| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
- dark-mode restore now travels nonced, the admin constants likewise
- AUTH_TOKEN deleted in favour of the csrf meta tag
- new report collector at /csp_reports
|
|
A muted glyph before each sentence makes both log surfaces
scannable by verb: create, publish, move, trash, restore,
destroy, and the discard pair. Rollback publishes get their own
icon (history) distinct from ordinary promotion. Unknown verbs
fall back to a dashed circle.
|
|
Each row in nodes#show's History section now carries terse badges
from the action log: created, published, and restored (rollback
re-promotion), with date and actor, rendered from entry metadata
alone. Backfilled entries wear the inferred marker, so
reconstructed provenance stays distinguishable from witnessed
history. A revision that was published and later restored shows
both badges chronologically -- its true biography.
Only create and publish entries carry page_id; trash, restore, and
destroy annotate the node's own log zoom instead of any single
revision, by design.
|
|
|
|
|
|
Sentences for trash (with the pre-trash path), restore_from_trash
(with the destination), and destroy (with the final path, read
from its flat create-symmetric key).
|
|
nodes#show gains a Trash section on trashed nodes: provenance from
the trash entry, a restore form whose parent picker pre-fills the
old parent while it still lives, and permanent deletion. A
Move-to-Trash button joins the status actions on living nodes.
nodes#trashed lists trashed subtree roots with weight, provenance,
and deletion; the dashboard housekeeping row links to it, and
trash/destroy redirect there.
Deletion from Trash now removes the whole subtree, deepest first,
each node through a real destroy! so every per-node cascade runs --
amending the never-recursive rule for this one sanctioned path
(both confirms state the count; the root entry carries
destroyed_descendants). Bare Node#destroy still refuses children.
|
|
|
|
Also update the node action contract to include the trash related verbs.
|
|
|
|
NestedTree's before_destroy silently delete_all'd the whole
subtree, bypassing every per-node cleanup. Nodes are never
destroyed recursively; descendants must be removed individually.
|
|
|
|
Page.aggregate interpolated order_by into SQL unchecked while
already allowlisting order_direction; the column is now normalized
and checked against the sortable columns, falling back to pages.id.
Its values arrive from editor-authored aggregate shortcodes, so
this was editor-gated, but the asymmetry was wrong regardless.
template_name and default_template_name now validate inclusion in
Page.custom_templates -- names render as filesystem paths, so only
names actually present in the template directory are acceptable.
Validated only on change: legacy rows whose template file has since
vanished stay saveable, and valid_template's render-time fallback
to standard_template continues to cover them. Two tests that wrote
fabricated template names through the front door now arrange their
state correctly (update_column for the stale-name fallback test, a
real template for the update-persists test).
|
|
password_required? treated a blank crypted_password as "needs a
password", but after bcrypt migration crypted_password is nil by
design -- every subsequent save of a migrated or new user failed
validation. The predicate now requires a password only when both
the legacy and the bcrypt credential are absent, or when one is
actually being set.
|
|
Add has_secure_password and bcrypt while retaining compatibility with
legacy SHA-1 password hashes. Existing users are upgraded to
password_digest on their next successful login.
Add regression tests covering both legacy and modern authentication
paths.
|
|
|
|
|
|
Weekdays appearing on a subset of the weeks of the month are now
aggregated into a simpler form.
|
|
Improve the humanizer to also understand weekly patters that
an editor can manually select, on top of the single week
rules.
Extend template and javascript controller to reflect these
changes.
|
|
Several minor improvements to the action log presentation:
* Now a table with date and human readable presentation as
rows is displayed
* If no changes in a title were detected, the old version
is omitted
* The "inferred" flag is demoted to the end of the line
* You can zoom in on the node's history directly from a
log line
* byline for the first publish action is preserved
* Revisions are directly linked to when a new one i
published
|
|
NodeActionsController#index lists entries newest-first, filterable
by node_id or user_id -- the two zoom shapes the log was designed
around. Rendering goes through NodeActionsHelper.action_summary,
which builds one sentence per entry from metadata alone, so entries
referencing deleted users or nodes render from their snapshots;
live associations only upgrade names to links. Unknown verbs
degrade to a generic sentence rather than an error, since the log
outlives its vocabulary. The helper is the escaping boundary:
every metadata value passes through h() before assembly.
Actor names link to the log's own user zoom rather than the unused
users page -- inspecting a suspicious user's other actions is the
intended workflow. Publish entries with a translation_diff expose
a collapsed per-locale change table linking out to the revision
itself. Sentences live in en.yml/de.yml following the existing
widget-string convention. nodes#show links to its node's zoomed
log.
|
|
A contract comment above NodeAction.record! now specifies every
verb's metadata shape. NodeAction.head_diff computes the publish
diff between an outgoing head and its replacement -- default-locale
title pair always, author/tags pairs and template/assets/abstract/
body flags only when changed, and a per-locale translation_diff
with added/removed/changed status. It is a pure function of its two
pages, shared by publish, rollback, and the future backfill, and
reads translation rows directly so fallbacks never masquerade as
content.
publish entries carry via ("draft" or "revision"); restore_revision!
is now transactional, takes the acting user, and logs through the
same diff. Staged slug/parent changes applied at publish log a move
entry with the path pair. Node creation logs a create entry with
initial title and path. The draft-scoped translation_destroy writer
is retired -- locale removal is recorded by the publish diff, where
it becomes public fact.
|
|
node_id/page_id/user_id are lookup and ordering only -- all three
nullify on delete, so an entry outlives its actor and its subject.
Everything that must survive those deletions lives in a mandatory
metadata jsonb written once at creation: the actor's username, the
node's human-readable name (pinned to the default locale), and
action-specific extras such as publish's title from/to.
NodeAction.record! is the single constructor, so every entry gets
the same baseline metadata without each call site re-implementing
it. occurred_at is one field for live and backfilled entries alike;
inferred_from distinguishes them -- nil means witnessed at the
moment it happened, populated names how a backfilled entry was
estimated.
Instrumented so far: publish (crediting the actual publisher,
threaded through from the controller -- previously nobody had the
act of publishing recorded anywhere), revert's discard_autosave and
destroy_draft branches, and translation destroy. publish_draft! now
runs in a transaction so the promotion and its log entry land
together. The remaining verbs follow once this mechanism has proven
itself.
|
|
mtime_busted_path only ever covered the four admin-only static
files. The public layout has the identical exposure -- ccc.css,
glightbox's vendored CSS/JS, and public.js are all served straight
from public/ with no pipeline fingerprinting either -- just never
got the same treatment.
|
|
Sized generously (1600x1600>) rather than against today's cramped
490px content column specifically -- shrinking an oversized source
to fit a narrow container is free and looks fine; a source too
small for a future, wider redesign has no fix short of re-uploading
every image. Purely additive: nothing reads :large yet, and
generate_variants already loops over the whole STYLES hash
generically, so no other code needed to change for new uploads to
start getting it. Test suite's own hardcoded style lists updated in
the same commit so they don't quietly drift out of sync with what
the app actually defines.
|
|
Both had already lost their reason to exist as production API:
wipe_draft!'s one remaining callsite (nodes#show) was removed two
sessions ago, and find_or_create_draft had zero production callers
left at all -- confirmed by a fresh grep, not assumed -- every one of
its ~65 call sites was test setup, unrelated to what those tests
actually cover.
wipe_draft! is deleted outright, along with its two tests -- the
lock/draft/autosave cleanup it silently performed already has
explicit, always-visible manual equivalents (Unlock, Discard
Autosave, Destroy Draft), so nothing real is lost.
find_or_create_draft moves to test_helper.rb as a plain method on
ActiveSupport::TestCase, alongside the create_node_with_draft/
create_node_with_published_page helpers already living there --
extending the framework's own designated test-extension point
rather than reopening the Node model from test code. Its three
tests of real dispatch behavior (idempotency on repeat calls, and
raising when a second user contends for the lock) are kept, since
~65 other tests depend on this helper actually working correctly;
only the call syntax changed, from node.find_or_create_draft(user)
to find_or_create_draft(node, user).
|
|
|
|
|
|
diff_against compared title/abstract/body under whatever I18n.locale
happened to be ambient, with no concept of 'diff this translation
specifically' -- so a change confined to one locale was invisible to
Diff Head vs. Draft regardless of which locale you were looking at
when you clicked it, exactly the 'yields nothing' complaint from
earlier this session.
Page#diff_against gains a locale: keyword, additive only -- nil
preserves the exact original ambient-locale behavior every existing
caller and test already depends on; passing a locale switches to
reading each side's actual PageTranslation row directly, same
fallback-free reasoning as Page#translation_summary. Page#locale_diff_summary
reports one entry per locale present on either side, so an added or
removed translation counts as a change even where content matches
everywhere it exists on both.
RevisionsController#diff now resolves a real locale before diffing --
defaulting to whichever locale actually changed, falling back to the
default locale only when nothing did -- and the view carries that
locale through every existing control (view toggle, layer-pair
buttons, the revision-select form) so it and the view/layer-pair axis
stay independently selectable rather than resetting each other.
|
|
Every promotion (autosave creation, draft creation, draft save) ran
every locale's translation through delete-all-then-recreate
unconditionally, giving every locale a fresh created_at/updated_at
regardless of whether its content actually changed. Confirmed via
Node.find(546).draft.translations -- both DE and EN shared one
identical timestamp despite only EN having been edited.
This silently defeated Page.find_with_outdated_translations' whole
staleness comparison: every save reset every locale back into
lockstep, so timestamps could never actually drift apart the way
the feature depends on. The existing test for it only ever passed
by manually backdating a timestamp with record_timestamps = false,
bypassing normal save flow entirely -- not something that happens
through ordinary editing.
Now updates a matching locale's translation only when its own
attributes actually differ, creates one genuinely new, and removes
one genuinely gone from the source -- same end state, but real
per-locale timestamps survive an unrelated save. search_vector is
excluded from both the comparison and the copied attributes: it's
DB-trigger-maintained from title/abstract, not real content, and
comparing a precomputed tsvector risked a false 'changed' from
representation noise alone.
Also reloads the source's translations association, not just
self's -- clone_attributes_from previously only guaranteed a fresh
self, leaving any caller holding a page reference across an earlier
mutation vulnerable to reading a stale cached association.
|
|
superseded/currently_public used published_at presence as a proxy
for page-row relevance, but published_at is carried forward onto
every new draft descended from previously-published content
(Page#clone_attributes_from, Node#save_draft!) -- so it read true
for any draft on an already-published node, whether or not that
specific draft had ever gone live itself. Compare draft_id/head_id
identity directly instead; no timestamp involved. A stale
(superseded) link now redirects to the live public page rather than
an admin URL an anonymous holder can't reach anyway.
|
|
The shared TinyMCE setup initializes cccms.setup_autosave() on any
page with a textarea.with_editor, unconditionally starting a
7-second interval that submits to the form's data-autosave-url --
which the translation edit form never set, so the interval PUT the
current page URL itself and 404'd on a nonexistent route.
Fixed by actually giving it something to talk to, rather than
suppressing it: a real autosave endpoint, and update now goes
through Node#autosave!/#save_draft! -- the same pipeline the
primary editor uses, fixed earlier this session for exactly this
kind of cross-locale carryover. This was the autosave-buffer parity
already flagged as due after the proof of concept; the shared JS
just forced the timing.
|
|
Replaces the old locale-switch-and-edit-the-same-screen workflow,
which conflated presentation locale with content locale and let an
editor silently drift into editing the wrong language with no
persistent signal that anything had changed. Non-default-locale
content now has its own explicit routes and screens, never sharing
a route param with the ambient chrome locale.
- PageTranslationsController: index/show/edit/update/destroy,
scoped to Page.non_default_locales; update handles first-time
creation too, so there's no separate new/create step.
- Reads go through the actual PageTranslation row
(Page#translation_summary), never through the Globalize
fallback-bearing accessor -- fallback is correct for public
rendering but wrong for editing, where a missing translation
needs to look empty, not borrowed from another locale.
- Translations ride on the page's own draft/head cycle; no
independent publish state.
- nodes#show gains a Translations section (per-locale Lock+Edit /
Create+Lock, Destroy, a link into the read-only Compare view) and
a locale indicator on its own default-locale content; nodes#edit,
nodes#update, and nodes#autosave are pinned to the default locale
via Globalize.with_locale regardless of the ambient route locale.
- nodes#show no longer double-loads the node or calls wipe_draft!
on every view (see previous commit for why that's now safe).
- .preview_link_row is renamed .aligned_action_row now that it has
a second real consumer.
|
|
Node#autosave! only copied assets forward onto a newly created
autosave, not translations. Invisible for single-locale nodes, but a
node with an existing multi-locale head/draft that gets locked and
edited in one locale would silently drop every other locale's
translation from the new autosave -- and from there into the draft
and eventually head, since save_draft! faithfully clones whatever
the autosave actually holds.
Fixed by using clone_attributes_from, which already does a complete
clone (translations, assets, tags, template, published_at) -- the
same method save_draft! already relies on two branches later, so no
new mechanism, just closing the one place still doing a partial,
hand-rolled version of it.
|
|
Replaces nodes#edit's old Images section -- a hidden panel dumping
every image asset in the system unfiltered (#image_browser) plus a
raw drag-and-drop box (#image_box) -- with a small search-and-click
picker built on the endpoint from the last two commits. Attaching
posts immediately and appends the new thumbnail via a cloned
<template> -- icons only render correctly through the Rails helper
server-side, so the template holds real, pre-rendered markup for JS
to clone rather than duplicating raw SVG in a JS string. Reordering is
jQuery UI sortable on the small attached list only, with a dedicated
drag handle rather than the whole thumbnail.
Two bugs caught while click-testing, fixed here rather than shipped
and patched after: the search panel never closed after attaching an
image, since the success handler re-triggered focus to keep it open
for attaching several in a row -- which meant it just re-populated
itself forever instead of signaling "done." Fixed to close explicitly;
a click-outside-closes handler was added alongside it, matching the
affordance the top-bar search already has.
A real, independent, pre-existing data bug surfaced during the same
testing: Node#autosave!'s first-time-creation branch never carried
related assets forward from whatever page was previously current --
attach an image, let autosave fire once, and it silently landed on a
fresh, assetless Page row. Long-dormant, not introduced by this work,
just finally exercised by something that made it visible. Fixed inside
the `unless self.autosave` guard specifically -- running this on every
call, not just creation, would overwrite anything attached directly to
an existing autosave in between, a worse bug than the one being fixed.
nodes#show gains a read-only Images section, rendered only when a page
actually has attached images, so an attachment can be confirmed
present without entering the edit/lock cycle -- useful on its own, and
specifically useful the next time an asset bug needs investigating.
Its thumbnail CSS is shared with the edit view's picker via a class
(.thumbnail_list) rather than duplicated under a second name.
|
|
initSearchPicker's input handler has always short-circuited on an empty
term -- correct for every existing picker, none of which want results
appearing before anything's typed. The asset picker does want exactly
that (show the last few uploaded images without requiring a search
first), so this adds an opt-in loadOnFocus option instead of changing
the shared default: it fires once when the input gains focus with no
term yet, reusing the same request/render path a real search uses
rather than a separate code path. The AJAX call itself was pulled out
into a named runSearch function so both triggers could share it without
duplicating the success/render logic.
RelatedAssetsController#search now treats a blank term as "show the
most recently created, unattached images" (limit 5) rather than
returning nothing, matching the new client behavior. A real search
term still returns up to 10 name-matched results, unchanged.
|
|
Backend for the asset-picker rebuild -- replaces the plan to dump
every image asset in the system into a hidden, unfiltered browse
panel on every node edit (the actual current behavior, confirmed by
reading nodes/edit.html.erb directly) with a small, name-scoped search
endpoint plus create/destroy/update for attach, detach, and reorder.
No schema change needed -- RelatedAsset already had everything this
requires (asset_id, page_id, an acts_as_list position). search excludes
assets already attached to the page, keeping results meaningfully small
given hundreds of assets total but only a handful per node in practice.
create is find_or_create_by! rather than a bare create!, guarding
against the same asset being attached twice from two separate search
results. update leans on acts_as_list's own insert_at rather than
custom position-shifting logic.
Node#editable_page (autosave || draft || head) is extracted since this
is now its third call site with identical logic -- deliberately not
touching nodes#show, which resolves draft || head without autosave on
purpose, a different and correct semantic for "current committed
state" versus "what's actively being edited."
|
|
"New X" becomes "Create X" throughout (users, events, assets, menu
items, nodes), matching the verb-first pattern the dashboard's own
signposts already established, with a shared plus icon rather than a
document-flavored one that only made sense next to "post". The
"Destroy"/"destroy"/"Delete" family is normalized to "Destroy"
everywhere, with a shared trash icon; occurrences#index and
pages#index also pick up the destructive button class they'd been
silently missing. Filter and Search convert from submit_tag to
button_tag, the only way either can hold an icon alongside its label.
Edit and the node editor's three dynamic labels (Continue Editing /
Edit Draft / Lock + Edit) share one icon without touching their
wording -- unlike Destroy's family, the state nuance in the text is
real information, not just inconsistent phrasing.
|
|
Replaces the old admin#index wizard -- accreted over years, never
designed as a whole -- with the dashboard settled on this session:
a three-icon nav (dashboard/search/log out, no locale selector), a
nodes-first search bar, four task signposts, and two symmetric
widgets (drafts/autosaves, recent changes) with a quiet housekeeping
row beneath them.
Node.recently_changed now filters and orders by the head page's own
updated_at instead of the node's blanket timestamp, so a lock/unlock
cycle with no actual publish no longer surfaces here, and the
original publisher is no longer misattributed to someone else's
housekeeping action. This also restores the "published" qualifier on
each entry, which the query previously couldn't guarantee was true.
@mynodes and its dedicated "My Work" table are retired along with the
old wizard -- "Continue my work" is a link to the existing, already-
correct NodesController#mine instead of a second, duplicate query.
Its one dedicated test (dedup across multiple revisions by the same
user) is ported to nodes_controller_test.rb, since mine already
carries the same .distinct protection the old query did; it just had
no test of its own until now.
|
|
Node.recently_changed and the new dashboard need "X ago" rendered
correctly in both locales. Rails' own distance_of_time_in_words/
time_ago_in_words can't do this on their own -- the scope option
changes which translation key is looked up, not the underlying
grammar, and German's "vor" requires dative case, which is a
different word form than the nominative plural Rails computes by
default. relative_time_phrase/relative_distance is a small, from-
scratch bucketing helper instead, with an explicit, hand-checked
translation table per unit per locale.
Also removes de.yml's datetime.distance_in_words and activerecord.
errors blocks. Both used the old {{count}}/{{model}} interpolation
syntax the current i18n gem no longer recognizes -- it silently
leaves the literal placeholder text in the rendered output rather
than erroring, which is why this went unnoticed until now. Both were
shadowing rails-i18n's own current, correct translations for exactly
these keys; deleting the local override lets the gem's version take
over instead of maintaining a second, broken copy.
|
|
AdminController#dashboard_search returns tags and nodes as separate,
labeled groups (via ActsAsTaggableOn::Tag.named_like and
Node.editor_search) rather than the flat per-node list every existing
picker returns. initSearchPicker gains a renderResults callback option
that, when given, replaces the default per-node rendering entirely --
lets the dashboard render its two labeled groups without a second,
parallel picker implementation.
resultsHeaderHtml (the "Press Enter to see all results" hint) is now
threaded through as a third argument to renderResults, so a picker with
custom rendering can still show it -- previously only the default
per-node branch ever did.
Tags link straight to the existing /admin/nodes/tags/:tag view rather
than becoming an in-place filter chip.
|
|
Node.drafts_and_autosaves and Node.recently_changed replace inline
query logic in NodesController#drafts/#recent -- pure refactor, no
behavior change for either action. The real reason: the dashboard's
upcoming abridged widgets need the exact same queries the full pages
already use, just limited and (for drafts) sorted with the current
user's own locked nodes first. Better to share one method than let a
widget and a full page quietly drift onto two versions of "what counts
as a current draft."
current_user_id: is an explicit, optional argument rather than a
separate method -- ordering by lock ownership only applies when a user
is given; omitted entirely, it's pure recency, exactly today's
behavior. Needed Arel.sql wrapping a sanitized CASE expression before
.order would accept it -- sanitize_sql_array only vouches for the
values inside the string, a separate Rails safety check still refuses
any string shaped like more than a plain column reference unless it's
explicitly marked as already-vetted.
|
|
Extracted from admin#index's inline table into NodesController#sitemap.
Nested <details>/<summary> per branch, one linear pass over the
existing flat [node, level] list (no added queries) -- each node's own
descendant count computed the same way, via a small stack rather than
re-walking the tree per node. Branches under updates/, club/erfas,
club/chaostreffs, and disclosure start collapsed by default
(CccConventions::SITEMAP_COLLAPSED_PATHS); any branch currently
collapsed, whether by that default or because someone just closed it,
is highlighted via a plain :not([open]) selector -- no state tracked
outside the DOM itself.
Dropped the update?-post exclusion this view used to rely on -- no
longer needed now that updates/ collapses instead of being filtered
out, so its real children (previously silently absent) now show up
correctly. admin#index's own, separate @sitemap query is unchanged;
that view has no collapse mechanism to compensate and wasn't part of
this.
|
|
Four NodesController actions -- drafts, recent, mine, chapters --
each building its own base scope, sharing one private method
(index_matching) for search narrowing and pagination. Wizard rewrite
to link into these instead of rendering its own tables is a separate,
later step.
Node.editor_search backs the shared "q" narrowing: an ILIKE substring
match against title/abstract on whichever of head or draft is
present, splitting the term on whitespace and requiring every word to
match somewhere independently, not as one phrase, since real words can
end up separated by markup in the underlying HTML. Deliberately
separate from Node.search, the public content search, which stays
tsvector-based and head-only.
chapters generalizes into /admin/nodes/tags/:tags for an arbitrary
tag list (OR'd, not AND'd), sharing the controller action but
rendering its own template rather than branching inside one view.
|
|
|
|
publish_draft! called move_to_child_of before validating the new
parent at all. Staging a node under one of its own descendants
created a real, if transient, cycle in parent_id the instant that
save landed -- and update_unique_names_of_children's after_save
callback, which recurses down through parent_id with no cycle check
and no depth limit, bounced between the two nodes forever, crashing
the whole process with a SystemStackError rather than just producing
bad data.
Now rejected up front with a normal ActiveRecord::RecordInvalid.
|
|
@mynodes used the bare @current_user ivar instead of the current_user
method (works today only because login_required already forces that
memoization; every other query and controller in this app uses the
method), and .uniq, which is plain Enumerable#uniq here, not
Relation#distinct -- it materializes every joined row into an Array
before deduplicating, rather than deduplicating in SQL.
@mypages had the same ivar pattern, no .order/.limit unlike every
sibling query in this action, and -- confirmed via a full grep -- is
referenced nowhere else in the codebase. Removed rather than bounded;
nothing was ever rendering it.
|
|
Both ran real code and checked nothing -- Minitest's "missing
assertions" warning has been firing on every run this session.
test_find_or_create_draft_if_draft_exists_and_is_owned_by_user called
the method twice but never checked what came back; now confirms the
second call returns the same draft rather than creating a new one,
and leaves the lock and page count untouched.
test_destroy_a_published_node destroyed a node and loaded the admin
index but never checked the response, unlike its two neighbors
covering the same destroy path (dangling pages, orphaned
occurrences). Added the assert_response :success its own shape
already implied.
No behavior changed -- both were passing before for the same reason
they're passing now, they just weren't proving anything.
|
