Rebuild: first end-to-end runs; prefetch per chunk, executable rule, wire fixes

End to end on the workstation (coordinator:run, the API, edge:node, real
NNTP): a.b.e-book.french and a.b.cd.image.playstation2.dvdiso ranges.

- Prefetcher: every connection takes the next missing chunk (leased
  ranges first, in order); one connection per range (~26k lines/s) kept
  the worker waiting
- executable rule: releases of bare executables (malware floods named
  after films and software) are dropped; they were 92% of the catalog rows
  of the test ranges, 21% of the segments
- fragments per file (details once) and gzipped batch bodies; the JSON
  limit (1 MB by default) raised to 256 MB, empty strings kept (a rule
  count's key is '')
- release_chunks.batch_from is a bigint (boneless article numbers)
- tests get their own database; edge:bench --store times BatchStore:
  1.3 s for a boneless range (1.24 ms per release chunk)

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

database/migrations/1770000000500_create_release_chunks.ts

@@ -17,7 +17,7 @@ export default class extends BaseSchema {
         id bigserial,
         release_id bigint NOT NULL,
         range_id bigint NOT NULL,
-        batch_seq integer NOT NULL,
+        batch_from bigint NOT NULL,
         files integer NOT NULL,
         segments integer NOT NULL,
         payload bytea NOT NULL,

database/migrations/1770000000600_create_fragments.ts

@@ -1,9 +1,9 @@
 import { BaseSchema } from '@adonisjs/lucid/schema'
 
 /**
- * Segments of files cut off by a range's edge (REBUILD.md, fragments): kept
- * by range, merged when the neighbouring range is done, then written as a
- * chunk.
+ * Files cut off by a range's edge (REBUILD.md, fragments): one row per file
+ * and range, with that range's segments; merged when the neighbouring range
+ * is done, then written as a chunk.
  */
 export default class extends BaseSchema {
   async up() {
@@ -11,8 +11,10 @@ export default class extends BaseSchema {
       table.bigIncrements('id')
       table.bigInteger('range_id').notNullable()
       table.string('file_key', 40).notNullable()
+      // The file's first article in this range.
       table.bigInteger('article_number').notNullable()
-      table.jsonb('segment').notNullable()
+      table.integer('segments').notNullable()
+      table.jsonb('fragment').notNullable()
       table.index(['file_key'])
       table.index(['range_id'])
     })

database/schema.ts

@@ -33,18 +33,20 @@ export class AuthAccessTokenSchema extends BaseModel {
 }
 
 export class FragmentSchema extends BaseModel {
-  static $columns = ['articleNumber', 'fileKey', 'id', 'rangeId', 'segment'] as const
+  static $columns = ['articleNumber', 'fileKey', 'fragment', 'id', 'rangeId', 'segments'] as const
   $columns = FragmentSchema.$columns
   @column()
   declare articleNumber: bigint | number
   @column()
   declare fileKey: string
+  @column()
+  declare fragment: any
   @column({ isPrimary: true })
   declare id: bigint | number
   @column()
   declare rangeId: bigint | number
   @column()
-  declare segment: any
+  declare segments: number
 }
 
 export class GroupSchema extends BaseModel {
@@ -165,10 +167,10 @@ export class RangeSchema extends BaseModel {
 }
 
 export class ReleaseChunkSchema extends BaseModel {
-  static $columns = ['batchSeq', 'createdAt', 'files', 'id', 'payload', 'rangeId', 'releaseId', 'segments'] as const
+  static $columns = ['batchFrom', 'createdAt', 'files', 'id', 'payload', 'rangeId', 'releaseId', 'segments'] as const
   $columns = ReleaseChunkSchema.$columns
   @column()
-  declare batchSeq: number
+  declare batchFrom: bigint | number
   @column.dateTime({ autoCreate: true })
   declare createdAt: DateTime
   @column()