oli/rdbms-playground
1
0
Fork 0
Code Issues 4 Pull Requests Actions Packages Projects Releases 1 Wiki Activity

ADR-0018 implementation: auto-fill contracts for serial and shortid

Browse Source 
Generalises serial and shortid beyond their previous restricted
forms:

- `serial` is no longer restricted to single-column PK. Non-PK
  serial columns get an emitted UNIQUE constraint and use
  application-side MAX(col)+1 at INSERT time (rowid alias still
  drives the PK case for free; per ADR-0010 worker-thread
  serialisation, the read-then-insert sequence is safe).
- `shortid` columns auto-fill existing null cells when the
  column is materialised — `add column T: x (shortid)` on a
  non-empty table no longer leaves rows in a not-really-valid
  NULL state.
- `int -> serial` joins the type-change matrix as always-clean
  identity (closes the asymmetry vs `text -> shortid`); other
  sources are refused with a route-via-int hint.
- `change column T: x (serial|shortid)` fills null source
  cells with sequence / generated values in the same rebuild
  transaction.

Internal infrastructure:

- ReadColumn gains `unique: bool`; read_schema detects single-
  column UNIQUE indexes via pragma_index_list /
  pragma_index_info; schema_to_ddl emits inline UNIQUE for
  non-PK columns.
- ColumnSchema (persistence) gains `unique: bool` so the flag
  survives YAML round-trip and rebuild-from-text reconstructs
  it faithfully — preserves the "serial -> int leaves UNIQUE
  in place" promise across save/load cycles.
- ChangeColumnTypeResult.client_side now carries `auto_filled`
  + `auto_fill_kind` alongside `transformed` + `lossy`; the
  app handler renders separate note lines when both apply.
- AddColumnResult is a new return type carrying pre-rendered
  [client-side] note lines for the auto-fill paths.

Tests: 519 -> 534 (+15). Clippy clean.
This commit is contained in:
claude@clouddev1
2026-05-08 14:32:19 +00:00
parent 7dfa718c6e
commit 5bb0a147f0
10 changed files with 1718 additions and 97 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/app.rs
+64 -24
View File
@@ -13,8 +13,8 @@ use tracing::{trace, warn};
use crate::action::Action;
use crate::db::{
CascadeEffect, ChangeColumnTypeResult, DataResult, DeleteResult, InsertResult,
TableDescription, UpdateResult,
AddColumnResult, CascadeEffect, ChangeColumnTypeResult, DataResult, DeleteResult,
InsertResult, TableDescription, UpdateResult,
};
use crate::dsl::{Command, ParseError, parse_command};
use crate::event::AppEvent;
@@ -317,6 +317,10 @@ impl App {
self.handle_dsl_change_column_success(&command, result);
Vec::new()
}
AppEvent::DslAddColumnSucceeded { command, result } => {
self.handle_dsl_add_column_success(&command, result);
Vec::new()
}
AppEvent::DslFailed { command, error } => {
self.handle_dsl_failure(&command, &error);
Vec::new()
@@ -782,6 +786,26 @@ impl App {
}
}
fn handle_dsl_add_column_success(
&mut self,
command: &Command,
result: AddColumnResult,
) {
let summary = format!("[ok] {} {}", command.verb(), command.display_subject());
self.note_system(summary);
// ADR-0018 §9: emit auto-fill note(s) before the
// structure render, so the pedagogical "the tool did
// this for you" line is in the user's eye-line next to
// the success summary.
for note in result.client_side_notes {
self.note_system(note);
}
for line in crate::output_render::render_structure(&result.description) {
self.note_system(line);
}
self.current_table = Some(result.description);
}
fn handle_dsl_change_column_success(
&mut self,
command: &Command,
@@ -790,28 +814,44 @@ impl App {
let summary = format!("[ok] {} {}", command.verb(), command.display_subject());
self.note_system(summary);
if let Some(note) = result.client_side {
// ADR-0017 §6: pedagogical hook telling the learner
// "the tool did this for you; raw SQL would need a
// CAST or application-level code." Lossy variant
// adds the lossy count when --force-conversion was
// used.
let line = if note.lossy > 0 {
format!(
"[client-side] {n} row(s) transformed; {l} of those discarded \
information (lossy). In raw SQL this would need an explicit \
`CAST` or application-level code.",
n = note.transformed,
l = note.lossy
)
} else {
format!(
"[client-side] {n} row(s) were transformed before being stored. \
In raw SQL this would need an explicit `CAST` or \
application-level code.",
n = note.transformed
)
};
self.note_system(line);
// ADR-0017 §6 + ADR-0018 §9: pedagogical hook
// telling the learner "the tool did this for you;
// raw SQL would need explicit CAST / UPDATE / etc."
//
// When both transformations and auto-fills happen
// in the same operation, both note lines are
// emitted in order (ADR-0018 §9 explicit rule).
if note.transformed > 0 {
let line = if note.lossy > 0 {
format!(
"[client-side] {n} row(s) transformed; {l} of those discarded \
information (lossy). In raw SQL this would need an explicit \
`CAST` or application-level code.",
n = note.transformed,
l = note.lossy
)
} else {
format!(
"[client-side] {n} row(s) were transformed before being stored. \
In raw SQL this would need an explicit `CAST` or \
application-level code.",
n = note.transformed
)
};
self.note_system(line);
}
if note.auto_filled > 0 {
let kind = match note.auto_fill_kind {
Some(crate::db::AutoFillKind::Serial) => "serial",
Some(crate::db::AutoFillKind::ShortId) => "shortid",
None => "auto-generated",
};
self.note_system(format!(
"[client-side] {m} null cell(s) given auto-generated {kind} values. \
In raw SQL this would need an explicit UPDATE to populate.",
m = note.auto_filled,
));
}
}
for line in crate::output_render::render_structure(&result.description) {
self.note_system(line);