The top-bar theme switch was bound twice — core/theme.js attached a click
handler to #theme-toggle-desktop while runtime/main.js delegated click on
.toggle-theme, and both ship in main.min.js, so one press toggled twice and
the button looked dead.
Also realign the control with the reference:
- Drop the far modifier from fa-toggle-theme so the sun/moon glyphs resolve
against the Solid face at the reference weight instead of Regular.
- Restore .toggle-radius .fa/.icon to the reference form (left/right 0 +
auto margin only). The extra position/top/transform/size/flex block and
the > i selector were overriding the shared abs-* centering rule.
- Replace the per-ID bindings with one delegated listener so the floating
rail, which carries no id, keeps working.
- Remove dead theme.js queries for selectors no markup or stylesheet ships
(#theme-toggle, #dropdown-theme-toggle, .theme-icon-moon/.theme-icon-sun,
[data-action='toggle-theme'], .is-dark).
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The theme was renamed Zibll/子比 → AKSRT → Jio, but upstream branding
survived the sweeps in CSS classes, the SVG sprite prefix, a persisted
widget field ID, JS events, admin Vue class names, test infrastructure
and docblocks.
Renames:
- zib-widget / zib-widget-wrap / zib-slider → jio-widget-box /
jio-widget-box-wrap / jio-slider-box (jio-widget is taken by the modern
widget shell, jio-box-* by the card primitives)
- zib-partners-scroll + zps-* → jio-partners-scroll + jps-*
- zib-combo-info-card + zci-* → jio-combo-info-card + jci-*
- zibsvg- → jiosvg-, including the exported picker helpers and the
fa-zibsvg- family-prefix guard
- zibpay_type → pay_type; the dual-read fallback in PostQuery collapses
to a single key
- [zibajax="submit"] dropped (no PHP emitter); zib_ajax.success →
jio_ajax.success
- aks-* admin Vue classes → jio-*
- zib_palette() → palette()
User-facing vendor traces are gone: the floating "more" button default no
longer ships a zibll.com link or 子比主题 marketing copy, the notice
widget ships no preset promo entries, and the four 官网教程 links are
removed from the post-list admin schema.
Test infrastructure renamed to neutral names (docker-compose.reference.yml,
Caddyfile, article-baseline.json, modern-widgets-contract.test.cjs) with
their DB/volume/env/slug identifiers following.
Kept intentionally, documented in tests/fixtures/README.md: the golden
fixture reference fingerprint (a SHA-256 record of an external directory,
read by validate-widget-golden / generate-widget-golden-fixture /
audit-widget-system), the doesNotMatch guards that prove the old tokens are
gone, and the reference theme's own meta key in the visual fixtures.
No backward-compat or data migration: saved widget configs holding old
icon keys or the old field ID silently lose them, per the project's
rewrite-over-compatibility rule.
Verified: 132/132 node contract tests, php tests/widget-system/run.php
(146 assertions), audit-admin-contracts, audit-psr4-paths and
validate-widget-golden all green; tracked bundles regenerated and in sync.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Align homepage and category discovery, modern widgets, admin controls, quotes, and comments with the reference theme. Remove superseded paths and add rendering and behavior coverage. Update stale asset and finance contracts; all 760 Node tests and 139 widget assertions pass.
- Render pending orders (max 5, >=3 min left) in a top-right float button
with hover list and countdown, gated by pay_wait_float_s
- Fix order countdown parsing expires_at as UTC
- Drop unused .paid-modal-lists-box thumb rule
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The old 售后 flow was an order-refund appendage: tickets were hard-bound to an
order number, users could only file them from 「我的订单」against a paid order,
and the admin action bottomed out at 「标记订单已退款」. It could not carry
account questions or feature requests that have nothing to do with an order.
Replace it with a general ticket system modelled on 魔方财务云's idcsmart_ticket:
title + department + description are required, linking an order is optional
(user's own paid orders only). Add departments, custom statuses, preset replies
and internal admin notes as four configuration surfaces. Conversation history
moves off the ticket row's JSON log into dedicated ticket_reply rows, which
unlocks quoted replies. Front end gets a 工单中心 panel under 我的服务; back end
moves from 支付中心 → 售后服务 to users.php → 工单管理.
Per CLAUDE.md's rewrite-over-compatibility rule: drop the old table, build new
ones, delete the old code. No dual paths, no data migration. The upstream
plugin's upstream/downstream delivery and internal-ticket conversion are
deliberately left out — both are multi-provider platform concerns with no
meaning on a single WordPress site.
Data: order_aftersale becomes six aksrt_ticket* tables (ticket, ticket_reply,
ticket_status, ticket_type, ticket_prereply, ticket_note). SCHEMA_VERSION
2.6.0 -> 3.0.0. Status and department seeds are idempotent; the default status
「待接单」is flagged as non-editable.
Domain: new src/Ticket/ (13 files, all under 460 lines). Attachments reuse
UserController::rest_upload_comment_image with a ticket-specific permission
callback, mirroring the existing /report/upload-image route rather than
growing a second upload path.
Notifications go through AKSRT\Messages\Message::send() and deep-link to the
right surface; email/SMS stay out of scope since the theme's mail and SMS
stacks are separate implementations.
Refund gate opens up: marking an order refunded no longer requires an in-flight
ticket, only that the order is paid.
Also generalise a hardcoded special case in tools/audit-admin-contracts.php —
schema domain inference now reads the parent directory instead of matching
"PostList" by name, so new leaf schema files need no per-name exception.
Verification:
php tools/audit-admin-contracts.php exit 0, errors=0
node --test (widget-system + admin) 664 pass / 0 fail
php -l over all new/changed files clean
node --check over all JS clean
npm run build + build:public clean
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The top-nav guest popover rendered its quick-login buttons as 29px circles
with zero gap: github, gitee and qq were touching. The user-card widget, which
uses the same .social_loginbar markup, was fine at 32px with 10px gaps.
header-userbox.css carries an AKSRT-specific rule, absent from the reference
theme, that gives every .toggle-radius inside .navbar-form a 29px midline
circle:
.navbar-top .navbar-form .toggle-radius { flex: 0 0 29px; margin: 0; }
It exists for the account-row .badg chips, the theme toggle and the language
toggle. The popover social buttons are .toggle-radius too and also sit inside
.navbar-form, so they were caught by it. Two independent mechanisms broke the
geometry: flex-basis outranks width in a flex container, so flex: 0 0 29px beat
the .social_loginbar .social-login-item width: 32px !important; and the rule
(0,3,0) specificity outranked the (0,2,0) margin: 3px 5px.
Exempt the social buttons with :not(). The chips, the theme toggle and the
language toggle still match, and the popover buttons fall back to the
.social_loginbar geometry ported 1:1 from the reference.
Verified on the live site by measuring computed styles in headless Edge: the
popover buttons go from 29px/0px margins/0px gap to 32px/5px margins/10px gap,
the user-card widget is unchanged, and the account-row chips keep 29px (theme
toggle) and 36px (.badg). Rebuilt the public bundle so the exemption survives
the build, and contract-locked the selector plus the bundle copy.
Verification: node --test across widget-system, navigation, admin,
asset-ledger, dev, front-submit and visual (615 tests).
sidebar.css paints .widget a with --color-ink-2 (specificity 0,1,1), which
out-specifies the .jb-* color group (0,1,0) the badge relies on for its white
ink. The user-card widget carries the .widget class, so the check-in bookmark
rendered dark-gray text on the orange gradient on widget pages. zibll sets no
link color on zib-widget anchors at all.
sidebar.css already restores .but colors for exactly this reason; extend that
restoration to .img-badge so the badge keeps its own --this-color. Rebuild the
homepage bundle and extend the badge contract test.
The legacy .jb-* compat layer in core/buttons.css restated .jb-yellow as #f99d4d -> #f7631d, and because the public bundle loads after main.min.css that value overrode the reference-accurate #f59f54 -> #ff6922 gradient, so the check-in bookmark (and every other .jb-yellow surface) painted the wrong orange. Correct the gradient and rebuild the public bundle.
Also drop the dead is-done class and the attribute-less span wrapper from the signed card badge, which the reference theme does not have. The header dropdown variant keeps its class, locked by the navigation contract.
The top-nav account row carried min-height: 50px on top of the 8px
margin-top that Bootstrap base .navbar-form rule keeps at >=768px (the
override only zeroes the horizontal margins). The 8px was counted twice,
so the row ran from 8px to 58px: its icon tiles landed 8px below the 25px
midline that the 50px menu row and the logo define, and the navbar grew
from 50px to 58px, which also broke the body.nav-fixed offset.
Drop the min-height so the row takes its natural 34px height from the
icon tiles, and give the theme/language group the same 34px cross-size so
its 29px toggle shares that midline instead of sitting 2.5px above it.
Measured against the reference theme, the navbar goes from 66px to
50.05px (reference 50.28px) and the logo, menu, publish chip, VIP chip,
theme toggle and account icons all centre on 25px.
The desktop navbar stays visible below 991px (only the menu folds into a
dropdown there), so the same fix applies at the mobile and tablet
breakpoints: the top bar drops from 68px to 52.05px.
Verified: node --test across tests/ (603 passing), and a stubbed navbar
render measured in a headless browser against the reference theme
stylesheet.
The 开通会员 chip rendered in the .jb-red orange gradient instead of the
legacy metallic .payvip-icon gradient. The markup always carried
payvip-icon, but core/buttons.css restates the legacy .but block and is
bundled into public.css, which loads after main.css. Same specificity,
later in the cascade, so its background / color / border / text-shadow
declarations outrank .payvip-icon and replace the payvip surface with
whatever .jb-* skin happens to be on the element.
The reference theme has no such compatibility layer, so there its
.payvip-icon declarations are the last word and the chip shows the
metallic gradient with #ece0e0 text, no border and no text shadow.
A previous fix restored the surface only for the user-card CTA, scoped to
.aksrt-header-userbox .pay-vip.but.payvip-icon, which left the top-nav
chip (and the guest CTA, and the user-card status bar) on the wrong
gradient. Re-assert the legacy .payvip-icon declarations unscoped in
header-userbox.css, which loads after core/buttons.css, and complete the
property set with border: none and text-shadow: none so the chip no
longer carries the 1px transparent border the .but block adds.
Verified with a cascade resolver that replays the real stylesheet load
order (bootstrap, main, public, header-userbox) against the reference
theme's own order: background, color, --this-color, border, text-shadow,
border-radius, padding, margin, font-size, display, overflow and position
now resolve identically for both the publish and the VIP chip.
The VIP chip never received the reference .but geometry. Commit f03d72c8
narrowed the legacy selector from a descendant to a child combinator as a
side effect of an unrelated user-popover fix, and the contract test then
pinned the narrowed form. Because the VIP anchor lives inside the
hover-show wrapper rather than being a direct child of .navbar-but, it
fell back to the bare .but defaults: 15px type instead of 0.8em, no
0.4em margin, and no padding override.
Restore the descendant selector so both chips share one geometry:
- .navbar-top .navbar-but .but is byte-identical to the reference again,
giving the VIP chip font-size 0.8em, padding 0.3em 1em !important and
margin 0.4em
- the publish and VIP anchors are plain inline-block .but elements again
(the merged flex container no longer blockifies them)
- ml10 is back on the VIP anchor; .navbar-top .navbar-but .but overrides
it exactly as it does in the reference, so it stays markup parity
without changing the computed margin
The right cluster was also merged into one .aksrt-navbar-actions flex
container with a 10px gap and zero padding, while the reference emits
three floated .navbar-form.navbar-right groups whose 15px padding is what
produces the inter-cluster spacing. Split them again: publish/VIP in
.navbar-but, theme toggle plus language switcher in a flex ac group, and
the account row in its own group. Measured against the reference, the
publish-to-VIP gap was 19.6px instead of 9.6px, and VIP-to-theme-toggle
was 14.8px instead of 34.8px.
The container rules in header-userbox.css are re-anchored onto the
reference structure: .aksrt-navbar-account owns the only remaining flex
context, .toggle-radius and .aksrt-navbar-language are scoped through
.navbar-form, and the hover-show rule targets .navbar-but. Coverage of
those selectors is unchanged, so the theme toggle, language switcher and
message icon keep their current geometry.
The .aksrt-navbar-actions wrapper and its gap are gone; a negative
assertion now guards against reintroducing them.
Verified: node --test across tests/ (603 passing), php -l on the changed
PHP files, and a stubbed render of aksrt_menu_button() for guest,
logged-in and layout 2.
The top-nav publish chip and open-membership chip carried an AKSRT-only
class layer (navbar-cta / navbar-cta-text / navbar-newposts /
navbar-vip-hover) plus a matching set of CSS overrides that replaced the
legacy .but geometry: margin 0.4em -> 0, an extra min-height and flex gap,
an FA5 pencil glyph that is not the reference pencil, and label text
wrapped in a span. The VIP hover card also had its own positioning
override instead of the reference static position.
Drop that layer and emit the reference class combination verbatim so the
already byte-identical legacy stylesheet owns the geometry again:
- publish chip renders newadd-btns but nowave <skin> [radius] btn-newadd
with the label echoed as saved (no injected icon, no span)
- VIP chip renders pay-vip|signin-loader but jb-red radius payvip-icon
with the vip_1 SVG at em12 mr3, inside span.hover-show.inline-block
- user-card CTA uses the vip_1 SVG (em12 mr6) instead of fa-gem and drops
vip-level, matching zib_get_header_payvip_icon()
- both buttons now honour Settings::close_sign(), matching the reference
zib_is_close_sign() gate and the floating publish button
- header-userbox.css keeps one rule that restores the hover wrapper to a
plain block container so .sub-vip-card takes its reference static
position; every other override is gone
- nav_newposts_text default drops the duplicated fas and uses the FA5
spelling of the reference pencil (fa-pencil-alt, not fa-pen)
Dead code removed: aksrt_normalize_header_cta_label(), the unreachable
multi-type dropdown branch of aksrt_render_new_add_btns() (only 'post' is
supported and only ['post'] is ever passed), and the
newadd-btns__toggle / newadd-btns__item / drop-newadd markup it emitted.
ml10 is no longer set on the VIP anchor: .aksrt-navbar-actions already
supplies the 10px gap the reference gets from ml10, and keeping both
would double it to 20px.
No migration for sites that already stored the old nav_newposts_text.
Verified: node --test across tests/ (603 passing), php -l on the changed
PHP files, and a stubbed render of every publish/VIP branch.
Make --radius-avatar the single source of truth for frontend user avatar shape. A new primitive in components-ext.css (img.avatar / avatar-mini / avatar-md / avatar-lg / .avatar-img img) forces a circle on every avatar image, which covers the bare avatar-mini in the ban and report modals, the classless img inside the VIP modal's .avatar-img, and the post-card meta avatar that previously depended on post-card.css being enqueued. The user center header frame, the public user home hero and the friends page avatar switch to the same token because route CSS loads after the public bundle and would otherwise override the primitive.
The VIP icon no longer rides the avatar corner. The user card widget, the after-article author card and the sign-in modal stop opting into aksrt_get_avatar_box()'s corner badge, and the icon renders first in the badge/level row instead, since aksrt_get_user_name already orders name -> VIP -> auth -> medal -> level. The sign-in modal appends it after the display name and keeps the legacy precedence where a ban badge takes the VIP slot. The user card widget name row now shrinks its display name so the extra icon cannot overflow a narrow sidebar.
Verified with the new frontend-avatar-shape-contract test plus post-visual, user-card-runtime, circular-button-icon-margin, user-features and page-display contracts, and php -l on every changed PHP file. The public bundle was rebuilt with tools/build-public-assets.mjs.
Clicking 加载更多 in the 多栏目文章 widget showed no skeleton at all — the
list simply stayed still until the response arrived, then the new cards
appeared. The homepage stream, archives and comments already had a working
skeleton; the widget's AJAX load goes through a different client and never
found its loader.
Root cause was DOM placement, not visibility: ContentRenderer emitted
.post_ajax_loader as a *sibling* of .widget-ajaxpager, while the widget REST
client resolves it with container.querySelector('.post_ajax_loader'). The
node existed but could never match, so .is-loading was never added and the
skeleton stayed display:none. Inactive tab panes had the same shape, and the
direct-load branch of ajaxpager() emitted no loader at all.
Changes, all following the reference theme's post_ajax() contract in
zibll/js/main.js (beforeSend anchors the loader to the end of the list;
success only hides it, never removes it):
- ContentRenderer: nest the loader inside .widget-ajaxpager for both the
direct-load and inactive-tab branches, and drop the sibling append.
- widgets/frontend.js: anchorLoader() moves the loader to the end of the
list before .ajax-pag on every request; appended cards are inserted before
it; a replace preserves the node instead of wiping it with innerHTML; a
fallback loader is created when the DOM has none; loadItems() now returns
its promise; drop the unreachable is-loading auto-load branch.
- Message centre, user-home panels and the search page switch from spinner /
opacity feedback to the same .posts-item-shaped skeleton, so all AJAX
surfaces share one loading contract.
- Remove the second, never-wired skeleton system (.aksrt-skeleton-* in
core-ui.css) and the dead spinner stylesheets/rules (skeleton.css,
ajax-spinner.css, .load-more__*, --skeleton-base/--skeleton-shine) plus the
unregistered assets/js/waterfall.js.
- Extend the widget, message-centre, user-home and asset-ledger contract
tests to pin the loader inside the container, the class-only visibility
switch, and the absence of the deleted dead code.
Verified: 470 node tests, 136 PHP widget assertions, admin-contract and
widget-system audits all green; rebuilt public assets after the CSS/JS
removals.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Loading indication now matches the reference theme's post_ajax() contract:
- Stop overwriting the 加载更多 button label with 正在加载...; the button
keeps its own text and the .post_ajax_loader skeleton carries the busy
state instead.
- Add the reference fallback loader (.loading ring + 加载中...) when no
skeleton exists in the DOM, and reinsert the skeleton after the last
card on every page load like beforeSend does.
- Retext the loader string to 加载中... (zib__('loading')) and make the
ajax_trigger default a translatable string.
- Route ajax-list.js / comments.js / widgets frontend.js through the
.is-loading class instead of inline display, which outranked the class
rule and could freeze the skeleton shut.
- Drop the search.php inline display:none that defeated .is-loading, and
remove the unproduced .post-ajax-loader / [data-aksrt-ajax-loader] shims.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Clicking 加载更多 showed no skeleton at all, and the placeholder shape did
not match the reference theme's post cards.
- Class-name mismatch: pagination.js queried .post-ajax-loader while every
PHP producer emits .post_ajax_loader, so showSkeleton() never found a
target and the loader stayed hidden. Accept both spellings.
- Inline display:none on the loader (pagination, comments, widget tab
panes) outranked .is-loading, so even the correct selector could not
reveal it. Drop the inline style; the reference uses a bare
<div class="post_ajax_loader"> and toggles visibility from JS.
- loadNextPage() set indicator.style.display = "flex", which forced the
.posts-item.card skeleton blocks (inline-block layout) into a single
row instead of stacking vertically — the visible symptom. Remove those
inline writes and drive .is-loading from CSS for every surface.
- Scope the hide/show rules on the class itself rather than requiring an
.ajaxpager ancestor: homepage tab panes, archive streams and comment
lists sit at three different depths.
- Rebuild aksrt_render_placeholder() on zib_get_post_placeholder's
structure — list = t1/k1/k2 + two s1, card = t1 item-heading/k2 + two
s1, thumb wrapped in post-graphic + radius8 item-thumbnail, all carrying
ajax-item so the loader is not mistaken for a real card.
- Add mini/mini_thumb modes so ContentRenderer::placeholder() stops
hand-rolling the same markup; normalize its 'lists' arg to 'list'.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The action bar used a locally invented pill-chip style that overrode the
reference theme's 80px icon-over-text columns, and the JS toggled a
class no stylesheet targeted — so likes/favorites showed no selected
state at all.
- restore reference markup: <svg class="icon"> + <text>/<count> and the
actived class, with data-stat-count/data-base-count preserved
- toggle actived plus a transient action-animation so the lovestar
keyframes finally reach the DOM
- drop the CSS overrides and now-dead action-btn/.action-btn--reward
and .post-actions-card rules; main.css already carried the reference
rules the overrides were shadowing
kate522andClaude Opus 5