mirror of
https://github.com/claunia/marechai.git
synced 2026-07-08 17:57:08 +00:00
Natalia Portillo
56ebccb010
- Implemented LinkBaseSoftwareDialog for linking orphan add-ons and DLCs to their base software. - Created SoftwareOrphanAddons page to display orphan add-ons with options to link them. - Added server-side methods in SoftwareService for fetching orphan add-ons and counting them. - Enhanced search functionality with filtering options for orphan add-ons. - Updated navigation menu to include a link to the orphan add-ons management page. - Added localization strings for new features and UI elements.
74 lines
3.2 KiB
C#
74 lines
3.2 KiB
C#
using Microsoft.EntityFrameworkCore.Migrations;
|
|
|
|
#nullable disable
|
|
|
|
namespace Marechai.Database.Migrations
|
|
{
|
|
/// <summary>
|
|
/// Create <c>NormalizeForMatch</c> — the prefix-comparison helper used by
|
|
/// the /admin/software/orphan-addons page. Equivalent to
|
|
/// <c>NormalizeForDuplicate</c> but additionally strips every
|
|
/// non-alphanumeric / non-whitespace character (including colons and
|
|
/// dashes), so <c>EyePet: Hospital</c> normalizes to <c>eyepet hospital</c>
|
|
/// and can be prefix-matched against the leading token <c>eyepet</c>.
|
|
/// </summary>
|
|
/// <remarks>
|
|
/// <para>
|
|
/// Pinned to <c>utf8mb4 / utf8mb4_general_ci</c> on the return type and
|
|
/// local variable for the same reason
|
|
/// <see cref="FixNormalizeForDuplicateCollation"/> does: MariaDB 11.5+
|
|
/// servers default the session collation to a UCA1400 variant which
|
|
/// otherwise produces <c>Illegal mix of collations</c> errors when the
|
|
/// function value is compared against parameter-bound prefixes inside a
|
|
/// <c>LIKE</c> predicate.
|
|
/// </para>
|
|
/// </remarks>
|
|
public partial class AddNormalizeForMatchFunction : Migration
|
|
{
|
|
/// <inheritdoc />
|
|
protected override void Up(MigrationBuilder migrationBuilder)
|
|
{
|
|
migrationBuilder.Sql("DROP FUNCTION IF EXISTS NormalizeForMatch");
|
|
|
|
migrationBuilder.Sql("""
|
|
CREATE FUNCTION NormalizeForMatch(val VARCHAR(255))
|
|
RETURNS VARCHAR(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci
|
|
DETERMINISTIC
|
|
NO SQL
|
|
BEGIN
|
|
DECLARE result VARCHAR(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci;
|
|
|
|
IF val IS NULL THEN
|
|
RETURN NULL;
|
|
END IF;
|
|
|
|
-- Strip every parenthesised group (e.g. "(Limited Edition)") and every
|
|
-- bracketed group (e.g. "[NTSC]"). Mirrors NormalizeForDuplicate so the
|
|
-- two functions agree on which trailing-tag content is irrelevant.
|
|
SET result = val;
|
|
SET result = REGEXP_REPLACE(result, '[[:space:]]*\\([^)]*\\)', '');
|
|
SET result = REGEXP_REPLACE(result, '[[:space:]]*\\[[^]]*]', '');
|
|
|
|
-- Replace every non-alphanumeric, non-whitespace character with a space.
|
|
-- This is the difference vs NormalizeForDuplicate: we want colons,
|
|
-- dashes, slashes, etc. treated as word separators so prefix matching
|
|
-- against the first N tokens is reliable.
|
|
SET result = REGEXP_REPLACE(result, '[^[:alnum:][:space:]]+', ' ');
|
|
|
|
-- Collapse runs of whitespace into single spaces.
|
|
SET result = REGEXP_REPLACE(result, '[[:space:]]+', ' ');
|
|
|
|
-- Trim and lowercase so the prefix LIKE comparison is case-insensitive.
|
|
RETURN LOWER(TRIM(result));
|
|
END
|
|
""");
|
|
}
|
|
|
|
/// <inheritdoc />
|
|
protected override void Down(MigrationBuilder migrationBuilder)
|
|
{
|
|
migrationBuilder.Sql("DROP FUNCTION IF EXISTS NormalizeForMatch");
|
|
}
|
|
}
|
|
}
|