
    /// Returns [`QueryError::UnknownField`] if any `set` or filter names an
    /// unknown field, and [`QueryError::TypeMismatch`] if any bound value's
    /// type doesn't match the field's declared type.
    pub fn compile(self) -> Result<UpdateQuery, QueryError> {
        let model: &'static ModelSchema = T::SCHEMA;

        let assignments = self
            .set
            .into_iter()
            .map(|p| match p {
                PendingAssignment::Raw(raw) => resolve_assignment(model, raw),
                PendingAssignment::Resolved(assignment) => Ok(assignment),
            })
            .collect::<Result<Vec<_>, _>>()?;

        let where_clause = resolve_pending(model, self.qs.pending)?;

        Ok(UpdateQuery {
            model,
            set: assignments,
            where_clause,
        })
    }
}

/// Convert `(field_name, desc)` pairs into `OrderClause`s by
/// resolving each field name on the schema. Slice 9.0b.
fn lower_order_by(
    model: &'static ModelSchema,
    items: &[(String, bool)],
) -> Result<Vec<crate::core::OrderClause>, QueryError> {
    let mut out = Vec::with_capacity(items.len());
    for (field_name, desc) in items {
        let field = model
            .field(field_name)
            .ok_or_else(|| QueryError::UnknownField {
                model: model.name,
                field: field_name.clone(),
            })?;
        out.push(crate::core::OrderClause {
            column: field.column,
            desc: *desc,
        });
    }
    Ok(out)
}

/// Convert `select_related` field names into `Join`s — slice 9.0d.
///
/// For each name: look up the field on `model`, verify it's a
/// `Relation::Fk`, find the target schema in inventory, build a
/// `Join` projecting all of the target's columns. Errors out on any
/// unresolvable name with a clear `SelectRelatedInvalid` reason.
fn lower_select_related(
    model: &'static ModelSchema,
    names: &[String],
) -> Result<Vec<crate::core::Join>, QueryError> {
    use crate::core::{inventory, Join, ModelEntry, Relation};
    let mut out: Vec<Join> = Vec::with_capacity(names.len());
    for name in names {
        let field = model
            .field(name)
            .ok_or_else(|| QueryError::SelectRelatedInvalid {
                model: model.name,
                field: name.clone(),
                reason: format!("no field `{name}` on this model"),
            })?;
        let (to, on) = match field.relation {
            Some(Relation::Fk { to, on }) | Some(Relation::O2O { to, on }) => (to, on),
            _ => {
                return Err(QueryError::SelectRelatedInvalid {
                    model: model.name,
                    field: name.clone(),
                    reason: "not a `ForeignKey<T>` field".into(),
                });
            }
        };
        let target = inventory::iter::<ModelEntry>
            .into_iter()
            .find(|e| e.schema.table == to)
            .map(|e| e.schema)
            .ok_or_else(|| QueryError::SelectRelatedInvalid {
                model: model.name,
                field: name.clone(),
                reason: format!(
                    "target table `{to}` is not registered (is the parent's `#[derive(Model)]` linked into the binary?)"
                ),
            })?;
        // Project every column on the target so the decoder has the
        // full row to rebuild a `Target` instance.
        let project: Vec<&'static str> = target.scalar_fields().map(|f| f.column).collect();
        out.push(Join {
            target,
            on_local: field.column,
            on_remote: on,
            // The Rust field name is unique within the model, so it
            // makes a clean alias prefix that doesn't collide with
            // other JOINs the writer or admin might add later.
            alias: field.name,
            project,
        });
    }
    Ok(out)
}

fn resolve_pending(
    model: &'static ModelSchema,
    pending: Vec<PendingFilter>,
) -> Result<WhereExpr, QueryError> {
    let mut nodes: Vec<WhereExpr> = Vec::with_capacity(pending.len());
    for entry in pending {
        match entry {
            PendingFilter::Raw(raw) => {
                nodes.push(WhereExpr::Predicate(resolve_filter(model, raw)?));
            }
            PendingFilter::Resolved(filter) => {
                nodes.push(WhereExpr::Predicate(filter));
            }
            PendingFilter::Expr(expr) => {
                nodes.push(expr);
            }
        }
    }
    Ok(WhereExpr::And(nodes))
}

fn resolve_filter(model: &'static ModelSchema, raw: RawFilter) -> Result<Filter, QueryError> {
    let field = model
        .field(&raw.field)
        .ok_or_else(|| QueryError::UnknownField {
            model: model.name,
            field: raw.field.clone(),
        })?;

    // `IsNull` carries a Bool sentinel (true = IS NULL, false = IS NOT NULL),
    // not a value to compare against the field — skip the type check.
    // `In` carries a List; element-by-element checking is a follow-up.
    let skip_type_check = matches!(raw.op, Op::IsNull | Op::In);

    if !skip_type_check {
        if let Some(value_ty) = raw.value.field_type() {
            if value_ty != field.ty {
                return Err(QueryError::TypeMismatch {
                    model: model.name,
                    field: raw.field,
                    expected: field.ty,
                    actual: value_ty,
                });
            }
        }
    }

    Ok(Filter {
        column: field.column,
        op: raw.op,
        value: raw.value,
    })
}

fn resolve_assignment(
    model: &'static ModelSchema,
    raw: RawAssignment,
) -> Result<Assignment, QueryError> {
    let field = model
        .field(&raw.field)
        .ok_or_else(|| QueryError::UnknownField {
            model: model.name,
            field: raw.field.clone(),
        })?;

    if let Some(value_ty) = raw.value.field_type() {
        if value_ty != field.ty {
            return Err(QueryError::TypeMismatch {
                model: model.name,
                field: raw.field,
                expected: field.ty,
                actual: value_ty,
            });
        }
    }

    Ok(Assignment {
        column: field.column,
        value: raw.value,
    })
}

// ------------------------------------------------------------------ AggregateBuilder

/// Fluent builder for [`AggregateQuery`]. Constructed via [`QuerySet::aggregate`].
pub struct AggregateBuilder<T: Model> {
    qs: QuerySet<T>,
    group_by: Vec<&'static str>,
    aggregates: Vec<(&'static str, AggregateExpr)>,
    having: Option<WhereExpr>,
    order_by: Vec<(&'static str, bool)>,
    limit: Option<i64>,
    offset: Option<i64>,
}

impl<T: Model> AggregateBuilder<T> {
    /// Add a `GROUP BY` column. Call multiple times to group by multiple columns.
    #[must_use]
    pub fn group_by(mut self, column: &'static str) -> Self {
        self.group_by.push(column);
        self
    }

    /// Add an aggregate expression under `alias` (e.g. `"post_count"`).
    #[must_use]
    pub fn annotate(mut self, alias: &'static str, expr: AggregateExpr) -> Self {
        self.aggregates.push((alias, expr));
        self
    }

    /// Add a `HAVING` predicate. Multiple calls AND-join.
    #[must_use]
    pub fn having<E: Into<crate::core::TypedExpr<T>>>(mut self, predicate: E) -> Self {
        let expr = predicate.into().into_expr();
        match self.having {
            None => self.having = Some(expr),
            Some(ref mut existing) => existing.push_and(expr),
        }
        self
    }

    /// Add `ORDER BY` columns. `desc = true` → DESC.
    #[must_use]
    pub fn order_by(mut self, items: &[(&'static str, bool)]) -> Self {
        self.order_by.extend_from_slice(items);
        self
    }

    /// Set `LIMIT`.
    #[must_use]
    pub fn limit(mut self, n: i64) -> Self {
        self.limit = Some(n);
        self
    }

    /// Set `OFFSET`.
    #[must_use]
    pub fn offset(mut self, n: i64) -> Self {
        self.offset = Some(n);
        self
    }

    /// Compile to an [`AggregateQuery`] IR.
    ///
    /// # Errors
    /// Returns [`QueryError`] if any filter or having clause names an unknown field.
    pub fn compile(self) -> Result<AggregateQuery, QueryError> {
        let model = T::SCHEMA;
        let where_clause = resolve_pending(model, self.qs.pending)?;
        let order_by = self
            .order_by
            .into_iter()
            .map(|(col, desc)| OrderClause { column: col, desc })
            .collect();
        Ok(AggregateQuery {
            model,
            where_clause,
            group_by: self.group_by,
            aggregates: self.aggregates,
            having: self.having,
            order_by,
            limit: self.limit,
            offset: self.offset,
        })
    }
}