litloop 1.18.3__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- litloop/__init__.py +89 -0
- litloop/__main__.py +6 -0
- litloop/_source_hygiene.py +158 -0
- litloop/agents/__init__.py +13 -0
- litloop/agents/client.py +164 -0
- litloop/agents/runner.py +123 -0
- litloop/agents/system.py +54 -0
- litloop/agents/tools.py +142 -0
- litloop/app_server.py +8 -0
- litloop/browser/__init__.py +1 -0
- litloop/browser/backend.py +1009 -0
- litloop/browser/camofox_api.py +368 -0
- litloop/browser/chrome_copy.py +136 -0
- litloop/browser/cloakbrowser_compat.py +72 -0
- litloop/browser/cookies.py +347 -0
- litloop/browser/engine.py +965 -0
- litloop/browser/login.py +636 -0
- litloop/browser/pdf.py +263 -0
- litloop/browser/profile_health.py +167 -0
- litloop/browser/runtime.py +210 -0
- litloop/cache.py +104 -0
- litloop/citations/__init__.py +1 -0
- litloop/citations/bib_import.py +102 -0
- litloop/citations/bibtex_crossref.py +143 -0
- litloop/citations/formats.py +225 -0
- litloop/citations/paper_list.py +200 -0
- litloop/citations/session_export.py +94 -0
- litloop/cli/__init__.py +12 -0
- litloop/cli/_utils.py +267 -0
- litloop/cli/commands.py +1098 -0
- litloop/cli/doctor.py +141 -0
- litloop/cli/main.py +1173 -0
- litloop/cli/web.py +190 -0
- litloop/config.py +413 -0
- litloop/data/VARCON-LICENSE.txt +62 -0
- litloop/data/elements.tsv +119 -0
- litloop/data/institutional_identity_policy.json +165 -0
- litloop/data/publisher_access_catalog.json +969 -0
- litloop/data/publisher_carsi.json +105 -0
- litloop/data/spelling_search.tsv.gz +0 -0
- litloop/data/spelling_us.tsv.gz +0 -0
- litloop/data/webvpn.dat +0 -0
- litloop/data/webvpn.json +603 -0
- litloop/db.py +1922 -0
- litloop/deps.py +128 -0
- litloop/desktop/__init__.py +1 -0
- litloop/desktop/agent_runner.py +537 -0
- litloop/desktop/download_jobs.py +571 -0
- litloop/desktop/error_text.py +192 -0
- litloop/desktop/server.py +999 -0
- litloop/desktop/settings.py +552 -0
- litloop/desktop/translate.py +103 -0
- litloop/diagnostics.py +126 -0
- litloop/download/__init__.py +1 -0
- litloop/download/doi_check.py +44 -0
- litloop/download/fetcher.py +975 -0
- litloop/download/markdown.py +130 -0
- litloop/download/pdf.py +454 -0
- litloop/download/pipeline.py +1166 -0
- litloop/download/progress.py +312 -0
- litloop/download/rename.py +211 -0
- litloop/extractors/__init__.py +0 -0
- litloop/extractors/html_extractor.py +31 -0
- litloop/extractors/pdf_extractor.py +151 -0
- litloop/extractors/publisher_adapters/__init__.py +0 -0
- litloop/extractors/publisher_adapters/acs.py +167 -0
- litloop/extractors/publisher_adapters/elsevier.py +178 -0
- litloop/extractors/publisher_adapters/generic.py +191 -0
- litloop/extractors/publisher_adapters/nature.py +204 -0
- litloop/extractors/publisher_adapters/rsc.py +154 -0
- litloop/extractors/publisher_adapters/tandfonline.py +154 -0
- litloop/extractors/publisher_adapters/wiley.py +160 -0
- litloop/http_utils.py +57 -0
- litloop/identifiers.py +134 -0
- litloop/institutional/__init__.py +1 -0
- litloop/institutional/auth.py +576 -0
- litloop/institutional/carsi_login.py +569 -0
- litloop/institutional/config_adapter.py +63 -0
- litloop/institutional/proxy_check.py +98 -0
- litloop/institutional/remote_assist.py +274 -0
- litloop/institutional/schools.py +102 -0
- litloop/institutional/session_heal.py +73 -0
- litloop/institutional/session_store.py +105 -0
- litloop/log.py +47 -0
- litloop/models.py +269 -0
- litloop/network.py +417 -0
- litloop/optional.py +310 -0
- litloop/publishers/__init__.py +1 -0
- litloop/publishers/access.py +232 -0
- litloop/publishers/pdf_router.py +451 -0
- litloop/publishers/profiles.py +899 -0
- litloop/publishers/strategies/__init__.py +122 -0
- litloop/publishers/strategies/_core.py +2340 -0
- litloop/publishers/strategies/_elsevier_api.py +578 -0
- litloop/publishers/strategies/acm.py +27 -0
- litloop/publishers/strategies/acs.py +42 -0
- litloop/publishers/strategies/aip.py +27 -0
- litloop/publishers/strategies/aps.py +29 -0
- litloop/publishers/strategies/asce.py +28 -0
- litloop/publishers/strategies/base.py +108 -0
- litloop/publishers/strategies/copernicus.py +24 -0
- litloop/publishers/strategies/elsevier.py +66 -0
- litloop/publishers/strategies/generic.py +39 -0
- litloop/publishers/strategies/ieee.py +28 -0
- litloop/publishers/strategies/iop.py +27 -0
- litloop/publishers/strategies/nature.py +27 -0
- litloop/publishers/strategies/oxford.py +27 -0
- litloop/publishers/strategies/registry.py +83 -0
- litloop/publishers/strategies/royalsociety.py +28 -0
- litloop/publishers/strategies/rsc.py +27 -0
- litloop/publishers/strategies/sage.py +30 -0
- litloop/publishers/strategies/science.py +27 -0
- litloop/publishers/strategies/springer.py +34 -0
- litloop/publishers/strategies/tandfonline.py +32 -0
- litloop/publishers/strategies/wiley.py +32 -0
- litloop/publishers/supplementary.py +219 -0
- litloop/quota.py +202 -0
- litloop/search/__init__.py +1 -0
- litloop/search/abstracts.py +291 -0
- litloop/search/deep.py +1216 -0
- litloop/search/quick.py +1102 -0
- litloop/search/scansci_find.py +410 -0
- litloop/search/screening.py +225 -0
- litloop/search/title_to_doi.py +163 -0
- litloop/search/watches.py +124 -0
- litloop/server.py +1731 -0
- litloop/sources/__init__.py +1339 -0
- litloop/sources/_batch.py +529 -0
- litloop/sources/_index.py +221 -0
- litloop/sources/_routes.py +290 -0
- litloop/sources/altcha.py +90 -0
- litloop/sources/arxiv.py +45 -0
- litloop/sources/attempt.py +91 -0
- litloop/sources/carsi.py +74 -0
- litloop/sources/core_api.py +46 -0
- litloop/sources/crossref.py +162 -0
- litloop/sources/elsevier_api.py +620 -0
- litloop/sources/elsevier_check.py +274 -0
- litloop/sources/europepmc.py +81 -0
- litloop/sources/ezproxy.py +317 -0
- litloop/sources/instsci.py +1295 -0
- litloop/sources/libgen.py +116 -0
- litloop/sources/nature.py +39 -0
- litloop/sources/oa_discovery.py +135 -0
- litloop/sources/openaire.py +122 -0
- litloop/sources/openalex.py +100 -0
- litloop/sources/publishers.py +684 -0
- litloop/sources/sage_cn.py +139 -0
- litloop/sources/scibban.py +10 -0
- litloop/sources/scihub.py +1379 -0
- litloop/sources/scihub_domains.py +203 -0
- litloop/sources/scihub_speed.py +187 -0
- litloop/sources/scoring.py +250 -0
- litloop/sources/semantic_scholar.py +146 -0
- litloop/sources/springer_tdm.py +172 -0
- litloop/sources/unpaywall.py +116 -0
- litloop/sources/vpnsci.py +38 -0
- litloop/store.py +264 -0
- litloop/templates/index.html +347 -0
- litloop/terms/__init__.py +1 -0
- litloop/terms/glossary.py +422 -0
- litloop/terms/spelling.py +89 -0
- litloop/terms/text.py +106 -0
- litloop/tor/__init__.py +1 -0
- litloop/tor/embedded.py +371 -0
- litloop/tor/proxy.py +110 -0
- litloop/updates.py +148 -0
- litloop/zotero/__init__.py +1 -0
- litloop/zotero/bridge.py +309 -0
- litloop/zotero/client.py +317 -0
- litloop-1.18.3.dist-info/METADATA +347 -0
- litloop-1.18.3.dist-info/RECORD +176 -0
- litloop-1.18.3.dist-info/WHEEL +5 -0
- litloop-1.18.3.dist-info/entry_points.txt +2 -0
- litloop-1.18.3.dist-info/licenses/LICENSE +197 -0
- litloop-1.18.3.dist-info/top_level.txt +1 -0
|
@@ -0,0 +1,999 @@
|
|
|
1
|
+
"""Local HTTP backend for the LitLoop desktop app.
|
|
2
|
+
|
|
3
|
+
The Tauri shell starts this as a child process:
|
|
4
|
+
|
|
5
|
+
python -m litloop.desktop.server --token <random>
|
|
6
|
+
|
|
7
|
+
It listens on 127.0.0.1 only, on a free port it picks itself, and prints one
|
|
8
|
+
line to stdout before serving, so the shell knows where to connect:
|
|
9
|
+
|
|
10
|
+
{"litloop_app_server": 1, "port": 51234}
|
|
11
|
+
|
|
12
|
+
Every request must carry the token (header ``X-LitLoop-Token``), so another
|
|
13
|
+
web page in the user's browser cannot call it. Data comes from the files
|
|
14
|
+
described in docs/APP_DATA.md, read and written through ``store``.
|
|
15
|
+
"""
|
|
16
|
+
|
|
17
|
+
from __future__ import annotations
|
|
18
|
+
|
|
19
|
+
import argparse
|
|
20
|
+
import json
|
|
21
|
+
import re
|
|
22
|
+
import secrets
|
|
23
|
+
import socket
|
|
24
|
+
import sys
|
|
25
|
+
import time
|
|
26
|
+
from typing import Any
|
|
27
|
+
|
|
28
|
+
from fastapi import FastAPI, HTTPException, Request
|
|
29
|
+
from fastapi.middleware.cors import CORSMiddleware
|
|
30
|
+
from fastapi.responses import JSONResponse
|
|
31
|
+
from pydantic import BaseModel
|
|
32
|
+
|
|
33
|
+
from .. import __version__, db, store
|
|
34
|
+
from ..search import screening
|
|
35
|
+
|
|
36
|
+
# Where the app's pages are served from: the Vite dev server, and Tauri's
|
|
37
|
+
# own origins (Windows uses http://tauri.localhost, macOS/Linux tauri://localhost).
|
|
38
|
+
ORIGINS = ["http://localhost:1420", "http://tauri.localhost", "https://tauri.localhost", "tauri://localhost"]
|
|
39
|
+
_SESSION_ID = re.compile(r"^[\w一-鿿.-]{1,200}$")
|
|
40
|
+
|
|
41
|
+
|
|
42
|
+
class Rating(BaseModel):
|
|
43
|
+
doi: str | None = None
|
|
44
|
+
title: str | None = None
|
|
45
|
+
relevance: str | None = None # None clears the user's rating: the agent's counts again
|
|
46
|
+
|
|
47
|
+
|
|
48
|
+
class Setting(BaseModel):
|
|
49
|
+
key: str
|
|
50
|
+
value: Any = None
|
|
51
|
+
|
|
52
|
+
|
|
53
|
+
class Paper(BaseModel):
|
|
54
|
+
doi: str
|
|
55
|
+
title: str | None = None
|
|
56
|
+
year: int | None = None
|
|
57
|
+
journal: str | None = None
|
|
58
|
+
|
|
59
|
+
|
|
60
|
+
class SkipPaper(BaseModel):
|
|
61
|
+
doi: str
|
|
62
|
+
|
|
63
|
+
|
|
64
|
+
class StartQueue(BaseModel):
|
|
65
|
+
publisher: str
|
|
66
|
+
|
|
67
|
+
|
|
68
|
+
class NewSearch(BaseModel):
|
|
69
|
+
"""The app's search form; agent_runner.build_prompt turns it into the message."""
|
|
70
|
+
question: str
|
|
71
|
+
criteria: str = ""
|
|
72
|
+
mode: str = "quick"
|
|
73
|
+
limit: int = 20
|
|
74
|
+
batches: int = 0
|
|
75
|
+
year_from: int | None = None
|
|
76
|
+
year_to: int | None = None
|
|
77
|
+
max_cas_zone: int | None = None
|
|
78
|
+
max_jcr_quartile: int | None = None
|
|
79
|
+
exclude_warning: bool = False
|
|
80
|
+
seeds: str = ""
|
|
81
|
+
seed_collection: str = ""
|
|
82
|
+
|
|
83
|
+
|
|
84
|
+
class ResumeSearch(BaseModel):
|
|
85
|
+
batches: int = 0
|
|
86
|
+
|
|
87
|
+
|
|
88
|
+
class NewDownload(BaseModel):
|
|
89
|
+
papers: list[Paper]
|
|
90
|
+
title: str
|
|
91
|
+
session: str | None = None
|
|
92
|
+
zotero_target: str | None = None
|
|
93
|
+
|
|
94
|
+
|
|
95
|
+
class PastedList(BaseModel):
|
|
96
|
+
text: str
|
|
97
|
+
|
|
98
|
+
|
|
99
|
+
class ReferenceLines(BaseModel):
|
|
100
|
+
lines: list[str]
|
|
101
|
+
|
|
102
|
+
|
|
103
|
+
class ZoteroCheck(BaseModel):
|
|
104
|
+
key: str
|
|
105
|
+
|
|
106
|
+
|
|
107
|
+
class ZoteroRepair(BaseModel):
|
|
108
|
+
item_key: str
|
|
109
|
+
doi: str | None = None
|
|
110
|
+
trash: list[str] = []
|
|
111
|
+
attach: bool = True # False: only move the empty attachments to the Trash
|
|
112
|
+
|
|
113
|
+
|
|
114
|
+
class ZoteroPush(BaseModel):
|
|
115
|
+
collection: str | None = None # None: <zotero_collection_root>/<job title>
|
|
116
|
+
tags: list[str] = []
|
|
117
|
+
|
|
118
|
+
|
|
119
|
+
class OpenPath(BaseModel):
|
|
120
|
+
path: str
|
|
121
|
+
|
|
122
|
+
|
|
123
|
+
class Onboarding(BaseModel):
|
|
124
|
+
done: bool = True
|
|
125
|
+
|
|
126
|
+
|
|
127
|
+
class GlossaryEntry(BaseModel):
|
|
128
|
+
terms: list[str] | None = None
|
|
129
|
+
note: str | None = None
|
|
130
|
+
confirm: bool = False # a suggestion becomes an entry in use
|
|
131
|
+
|
|
132
|
+
|
|
133
|
+
class BuiltinSwitch(BaseModel):
|
|
134
|
+
id: str # "element:<symbol>" or "spelling:<American word>"
|
|
135
|
+
enabled: bool
|
|
136
|
+
|
|
137
|
+
|
|
138
|
+
class Translate(BaseModel):
|
|
139
|
+
doi: str | None = None
|
|
140
|
+
title: str | None = None
|
|
141
|
+
|
|
142
|
+
|
|
143
|
+
class Export(BaseModel):
|
|
144
|
+
format: str = "bibtex" # bibtex | ris | doi
|
|
145
|
+
which: str = "selected" # selected | high
|
|
146
|
+
|
|
147
|
+
|
|
148
|
+
class Watch(BaseModel):
|
|
149
|
+
session: str
|
|
150
|
+
query: str = ""
|
|
151
|
+
|
|
152
|
+
|
|
153
|
+
class Decision(BaseModel):
|
|
154
|
+
doi: str | None = None
|
|
155
|
+
title: str | None = None
|
|
156
|
+
decision: str | None = None # "selected", "rejected" or None
|
|
157
|
+
reason: str | None = None # why, when rejecting; optional
|
|
158
|
+
|
|
159
|
+
|
|
160
|
+
class BatchOp(BaseModel):
|
|
161
|
+
doi: str | None = None
|
|
162
|
+
title: str | None = None
|
|
163
|
+
relevance: str | None = None
|
|
164
|
+
decision: str | None = None
|
|
165
|
+
reject_reason: str | None = None
|
|
166
|
+
clear_user_relevance: bool = False
|
|
167
|
+
delete_if_unrated: bool = False
|
|
168
|
+
|
|
169
|
+
|
|
170
|
+
class BatchUpdate(BaseModel):
|
|
171
|
+
ops: list[BatchOp]
|
|
172
|
+
|
|
173
|
+
|
|
174
|
+
_OPENALEX: dict[str, Any] = {"at": 0.0, "data": None} # see /api/quota/openalex
|
|
175
|
+
_APP_UPDATE: dict[str, Any] = {"at": 0.0, "data": None} # see /api/app-update
|
|
176
|
+
_APP_UPDATE_URL = "https://api.github.com/repos/dysprosium231/litloop-paper-agent/releases/latest"
|
|
177
|
+
_WRITE_MODE: dict[str, Any] = {} # zotero-mcp's write mode, cached until the user authorizes
|
|
178
|
+
|
|
179
|
+
|
|
180
|
+
def _parse_version_tuple(version: str) -> tuple[int, ...] | None:
|
|
181
|
+
"""Extract numeric version parts (e.g. 'v1.18.3' -> (1, 18, 3))."""
|
|
182
|
+
cleaned = version.strip().lstrip("vV")
|
|
183
|
+
if not cleaned:
|
|
184
|
+
return None
|
|
185
|
+
parts: list[int] = []
|
|
186
|
+
for piece in re.split(r"[.+_-]", cleaned):
|
|
187
|
+
if piece.isdigit():
|
|
188
|
+
parts.append(int(piece))
|
|
189
|
+
else:
|
|
190
|
+
m = re.match(r"^(\d+)", piece)
|
|
191
|
+
if m:
|
|
192
|
+
parts.append(int(m.group(1)))
|
|
193
|
+
break
|
|
194
|
+
return tuple(parts) if parts else None
|
|
195
|
+
|
|
196
|
+
|
|
197
|
+
def _explain(d: dict[str, Any] | None, action: str = "下载", full: bool = False) -> None:
|
|
198
|
+
"""Add ``error_text`` (docs/UI_COPY.md section 7) next to a raw ``error``; the raw text stays for the tooltip."""
|
|
199
|
+
from . import error_text
|
|
200
|
+
if d and d.get("error"):
|
|
201
|
+
d["error_text"] = error_text.message(d["error"], action) if full else error_text.reason(d["error"])
|
|
202
|
+
if not full:
|
|
203
|
+
hint = error_text.next_step(d["error"])
|
|
204
|
+
if hint:
|
|
205
|
+
d["error_hint"] = hint
|
|
206
|
+
act = error_text.action_for(d["error"])
|
|
207
|
+
if act:
|
|
208
|
+
d["error_action"] = act
|
|
209
|
+
|
|
210
|
+
|
|
211
|
+
def _fail(status: int, raw: str | None, action: str, fallback: str) -> HTTPException:
|
|
212
|
+
from . import error_text
|
|
213
|
+
return HTTPException(status, error_text.message(raw or fallback, action))
|
|
214
|
+
|
|
215
|
+
|
|
216
|
+
def _record_path(session: str) -> str:
|
|
217
|
+
if not _SESSION_ID.match(session) or session.startswith("."):
|
|
218
|
+
raise HTTPException(400, "无法打开检索会话:会话编号不正确")
|
|
219
|
+
if not db.session_exists(session):
|
|
220
|
+
raise HTTPException(404, "无法打开检索会话:此会话已不存在")
|
|
221
|
+
return session
|
|
222
|
+
|
|
223
|
+
|
|
224
|
+
def _find(items: list[dict[str, Any]], doi: str | None, title: str | None) -> dict[str, Any]:
|
|
225
|
+
key = screening._key({"doi": doi, "title": title})
|
|
226
|
+
for it in items:
|
|
227
|
+
if screening._key(it) == key:
|
|
228
|
+
return it
|
|
229
|
+
raise HTTPException(404, "此检索会话中没有这篇文献")
|
|
230
|
+
|
|
231
|
+
|
|
232
|
+
def create_app(token: str) -> FastAPI:
|
|
233
|
+
app = FastAPI(title="LitLoop app backend", version=__version__, docs_url=None, redoc_url=None)
|
|
234
|
+
app.add_middleware(CORSMiddleware, allow_origins=ORIGINS, allow_methods=["GET", "POST"],
|
|
235
|
+
allow_headers=["X-LitLoop-Token", "Content-Type"])
|
|
236
|
+
|
|
237
|
+
@app.middleware("http")
|
|
238
|
+
async def require_token(request: Request, call_next):
|
|
239
|
+
if request.method != "OPTIONS" and not secrets.compare_digest(
|
|
240
|
+
request.headers.get("X-LitLoop-Token", ""), token):
|
|
241
|
+
return JSONResponse({"detail": "missing or wrong token"}, status_code=401)
|
|
242
|
+
return await call_next(request)
|
|
243
|
+
|
|
244
|
+
@app.get("/api/health")
|
|
245
|
+
def health() -> dict[str, Any]:
|
|
246
|
+
return {"version": __version__, "schema": store.SCHEMA}
|
|
247
|
+
|
|
248
|
+
@app.get("/api/sessions")
|
|
249
|
+
def sessions(limit: int = 50) -> list[dict[str, Any]]:
|
|
250
|
+
from ..search import deep as deep_search
|
|
251
|
+
out = screening.list_sessions(limit)
|
|
252
|
+
for s in out:
|
|
253
|
+
s["deep"] = (db.load_summary(s["session"]) is not None) or deep_search.has_pool(s["session"])
|
|
254
|
+
return out
|
|
255
|
+
|
|
256
|
+
@app.get("/api/sessions/{session}")
|
|
257
|
+
def session(session: str) -> dict[str, Any]:
|
|
258
|
+
from ..search import deep as deep_search
|
|
259
|
+
_record_path(session)
|
|
260
|
+
record = db.load_screening(session)
|
|
261
|
+
summary = db.load_summary(session)
|
|
262
|
+
if summary is not None:
|
|
263
|
+
# The file keeps the agent's English notes; the app shows today's, in Chinese.
|
|
264
|
+
from .. import quota
|
|
265
|
+
summary["api_warnings"] = quota.warnings("zh")
|
|
266
|
+
return {"record": record, "summary": summary, "resumable": deep_search.has_pool(session)}
|
|
267
|
+
|
|
268
|
+
@app.get("/api/sessions/{session}/progress")
|
|
269
|
+
def progress(session: str) -> dict[str, Any] | None:
|
|
270
|
+
_record_path(session)
|
|
271
|
+
p = store.read_progress(session)
|
|
272
|
+
if p and p.get("error"):
|
|
273
|
+
_explain(p, "检索", full=True)
|
|
274
|
+
return p
|
|
275
|
+
|
|
276
|
+
@app.get("/api/sessions/{session}/prisma")
|
|
277
|
+
def prisma_view(session: str) -> dict[str, Any]:
|
|
278
|
+
_record_path(session)
|
|
279
|
+
return screening.prisma(session)
|
|
280
|
+
|
|
281
|
+
@app.post("/api/sessions/{session}/rating")
|
|
282
|
+
def rate(session: str, body: Rating) -> dict[str, Any]:
|
|
283
|
+
if body.relevance is not None and body.relevance not in screening.RELEVANCE:
|
|
284
|
+
raise HTTPException(400, f"无法评分:评分只能是{'、'.join(screening.RELEVANCE)}")
|
|
285
|
+
_record_path(session)
|
|
286
|
+
res = db.update_item_rating(session, body.doi, body.title, body.relevance)
|
|
287
|
+
if res is None:
|
|
288
|
+
if body.relevance is None:
|
|
289
|
+
raise HTTPException(404, "此检索会话中没有这篇文献")
|
|
290
|
+
from ..search import deep as deep_search
|
|
291
|
+
try:
|
|
292
|
+
paper = deep_search.batch_paper(session, body.doi, body.title)
|
|
293
|
+
except FileNotFoundError:
|
|
294
|
+
paper = None
|
|
295
|
+
if not paper:
|
|
296
|
+
raise HTTPException(404, "此检索会话中没有这篇文献")
|
|
297
|
+
res = db.update_item_rating(
|
|
298
|
+
session, body.doi, body.title, body.relevance, create_paper=paper
|
|
299
|
+
)
|
|
300
|
+
if res is None:
|
|
301
|
+
raise HTTPException(404, "此检索会话中没有这篇文献")
|
|
302
|
+
return res
|
|
303
|
+
|
|
304
|
+
@app.post("/api/sessions/{session}/next")
|
|
305
|
+
def next_round(session: str) -> dict[str, Any]:
|
|
306
|
+
"""The app's continue button: serve the next batch here, without the agent (it rates the rest when it is back)."""
|
|
307
|
+
import threading
|
|
308
|
+
import time
|
|
309
|
+
|
|
310
|
+
from ..search import deep as deep_search
|
|
311
|
+
_record_path(session)
|
|
312
|
+
if not deep_search.has_pool(session):
|
|
313
|
+
raise HTTPException(400, "无法继续检索:这不是深度检索会话。")
|
|
314
|
+
p = store.read_progress(session)
|
|
315
|
+
if p and p.get("state") == "running" and (p.get("pid") or time.time() - float(p.get("updated") or 0) < 300):
|
|
316
|
+
raise HTTPException(409, "正在生成下一批,完成后再继续。")
|
|
317
|
+
|
|
318
|
+
def run() -> None:
|
|
319
|
+
try:
|
|
320
|
+
deep_search.next_batch(session, force=True)
|
|
321
|
+
except Exception: # the progress file says it failed; the app shows that
|
|
322
|
+
pass
|
|
323
|
+
threading.Thread(target=run, name=f"next-{session}", daemon=True).start()
|
|
324
|
+
return {"started": True}
|
|
325
|
+
|
|
326
|
+
@app.post("/api/sessions/{session}/resume")
|
|
327
|
+
def resume_session(session: str, body: ResumeSearch | None = None) -> dict[str, Any]:
|
|
328
|
+
"""Resume a deep-search session through the configured agent."""
|
|
329
|
+
from . import agent_runner
|
|
330
|
+
_record_path(session)
|
|
331
|
+
try:
|
|
332
|
+
return agent_runner.resume(session, batches=body.batches if body else 0)
|
|
333
|
+
except ValueError as exc:
|
|
334
|
+
raise HTTPException(400, str(exc)) from exc
|
|
335
|
+
except RuntimeError as exc:
|
|
336
|
+
raise HTTPException(409, str(exc)) from exc
|
|
337
|
+
|
|
338
|
+
@app.get("/api/updates")
|
|
339
|
+
def updates_available() -> list[dict[str, Any]]:
|
|
340
|
+
"""Components with a newer version (from the weekly check; a due check starts in the background)."""
|
|
341
|
+
from .. import updates
|
|
342
|
+
return updates.available()
|
|
343
|
+
|
|
344
|
+
@app.get("/api/history")
|
|
345
|
+
def history() -> list[dict[str, Any]]:
|
|
346
|
+
"""Every screening session with its counts and its final picks (the app's history page)."""
|
|
347
|
+
return screening.history()
|
|
348
|
+
|
|
349
|
+
@app.get("/api/watches")
|
|
350
|
+
def watches() -> list[dict[str, Any]]:
|
|
351
|
+
"""Saved searches, newest check first (定期追踪)."""
|
|
352
|
+
from ..search import watches as saved
|
|
353
|
+
return saved.list_watches()
|
|
354
|
+
|
|
355
|
+
@app.post("/api/watches")
|
|
356
|
+
def add_watch(body: Watch) -> dict[str, Any]:
|
|
357
|
+
"""Remember this session's last search so later checks report only new papers."""
|
|
358
|
+
from ..search import watches as saved
|
|
359
|
+
record = screening.load(body.session)
|
|
360
|
+
searches = record.get("searches") or []
|
|
361
|
+
last = searches[-1] if searches else {}
|
|
362
|
+
core = list((last.get("params") or {}).get("core_terms") or [])
|
|
363
|
+
query = body.query or str(last.get("query") or "") or str(record.get("question") or "")
|
|
364
|
+
try:
|
|
365
|
+
return saved.add(record.get("question", ""), query,
|
|
366
|
+
core, record.get("items") or [], body.session)
|
|
367
|
+
except ValueError as exc:
|
|
368
|
+
raise HTTPException(400, str(exc)) from exc
|
|
369
|
+
|
|
370
|
+
@app.delete("/api/watches/{wid}")
|
|
371
|
+
def delete_watch(wid: str) -> dict[str, Any]:
|
|
372
|
+
from ..search import watches as saved
|
|
373
|
+
if not saved.remove(wid):
|
|
374
|
+
raise HTTPException(404, "没有这个追踪")
|
|
375
|
+
return {"removed": True}
|
|
376
|
+
|
|
377
|
+
@app.post("/api/watches/check-due")
|
|
378
|
+
def check_due_watches() -> dict[str, Any]:
|
|
379
|
+
"""Check the watches that are older than watch_check_days (the app calls this once on open)."""
|
|
380
|
+
from ..search import watches
|
|
381
|
+
return watches.check_due()
|
|
382
|
+
|
|
383
|
+
@app.post("/api/watches/{wid}/check")
|
|
384
|
+
def check_watch(wid: str) -> dict[str, Any]:
|
|
385
|
+
"""Run the saved search now; only papers this watch has not seen come back."""
|
|
386
|
+
from ..search import watches as saved
|
|
387
|
+
try:
|
|
388
|
+
return saved.check(wid)
|
|
389
|
+
except ValueError as exc:
|
|
390
|
+
raise HTTPException(404, str(exc)) from exc
|
|
391
|
+
except Exception as exc: # network, quota, source outage
|
|
392
|
+
raise HTTPException(502, f"无法检查追踪:{exc}") from exc
|
|
393
|
+
|
|
394
|
+
@app.post("/api/sessions/{session}/prisma/export")
|
|
395
|
+
def export_prisma(session: str) -> dict[str, Any]:
|
|
396
|
+
"""Write the session's flow counts as Markdown under <output_dir>/exports."""
|
|
397
|
+
from ..citations import session_export as export
|
|
398
|
+
_record_path(session)
|
|
399
|
+
return export.prisma_export(session)
|
|
400
|
+
|
|
401
|
+
@app.post("/api/sessions/{session}/export")
|
|
402
|
+
def export_session(session: str, body: Export) -> dict[str, Any]:
|
|
403
|
+
"""Write the session's picks (or its 高 papers) as BibTeX / RIS / a DOI list under <output_dir>/exports."""
|
|
404
|
+
from ..citations import session_export as export
|
|
405
|
+
_record_path(session)
|
|
406
|
+
if body.which not in ("selected", "high"):
|
|
407
|
+
raise HTTPException(400, "无法导出:只能导出待下载或评为高的文献")
|
|
408
|
+
try:
|
|
409
|
+
return export.export(session, body.format, body.which)
|
|
410
|
+
except ValueError as exc:
|
|
411
|
+
raise HTTPException(400, "无法导出:没有可导出的文献" if "no papers" in str(exc) else "无法导出:不支持此格式") from exc
|
|
412
|
+
|
|
413
|
+
@app.get("/api/translation")
|
|
414
|
+
def translation() -> dict[str, Any]:
|
|
415
|
+
"""What the detail pane needs: who translates, which ratings without asking, whether it can."""
|
|
416
|
+
from . import translate
|
|
417
|
+
from ..config import load_config
|
|
418
|
+
config = load_config()
|
|
419
|
+
name = translate.provider(config)
|
|
420
|
+
return {"provider": name, "auto": screening.translate_wanted(config),
|
|
421
|
+
"ready": name != "agent" and bool(str(config.get(f"{name}_api_key") or "").strip())}
|
|
422
|
+
|
|
423
|
+
@app.post("/api/sessions/{session}/translate")
|
|
424
|
+
def translate_abstract(session: str, body: Translate) -> dict[str, Any]:
|
|
425
|
+
"""Translate one paper's abstract with DeepSeek / DeepL and keep it in the record."""
|
|
426
|
+
from . import translate
|
|
427
|
+
_record_path(session)
|
|
428
|
+
item = _find(db.load_screening(session)["items"], body.doi, body.title)
|
|
429
|
+
if item.get("abstract_zh"):
|
|
430
|
+
return {"abstract_zh": item["abstract_zh"]}
|
|
431
|
+
if not item.get("abstract"):
|
|
432
|
+
raise HTTPException(400, "无法翻译:此文献暂无摘要")
|
|
433
|
+
try:
|
|
434
|
+
text = translate.translate_text(item["abstract"])
|
|
435
|
+
except translate.TranslationError as exc:
|
|
436
|
+
raise HTTPException(502, f"无法翻译:{exc}。") from exc
|
|
437
|
+
if not db.set_item_abstract_zh(session, body.doi, body.title, text):
|
|
438
|
+
raise HTTPException(404, "此检索会话中没有这篇文献")
|
|
439
|
+
return {"abstract_zh": text}
|
|
440
|
+
|
|
441
|
+
@app.post("/api/sessions/{session}/decision")
|
|
442
|
+
def decide(session: str, body: Decision) -> dict[str, Any]:
|
|
443
|
+
if body.decision not in (None, "selected", "rejected"):
|
|
444
|
+
raise HTTPException(400, "无法保存:只能加入下载或排除")
|
|
445
|
+
_record_path(session)
|
|
446
|
+
res = db.update_item_decision(session, body.doi, body.title, body.decision, body.reason)
|
|
447
|
+
if res is None:
|
|
448
|
+
raise HTTPException(404, "此检索会话中没有这篇文献")
|
|
449
|
+
return res
|
|
450
|
+
|
|
451
|
+
@app.post("/api/sessions/{session}/batch")
|
|
452
|
+
def batch_update(session: str, body: BatchUpdate) -> dict[str, Any]:
|
|
453
|
+
_record_path(session)
|
|
454
|
+
if not body.ops:
|
|
455
|
+
raise HTTPException(400, "无法批量保存:未选择文献")
|
|
456
|
+
for op in body.ops:
|
|
457
|
+
if "relevance" in op.model_fields_set and op.relevance is not None and op.relevance not in screening.RELEVANCE:
|
|
458
|
+
raise HTTPException(400, f"无法评分:评分只能是{'、'.join(screening.RELEVANCE)}")
|
|
459
|
+
if "decision" in op.model_fields_set and op.decision not in (None, "selected", "rejected"):
|
|
460
|
+
raise HTTPException(400, "无法保存:只能加入下载或排除")
|
|
461
|
+
|
|
462
|
+
record = db.load_screening(session)
|
|
463
|
+
existing_keys = {screening._key(it) for it in record["items"]}
|
|
464
|
+
create_papers: dict[str, dict[str, Any]] = {}
|
|
465
|
+
for op in body.ops:
|
|
466
|
+
if op.delete_if_unrated or op.clear_user_relevance:
|
|
467
|
+
continue
|
|
468
|
+
if "relevance" not in op.model_fields_set or op.relevance is None:
|
|
469
|
+
continue
|
|
470
|
+
key = screening._key({"doi": op.doi, "title": op.title})
|
|
471
|
+
if key in existing_keys or key in create_papers:
|
|
472
|
+
continue
|
|
473
|
+
from ..search import deep as deep_search
|
|
474
|
+
try:
|
|
475
|
+
paper = deep_search.batch_paper(session, op.doi, op.title)
|
|
476
|
+
except FileNotFoundError:
|
|
477
|
+
paper = None
|
|
478
|
+
if paper:
|
|
479
|
+
create_papers[key] = paper
|
|
480
|
+
|
|
481
|
+
return db.batch_update_items(session, body.ops, create_papers=create_papers)
|
|
482
|
+
|
|
483
|
+
@app.get("/api/glossary")
|
|
484
|
+
def glossary_list() -> dict[str, Any]:
|
|
485
|
+
from ..terms import glossary
|
|
486
|
+
return {"active": glossary.entries("active"), "suggested": glossary.entries("suggested")}
|
|
487
|
+
|
|
488
|
+
@app.post("/api/glossary")
|
|
489
|
+
def glossary_add(body: GlossaryEntry) -> dict[str, Any]:
|
|
490
|
+
from ..terms import glossary
|
|
491
|
+
try:
|
|
492
|
+
return glossary.add(body.terms or [], note=body.note or "")
|
|
493
|
+
except ValueError as exc:
|
|
494
|
+
raise HTTPException(400, str(exc)) from exc
|
|
495
|
+
|
|
496
|
+
@app.get("/api/glossary/builtin")
|
|
497
|
+
def glossary_builtin(q: str = "") -> dict[str, Any]:
|
|
498
|
+
from ..terms import glossary
|
|
499
|
+
return glossary.builtin(q)
|
|
500
|
+
|
|
501
|
+
@app.post("/api/glossary/builtin")
|
|
502
|
+
def glossary_builtin_switch(body: BuiltinSwitch) -> dict[str, Any]:
|
|
503
|
+
from ..terms import glossary
|
|
504
|
+
try:
|
|
505
|
+
glossary.set_builtin(body.id, body.enabled)
|
|
506
|
+
except KeyError as exc:
|
|
507
|
+
raise HTTPException(404, str(exc).strip("'\"")) from exc
|
|
508
|
+
return {"id": body.id, "disabled": not body.enabled}
|
|
509
|
+
|
|
510
|
+
@app.post("/api/glossary/{entry_id}")
|
|
511
|
+
def glossary_update(entry_id: str, body: GlossaryEntry) -> dict[str, Any]:
|
|
512
|
+
from ..terms import glossary
|
|
513
|
+
try:
|
|
514
|
+
return glossary.update(entry_id, terms=body.terms, note=body.note, confirm=body.confirm)
|
|
515
|
+
except KeyError as exc:
|
|
516
|
+
raise HTTPException(404, str(exc).strip("'\"")) from exc
|
|
517
|
+
except ValueError as exc:
|
|
518
|
+
raise HTTPException(400, str(exc)) from exc
|
|
519
|
+
|
|
520
|
+
@app.post("/api/glossary/{entry_id}/delete")
|
|
521
|
+
def glossary_delete(entry_id: str) -> dict[str, Any]:
|
|
522
|
+
"""POST: the app's CORS allows GET and POST only. A deleted suggestion is not suggested again."""
|
|
523
|
+
from ..terms import glossary
|
|
524
|
+
try:
|
|
525
|
+
glossary.delete(entry_id)
|
|
526
|
+
except KeyError as exc:
|
|
527
|
+
raise HTTPException(404, str(exc).strip("'\"")) from exc
|
|
528
|
+
return {"deleted": entry_id}
|
|
529
|
+
|
|
530
|
+
@app.get("/api/quota")
|
|
531
|
+
def quota_summary() -> dict[str, Any]:
|
|
532
|
+
from .. import quota
|
|
533
|
+
return quota.summary()
|
|
534
|
+
|
|
535
|
+
@app.get("/api/quota/openalex")
|
|
536
|
+
def quota_openalex(fresh: bool = False) -> dict[str, Any] | None:
|
|
537
|
+
"""OpenAlex's own count (costs no credits, but takes seconds): kept two minutes."""
|
|
538
|
+
import time
|
|
539
|
+
|
|
540
|
+
from .. import quota
|
|
541
|
+
if fresh or time.time() - _OPENALEX["at"] > 120:
|
|
542
|
+
_OPENALEX.update(at=time.time(), data=quota.openalex_live())
|
|
543
|
+
return _OPENALEX["data"]
|
|
544
|
+
|
|
545
|
+
@app.get("/api/settings")
|
|
546
|
+
def settings_list() -> list[dict[str, Any]]:
|
|
547
|
+
from . import settings as app_settings
|
|
548
|
+
return app_settings.settings()
|
|
549
|
+
|
|
550
|
+
@app.post("/api/settings")
|
|
551
|
+
def settings_update(body: Setting) -> dict[str, Any]:
|
|
552
|
+
from . import settings as app_settings
|
|
553
|
+
try:
|
|
554
|
+
return app_settings.update(body.key, body.value)
|
|
555
|
+
except KeyError as exc:
|
|
556
|
+
raise HTTPException(400, "无法保存:此项暂不支持在 app 中修改") from exc
|
|
557
|
+
except ValueError as exc:
|
|
558
|
+
raise HTTPException(400, str(exc)) from exc
|
|
559
|
+
|
|
560
|
+
@app.get("/api/onboarding")
|
|
561
|
+
def onboarding() -> dict[str, Any]:
|
|
562
|
+
from . import settings as app_settings
|
|
563
|
+
return app_settings.onboarding()
|
|
564
|
+
|
|
565
|
+
@app.post("/api/onboarding")
|
|
566
|
+
def onboarding_done(body: Onboarding) -> dict[str, Any]:
|
|
567
|
+
"""done=true: finished or skipped; done=false: show the guide again (settings page)."""
|
|
568
|
+
from . import settings as app_settings
|
|
569
|
+
return app_settings.finish_onboarding(body.done)
|
|
570
|
+
|
|
571
|
+
@app.post("/api/settings/test/{group}")
|
|
572
|
+
def settings_test(group: str) -> dict[str, Any]:
|
|
573
|
+
from . import settings as app_settings
|
|
574
|
+
try:
|
|
575
|
+
return app_settings.run_test(group)
|
|
576
|
+
except KeyError as exc:
|
|
577
|
+
raise HTTPException(404, "无法完成检测:此分组没有检测项") from exc
|
|
578
|
+
|
|
579
|
+
@app.get("/api/components/{key}/status")
|
|
580
|
+
def component_status(key: str) -> dict[str, Any]:
|
|
581
|
+
"""Return the current install progress and state for an optional component."""
|
|
582
|
+
from .. import optional
|
|
583
|
+
try:
|
|
584
|
+
return optional.install_status(key)
|
|
585
|
+
except KeyError as exc:
|
|
586
|
+
raise HTTPException(404, "没有这个可选组件") from exc
|
|
587
|
+
|
|
588
|
+
@app.post("/api/components/{key}/install")
|
|
589
|
+
def component_install(key: str) -> dict[str, Any]:
|
|
590
|
+
"""Download an optional component (the layout model) into the data dir."""
|
|
591
|
+
from .. import optional
|
|
592
|
+
if key not in optional.COMPONENTS:
|
|
593
|
+
raise HTTPException(404, "没有这个组件")
|
|
594
|
+
from ..config import load_config
|
|
595
|
+
try:
|
|
596
|
+
return optional.install(key, index_url=str(load_config().get("package_index_url") or ""))
|
|
597
|
+
except KeyError as exc:
|
|
598
|
+
raise HTTPException(404, "没有这个组件") from exc
|
|
599
|
+
|
|
600
|
+
@app.get("/api/app-update")
|
|
601
|
+
def app_update(fresh: bool = False) -> dict[str, Any]:
|
|
602
|
+
"""Check GitHub Releases for a newer LitLoop version (cached 10 minutes)."""
|
|
603
|
+
now = time.time()
|
|
604
|
+
cached = _APP_UPDATE.get("data")
|
|
605
|
+
if not fresh and cached is not None and now - float(_APP_UPDATE.get("at") or 0.0) <= 600:
|
|
606
|
+
return cached
|
|
607
|
+
|
|
608
|
+
base: dict[str, Any] = {
|
|
609
|
+
"current": __version__,
|
|
610
|
+
"latest": "",
|
|
611
|
+
"available": False,
|
|
612
|
+
"url": "",
|
|
613
|
+
"notes": "",
|
|
614
|
+
"error": "",
|
|
615
|
+
}
|
|
616
|
+
import requests
|
|
617
|
+
try:
|
|
618
|
+
resp = requests.get(
|
|
619
|
+
_APP_UPDATE_URL,
|
|
620
|
+
headers={
|
|
621
|
+
"Accept": "application/vnd.github+json",
|
|
622
|
+
"User-Agent": f"LitLoop/{__version__}",
|
|
623
|
+
},
|
|
624
|
+
timeout=5,
|
|
625
|
+
)
|
|
626
|
+
except requests.Timeout:
|
|
627
|
+
return {**base, "error": "无法检查更新:请求超时"}
|
|
628
|
+
except requests.RequestException:
|
|
629
|
+
return {**base, "error": "无法检查更新:网络连接失败,请检查网络或代理设置"}
|
|
630
|
+
|
|
631
|
+
if resp.status_code != 200:
|
|
632
|
+
return {**base, "error": f"无法检查更新:GitHub 返回 HTTP {resp.status_code}"}
|
|
633
|
+
|
|
634
|
+
try:
|
|
635
|
+
payload = resp.json()
|
|
636
|
+
except ValueError:
|
|
637
|
+
return {**base, "error": "无法检查更新:返回数据格式异常"}
|
|
638
|
+
if not isinstance(payload, dict):
|
|
639
|
+
return {**base, "error": "无法检查更新:返回数据格式异常"}
|
|
640
|
+
|
|
641
|
+
tag = str(payload.get("tag_name") or "").strip()
|
|
642
|
+
latest = tag.lstrip("vV")
|
|
643
|
+
if not latest:
|
|
644
|
+
return {**base, "error": "无法检查更新:未找到版本号"}
|
|
645
|
+
|
|
646
|
+
cur_ver = _parse_version_tuple(__version__)
|
|
647
|
+
lat_ver = _parse_version_tuple(latest)
|
|
648
|
+
if cur_ver is not None and lat_ver is not None:
|
|
649
|
+
available = lat_ver > cur_ver
|
|
650
|
+
else:
|
|
651
|
+
available = latest != __version__
|
|
652
|
+
|
|
653
|
+
result = {
|
|
654
|
+
"current": __version__,
|
|
655
|
+
"latest": latest,
|
|
656
|
+
"available": available,
|
|
657
|
+
"url": str(payload.get("html_url") or ""),
|
|
658
|
+
"notes": str(payload.get("body") or ""),
|
|
659
|
+
"error": "",
|
|
660
|
+
}
|
|
661
|
+
_APP_UPDATE.update(at=time.time(), data=result)
|
|
662
|
+
return result
|
|
663
|
+
|
|
664
|
+
@app.get("/api/data-dir")
|
|
665
|
+
def data_dir() -> dict[str, Any]:
|
|
666
|
+
"""Where the papers, sessions and settings live (not inside the install folder)."""
|
|
667
|
+
from ..config import DATA_DIR
|
|
668
|
+
return {"path": str(DATA_DIR), "exists": DATA_DIR.exists()}
|
|
669
|
+
|
|
670
|
+
@app.post("/api/diagnostics")
|
|
671
|
+
def diagnostics_export() -> dict[str, Any]:
|
|
672
|
+
"""Write the diagnostic bundle and return where it landed."""
|
|
673
|
+
from ..diagnostics import write_bundle
|
|
674
|
+
try:
|
|
675
|
+
path = write_bundle()
|
|
676
|
+
except OSError as exc:
|
|
677
|
+
raise HTTPException(500, f"无法写诊断包:{exc}") from exc
|
|
678
|
+
return {"path": str(path), "size": path.stat().st_size if path.exists() else 0}
|
|
679
|
+
|
|
680
|
+
# Downloads: jobs started from the app (download_jobs), one at a time.
|
|
681
|
+
@app.get("/api/downloads")
|
|
682
|
+
def downloads() -> dict[str, Any]:
|
|
683
|
+
from . import download_jobs
|
|
684
|
+
from ..download import progress as progress_reporter
|
|
685
|
+
from ..config import load_config
|
|
686
|
+
live = progress_reporter.read_state() if download_jobs.running() else None
|
|
687
|
+
# A download the agent or the command line runs: shown on the page too.
|
|
688
|
+
external = None if download_jobs.running() else progress_reporter.running_elsewhere()
|
|
689
|
+
return {"jobs": download_jobs.list_jobs(), "running": download_jobs.running(), "live": live,
|
|
690
|
+
"external": external,
|
|
691
|
+
"pushing": download_jobs.pushing(), "output_dir": str(load_config().get("output_dir") or "")}
|
|
692
|
+
|
|
693
|
+
@app.get("/api/downloads/notice")
|
|
694
|
+
def downloads_notice() -> dict[str, Any] | None:
|
|
695
|
+
"""Polled on every page for the system notifications (the app's own jobs and the agent's)."""
|
|
696
|
+
from ..download import progress as progress_reporter
|
|
697
|
+
return progress_reporter.latest_notice()
|
|
698
|
+
|
|
699
|
+
@app.get("/api/screening/notice")
|
|
700
|
+
def screening_notice() -> dict[str, Any] | None:
|
|
701
|
+
"""Polled on every page for the system notifications (a deep-search round finished)."""
|
|
702
|
+
return store.latest_round_notice()
|
|
703
|
+
|
|
704
|
+
@app.get("/api/downloads/pending")
|
|
705
|
+
def downloads_pending() -> list[dict[str, Any]]:
|
|
706
|
+
"""Papers picked 选中下载 in screening that no job has downloaded yet (or holds in a queue), by session."""
|
|
707
|
+
from . import download_jobs
|
|
708
|
+
got = {k for d in download_jobs.downloaded_dois() | download_jobs.queued_dois()
|
|
709
|
+
for k in download_jobs.same_paper(d)}
|
|
710
|
+
out = []
|
|
711
|
+
for s in db.pending_selected_by_session(): # newest first: a paper picked twice shows in the newest
|
|
712
|
+
papers = []
|
|
713
|
+
for it in s["papers"]:
|
|
714
|
+
keys = download_jobs.same_paper(it["doi"])
|
|
715
|
+
if keys & got:
|
|
716
|
+
continue
|
|
717
|
+
got |= keys
|
|
718
|
+
papers.append(it)
|
|
719
|
+
if papers:
|
|
720
|
+
out.append({"session": s["session"], "question": s["question"], "papers": papers})
|
|
721
|
+
return out
|
|
722
|
+
|
|
723
|
+
@app.get("/api/downloads/{job}")
|
|
724
|
+
def download_job(job: str) -> dict[str, Any]:
|
|
725
|
+
from . import download_jobs
|
|
726
|
+
if not re.fullmatch(r"[\w-]{1,40}", job):
|
|
727
|
+
raise HTTPException(400, "无法打开下载任务:任务编号不正确")
|
|
728
|
+
try:
|
|
729
|
+
d = download_jobs.get(job)
|
|
730
|
+
except FileNotFoundError as exc:
|
|
731
|
+
raise HTTPException(404, "无法打开下载任务:此任务已不存在") from exc
|
|
732
|
+
d["zotero_default"] = download_jobs.default_collection(d["title"])
|
|
733
|
+
_explain(d, full=True)
|
|
734
|
+
_explain(d.get("zotero"), "推送", full=True)
|
|
735
|
+
for it in d.get("items") or []:
|
|
736
|
+
_explain(it)
|
|
737
|
+
_explain(it.get("zotero"), "推送")
|
|
738
|
+
return d
|
|
739
|
+
|
|
740
|
+
@app.post("/api/downloads/{job}/queue")
|
|
741
|
+
def download_queue(job: str, body: StartQueue) -> dict[str, Any]:
|
|
742
|
+
"""Start one publisher's queue of a long list (download_jobs.start_queue)."""
|
|
743
|
+
from . import download_jobs
|
|
744
|
+
if not re.fullmatch(r"[\w-]{1,40}", job):
|
|
745
|
+
raise HTTPException(400, "无法开始:任务编号不正确")
|
|
746
|
+
try:
|
|
747
|
+
return download_jobs.start_queue(job, body.publisher)
|
|
748
|
+
except FileNotFoundError as exc:
|
|
749
|
+
raise HTTPException(404, "无法开始:此下载任务已不存在") from exc
|
|
750
|
+
except download_jobs.Busy as exc:
|
|
751
|
+
raise HTTPException(409, "已有下载任务在进行,完成后再开始新的下载") from exc
|
|
752
|
+
except ValueError as exc:
|
|
753
|
+
raise HTTPException(400, str(exc)) from exc
|
|
754
|
+
|
|
755
|
+
@app.post("/api/downloads/{job}/zotero")
|
|
756
|
+
def download_push(job: str, body: ZoteroPush) -> dict[str, Any]:
|
|
757
|
+
from . import download_jobs
|
|
758
|
+
if not re.fullmatch(r"[\w-]{1,40}", job):
|
|
759
|
+
raise HTTPException(400, "无法推送:任务编号不正确")
|
|
760
|
+
try:
|
|
761
|
+
return download_jobs.push_zotero(job, body.collection, body.tags)
|
|
762
|
+
except FileNotFoundError as exc:
|
|
763
|
+
raise HTTPException(404, "无法推送:此下载任务已不存在") from exc
|
|
764
|
+
except download_jobs.Busy as exc:
|
|
765
|
+
raise HTTPException(409, "已有推送任务在进行,完成后再开始新的推送") from exc
|
|
766
|
+
except ValueError as exc:
|
|
767
|
+
raise HTTPException(400, str(exc)) from exc
|
|
768
|
+
|
|
769
|
+
@app.get("/api/zotero")
|
|
770
|
+
def zotero_overview() -> dict[str, Any]:
|
|
771
|
+
"""Is Zotero up, which zotero-mcp is used, what waits to be pushed, what was pushed."""
|
|
772
|
+
import requests
|
|
773
|
+
|
|
774
|
+
from . import download_jobs
|
|
775
|
+
from ..config import load_config
|
|
776
|
+
from ..zotero.client import discover_zotero_mcp
|
|
777
|
+
try:
|
|
778
|
+
running = requests.get("http://127.0.0.1:23119/connector/ping", timeout=2,
|
|
779
|
+
proxies={"http": None, "https": None}).status_code == 200
|
|
780
|
+
except requests.RequestException:
|
|
781
|
+
running = False
|
|
782
|
+
found = discover_zotero_mcp(load_config())
|
|
783
|
+
history = download_jobs.zotero_history()
|
|
784
|
+
for h in history:
|
|
785
|
+
_explain(h, "推送", full=True)
|
|
786
|
+
return {"running": running, "mcp": found["python"] if found else None,
|
|
787
|
+
"pending": download_jobs.zotero_pending(), "history": history,
|
|
788
|
+
"pushing": download_jobs.pushing()}
|
|
789
|
+
|
|
790
|
+
@app.get("/api/zotero/write-mode")
|
|
791
|
+
def zotero_write_mode() -> dict[str, Any]:
|
|
792
|
+
from ..config import load_config
|
|
793
|
+
from ..zotero.client import write_mode
|
|
794
|
+
# Asking costs a zotero-mcp start (2-3 s) and the mode only changes when the
|
|
795
|
+
# user authorizes, so the first good answer is kept until then.
|
|
796
|
+
if "mode" not in _WRITE_MODE:
|
|
797
|
+
r = write_mode(load_config())
|
|
798
|
+
if not r.get("success"):
|
|
799
|
+
return {"mode": None, "local_supported": None}
|
|
800
|
+
_WRITE_MODE.update(mode=r.get("mode"), local_supported=r.get("local_supported"))
|
|
801
|
+
return dict(_WRITE_MODE)
|
|
802
|
+
|
|
803
|
+
@app.post("/api/zotero/authorize")
|
|
804
|
+
def zotero_authorize() -> dict[str, Any]:
|
|
805
|
+
"""Zotero asks the user (a dialog in the Zotero app); the answer decides the mode."""
|
|
806
|
+
from ..config import load_config
|
|
807
|
+
from ..zotero.client import authorize_local_writes
|
|
808
|
+
r = authorize_local_writes(load_config())
|
|
809
|
+
_WRITE_MODE.clear()
|
|
810
|
+
from . import error_text
|
|
811
|
+
ok = bool(r.get("success"))
|
|
812
|
+
return {"mode": r.get("mode"), "ok": ok,
|
|
813
|
+
"message": r.get("message") if ok else error_text.message(r.get("error") or r.get("message"), "授权")}
|
|
814
|
+
|
|
815
|
+
@app.post("/api/zotero/check")
|
|
816
|
+
def zotero_check(body: ZoteroCheck) -> dict[str, Any]:
|
|
817
|
+
"""The PDF state of every item in a collection, and for problem items a local PDF if one exists."""
|
|
818
|
+
from . import download_jobs
|
|
819
|
+
from ..config import load_config
|
|
820
|
+
from ..zotero.client import check_collection
|
|
821
|
+
if not re.fullmatch(r"[A-Z0-9]{8}", body.key):
|
|
822
|
+
raise HTTPException(400, "无法检查附件:分类编号不正确")
|
|
823
|
+
r = check_collection(body.key, load_config())
|
|
824
|
+
if not r.get("success"):
|
|
825
|
+
raise _fail(503, r.get("error"), "检查附件", "Zotero is not reachable")
|
|
826
|
+
for it in r["items"]:
|
|
827
|
+
pdf = download_jobs.find_local_pdf(it["doi"]) if it["state"] != "ok" and it["doi"] else None
|
|
828
|
+
it["local_pdf"] = str(pdf) if pdf else None
|
|
829
|
+
return {"items": r["items"]}
|
|
830
|
+
|
|
831
|
+
@app.post("/api/zotero/repair")
|
|
832
|
+
def zotero_repair(body: ZoteroRepair) -> dict[str, Any]:
|
|
833
|
+
"""Repair one item: trash its empty attachments, attach the local PDF found for its DOI."""
|
|
834
|
+
from . import download_jobs
|
|
835
|
+
from ..config import load_config
|
|
836
|
+
from ..zotero.client import repair_attachment
|
|
837
|
+
if not re.fullmatch(r"[A-Z0-9]{8}", body.item_key) or any(not re.fullmatch(r"[A-Z0-9]{8}", k) for k in body.trash):
|
|
838
|
+
raise HTTPException(400, "无法重新附加:条目编号不正确")
|
|
839
|
+
pdf = download_jobs.find_local_pdf(body.doi or "") if body.attach else None # never a path from the client
|
|
840
|
+
if body.attach and not pdf:
|
|
841
|
+
raise _fail(404, "no local PDF", "重新附加", "")
|
|
842
|
+
r = repair_attachment(body.item_key, pdf, body.trash, load_config())
|
|
843
|
+
if not r.get("success"):
|
|
844
|
+
raise _fail(502, r.get("error"), "重新附加", "repair did not complete")
|
|
845
|
+
return r
|
|
846
|
+
|
|
847
|
+
@app.get("/api/zotero/collections")
|
|
848
|
+
def zotero_collections() -> list[dict[str, Any]]:
|
|
849
|
+
"""The library's collections as '/'-paths (read through zotero-mcp)."""
|
|
850
|
+
from ..config import load_config
|
|
851
|
+
from ..zotero.client import zotero_collections as read
|
|
852
|
+
r = read(load_config())
|
|
853
|
+
if not r.get("success"):
|
|
854
|
+
raise _fail(503, r.get("error"), "读取分类", "Zotero is not reachable")
|
|
855
|
+
return r["collections"]
|
|
856
|
+
|
|
857
|
+
@app.post("/api/downloads")
|
|
858
|
+
def download_start(body: NewDownload) -> dict[str, Any]:
|
|
859
|
+
from . import download_jobs
|
|
860
|
+
try:
|
|
861
|
+
return download_jobs.start([p.model_dump() for p in body.papers], body.title, body.session,
|
|
862
|
+
body.zotero_target)
|
|
863
|
+
except download_jobs.Busy as exc:
|
|
864
|
+
raise HTTPException(409, "已有下载任务在进行,完成后再开始新的下载") from exc
|
|
865
|
+
except ValueError as exc:
|
|
866
|
+
raise HTTPException(400, str(exc)) from exc
|
|
867
|
+
|
|
868
|
+
@app.post("/api/searches")
|
|
869
|
+
def search_start(body: NewSearch) -> dict[str, Any]:
|
|
870
|
+
from . import agent_runner
|
|
871
|
+
try:
|
|
872
|
+
return agent_runner.start(body.model_dump())
|
|
873
|
+
except ValueError as exc:
|
|
874
|
+
raise HTTPException(400, str(exc)) from exc
|
|
875
|
+
except RuntimeError as exc:
|
|
876
|
+
raise HTTPException(409, str(exc)) from exc
|
|
877
|
+
|
|
878
|
+
@app.get("/api/searches")
|
|
879
|
+
def search_list() -> list[dict[str, Any]]:
|
|
880
|
+
from . import agent_runner
|
|
881
|
+
return agent_runner.list_runs()
|
|
882
|
+
|
|
883
|
+
@app.get("/api/searches/{run}")
|
|
884
|
+
def search_get(run: str) -> dict[str, Any]:
|
|
885
|
+
from . import agent_runner
|
|
886
|
+
try:
|
|
887
|
+
d = agent_runner.get(run)
|
|
888
|
+
except ValueError as exc:
|
|
889
|
+
raise HTTPException(400, str(exc)) from exc
|
|
890
|
+
if not d:
|
|
891
|
+
raise HTTPException(404, "没有这次检索")
|
|
892
|
+
return d
|
|
893
|
+
|
|
894
|
+
@app.post("/api/searches/{run}/cancel")
|
|
895
|
+
def search_cancel(run: str) -> dict[str, Any]:
|
|
896
|
+
from . import agent_runner
|
|
897
|
+
return {"cancelled": agent_runner.cancel(run)}
|
|
898
|
+
|
|
899
|
+
@app.post("/api/downloads/parse")
|
|
900
|
+
def downloads_parse(body: PastedList) -> dict[str, Any]:
|
|
901
|
+
"""DOIs, doi.org links and arXiv ids in pasted text, for a download list.
|
|
902
|
+
|
|
903
|
+
``downloaded``: those whose PDF is still in the save folder; the job skips
|
|
904
|
+
exactly these (2026-10-20: the job history counted too, so the app had to
|
|
905
|
+
say "downloaded before, skipped if the file is still there")."""
|
|
906
|
+
from . import download_jobs
|
|
907
|
+
from ..identifiers import extract_identifiers
|
|
908
|
+
dois, unrecognized = extract_identifiers(body.text[:200_000])
|
|
909
|
+
got = set(download_jobs.local_files())
|
|
910
|
+
return {"papers": [{"doi": d} for d in dois[:2000]], "unrecognized": unrecognized[:100],
|
|
911
|
+
"unrecognized_count": len(unrecognized), "too_many": len(dois) > 2000,
|
|
912
|
+
"downloaded": [d for d in dois if download_jobs.same_paper(d) & got]}
|
|
913
|
+
|
|
914
|
+
@app.post("/api/downloads/resolve")
|
|
915
|
+
def downloads_resolve(body: ReferenceLines) -> dict[str, Any]:
|
|
916
|
+
"""Pasted lines with no DOI (reference-list entries, titles): look each up
|
|
917
|
+
on Crossref. Only when the user asks (a button), at most 100 lines."""
|
|
918
|
+
from . import download_jobs
|
|
919
|
+
from ..config import load_config
|
|
920
|
+
from ..search.title_to_doi import resolve_references
|
|
921
|
+
lines = [ln.strip() for ln in body.lines if ln.strip()][:100]
|
|
922
|
+
rows = resolve_references(lines, load_config())
|
|
923
|
+
got = set(download_jobs.local_files())
|
|
924
|
+
for r in rows:
|
|
925
|
+
if r.get("doi"):
|
|
926
|
+
r["downloaded"] = bool(download_jobs.same_paper(r["doi"]) & got)
|
|
927
|
+
return {"results": rows}
|
|
928
|
+
|
|
929
|
+
@app.post("/api/downloads/skip")
|
|
930
|
+
def download_skip(body: SkipPaper) -> dict[str, Any]:
|
|
931
|
+
"""The user will not log in for this paper: give it up, go on with the next."""
|
|
932
|
+
from ..sources import skip_download
|
|
933
|
+
return {"skipped": skip_download(body.doi)}
|
|
934
|
+
|
|
935
|
+
@app.post("/api/downloads/cancel")
|
|
936
|
+
def download_cancel() -> dict[str, Any]:
|
|
937
|
+
from . import download_jobs
|
|
938
|
+
return {"cancelled": download_jobs.cancel()}
|
|
939
|
+
|
|
940
|
+
@app.post("/api/open")
|
|
941
|
+
def open_path(body: OpenPath) -> dict[str, Any]:
|
|
942
|
+
"""Open a downloaded PDF or a job's folder with the system's default program.
|
|
943
|
+
|
|
944
|
+
Only paths a download job recorded (its files and output folder) and
|
|
945
|
+
the configured save folder are opened, never an arbitrary path:
|
|
946
|
+
opening runs whatever handles it.
|
|
947
|
+
"""
|
|
948
|
+
from pathlib import Path
|
|
949
|
+
|
|
950
|
+
from . import download_jobs
|
|
951
|
+
from ..config import load_config
|
|
952
|
+
target = Path(body.path).resolve()
|
|
953
|
+
allowed = set()
|
|
954
|
+
if load_config().get("output_dir"):
|
|
955
|
+
allowed.add(Path(load_config()["output_dir"]).expanduser().resolve())
|
|
956
|
+
allowed.add((Path(load_config()["output_dir"]).expanduser() / "exports").resolve())
|
|
957
|
+
for j in download_jobs.list_jobs(200):
|
|
958
|
+
d = download_jobs.get(j["job"])
|
|
959
|
+
if d.get("output_dir"):
|
|
960
|
+
allowed.add(Path(d["output_dir"]).resolve())
|
|
961
|
+
allowed |= {Path(i["file"]).resolve() for i in d["items"] if i.get("file")}
|
|
962
|
+
ok_file = target in allowed and target.suffix.lower() == ".pdf" and target.is_file()
|
|
963
|
+
ok_dir = target.is_dir() and (target in allowed or any(target == a.parent for a in allowed if a.suffix))
|
|
964
|
+
if not (ok_file or ok_dir):
|
|
965
|
+
raise HTTPException(403, "无法打开:只能打开 LitLoop 下载的 PDF 和它们的文件夹")
|
|
966
|
+
_open_with_system(target)
|
|
967
|
+
return {"opened": str(target)}
|
|
968
|
+
|
|
969
|
+
return app
|
|
970
|
+
|
|
971
|
+
|
|
972
|
+
def _open_with_system(path) -> None:
|
|
973
|
+
import os
|
|
974
|
+
import subprocess
|
|
975
|
+
if os.name == "nt":
|
|
976
|
+
os.startfile(str(path)) # noqa: S606 - a vetted PDF or folder
|
|
977
|
+
else:
|
|
978
|
+
subprocess.Popen(["open" if sys.platform == "darwin" else "xdg-open", str(path)])
|
|
979
|
+
|
|
980
|
+
|
|
981
|
+
def _free_port() -> int:
|
|
982
|
+
with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s:
|
|
983
|
+
s.bind(("127.0.0.1", 0))
|
|
984
|
+
return s.getsockname()[1]
|
|
985
|
+
|
|
986
|
+
|
|
987
|
+
def main(argv: list[str] | None = None) -> None:
|
|
988
|
+
ap = argparse.ArgumentParser(description="LitLoop desktop app backend (127.0.0.1 only)")
|
|
989
|
+
ap.add_argument("--token", required=True, help="random token every request must carry")
|
|
990
|
+
ap.add_argument("--port", type=int, default=0, help="0 = pick a free port")
|
|
991
|
+
args = ap.parse_args(argv)
|
|
992
|
+
import uvicorn
|
|
993
|
+
port = args.port or _free_port()
|
|
994
|
+
print(json.dumps({"litloop_app_server": 1, "port": port}), flush=True)
|
|
995
|
+
uvicorn.run(create_app(args.token), host="127.0.0.1", port=port, log_level="warning")
|
|
996
|
+
|
|
997
|
+
|
|
998
|
+
if __name__ == "__main__":
|
|
999
|
+
main(sys.argv[1:])
|