MySQL CONCAT() returns NULL if any argument is NULL, causing SHA2(CONCAT(...))
to produce NULL for any row with nullable columns. Two genuinely different
rows both hash to NULL and compare as equal — silently dropping real
data differences from the diff output.

Fix: wrap each column in IFNULL(col, '\0') inside the CAST before
concatenation, and add a NULL-presence bitmap (IF(col IS NULL, '1', '0'))
as a secondary comparison column. The WHERE clause now checks both
hash1 <> hash2 OR nullmap1 <> nullmap2, so:
- NULL vs empty string are distinguished (different bitmaps)
- NULL vs non-NULL values are caught (different hashes + bitmaps)
- Regular value changes continue to work via hash comparison

Refs: bugs.md Bug #7, PR #77, PR #63

MySQL CONCAT() returns NULL if any argument is NULL, causing SHA2(CONCAT(...))
to produce NULL for any row with nullable columns. Two genuinely different
rows both hash to NULL and compare as equal — silently dropping real
data differences from the diff output.

Fix: wrap each column in IFNULL(col, '\0') inside the CAST before
concatenation, and add a NULL-presence bitmap (IF(col IS NULL, '1', '0'))
as a secondary comparison column. The WHERE clause now checks both
hash1 <> hash2 OR nullmap1 <> nullmap2, so:
- NULL vs empty string are distinguished (different bitmaps)
- NULL vs non-NULL values are caught (different hashes + bitmaps)
- Regular value changes continue to work via hash comparison

Related contributor PRs: #77, #63

## Fix 10 bugs — data integrity, SQL generation, and robustness

This PR fixes 10 bugs across the SQL generator, data diff pipeline, and
core runtime. Each fix has independent unit tests and, where applicable,
dedicated end-to-end scenarios.

---

### Changes

#### Bug 1 — Empty constraint name generates invalid SQL
`AlterTableDropConstraintSQL::getUp()` now throws a dedicated
`InvalidConstraintException` (extending `BaseException`) instead of
emitting `DROP CONSTRAINT \`\``.
New custom exception class:
`src/Exceptions/InvalidConstraintException.php`.

#### Bug 2 & 9 — `UpdateDataSQL` crashes on `DiffOpAdd` / emits empty
string for NULL
`getDown()` crashed with a fatal error when the diff contained
`DiffOpAdd` objects (which have no `getOldValue()`). Also, both
directions wrote empty string instead of SQL `NULL` for null column
values.
Closes contributor PR #93.

#### Bug 3 — Hyphenated database names cause MySQL syntax error 1064
`LocalTableData` data-diff queries now use backtick-quoted ``
`db`.`table` `` references. MySQL was interpreting hyphens as
subtraction operators.
Closes contributor PR #92.

#### Bug 4 — `TableIterator::next()` passes Illuminate `Collection` to
`array_merge()`
`->get()` returns a `Collection`; it is now converted with `->toArray()`
and `stdClass` rows are normalised to associative arrays before being
passed to `ArrayDiff`.

#### Bug 5 — Null engine value generates `ALTER TABLE … ENGINE = ;`
`AlterTableEngineSQL::getUp()` and `getDown()` return empty string when
either engine value is empty/null. `TableSchema::getDiff()` also skips
engine diff creation in this case.

#### Bug 6 — Views appear in table diff and generate `DROP TABLE`
instead of `DROP VIEW`
`MySQLAdapter::getTables()` now uses `SHOW FULL TABLES WHERE Table_type
= 'BASE TABLE'`. PostgreSQL and SQLite adapters were already correct.
Closes contributor PR #123.

#### Bug 7 — Rows with NULL columns are silently dropped from data diff
MySQL `CONCAT()` returns `NULL` if any argument is `NULL`, collapsing
all such rows to an identical hash. Fix wraps each column in
`IFNULL(col, '\0')` and adds a NULL-presence bitmap as a secondary
comparison field.
Closes contributor PRs #77 and #63.

#### Bug 8 — INSERT statements use positional `VALUES(...)` instead of
named columns
`InsertDataSQL::getUp()` and `DeleteDataSQL::getDown()` now emit
explicit column lists: `` INSERT INTO `t` (`col1`,`col2`) VALUES(...)
``. This prevents silent data corruption when column order differs
between source and target.

#### Bug 10 — Library hardcodes `memory_limit = 512M` via `ini_set`
Both `ini_set('memory_limit', '512M')` calls removed from
`DBDiff::run()` and `DBDiff::getDiffResult()`. The CLI entry points now
set a sensible 1G default instead (see below).

---

### Tests

| Area | What was added |
|------|---------------|
| Unit | `AlterTableDropConstraintSQLTest`, `UpdateDataSQLTest`,
`MySQLDialectQuoteTest`, `ArrayDiffTest`, `AlterTableEngineSQLTest`,
`InsertDataSQLTest`, `DropTableSQLTest`, `AddTableSQLTest`,
`MemoryLimitTest` |
| E2E | `End2EndTest::testHyphenatedDatabaseNames` (Bug 3) |
| Comprehensive | `AbstractComprehensiveTest::testViewsExcludedFromDiff`
(Bug 6), `::testNullableColumnDataDetected` (Bug 7) |
| Baselines | All PostgreSQL 14–18 and SQLite comprehensive/e2e
baselines updated for Bug 8's new column-list INSERT format |

---

### Configurable memory limit

The CLI entry points (`dbdiff`, `dbdiff.php`, PHAR) now set a default
PHP memory limit of **1G** on startup. PHP's built-in default of 128M is
too low for real-world database sizes. The limit is fully configurable
at three levels (highest wins):

1. `--memory-limit=<value>` CLI flag (e.g. `--memory-limit=2G`)
2. `memory_limit: <value>` top-level key in `.dbdiff` / `dbdiff.yml`
3. `1G` hard default in the entry point scripts

Any PHP shorthand is accepted (`512M`, `1G`, `2G`, `-1` for unlimited).
The `ini_set` lives only in the CLI entry points — library consumers
embedding DBDiff via Composer are unaffected.

---

### SonarQube
- Trailing whitespace removed from `LocalTableData.php` L413–414
- `\RuntimeException` replaced with dedicated
`InvalidConstraintException` in `AlterTableDropConstraintSQL`
- `$memory_limit` renamed to `$memoryLimit` in `DefaultParams` to match
camelCase convention