admin/PlantAI-rag-new
1
0
Fork 1
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(review): official format rules for 5 featured journals

Browse Source 
- Replace featured list with accessible journals: Genome Biology,
  BMC Plant Biology, Molecular Plant, Frontiers in Plant Science, Rice
- Per-journal format_rules verified against official author guidelines
  (rules_source=official, source_url, last_verified=2026-10-10)
- Update recommend tests to the new seed set
This commit is contained in:
potaser
2026-10-11 12:19:17 +08:00
parent 5a71e29cf7
commit c747ebd77a
2 changed files with 137 additions and 82 deletions
Download Patch File Download Diff File Expand all files Collapse all files
review_service/data/journals/journals.json
+111 -56
View File
@@ -1,7 +1,7 @@
{ {
"version": 1, "version": 1,
"updated": "2026-10-09", "updated": "2026-10-10",
"notes": "植物基因组学领域种子期刊。acceptance_rate 无公开权威数据一律为 null（不编造）；impact_factor / review_time_weeks 为公开 JCR/期刊官网常识近似值，以 last_updated 标注刷新时间，接入真实数据源前需人工核验。scope_desc 为英文范围描述，供 vstore 向量匹配使用。 format_rules_default 为通用默认规则（rules_source=default）；各条目 format_rules 暂为占位（rules_source=placeholder），接入真实投稿指南核验前不填期刊专属数字。 新增标签字段（不进 SQLite，接口按 id 从本文件合并）：scope_tags / application_tags / cross_domain 按 scope_desc 归类的标签；cas_partition / jcr_quartile 首版占位（partition_source=placeholder），核实不确定的一律 null、不猜测，接入真实分区数据前需人工核验。", "notes": "植物基因组学领域种子期刊。acceptance_rate 无公开权威数据一律为 null（不编造）；impact_factor / review_time_weeks 为公开 JCR/期刊官网常识近似值，以 last_updated 标注刷新时间，接入真实数据源前需人工核验。scope_desc 为英文范围描述，供 vstore 向量匹配使用。 format_rules_default 为通用默认规则（rules_source=default）；精选 5 本 format_rules 已于 2026-10-10 按各刊官网投稿指南人工核验填入（rules_source=official，附 source_url / last_verified / notes），官网未规定的字段一律 null 不猜测。 新增标签字段（不进 SQLite，接口按 id 从本文件合并）：scope_tags / application_tags / cross_domain 按 scope_desc 归类的标签；cas_partition / jcr_quartile 为占位（partition_source=placeholder），核实不确定的一律 null、不猜测，接入真实分区数据前需人工核验。 精选列表已更换为本环境可访问官网的 5 本（Genome Biology / BMC Plant Biology / Molecular Plant / Frontiers in Plant Science / Rice）；原 OUP/Wiley/Springer 4 本（The Plant Cell / The Plant Journal / Theoretical and Applied Genetics / The Plant Genome）移出精选，仍保留在 journal_directory 1056 条全量目录中。",
"format_rules_default": { "format_rules_default": {
"rules_source": "default", "rules_source": "default",
"max_pages": 10, "max_pages": 10,
@@ -13,26 +13,6 @@
"refs_style": "apa" "refs_style": "apa"
}, },
"journals": [ "journals": [
{
"id": "plant-cell",
"name": "The Plant Cell",
"scope_desc": "Broad-interest plant biology with emphasis on molecular, cellular and physiological mechanisms, including genomics, systems biology, gene regulation, signaling, and plant-microbe interactions. Publishes genome-scale and functional genomics studies with strong mechanistic insight.",
"impact_factor": 11.6,
"review_time_weeks": 6.0,
"acceptance_rate": null,
"submission_deadline": null,
"official_url": "https://academic.oup.com/plcell",
"last_updated": "2026-10-09",
"scope_tags": ["植物基因组学", "分子生物学", "细胞生物学", "信号转导"],
"application_tags": ["基础研究"],
"cas_partition": "1区",
"jcr_quartile": "Q1",
"partition_source": "placeholder",
"cross_domain": ["系统生物学", "植物-微生物互作"],
"format_rules": {
"rules_source": "placeholder"
}
},
{ {
"id": "genome-biology", "id": "genome-biology",
"name": "Genome Biology", "name": "Genome Biology",
@@ -50,67 +30,142 @@
"partition_source": "placeholder", "partition_source": "placeholder",
"cross_domain": ["进化生物学", "单细胞组学"], "cross_domain": ["进化生物学", "单细胞组学"],
"format_rules": { "format_rules": {
"rules_source": "placeholder" "max_pages": null,
"min_font_pt": null,
"min_margin_mm": null,
"min_line_spacing": 2.0,
"max_fonts": null,
"min_figures": null,
"refs_style": "vancouver",
"anonymous": false,
"rules_source": "official",
"source_url": "https://genomebiology.biomedcentral.com/submission-guidelines",
"last_verified": "2026-10-10",
"notes": "官网要求双倍行距（'Use double-line spacing'）、行页码编号；参考文献 'Journal abbreviations follow Index Medicus/MEDLINE'（顺序编号制→vancouver）；BMC 出版社 policy 页明确 single-anonymous（审稿人知道作者身份，作者无需匿名）；字号/边距/页数上限该页未规定。BMC editorial policy 辅助来源：https://www.biomedcentral.com/getpublished/editorial-policies"
} }
}, },
{ {
"id": "plant-journal", "id": "bmc-plant-biology",
"name": "The Plant Journal", "name": "BMC Plant Biology",
"scope_desc": "Molecular and physiological plant sciences, including genome biology and genomics, gene function, cell and developmental biology, stress and environmental responses, and plant biotechnology; favors experimental studies with clear mechanistic advance.", "scope_desc": "Open-access journal covering all aspects of plant biology, from molecular and cellular studies to the physiology and ecology of whole plants and communities, including plant-microbe interactions, biotechnology and applied plant science.",
"impact_factor": 7.2, "impact_factor": null,
"review_time_weeks": 5.0, "review_time_weeks": null,
"acceptance_rate": null, "acceptance_rate": null,
"submission_deadline": null, "submission_deadline": null,
"official_url": "https://onlinelibrary.wiley.com/journal/1365313X", "official_url": "https://bmcplantbiol.biomedcentral.com",
"last_updated": "2026-10-09", "last_updated": "2026-10-10",
"scope_tags": ["分子生物学", "植物基因组学", "细胞与发育生物学", "逆境生物学"], "scope_tags": ["植物生物学", "分子生物学", "逆境生物学", "植物-微生物互作"],
"application_tags": ["基础研究"],
"cas_partition": null,
"jcr_quartile": null,
"partition_source": "placeholder",
"cross_domain": ["植物生物技术"],
"format_rules": {
"max_pages": null,
"min_font_pt": null,
"min_margin_mm": null,
"min_line_spacing": 2.0,
"max_fonts": null,
"min_figures": null,
"refs_style": "vancouver",
"anonymous": false,
"rules_source": "official",
"source_url": "https://bmcplantbiol.biomedcentral.com/submission-guidelines",
"last_verified": "2026-10-10",
"notes": "同 BMC 系列模板页：双倍行距+行页码编号；参考文献 MEDLINE 缩写（→vancouver）；BMC 系列 editorial policy 为 single-anonymous（作者无需匿名）；字号/边距/页数上限该页未规定。"
}
},
{
"id": "molecular-plant",
"name": "Molecular Plant",
"scope_desc": "High-impact journal for fundamental and translational plant science at the molecular level, covering genomics, cell and developmental biology, signal transduction, stress and metabolic responses, and plant biotechnology.",
"impact_factor": null,
"review_time_weeks": null,
"acceptance_rate": null,
"submission_deadline": null,
"official_url": "https://www.cell.com/molecular-plant",
"last_updated": "2026-10-10",
"scope_tags": ["分子生物学", "基因组学", "多组学"],
"application_tags": ["基础研究", "分子育种"], "application_tags": ["基础研究", "分子育种"],
"cas_partition": null, "cas_partition": null,
"jcr_quartile": "Q1", "jcr_quartile": null,
"partition_source": "placeholder", "partition_source": "placeholder",
"cross_domain": ["表观遗传学", "植物生物技术"], "cross_domain": ["植物生物技术"],
"format_rules": { "format_rules": {
"rules_source": "placeholder" "max_pages": null,
"min_font_pt": 12,
"min_margin_mm": null,
"min_line_spacing": 1.5,
"max_fonts": null,
"min_figures": null,
"refs_style": "author_year",
"anonymous": null,
"rules_source": "official",
"source_url": "https://www.cell.com/molecular-plant/authors",
"last_verified": "2026-10-10",
"notes": "官网作者指南（本环境 cell.com 403，依据作者浏览器核对摘录）：正文 12 point 字号、1.5 行距，参考文献 'cited in the text by name and year in parentheses'（作者-年份制→author_year）；页数上限/边距/图表下限/匿名审稿政策该摘录未覆盖→null"
} }
}, },
{ {
"id": "theoretical-applied-genetics", "id": "frontiers-in-plant-science",
"name": "Theoretical and Applied Genetics", "name": "Frontiers in Plant Science",
"scope_desc": "Plant breeding and genetics research combining theoretical and applied approaches, including quantitative genetics, QTL mapping, genome-wide association studies, genomic selection, population genetics, and molecular marker applications for crop improvement.", "scope_desc": "Open-access journal spanning fundamental and applied plant sciences: molecular and cellular mechanisms, physiology, ecology, evolution, and applications in agriculture and biotechnology.",
"impact_factor": 5.3, "impact_factor": null,
"review_time_weeks": 5.0, "review_time_weeks": null,
"acceptance_rate": null, "acceptance_rate": null,
"submission_deadline": null, "submission_deadline": null,
"official_url": "https://link.springer.com/journal/338", "official_url": "https://www.frontiersin.org/journals/plant-science",
"last_updated": "2026-10-09", "last_updated": "2026-10-10",
"scope_tags": ["遗传学", "作物育种", "数量遗传学", "基因组选择"], "scope_tags": ["植物生物学", "分子生物学", "生理学", "生态学"],
"application_tags": ["种业", "分子育种", "基础研究"], "application_tags": ["基础研究"],
"cas_partition": null, "cas_partition": null,
"jcr_quartile": null, "jcr_quartile": null,
"partition_source": "placeholder", "partition_source": "placeholder",
"cross_domain": ["统计遗传学", "全基因组关联分析"], "cross_domain": ["农业科学"],
"format_rules": { "format_rules": {
"rules_source": "placeholder" "max_pages": null,
"min_font_pt": null,
"min_margin_mm": null,
"min_line_spacing": 1.0,
"max_fonts": null,
"min_figures": null,
"refs_style": null,
"anonymous": null,
"rules_source": "official",
"source_url": "https://www.frontiersin.org/journals/plant-science/for-authors/author-guidelines",
"last_verified": "2026-10-10",
"notes": "官网 author-guidelines：'The entire document should be single-spaced and must contain page and line numbers'（单倍行距→1.0）；页数按 Article types 页 word count 而非数字页限→max_pages=null；字号/边距/图表下限该页未规定。Frontiers 为 open peer review；官网未明文规定作者是否匿名，建议人工核实 review-guidelines 页。Frontiers 官网明确两种制式并存：Harvard (author-date) 或 Vancouver (numbered)，具体到 Plant Science 需查 help center（当前抓取失败），建议人工核实 https://www.frontiersin.org/help/"
} }
}, },
{ {
"id": "plant-genome", "id": "rice",
"name": "The Plant Genome", "name": "Rice",
"scope_desc": "Plant and crop genome research with practical breeding applications, including genome assembly and annotation, structural variation, GWAS and QTL analysis, marker-assisted and genomic selection, and translational genomics for crop improvement.", "scope_desc": "Open-access journal publishing research on all aspects of rice science, from molecular biology and genetics to breeding, physiology, ecology and production systems, with relevance to cereal crop improvement.",
"impact_factor": 3.3, "impact_factor": null,
"review_time_weeks": 4.0, "review_time_weeks": null,
"acceptance_rate": null, "acceptance_rate": null,
"submission_deadline": null, "submission_deadline": null,
"official_url": "https://acsess.onlinelibrary.wiley.com/journal/1940-3372", "official_url": "https://thericejournal.springeropen.com",
"last_updated": "2026-10-09", "last_updated": "2026-10-10",
"scope_tags": ["植物基因组学", "作物基因组学", "结构变异", "关联分析"], "scope_tags": ["水稻", "作物基因组学", "遗传育种"],
"application_tags": ["种业", "分子育种"], "application_tags": ["基础研究", "分子育种", "种业"],
"cas_partition": null, "cas_partition": null,
"jcr_quartile": null, "jcr_quartile": null,
"partition_source": "placeholder", "partition_source": "placeholder",
"cross_domain": ["转化基因组学", "标记辅助选择"], "cross_domain": ["作物科学"],
"format_rules": { "format_rules": {
"rules_source": "placeholder" "max_pages": null,
"min_font_pt": null,
"min_margin_mm": null,
"min_line_spacing": 2.0,
"max_fonts": null,
"min_figures": null,
"refs_style": "vancouver",
"anonymous": false,
"rules_source": "official",
"source_url": "https://thericejournal.springeropen.com/submission-guidelines",
"last_verified": "2026-10-10",
"notes": "SpringerOpen 系列模板页：'Use double-line spacing'（双倍行距→2.0）、行页码编号；参考文献 'Journal abbreviations follow Index Medicus/MEDLINE'（→vancouver）；审稿明文 single-anonymous（审稿人知道作者身份，作者无需匿名）；字号/边距/页数上限/图表下限该页未规定。"
} }
} }
] ]
review_service/tests/test_recommend.py
+26 -26
View File
@@ -24,20 +24,20 @@ ABSTRACT = (
) )
SEED_NAMES = { SEED_NAMES = {
"The Plant Cell",
"Genome Biology", "Genome Biology",
"The Plant Journal", "BMC Plant Biology",
"Theoretical and Applied Genetics", "Molecular Plant",
"The Plant Genome", "Frontiers in Plant Science",
"Rice",
} }
# 覆盖全部 5 本种子期刊的合法重排结果 # 覆盖全部 5 本种子期刊的合法重排结果
ALL_RANKINGS = [ ALL_RANKINGS = [
{"id": "plant-cell", "tier": "challenge", "match_reason": "功能基因组机制研究契合", "risk_note": "竞争激烈"},
{"id": "genome-biology", "tier": "challenge", "match_reason": "组装与比较基因组学范围契合", "risk_note": "IF 门槛高"}, {"id": "genome-biology", "tier": "challenge", "match_reason": "组装与比较基因组学范围契合", "risk_note": "IF 门槛高"},
{"id": "plant-journal", "tier": "fit", "match_reason": "植物分子生物学范围契合", "risk_note": "无"}, {"id": "molecular-plant", "tier": "challenge", "match_reason": "分子机制研究契合", "risk_note": "竞争激烈"},
{"id": "theoretical-applied-genetics", "tier": "fit", "match_reason": "GWAS/驯化分析契合", "risk_note": "偏重育种"}, {"id": "frontiers-in-plant-science", "tier": "fit", "match_reason": "植物生物学范围契合", "risk_note": "无"},
{"id": "plant-genome", "tier": "safe", "match_reason": "作物基因组组装契合", "risk_note": "IF 较低"}, {"id": "rice", "tier": "fit", "match_reason": "水稻基因组学契合", "risk_note": "无"},
{"id": "bmc-plant-biology", "tier": "safe", "match_reason": "植物生物学全领域契合", "risk_note": "无"},
] ]
@@ -99,8 +99,8 @@ def test_recommend_happy_path(client: TestClient):
assert it["risk_note"], f"缺少风险提示：{it['name']}" assert it["risk_note"], f"缺少风险提示：{it['name']}"
assert it["similarity"] == 0.5 assert it["similarity"] == 0.5
# 三档内容与 LLM 输出一致 # 三档内容与 LLM 输出一致
assert [it["name"] for it in body["tiers"]["challenge"]["items"]][0] == "The Plant Cell" assert [it["name"] for it in body["tiers"]["challenge"]["items"]][0] == "Genome Biology"
assert body["tiers"]["safe"]["items"][0]["name"] == "The Plant Genome" assert body["tiers"]["safe"]["items"][0]["name"] == "BMC Plant Biology"
meta = body["meta"] meta = body["meta"]
assert meta["llm_ok"] is True and meta["vector_ok"] is True assert meta["llm_ok"] is True and meta["vector_ok"] is True
@@ -144,7 +144,7 @@ def test_hallucinated_journal_dropped(client: TestClient, monkeypatch):
def test_llm_name_tamper_ignored(client: TestClient, monkeypatch): def test_llm_name_tamper_ignored(client: TestClient, monkeypatch):
"""LLM 篡改期刊名不采信：输出名称一律取 DB。""" """LLM 篡改期刊名不采信：输出名称一律取 DB。"""
tampered = [ tampered = [
{"id": "plant-cell", "tier": "fit", "name": "Fake Journal X", {"id": "genome-biology", "tier": "fit", "name": "Fake Journal X",
"match_reason": "被篡改的名称", "risk_note": "无"} "match_reason": "被篡改的名称", "risk_note": "无"}
] ]
monkeypatch.setattr(llm, "complete_json", make_fake_llm(tampered)) monkeypatch.setattr(llm, "complete_json", make_fake_llm(tampered))
@@ -152,8 +152,8 @@ def test_llm_name_tamper_ignored(client: TestClient, monkeypatch):
assert body["ok"] is True assert body["ok"] is True
raw = str(body) raw = str(body)
assert "Fake Journal X" not in raw assert "Fake Journal X" not in raw
plant_cell = [it for it in _all_items(body) if it["id"] == "plant-cell"] gb = [it for it in _all_items(body) if it["id"] == "genome-biology"]
assert plant_cell and plant_cell[0]["name"] == "The Plant Cell" assert gb and gb[0]["name"] == "Genome Biology"
def test_llm_failure_fallback(client: TestClient, monkeypatch): def test_llm_failure_fallback(client: TestClient, monkeypatch):
@@ -165,9 +165,9 @@ def test_llm_failure_fallback(client: TestClient, monkeypatch):
items = _all_items(body) items = _all_items(body)
assert len(items) == 5 assert len(items) == 5
assert {it["name"] for it in items} == SEED_NAMES assert {it["name"] for it in items} == SEED_NAMES
# 规则分档：IF≥9 → challenge（The Plant Cell 11.6 / Genome Biology 10.1） # 规则分档：IF≥9 → challenge（仅 Genome Biology 10.1 有公开 IF）
challenge = {it["name"] for it in body["tiers"]["challenge"]["items"]} challenge = {it["name"] for it in body["tiers"]["challenge"]["items"]}
assert challenge == {"The Plant Cell", "Genome Biology"} assert challenge == {"Genome Biology"}
for it in items: for it in items:
assert it["match_reason"] and it["risk_note"] assert it["match_reason"] and it["risk_note"]
@@ -187,16 +187,16 @@ def test_vector_failure_degrades(client: TestClient, monkeypatch):
def test_invalid_tier_repaired(client: TestClient, monkeypatch): def test_invalid_tier_repaired(client: TestClient, monkeypatch):
"""非法 tier 按规则补位，不丢期刊。""" """非法 tier 按规则补位，不丢期刊。"""
bad = [ bad = [
{"id": "plant-genome", "tier": "超越顶刊", "match_reason": "tier 非法", "risk_note": "无"}, {"id": "rice", "tier": "超越顶刊", "match_reason": "tier 非法", "risk_note": "无"},
{"id": "plant-cell", "tier": "challenge", "match_reason": "合法", "risk_note": "无"}, {"id": "genome-biology", "tier": "challenge", "match_reason": "合法", "risk_note": "无"},
] ]
monkeypatch.setattr(llm, "complete_json", make_fake_llm(bad)) monkeypatch.setattr(llm, "complete_json", make_fake_llm(bad))
body = client.post("/api/review/recommend", json={"abstract": ABSTRACT}).json() body = client.post("/api/review/recommend", json={"abstract": ABSTRACT}).json()
assert body["ok"] is True assert body["ok"] is True
assert len(_all_items(body)) == 5 assert len(_all_items(body)) == 5
# plant-genome 规则分档 → IF 3.3 ≤ 4 → safe # rice 无公开 IF/周期 → 规则分档 fit
safe_ids = [it["id"] for it in body["tiers"]["safe"]["items"]] fit_ids = [it["id"] for it in body["tiers"]["fit"]["items"]]
assert "plant-genome" in safe_ids assert "rice" in fit_ids
assert body["meta"]["repaired"] >= 1 assert body["meta"]["repaired"] >= 1
@@ -223,14 +223,14 @@ def test_soft_preference_risk_note(client: TestClient):
"""软偏好：超周期偏好只加风险提示，不删条目。""" """软偏好：超周期偏好只加风险提示，不删条目。"""
body = client.post( body = client.post(
"/api/review/recommend", "/api/review/recommend",
json={"abstract": ABSTRACT, "preferences": {"max_review_weeks": 5}}, json={"abstract": ABSTRACT, "preferences": {"max_review_weeks": 3}},
).json() ).json()
assert body["ok"] is True assert body["ok"] is True
items = _all_items(body) items = _all_items(body)
assert len(items) == 5 # 不删 assert len(items) == 5 # 不删
# The Plant Cell 审稿约 6 周 > 5 周 → 带 ⚠ 提示 # Genome Biology 审稿约 4 周 > 3 周 → 带 ⚠ 提示
pc = next(it for it in items if it["id"] == "plant-cell") gb = next(it for it in items if it["id"] == "genome-biology")
assert "超出你 5 周" in pc["risk_note"] assert "超出你 3 周" in gb["risk_note"]
def test_seed_idempotent(tmp_path): def test_seed_idempotent(tmp_path):
@@ -248,8 +248,8 @@ def test_seed_idempotent(tmp_path):
rows = asyncio.run(work()) rows = asyncio.run(work())
assert len(rows) == 5 assert len(rows) == 5
assert {r["id"] for r in rows} == { assert {r["id"] for r in rows} == {
"plant-cell", "genome-biology", "plant-journal", "genome-biology", "bmc-plant-biology", "molecular-plant",
"theoretical-applied-genetics", "plant-genome", "frontiers-in-plant-science", "rice",
} }
assert all(r["acceptance_rate"] is None for r in rows) assert all(r["acceptance_rate"] is None for r in rows)
assert all(r["scope_desc"] for r in rows) assert all(r["scope_desc"] for r in rows)