mirror of
https://git.mirrors.martin98.com/https://github.com/langgenius/dify.git
synced 2025-08-12 17:58:59 +08:00
add CrossRef builtin tool: doi query and title query (#7406)
This commit is contained in:
RookieAgent
GitHub
parent
53cf756207
commit
4ff4859036
49
api/core/tools/provider/builtin/crossref/_assets/icon.svg
Normal file
49
api/core/tools/provider/builtin/crossref/_assets/icon.svg
Normal file
@ -0,0 +1,49 @@
|
||||
<?xml version="1.0" encoding="utf-8"?>
|
||||
<!-- Generator: Adobe Illustrator 19.2.1, SVG Export Plug-In . SVG Version: 6.00 Build 0) -->
|
||||
<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN" "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">
|
||||
<svg version="1.1" id="Layer_1" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" x="0px" y="0px"
|
||||
viewBox="0 0 200 130.2" style="enable-background:new 0 0 200 130.2;" xml:space="preserve">
|
||||
<style type="text/css">
|
||||
.st0{fill:#3EB1C8;}
|
||||
.st1{fill:#D8D2C4;}
|
||||
.st2{fill:#4F5858;}
|
||||
.st3{fill:#FFC72C;}
|
||||
.st4{fill:#EF3340;}
|
||||
</style>
|
||||
<g>
|
||||
<polygon class="st0" points="111.8,95.5 111.8,66.8 135.4,59 177.2,73.3 "/>
|
||||
<polygon class="st1" points="153.6,36.8 111.8,51.2 135.4,59 177.2,44.6 "/>
|
||||
<polygon class="st2" points="135.4,59 177.2,44.6 177.2,73.3 "/>
|
||||
<polygon class="st3" points="177.2,0.3 177.2,29 153.6,36.8 111.8,22.5 "/>
|
||||
<polygon class="st4" points="153.6,36.8 111.8,51.2 111.8,22.5 "/>
|
||||
<g>
|
||||
<g>
|
||||
<g>
|
||||
<g>
|
||||
<path class="st2" d="M26.3,104.8c-0.5-3.7-4.1-6.5-8.1-6.5c-7.3,0-10.1,6.2-10.1,12.7c0,6.2,2.8,12.4,10.1,12.4
|
||||
c5,0,7.8-3.4,8.4-8.3h7.9c-0.8,9.2-7.2,15.2-16.3,15.2C6.8,130.2,0,121.7,0,111c0-11,6.8-19.6,18.2-19.6c8.2,0,15,4.8,16,13.3
|
||||
H26.3z"/>
|
||||
<path class="st2" d="M37.4,102.5h7v5h0.1c1.4-3.4,5-5.7,8.6-5.7c0.5,0,1.1,0.1,1.6,0.3v6.9c-0.7-0.2-1.8-0.3-2.6-0.3
|
||||
c-5.4,0-7.3,3.9-7.3,8.6v12.1h-7.4V102.5z"/>
|
||||
<path class="st2" d="M68.7,101.8c8.5,0,13.9,5.6,13.9,14.2c0,8.5-5.5,14.1-13.9,14.1c-8.4,0-13.9-5.6-13.9-14.1
|
||||
C54.9,107.4,60.3,101.8,68.7,101.8z M68.7,124.5c5,0,6.5-4.3,6.5-8.6c0-4.3-1.5-8.6-6.5-8.6c-5,0-6.5,4.3-6.5,8.6
|
||||
C62.2,120.2,63.8,124.5,68.7,124.5z"/>
|
||||
<path class="st2" d="M91.2,120.6c0.1,3.2,2.8,4.5,5.7,4.5c2.1,0,4.8-0.8,4.8-3.4c0-2.2-3.1-3-8.4-4.2c-4.3-0.9-8.5-2.4-8.5-7.2
|
||||
c0-6.9,5.9-8.6,11.7-8.6c5.9,0,11.3,2,11.8,8.6h-7c-0.2-2.9-2.4-3.6-5-3.6c-1.7,0-4.1,0.3-4.1,2.5c0,2.6,4.2,3,8.4,4
|
||||
c4.3,1,8.5,2.5,8.5,7.5c0,7.1-6.1,9.3-12.3,9.3c-6.2,0-12.3-2.3-12.6-9.5H91.2z"/>
|
||||
<path class="st2" d="M118.1,120.6c0.1,3.2,2.8,4.5,5.7,4.5c2.1,0,4.8-0.8,4.8-3.4c0-2.2-3.1-3-8.4-4.2
|
||||
c-4.3-0.9-8.5-2.4-8.5-7.2c0-6.9,5.9-8.6,11.7-8.6c5.9,0,11.3,2,11.8,8.6h-7c-0.2-2.9-2.4-3.6-5-3.6c-1.7,0-4.1,0.3-4.1,2.5
|
||||
c0,2.6,4.2,3,8.4,4c4.3,1,8.5,2.5,8.5,7.5c0,7.1-6.1,9.3-12.3,9.3c-6.2,0-12.3-2.3-12.6-9.5H118.1z"/>
|
||||
<path class="st2" d="M138.4,102.5h7v5h0.1c1.4-3.4,5-5.7,8.6-5.7c0.5,0,1.1,0.1,1.6,0.3v6.9c-0.7-0.2-1.8-0.3-2.6-0.3
|
||||
c-5.4,0-7.3,3.9-7.3,8.6v12.1h-7.4V102.5z"/>
|
||||
<path class="st2" d="M163.7,117.7c0.2,4.7,2.5,6.8,6.6,6.8c3,0,5.3-1.8,5.8-3.5h6.5c-2.1,6.3-6.5,9-12.6,9
|
||||
c-8.5,0-13.7-5.8-13.7-14.1c0-8,5.6-14.2,13.7-14.2c9.1,0,13.6,7.7,13,15.9H163.7z M175.7,113.1c-0.7-3.7-2.3-5.7-5.9-5.7
|
||||
c-4.7,0-6,3.6-6.1,5.7H175.7z"/>
|
||||
<path class="st2" d="M187.2,107.5h-4.4v-4.9h4.4v-2.1c0-4.7,3-8.2,9-8.2c1.3,0,2.6,0.2,3.9,0.2V98c-0.9-0.1-1.8-0.2-2.7-0.2
|
||||
c-2,0-2.8,0.8-2.8,3.1v1.6h5.1v4.9h-5.1v21.9h-7.4V107.5z"/>
|
||||
</g>
|
||||
</g>
|
||||
</g>
|
||||
</g>
|
||||
</g>
|
||||
</svg>
|
||||
|
After Width: | Height: | Size: 3.0 KiB |
20
api/core/tools/provider/builtin/crossref/crossref.py
Normal file
20
api/core/tools/provider/builtin/crossref/crossref.py
Normal file
@ -0,0 +1,20 @@
|
||||
from core.tools.errors import ToolProviderCredentialValidationError
|
||||
from core.tools.provider.builtin.crossref.tools.query_doi import CrossRefQueryDOITool
|
||||
from core.tools.provider.builtin_tool_provider import BuiltinToolProviderController
|
||||
|
||||
|
||||
class CrossRefProvider(BuiltinToolProviderController):
|
||||
def _validate_credentials(self, credentials: dict) -> None:
|
||||
try:
|
||||
CrossRefQueryDOITool().fork_tool_runtime(
|
||||
runtime={
|
||||
"credentials": credentials,
|
||||
}
|
||||
).invoke(
|
||||
user_id='',
|
||||
tool_parameters={
|
||||
"doi": '10.1007/s00894-022-05373-8',
|
||||
},
|
||||
)
|
||||
except Exception as e:
|
||||
raise ToolProviderCredentialValidationError(str(e))
|
||||
29
api/core/tools/provider/builtin/crossref/crossref.yaml
Normal file
29
api/core/tools/provider/builtin/crossref/crossref.yaml
Normal file
@ -0,0 +1,29 @@
|
||||
identity:
|
||||
author: Sakura4036
|
||||
name: crossref
|
||||
label:
|
||||
en_US: CrossRef
|
||||
zh_Hans: CrossRef
|
||||
description:
|
||||
en_US: Crossref is a cross-publisher reference linking registration query system using DOI technology created in 2000. Crossref establishes cross-database links between the reference list and citation full text of papers, making it very convenient for readers to access the full text of papers.
|
||||
zh_Hans: Crossref是于2000年创建的使用DOI技术的跨出版商参考文献链接注册查询系统。Crossref建立了在论文的参考文献列表和引文全文之间的跨数据库链接，使得读者能够非常便捷地获取文献全文。
|
||||
icon: icon.svg
|
||||
tags:
|
||||
- search
|
||||
credentials_for_provider:
|
||||
mailto:
|
||||
type: text-input
|
||||
required: true
|
||||
label:
|
||||
en_US: email address
|
||||
zh_Hans: email地址
|
||||
pt_BR: email address
|
||||
placeholder:
|
||||
en_US: Please input your email address
|
||||
zh_Hans: 请输入你的email地址
|
||||
pt_BR: Please input your email address
|
||||
help:
|
||||
en_US: According to the requirements of Crossref, an email address is required
|
||||
zh_Hans: 根据Crossref的要求，需要提供一个邮箱地址
|
||||
pt_BR: According to the requirements of Crossref, an email address is required
|
||||
url: https://api.crossref.org/swagger-ui/index.html
|
||||
25
api/core/tools/provider/builtin/crossref/tools/query_doi.py
Normal file
25
api/core/tools/provider/builtin/crossref/tools/query_doi.py
Normal file
@ -0,0 +1,25 @@
|
||||
from typing import Any, Union
|
||||
|
||||
import requests
|
||||
|
||||
from core.tools.entities.tool_entities import ToolInvokeMessage
|
||||
from core.tools.errors import ToolParameterValidationError
|
||||
from core.tools.tool.builtin_tool import BuiltinTool
|
||||
|
||||
|
||||
class CrossRefQueryDOITool(BuiltinTool):
|
||||
"""
|
||||
Tool for querying the metadata of a publication using its DOI.
|
||||
"""
|
||||
def _invoke(self, user_id: str, tool_parameters: dict[str, Any]) -> Union[ToolInvokeMessage, list[ToolInvokeMessage]]:
|
||||
doi = tool_parameters.get('doi')
|
||||
if not doi:
|
||||
raise ToolParameterValidationError('doi is required.')
|
||||
# doc: https://github.com/CrossRef/rest-api-doc
|
||||
url = f"https://api.crossref.org/works/{doi}"
|
||||
response = requests.get(url)
|
||||
response.raise_for_status()
|
||||
response = response.json()
|
||||
message = response.get('message', {})
|
||||
|
||||
return self.create_json_message(message)
|
||||
@ -0,0 +1,23 @@
|
||||
identity:
|
||||
name: crossref_query_doi
|
||||
author: Sakura4036
|
||||
label:
|
||||
en_US: CrossRef Query DOI
|
||||
zh_Hans: CrossRef DOI 查询
|
||||
pt_BR: CrossRef Query DOI
|
||||
description:
|
||||
human:
|
||||
en_US: A tool for searching literature information using CrossRef by DOI.
|
||||
zh_Hans: 一个使用CrossRef通过DOI获取文献信息的工具。
|
||||
pt_BR: A tool for searching literature information using CrossRef by DOI.
|
||||
llm: A tool for searching literature information using CrossRef by DOI.
|
||||
parameters:
|
||||
- name: doi
|
||||
type: string
|
||||
required: true
|
||||
label:
|
||||
en_US: DOI
|
||||
zh_Hans: DOI
|
||||
pt_BR: DOI
|
||||
llm_description: DOI for searching in CrossRef
|
||||
form: llm
|
||||
120
api/core/tools/provider/builtin/crossref/tools/query_title.py
Normal file
120
api/core/tools/provider/builtin/crossref/tools/query_title.py
Normal file
@ -0,0 +1,120 @@
|
||||
import time
|
||||
from typing import Any, Union
|
||||
|
||||
import requests
|
||||
|
||||
from core.tools.entities.tool_entities import ToolInvokeMessage
|
||||
from core.tools.tool.builtin_tool import BuiltinTool
|
||||
|
||||
|
||||
def convert_time_str_to_seconds(time_str: str) -> int:
|
||||
"""
|
||||
Convert a time string to seconds.
|
||||
example: 1s -> 1, 1m30s -> 90, 1h30m -> 5400, 1h30m30s -> 5430
|
||||
"""
|
||||
time_str = time_str.lower().strip().replace(' ', '')
|
||||
seconds = 0
|
||||
if 'h' in time_str:
|
||||
hours, time_str = time_str.split('h')
|
||||
seconds += int(hours) * 3600
|
||||
if 'm' in time_str:
|
||||
minutes, time_str = time_str.split('m')
|
||||
seconds += int(minutes) * 60
|
||||
if 's' in time_str:
|
||||
seconds += int(time_str.replace('s', ''))
|
||||
return seconds
|
||||
|
||||
|
||||
class CrossRefQueryTitleAPI:
|
||||
"""
|
||||
Tool for querying the metadata of a publication using its title.
|
||||
Crossref API doc: https://github.com/CrossRef/rest-api-doc
|
||||
"""
|
||||
query_url_template: str = "https://api.crossref.org/works?query.bibliographic={query}&rows={rows}&offset={offset}&sort={sort}&order={order}&mailto={mailto}"
|
||||
rate_limit: int = 50
|
||||
rate_interval: float = 1
|
||||
max_limit: int = 1000
|
||||
|
||||
def __init__(self, mailto: str):
|
||||
self.mailto = mailto
|
||||
|
||||
def _query(self, query: str, rows: int = 5, offset: int = 0, sort: str = 'relevance', order: str = 'desc', fuzzy_query: bool = False) -> list[dict]:
|
||||
"""
|
||||
Query the metadata of a publication using its title.
|
||||
:param query: the title of the publication
|
||||
:param rows: the number of results to return
|
||||
:param sort: the sort field
|
||||
:param order: the sort order
|
||||
:param fuzzy_query: whether to return all items that match the query
|
||||
"""
|
||||
url = self.query_url_template.format(query=query, rows=rows, offset=offset, sort=sort, order=order, mailto=self.mailto)
|
||||
response = requests.get(url)
|
||||
response.raise_for_status()
|
||||
rate_limit = int(response.headers['x-ratelimit-limit'])
|
||||
# convert time string to seconds
|
||||
rate_interval = convert_time_str_to_seconds(response.headers['x-ratelimit-interval'])
|
||||
|
||||
self.rate_limit = rate_limit
|
||||
self.rate_interval = rate_interval
|
||||
|
||||
response = response.json()
|
||||
if response['status'] != 'ok':
|
||||
return []
|
||||
|
||||
message = response['message']
|
||||
if fuzzy_query:
|
||||
# fuzzy query return all items
|
||||
return message['items']
|
||||
else:
|
||||
for paper in message['items']:
|
||||
title = paper['title'][0]
|
||||
if title.lower() != query.lower():
|
||||
continue
|
||||
return [paper]
|
||||
return []
|
||||
|
||||
def query(self, query: str, rows: int = 5, sort: str = 'relevance', order: str = 'desc', fuzzy_query: bool = False) -> list[dict]:
|
||||
"""
|
||||
Query the metadata of a publication using its title.
|
||||
:param query: the title of the publication
|
||||
:param rows: the number of results to return
|
||||
:param sort: the sort field
|
||||
:param order: the sort order
|
||||
:param fuzzy_query: whether to return all items that match the query
|
||||
"""
|
||||
rows = min(rows, self.max_limit)
|
||||
if rows > self.rate_limit:
|
||||
# query multiple times
|
||||
query_times = rows // self.rate_limit + 1
|
||||
results = []
|
||||
|
||||
for i in range(query_times):
|
||||
result = self._query(query, rows=self.rate_limit, offset=i * self.rate_limit, sort=sort, order=order, fuzzy_query=fuzzy_query)
|
||||
if fuzzy_query:
|
||||
results.extend(result)
|
||||
else:
|
||||
# fuzzy_query=False, only one result
|
||||
if result:
|
||||
return result
|
||||
time.sleep(self.rate_interval)
|
||||
return results
|
||||
else:
|
||||
# query once
|
||||
return self._query(query, rows, sort=sort, order=order, fuzzy_query=fuzzy_query)
|
||||
|
||||
|
||||
class CrossRefQueryTitleTool(BuiltinTool):
|
||||
"""
|
||||
Tool for querying the metadata of a publication using its title.
|
||||
"""
|
||||
def _invoke(self, user_id: str, tool_parameters: dict[str, Any]) -> Union[ToolInvokeMessage, list[ToolInvokeMessage]]:
|
||||
query = tool_parameters.get('query')
|
||||
fuzzy_query = tool_parameters.get('fuzzy_query', False)
|
||||
rows = tool_parameters.get('rows', 3)
|
||||
sort = tool_parameters.get('sort', 'relevance')
|
||||
order = tool_parameters.get('order', 'desc')
|
||||
mailto = self.runtime.credentials['mailto']
|
||||
|
||||
result = CrossRefQueryTitleAPI(mailto).query(query, rows, sort, order, fuzzy_query)
|
||||
|
||||
return [self.create_json_message(r) for r in result]
|
||||
105
api/core/tools/provider/builtin/crossref/tools/query_title.yaml
Normal file
105
api/core/tools/provider/builtin/crossref/tools/query_title.yaml
Normal file
@ -0,0 +1,105 @@
|
||||
identity:
|
||||
name: crossref_query_title
|
||||
author: Sakura4036
|
||||
label:
|
||||
en_US: CrossRef Title Query
|
||||
zh_Hans: CrossRef 标题查询
|
||||
pt_BR: CrossRef Title Query
|
||||
description:
|
||||
human:
|
||||
en_US: A tool for querying literature information using CrossRef by title.
|
||||
zh_Hans: 一个使用CrossRef通过标题搜索文献信息的工具。
|
||||
pt_BR: A tool for querying literature information using CrossRef by title.
|
||||
llm: A tool for querying literature information using CrossRef by title.
|
||||
parameters:
|
||||
- name: query
|
||||
type: string
|
||||
required: true
|
||||
label:
|
||||
en_US: 标题
|
||||
zh_Hans: 查询语句
|
||||
pt_BR: 标题
|
||||
human_description:
|
||||
en_US: Query bibliographic information, useful for citation look up. Includes titles, authors, ISSNs and publication years
|
||||
zh_Hans: 用于搜索文献信息，有助于查找引用。包括标题，作者，ISSN和出版年份
|
||||
pt_BR: Query bibliographic information, useful for citation look up. Includes titles, authors, ISSNs and publication years
|
||||
llm_description: key words for querying in Web of Science
|
||||
form: llm
|
||||
- name: fuzzy_query
|
||||
type: boolean
|
||||
default: false
|
||||
label:
|
||||
en_US: Whether to fuzzy search
|
||||
zh_Hans: 是否模糊搜索
|
||||
pt_BR: Whether to fuzzy search
|
||||
human_description:
|
||||
en_US: used for selecting the query type, fuzzy query returns more results, precise query returns 1 or none
|
||||
zh_Hans: 用于选择搜索类型，模糊搜索返回更多结果，精确搜索返回1条结果或无
|
||||
pt_BR: used for selecting the query type, fuzzy query returns more results, precise query returns 1 or none
|
||||
form: form
|
||||
- name: limit
|
||||
type: number
|
||||
required: false
|
||||
label:
|
||||
en_US: max query number
|
||||
zh_Hans: 最大搜索数
|
||||
pt_BR: max query number
|
||||
human_description:
|
||||
en_US: max query number(fuzzy search returns the maximum number of results or precise search the maximum number of matches)
|
||||
zh_Hans: 最大搜索数(模糊搜索返回的最大结果数或精确搜索最大匹配数)
|
||||
pt_BR: max query number(fuzzy search returns the maximum number of results or precise search the maximum number of matches)
|
||||
form: llm
|
||||
default: 50
|
||||
- name: sort
|
||||
type: select
|
||||
required: true
|
||||
options:
|
||||
- value: relevance
|
||||
label:
|
||||
en_US: relevance
|
||||
zh_Hans: 相关性
|
||||
pt_BR: relevance
|
||||
- value: published
|
||||
label:
|
||||
en_US: publication date
|
||||
zh_Hans: 出版日期
|
||||
pt_BR: publication date
|
||||
- value: references-count
|
||||
label:
|
||||
en_US: references-count
|
||||
zh_Hans: 引用次数
|
||||
pt_BR: references-count
|
||||
default: relevance
|
||||
label:
|
||||
en_US: sorting field
|
||||
zh_Hans: 排序字段
|
||||
pt_BR: sorting field
|
||||
human_description:
|
||||
en_US: Sorting of query results
|
||||
zh_Hans: 检索结果的排序字段
|
||||
pt_BR: Sorting of query results
|
||||
form: form
|
||||
- name: order
|
||||
type: select
|
||||
required: true
|
||||
options:
|
||||
- value: desc
|
||||
label:
|
||||
en_US: descending
|
||||
zh_Hans: 降序
|
||||
pt_BR: descending
|
||||
- value: asc
|
||||
label:
|
||||
en_US: ascending
|
||||
zh_Hans: 升序
|
||||
pt_BR: ascending
|
||||
default: desc
|
||||
label:
|
||||
en_US: Order
|
||||
zh_Hans: 排序
|
||||
pt_BR: Order
|
||||
human_description:
|
||||
en_US: Order of query results
|
||||
zh_Hans: 检索结果的排序方式
|
||||
pt_BR: Order of query results
|
||||
form: form
|
||||
Loading…
x
Reference in New Issue
Block a user