Records the decisions the 参考文献 list now rests on: a citation is a place in
the text and an entry is a work, so a paper with 86 citations to 65 works lists
65 entries; the entries are written by `app/citation_styles` and a client prints
them verbatim; a gap is reported through `missing` rather than quietly rendered
as a shorter entry.
Also documents `scripts/check_citation_styles.py` — the formatter is checked
with no server and no database, because a wrong branch prints something that
looks like a bibliography entry instead of failing loudly — and why the
reference smoke test now stamps a token into everything it creates.
The card listed the quoted passage where the bibliography entry belongs, under
a `引用 #53` tag that named a database id rather than the work. It now prints
the server's `references` verbatim — one line per work, already formatted in
Springer's basic style — with the fields a complete entry still lacks shown as
「缺 责任者、出版年」, and the paragraphs that cite it underneath:
1 引言 / Introduction
2.1 技术主题识别 / Technology Topic Identification ×2
Each of those links carries the quoted passages for that paragraph in its
tooltip, which is where the quote belongs: it is evidence for a citation, not
the citation itself. The client formats nothing and groups nothing but the
paragraph labels — a second implementation of "which work is number 7" would
eventually disagree with the first.
The marker tooltip in the text stops saying `引用 #53` too and shows the entry
it points at, so hovering a `[7]` answers the question the number raises.
It asserted that a search for "Wang" found one row and that the library was
empty when it finished — both true only of an empty database. The library it
runs against now holds 65 real references and 10 labels, so the run failed on
its own baseline rather than on anything it was testing.
Every row the run creates now carries a per-run token in its title, in one
contributor's name and in its label names, and each count is asserted against a
search for that token. A second, distinct token rides on the author name alone,
because searching for the first one finds every row the run made and "the
search matches an author" needs a string that is in exactly one place.
What the test means does not change: "the search found one row" is a statement
about this run, and the token is what makes it one.
The list was printing each citation's quote under a `引用 #53` tag: no authors,
no year, no pages, no DOI — the quoted passage where the entry should be — and
it numbered *citations*, so a paper with 86 citations to 65 works listed 21 of
them two or three times under different numbers. The same article was `[5]` in
one paragraph and `[17]` in another.
Two rules were wrong, so two are stated here:
**A work is numbered where the reader first meets it.** A citation is a place
in the text; an entry is a work. `citations` still has one row per place, and
each row now carries the number of its *work*, so the marker in the text, the
tooltip on it and the entry in the list are the same number. The document grew
a `references` list — the entries, in order — and `reference_style`, which
names the style that wrote them.
**How a reference prints is a property of a style, not of a citation.** The new
`app/citation_styles` package holds one module per style; `springer.py`
implements Springer's basic style, whose shape is
`Names: Title. <where it sits> (year). <identifier>` with a branch per document
type — a journal with its volume(issue) and pages, a book with publisher and
place, a chapter inside somebody else's book, a patent with its number, a page
whose URL is its source.
Three of its rules are decisions rather than transcription, and each is
documented where it applies: up to six contributors are named and more become
three plus `et al.`; a journal prints under its 刊名缩写 when the writer stored
one, because no algorithm abbreviates an arbitrary title correctly; and a CJK
name keeps its printed form even with halves filled in, because dotting 「三」
into an initial is wrong in Chinese and insufficient in English.
Nothing is invented to fill a gap. An entry says what it knows and reports the
rest as `missing` (「缺 责任者、出版年」), and the document warns about the
incomplete ones by number: an entry that is short because fields are empty
looks like a fact about the document rather than a library that needs attention.
`document_text` had the same bug in the MCP surface and now prints the entries,
with in-text markers taken from the document's numbering rather than from each
sentence's local position — which is what made the text an agent reads disagree
with the list under it.
`scripts/check_citation_styles.py` checks the formatter with no server and no
database: one reference of every type, and the exact entry the style calls for.
The reference library landed with a full REST API and no MCP surface at all:
an agent could cite `reference_id: 42` — the citation is validated now — but
could not look up what 42 was, create it, or find out whether the library
already had it. That is the citation half of "generate the paper, then write it
in" missing, and it is the half an agent is worst at doing by hand.
Twelve tools, mirroring the REST surface plus one addition each way:
- `reference_*` — search, read, create, update, delete, plus
`reference_import`, which takes a whole pasted bibliography in one call and
reports what it skipped (an existing DOI or an identical title) and what it
could not read, one bad entry at a time rather than losing the other nineteen;
and `reference_citations`, which answers "what is this cited for" before a
rename or a delete. `reference_list`'s keyword search across title, container,
DOI and author names is the de-duplication path that `reference_import` uses.
- `reference_tag_*` — the five CRUD tools. Deleting a tag drops the labelling
and leaves the literature, which is the opposite of what the field library
needs and the same choice the REST route makes.
Two things had to be fixed to get here, both found by testing rather than
reading:
`template_list` was crashing. The rename that turned a heading into a title and
its use into a placement changed `TemplateListItem.field_count` to
`placement_count` and left the tool reading the old attribute, so the call died
with `Error executing tool template_list` — unreadable, and live in the
cluster. The suite passed 47 checks while this was broken because it never
called `template_list`: the uncovered tools were exactly the read-only ones, so
the write paths were all green and the browse path was not. The suite now
covers all 41 tools — 76 checks — and step 5 asserts the shape each read tool
returns, not merely that it returned.
`DOMAIN_REFUSALS` in the tool registry held one class, `UnknownReference`. The
tag rules arrived with the tag tools and raise two more, so 「标签不存在:9」 and
「标签「必读」已存在」 would have reached a model as `Error executing tool …`.
The tuple is now the list of every refusal a tool is expected to translate, and
a new one belongs in it on the day it is written.
`reference_update` takes its changes as one nested object rather than 38 keyword
parameters: a nested model keeps `model_fields_set`, so an omitted field stays
absent while an explicit null clears the column. Flattened, the two are
indistinguishable — every omitted field arrives as None — which is the same
problem `paper_update` solves by rebuilding its payload from what it was given.
Verified against the deployed service, not a local process: 76 checks pass over
the NodePort, then the same 76 over stdio and against a local HTTP server, and
the REST suite still passes its 55. The image is rebuilt as v4 and the
Deployment follows it, because `imagePullPolicy: IfNotPresent` means re-pushing
a tag the node already holds changes nothing.
Ten tables now, and the two new ones are argued for where the other
schema decisions are: why a tag is not a keyword, why the link is bare,
why a tag is made before it is applied, and why deleting one is not
refused while deleting a cited reference is.
Also corrects the table count — the section claimed nine while listing
eight — and the shell diagram, which had a header with two entries and
named the wrong side for the second-level menu.
参考文献 now sits in the header to the right of 论文 and 模板 — the library is
drawn *from* rather than worked in, so it is a button one step away. Its
own second-level menu marks out 文献列表 / 新增文献 / 标签管理 and then lists
the tags themselves, read from their store, so a label can be picked
straight from the rail.
A tag in the rail links to `/references?tag=<id>`, which makes the query
parameter part of the view rather than hidden filter state: the table's own
selector writes the same parameter, so the two cannot disagree and a
filtered view is a link someone can be sent.
The form keeps the order the feature intends — a tag is created before it
is applied, so no stored reference can point at a label that is not there —
and the citation count in a row opens the answer it exists to give (which
paper, which paragraph, which quoted passage).
Tag colours are painted from the row rather than a class map: the palette
is the writer's, and the chip has to be that hue in the table, in the rail
and in the picker.
A tag is the writer's filing of a row — 「综述」, 「必读」 — and it is a
different fact from `reference.keywords`, which are the 关键词 printed on
the article. A tag never prints, one document carries several, and the same
label is on many documents, so it gets two tables rather than a column:
reference_tag name (unique), colour, note
reference_tag_link reference_id + tag_id — the pair is the primary key
The link is a bare link, which is why it is `_link` and not `_placement`:
there is no `sort` to own, so the pair *is* the row and the same label
cannot be applied twice.
A tag is made first and applied second: `tag_ids`, never tag names, so a
typo cannot invent a label — an unknown id is a 400 naming it, and nothing
is written. Deleting a tag is deliberately *not* guarded the way deleting a
cited reference is: both foreign keys cascade, so the labels go and the
literature stays.
`GET /references?tag=1&tag=2` narrows rather than widens — the rows carrying
both labels — because that is what a second filter reads as.
The smoke test now walks the whole flow in the order the feature intends:
labels first, then tagged references, then a rename that follows every row.
One table per document the writer has read, and one for its contributors:
reference id, type, title, subtitle, title_translated,
container_title, volume, issue, pages, publisher,
publisher_place, event_*, publication_year/month/day,
accessed_date, doi, isbn, issn, url, patent_*,
standard_number, report_number, arxiv_id, abstract,
keywords, note, citation_key, language
reference_author reference_id, sort, name, family_name, given_name, role
The field set is the intersection of the formats a paper has to survive, taken
from the CSL JSON schema (103 fields, 45 types — what Zotero, Mendeley and
Pandoc exchange) against GB/T 7714-2015, BibTeX and RIS. The mapping is on
`app.models.reference` rather than in a commit message, because the next person
to add a column needs it more than this one does. Three decisions worth naming:
* **Partial dates.** Year, month and day are separate nullable integers, which
is CSL's `issued`: a book has a year, a newspaper article a day, a patent its
公告日期, and nothing invents the rest.
* **Contributors are rows, not a column.** A citation prints an ordered list
with 编 / 译 marks, and styles rewrite names (「张三」 / "Zhang, S." /
"S. Zhang"), so the printed form and its two halves are stored separately —
display reads `name`, a reformatting style reads the halves, and an
institutional author has neither.
* **`volume` / `issue` / `pages` are strings.** Journals print `12(Suppl.)`,
`S1` and `e12345`, and a citation reproduces what is printed.
The library is global: the same article is cited from several papers, so the
entry cannot belong to one of them. The citation stays on the paper side in
`paper_sentence_reference`, which now carries a real foreign key to `reference`
(`ON DELETE RESTRICT`) — the column had been a loose integer waiting for this
table. A sentence may quote several documents and a document may be quoted by
many sentences, which is why the relation is a table and not a column on
`paper_sentence`.
Refusals are raised once, in the data layer, as `UnknownReference`, and
translated where each caller speaks: `400` with the offending ids for REST (one
exception handler, no route carries the check), a readable tool error for MCP
(one wrapper in the tool registry). Neither entry point can skip it, so the
foreign key never has to be the thing that complains.
API: `/references` CRUD plus batch delete, and `/references/{id}/citations` for
the reverse lookup — "where have I used this article?" — which is what tells a
writer whether an entry is safe to delete.
Verified: 29 new checks in `scripts/smoke_references.py` (partial dates, DOI
normalisation, author order and roles, search by author/type/year/cited, the
citation round trip, 400 on a phantom id, 409 on a cited entry), the paper
suite's 45 and the MCP suite's 47 still pass. The MCP suite now cites a real
entry instead of a phantom id, which is what an agent has to do now.
The running pod predates the table rename and fails every tool that touches the
template tables, so the image is rebuilt as v3 and the Deployment follows it.
The manifest is bumped with it — `imagePullPolicy: IfNotPresent` means
re-pushing the same tag would keep serving the cached layers, so a rebuild needs
a new tag to reach the cluster at all.
Verified by handshake against the running server: `field_list` answers with the
45 library titles.
`template_field_library` and `template_field` were one word apart and meant
opposite kinds of thing — a catalogue of rows versus one row — and neither name
said what a row actually is. Every row of the library is a **heading**: a piece
of text like 「1. Introduction」 with a level and typography, which is what the
reader sees at the top of a paragraph; and every row of the join table is that
heading *placed* in a template, `sort` included.
template_field_library -> template_title
template_field -> template_title_placement
field_id (the column) -> title_id
TemplateFieldLibrary -> TemplateTitle
TemplateField -> TemplateTitlePlacement
The join table keeps its own name rather than becoming `_link`, because it is
not a bare link: it carries `sort`, the only place display order is stored, and
it deliberately allows the same title twice in one template. `_link` would
describe two foreign keys and hide the column the ordering rule rests on.
Carried through every level, because a half-renamed layer is how the confusing
pair got there in the first place: HTTP paths (`/template-field-library` ->
`/template-titles`), the template payload (`fields`/`field_id` ->
`placements`/`title_id`, and `field_count` -> `placement_count`), the Pydantic
schemas, the CRUD modules, the seed script, the frontend API module and types,
and `docs/OVERVIEW.md`, which now states the naming rules outright.
Two deliberate exceptions:
* `paper_template_filed_sort` is untouched — it is a column of the paper module
and is spelled as the feature was specified.
* The MCP tools keep their agent-facing names (`field_list`, `field_ids`, and
the `fields`/`field_id` keys in their responses). Agents already call them by
those names; the storage names they read underneath have changed.
The migration renames in place (data preserved: 4 templates, 45 titles, 63
placements) and is written to tolerate a database where part of it already ran,
which is not hypothetical — the first attempt renamed the tables and then failed
on the column, because Alembic's MySQL dialect wants an explicit type for
`alter_column` and will not rename a column without one. TiDB carries the
foreign key along with both the renamed table and the renamed column, rehearsed
against the cluster first, and it stays enforced.
Verified: seed re-runs clean, the paper smoke test's 45 checks and the MCP
smoke test's 47 pass, `vue-tsc` and the build are clean, and the UI was driven
in a headless browser through the template list, its preview drawer, the title
library page, the template form (which now posts `placements` with `title_id`)
and the paper document with no console errors.
The server has to be reachable by the Harness, Claude Code and Codex, and all
three of those are clients of an address rather than of a source tree — a
stdio child process only serves whichever machine spawned it. So it is packaged
and deployed: namespace `paper-doc`, one Deployment, one NodePort Service at
`http://192.168.1.88:30931/mcp` behind a bearer token, talking to the
in-cluster TiDB Service rather than the node's NodePort.
The namespace is `paper-doc` and not `paper_doc`, which is not a preference:
a namespace name is an RFC 1123 label, underscores are outside that alphabet,
and the API server rejects it outright. The database keeps its underscore.
Two routes were added to the server for the deployment, and they are the only
unauthenticated ones. `/healthz` answers "is this process alive" without
touching anything, `/readyz` answers "can it serve" by running `SELECT 1`. The
split is the point: a failing readiness probe takes the pod out of the
Service's endpoints and puts it back when TiDB returns, while a failing
liveness probe kills a process that was working correctly.
There is no container runtime on this host — k3s runs containerd and neither
`ctr` nor `crictl` builds from a Dockerfile — so `deploy/build-mcp-image.sh`
builds inside a throwaway privileged docker-in-docker pod, mounting the working
tree through a hostPath so the build context is the tree you are looking at,
uncommitted edits included. It reads the registry credentials out of the
cluster's own pull secret, so no password is in the script or in a shell
history. One Dockerfile serves both entry points, because the API and the MCP
server share the domain layer and differ by one argv.
Verified against the deployed Service, not against a local process: 47 MCP
checks pass over the NodePort, and the Node MCP SDK — the one the Harness's
mcp-client plugin uses — lists all 29 tools and calls them. The same 47 pass in
stdio and against a local HTTP server, and the REST suite still passes its 45.
`docs/MCP.md` is the manual a person needs to put this in front of an agent:
the 29 tools with their arguments, the workflow ("generate, then write it in"
in one call, versus paragraph by paragraph), the error contract, and the exact
configuration for the three clients in play — the Harness's
`cordis.patch.yml`, Claude Code's `claude mcp add` and `.mcp.json`, and Codex's
`config.toml` — including the note that a new Harness entry has to sit inside
the `insert:` list or it is silently treated as an override of an entry that
does not exist.
`overview.md` gains the section that belongs in a design document rather than a
manual: why the MCP surface is shaped differently from the REST one, and the
two rules a new tool has to follow — keep the docstring to a line or two,
because it is sent with every request, and register with
`structured_output=False`, because the inferred envelope sends the same JSON
twice and clients disagree about which copy counts.
The paper is written by a model now. A browser is the wrong client for that:
the work is "generate the paper, then put it in", and doing it through a form
means a person retyping what a model already produced. So the same domain is
served over the Model Context Protocol, which Claude Code, Codex and the
DeepSeek Harness all speak.
It is a third front door, not a second implementation. Every tool is three
lines around an `app.crud` call and validates through `app.schemas`, exactly
as the REST routes do, so a rule fixed in the CRUD layer is fixed on both
surfaces and a paper written by an agent is indistinguishable from one written
by hand. What `app/mcp/` adds is only what a model needs and a browser does
not:
- 29 tools, prefixed `paper_` `paragraph_` `sentence_` `template_` `field_`,
because a model picks a tool out of a list by name rather than by reading 29
descriptions;
- results as compact `None`-free JSON, since a tool result is paid for in
context tokens and `PaperRead.model_dump()` carries four counts and two
timestamps into every list row;
- paragraphs addressed by **heading** as well as by position. Storage is
correct as it stands — a sentence remembers the position it sits at, which is
what makes a template switch non-destructive — but nobody writing
"1. Introduction" knows the template places it at `sort = 20`. The server
translates, and refuses with the real heading list when it cannot, so a model
that guessed wrong corrects itself in one retry;
- `paper_write` and `paper_write_text`: one intention, one call. The latter
finds its own sections from Markdown headings or from lines that name a
template heading, and reports every heading it could not place instead of
writing half a paper;
- `sentence_search` across papers, for consistency rather than retrieval — a
paper that says 洪水损失 should not be joined by one that says GUL;
- `paper_delete` refuses once, naming what would go with it. A cascading delete
has no undo in a tool call.
Two transports, one build. `stdio` is what a client spawns — so nothing in the
process may print to stdout, and diagnostics go to stderr. `streamable-http` is
what a client on another machine connects to, optionally behind a bearer token;
binding a non-loopback address disables the SDK's DNS-rebinding allow-list,
because a LAN client sends whatever Host it knows the server by.
Tools register with `structured_output=False` on purpose: inferred from a
`-> str` annotation the SDK publishes a `{"result": ...}` envelope and sends
the JSON twice, once as `structuredContent` and once as text, and clients that
read only one of the two then disagree about what came back.
`scripts/smoke_mcp.py` drives the whole loop through a real MCP client — the
child process and JSON-RPC over stdin/stdout a client actually uses — and runs
unchanged against a running HTTP server via `--url`. 47 checks pass on both
transports; the REST suite still passes its 45.
An empty paragraph printed 「(本段暂无内容)」 under its heading. A paragraph
with no sentences is not an error to report: it is a position in an outline, and
some of those are meant to be empty — a separator, a position the template
reserves, a section the writer has not reached yet. The words also go with the
paper on export, which is worse: a placeholder is the tool talking, inside the
document.
The structure stays exactly as it was (the heading, the count, the edit button);
only the annotation is gone. Measured on a paper with 19 empty paragraphs: no
placeholder element anywhere, the text absent from the page, and each empty
paragraph renders as its heading alone at 56px.
A paragraph is its sentences concatenated. Nothing was put between them, which
is right for Chinese — 「。」 already separates — and wrong for English, where
`Adaptive capacity rises.` followed by `Relocation follows.` printed as
`...rises.Relocation...`: a sentence boundary the reader cannot see. CJK output
hid it, so it would have surfaced as an English bug later rather than now.
`sentence_separator` now returns "" or a single space per seam, and the document
carries it as `SentenceRead.separator_before`. The rule is about the seam, not
the language: a space goes in unless both sides are CJK. Mixed seams take the
space. Empty sentences take none. It is derived on every read and never stored,
so it cannot drift from the text, and the client prints `separator_before +
content` and adds no spacing of its own.
No splitting was added anywhere, and none exists: a sentence is one line in the
editor and nothing parses it. The single split this project has ever performed
was the one-time move of the old abstract column, which cut after 「。」 only —
conservative on purpose, since an English abstract is better left in one row
than cut at the first `et al.`.
Five smoke checks cover the seam (Chinese, English, the assembled paragraph,
and that no separator is stored inside the content). 45 checks pass.
Each heading carried a small chip with the paragraph's `sort` before its name.
That number is how a sentence is addressed and how a template switch lands
content on the right paragraph, but it is a writing detail: in the document it
is an implementation number sitting in the middle of the paper, in front of
every heading.
The reading view now prints the heading and nothing else. The editor keeps it
where it is acted on: the paragraph dialog names it in its title (`0 Abstract
sort = 1`), every sentence row carries its own `sort` input, and the 所属段落
select lists the positions to move between.
A template's abstract paragraph (`0 Abstract`) *is* the paper's abstract: it has
a position in the document, the heading and typography the template gives it,
and the same sentence-by-sentence editing as everything else. `paper.abstract`
was a second home for that text — one the document never reads, so a paper
could show two different abstracts and the column could drift from the body.
The column is gone, from the table, the model, the schemas, the API payloads,
the paper form, the list subtitle and the paper page. Nothing else changed.
The text already written into it is not gone. Revision a83f5c21d7b6 writes each
stored abstract into the paper's body first — one sentence per 。 at the
paragraph carrying the abstract heading, appended after anything already there
rather than replacing it. A template without such a heading keeps the text too,
one position above its first paragraph, where the document renders it under
未设定. Verified against the one paper that had an abstract: 450 characters in,
450 out, identical including `|J| ≤ α · I⁻ᵝ`, split into six sentences in the
`0 Abstract` paragraph, still with its own edit button.
The smoke test no longer assumes an empty database: it records the paper total
before it starts and compares against that, so it can run on a real one.
The writing surface was capped at 900px and centred, on the reasoning that a
narrower text column reads better. That is true of a document viewer, but this
page is also the editing surface: each paragraph's heading sits on the left of
its row and its edit button on the right, and capping the sheet left those two
floating in the middle of the panel — 385px of nothing on each side at 1920,
and worse on a wider screen. It reads as a broken layout, not a measure.
The sheet now fills the card. Measured in a browser at 2560/1920/1440/1280/1024
wide, the row spans the panel with a 37px inset on both sides at every width,
the edit button lands on the right inset, and nothing overflows horizontally.
Note for later: prose now uses the full width too, which is 106 字/行 at 1920
and 149 字/行 at 2560. If that proves too long to read comfortably, the fix is
a max-width on `.paragraph-body` alone — the heading, the edit button and the
row's hover highlight would stay full width.
A list page is a table with controls around it, and the table was as tall as
its rows: on a large screen most of the window sat empty below the last row
while the page stayed the thing that scrolled. The three list pages now stretch
the table to the height the window offers — the toolbar and the pagination stay
put and the rows scroll inside the table — and give the height back on a small
screen.
Pure CSS, no resize listener and no measured pixel height, so nothing goes
stale when the rail collapses or the window is resized. Four links carry it,
each load-bearing: `.page--fill` is `min-height: 100%` (a floor, not a cap),
`.card--fill` makes the card a column, its body gets `min-height: 0` so a flex
child may shrink below its content, and `.table-fill` is `flex: 1 1 0` so the
table takes the leftover height rather than claiming its content height as its
basis. Element Plus then pins the header and scrolls the rows.
Verified by rendering the pages in a headless browser at 1920x1080, 1440x900
and 1280x620: the table measures 781/601/321px, the page never scrolls, the
rows scroll inside it, header and body columns stay aligned to 0px through a
400px horizontal scroll, the pinned action column stays pinned, and the empty
state is centred in the full-height body rather than stranded at the top. A
1280x420 window hits the 240px floor and falls back to the page scrolling, and
the paper document page is untouched.
Two tables were named after the concept they came from rather than the module
they belong to, so the schema read as if the template tables were part of the
paper module. Renamed (data preserved, `RENAME TABLE` moves rows in place):
paper_template -> template the 模板 module
section_field -> template_field_library the 字段库 the 模板 module owns
The paper tables and `template_field` already followed the rule. The rename
carries through everything that named a module:
models Template, TemplateField, TemplateFieldLibrary
schemas Template*, TemplateFieldLibrary*
crud app/crud/template.py, app/crud/template_field_library.py
API /template-field-library (was /section-fields); handlers are now
named after library entries, which removes the ambiguity with
TemplateField — a placement, a different thing entirely
client src/api/templateFieldLibrary.ts
`paper_template_filed_sort` is deliberately untouched: it is a column of the
paper module, spelled as the feature was specified.
TiDB v8.5 with tidb_enable_foreign_key on — as this cluster runs — enforces
foreign keys rather than ignoring them, so the docs' "TiDB does not enforce
foreign keys" was wrong. Corrected, with what actually follows from it: the
rename was rehearsed (RENAME TABLE carries a referencing constraint along), the
API keeps checking first so a violation names the row instead of surfacing a
driver error, and the ORM cascades stay so behaviour does not depend on a
cluster setting.
Revision f27a1c6d9e04 verified both ways; 40 smoke checks, type-check and build
all pass.
Records the two decisions the paper feature rests on: a sentence is addressed
by the template placement's `sort` rather than by a row id, and citations are
a table rather than a column. Both explain behaviour that otherwise looks
arbitrary — why an empty paper already has its structure, why a template switch
destroys nothing, and why content with no matching heading is still rendered.
Also documents the new endpoints, the /papers/:id route, the paper view's three
jobs, and the smoke test.
The 论文 menu now carries the papers themselves, read from a pinia store, so a
paper opens straight from the rail; every page that changes the list reloads
that store, and a filter box appears once there are more than six.
PapersView is the library — search, create, edit, open, single and batch
delete. PaperDetailView is the writing surface: it renders the server's
document verbatim, with an edit button beside every paragraph (written or not,
because content only ever enters a paper through the editor), citation markers
numbered in reading order to match the 参考文献 list, and a 切换模板 dialog that
previews the impact before re-shaping the document.
The paragraph editor shows one paragraph as the paper will read it — every
sentence on its own line, in sort order, each editable, each able to carry
citations — and writes the paragraph back whole. Whitespace-only lines are
dropped on save; a citation with an empty 引用内容 is refused; and 所属段落
moves the paragraph to another position, appended after what is already there.
A paper is a document written against a template. Its structure is never
copied into it: `paper` points at a template and the outline is read live on
every render, so switching `template_id` re-shapes the whole document in one
write.
Sentences are addressed by *position* rather than by a template row id:
`paper_sentence.paper_template_filed_sort` holds the sort of the placement the
sentence belongs to, and `sort` holds its place inside that paragraph. That
indirection is what makes a template switch non-destructive — a sentence that
remembers "position 7" lands on whatever the new template puts at position 7 —
and it is why a paragraph is any position either the template or the content
mentions: the structure survives with no content, and content whose position
the template does not define is still rendered, in order, under 未设定.
Citations are a table rather than a column, since one sentence may quote
several references. `quote` is required — a citation that does not say what it
quotes is refused with 422 — while `reference_id` is a plain nullable integer
with no foreign key, because the reference library does not exist yet.
Deleting a template a paper is written against is refused with 409 and a count,
matching how the field library refuses to drop a field still in use.
scripts/smoke_papers.py walks the whole loop — create, empty structure, write a
paragraph with citations, switch templates, keep unmatched positions, move a
paragraph, delete — in 40 checks, and cleans up after itself.
The logo block stopped wherever its text happened to end (~140px) while the
rail below it is 208px, so the left edge was a step rather than a column.
Both now read `asideWidth()` from src/layout.ts, so they cannot drift apart,
and collapsing the rail narrows the logo block with it — the wordmark is
dropped and the mark centres in the 56px strip. The block also carries the
rail's separator up through the header, so the column reads as one piece.
The header's own 20px padding had to go for this to work: the logo block has
to start at x=0 to be flush with the rail, so the nav and the right-hand mark
now own their insets. The mark's inset comes from the same `ASIDE_INSET` as
the rail's section title, so the two line up.
A side effect worth keeping: the rail is 208px and the top nav's first item
and .el-main are each inset a further 20px by their own Element Plus defaults,
so the nav text, the page body and the page header all start on one vertical
line.
The widths live in a module rather than CSS custom properties because el-aside
takes its width through a prop and would otherwise fight an inline style.
The rail was on the right. It now leads the row instead, so the section menu
reads as a conventional left-hand navigation column.
Which side it lands on was never a stylesheet decision: `.app-body` is a flex
row and the aside is simply its first child now, with `.app-main` after it.
The only style change that had to follow is the rail's separator, which moves
from its left edge to its right, plus the collapse tooltip's placement.
The global shell contract in base.css now says this explicitly, so the next
person to move the rail knows the side comes from DOM order rather than from a
rule in that file.
Records the decisions that are not visible from the code alone:
- why section_field has no parent_id, and what reuse that buys;
- that template_field.sort is the single source of display order, and
that repeats and ties are deliberately legal;
- that templates reference library fields rather than copying them, which
is what makes a rename propagate;
- that TiDB parses FOREIGN KEY and ignores it, so the integrity lives in
the application layer.
Also adds the API table, the frontend route table, and the seed script.
The shell: the app mark at both ends of the header, the two top-level
entries between them, and the second-level menu on the right. That menu is
driven entirely by `route.meta.section`, so a route declares which menu it
belongs to and a deep link renders the right one on first paint; routes with
no section (the welcome page) show none.
Screens:
/ welcome, with live template and field counts so the
page doubles as a connectivity check
/papers 论文, content still to be decided
/templates/list the template table: search, create, edit, preview,
delete, batch delete, pagination
/templates/fields the field library: the same CRUD, plus level filter
The template form is the centre of it. Fields are chosen from the flat
library by clicking — any subset, any order, the same field more than once —
and the selection is always rendered sorted by `sort`, so editing a number
reorders the outline immediately. Repeats and duplicate sorts are surfaced as
warnings rather than blocked, because a repeat is often deliberate and a tie
is legal; a "自动排序" button renumbers the selection 1..N. The panel on the
right of the picker opens 字段管理 in a second tab, so a half-filled form is
never lost to a navigation.
The preview drawer renders the outline exactly as it will read, at each
field's own size and colour, with the sort value shown alongside — the
quickest way to confirm the ordering before writing against a template.
Also in this change:
- ApiError now extends Error. It was a plain object, so the common
`error instanceof Error ? error.message : ...` idiom fell through to a
generic message and threw away what the server actually said — such as
which template name is already taken. Server messages now reach the user.
- font colours are normalised client-side on blur, matching the backend,
so a hand-typed rgb(255, 0, 0) is tidied up rather than rejected later.
- api/health.ts is removed: the rewritten welcome page was its only
consumer, and the counts already prove connectivity.
Three tables behind the template configuration feature:
section_field the reusable heading library (name, level, font size,
colour)
paper_template a named outline (name, abstract)
template_field the join, and the only home of display order (sort)
The field library is deliberately flat. A `parent_id` would tie a level-2
heading to exactly one level-1 heading, and the point of the feature is that
a field such as "Background" can sit under both "1. Introduction" and
"2. Related Work" — and in any number of templates, at a different position
in each. Hierarchy is expressed only by `level`, which is a rendering hint.
Display order lives entirely in `template_field.sort`. The order fields were
picked in is never stored, so selecting fields out of order and assigning
sorts renders in sort order. Two consequences are intentional and documented
on the model: repeats are allowed (no unique constraint on template+field)
and ties are legal (broken by insertion order, so the ordering is total).
Templates reference library fields rather than copying them, so renaming or
restyling a field updates every template that places it.
TiDB parses FOREIGN KEY and then ignores it, so the constraints are declared
for documentation and the integrity is enforced in the application layer:
deleting a field still in use is refused with the field names, creating a
template against a missing field is refused, and deleting a template removes
its join rows through the ORM's delete-orphan cascade.
Also normalises font_color to #RRGGBB on write (accepting rgb() and
shorthand) and emits font_size as a JSON number rather than pydantic's
default Decimal string.
scripts/seed.py is idempotent and fills the library with a standard academic
outline plus three starter templates.
Element Plus ships no html/body reset, so the browser default
`body { margin: 8px }` applied, while the app shell was declared
`min-height: 100vh`. The document therefore measured one viewport plus
8px: a vertical scrollbar was permanently visible on the right edge even
for short pages, every page was pushed 8px off the intended 20px gutter,
and the window and .el-main both scrolled.
Add src/styles/base.css, imported once from main.ts after Element Plus,
which now owns the reset and the shell contract:
- html/body/#app/.app-shell are exactly one viewport tall with no margin;
body never scrolls, so the window has no scrollbar
- .app-main is the single scroll container (min-height: 0, overflow-y
auto, overflow-x hidden) with a slim low-contrast scrollbar that only
shows when content truly exceeds the viewport
- global guards against content escaping the gutter: media max-width,
wrapping for long code strings
- shared .page helper for a consistent per-view vertical rhythm
Also let the header menu track its content box, so its active underline
stays flush with the header separator instead of spilling past it, and
drop the now-unused RouterLink import.
Backend (backend/):
- FastAPI app with layered app/{api,core,crud,db,models,schemas} structure
- TiDB connection via SQLAlchemy. The URL is built with URL.create rather
than string formatting: the password contains '@', which would otherwise be
parsed as the user/host separator and silently truncate the credential.
- Alembic environment wired to Base.metadata and the app settings, so
backend/.env stays the single source of truth for credentials
- /api/health endpoint reporting database reachability
- models/ and crud/ are intentionally empty: no model classes are defined and
no tables are created, at import time or otherwise
Frontend (frontend/):
- Vue 3 + Vite + TypeScript scaffold (create-vue, --bare)
- axios instance with a normalized error shape
- vue-router with a home route and a catch-all 404
- pinia + pinia-plugin-persistedstate; the app store persists selected keys
- Element Plus with its icons registered globally
- dev proxy forwards /api to the FastAPI service on port 8000
The database paper_doc was created in TiDB out of band; no table exists yet.