APOC custom functions and procedures

@TOC[1]

specifies

APOC Provide related procedures to create user-defined functions and procedures . These functions and procedures are actually parameterized Cypher Language query , Similar macro （Macro） The concept of . In the following case, the library version is ongdb-3.5.22. Baidu concept explanation [2] macro （ English ：Macro） It is a title of batch processing . The macro in computer science is an abstraction （Abstraction）, It replaces certain text patterns according to a set of predefined rules . The interpreter or the compiler automatically performs the first mock exam when it meets macros . For compiled languages , Macro expansion occurs at compile time , The tool for macro development is often called macro developer . The term macro is often used in many similar environments , They are concepts derived from macro expansion , This includes keyboard macros and macro languages . In most cases ,“ macro ” The use of this word implies the conversion of small commands or actions into a series of instructions .

View custom functions and procedures

CALL dbms.functions() YIELD name,signature,description,roles WHERE name CONTAINS 'custom' RETURN name,signature,description,roles
UNION
CALL dbms.procedures() YIELD name,signature,description,roles WHERE name CONTAINS 'custom' RETURN name,signature,description,roles

Copy

View the stored procedures for building custom functions and procedures

CALL dbms.functions() YIELD name,signature,description,roles WHERE name CONTAINS 'apoc.custom' RETURN name,signature,description,roles
UNION
CALL dbms.procedures() YIELD name,signature,description,roles WHERE name CONTAINS 'apoc.custom' RETURN name,signature,description,roles

Copy

Register a custom function

#  Input and output fields and their types , The format is as follows ：
[ ['item1','type1'],
  ['item2','type2'],
  …
]
# String::output
# List<List<String>>::inputs
# String::description

apoc.custom.asFunction(name, statement, outputs, inputs, forceSingle, description)
apoc.custom.declareFunction(signature, statement, forceSingle, description)

Copy

Register a custom process

• Support to return more complex data types

#  Input and output fields and their types , The format is as follows ：
[ ['item1','type1'],
['item2','type2'],
…
]
# List<List<String>>::output
# List<List<String>>::inputs
# String::description
# mode Patterns supported by the process :
  /** This procedure will only perform read operations against the graph */
  READ,
  /** This procedure may perform both read and write operations against the graph */
  WRITE,
  /** This procedure will perform operations against the schema */
  SCHEMA,
  /** This procedure will perform system operations - i.e. not against the graph */
  DBMS,
  /** This procedure will only perform read operations against the graph */
  DEFAULT

apoc.custom.asProcedure(name, statement, mode, outputs, inputs, description)
apoc.custom.declareProcedure(signature, statement, mode, description)

Copy

Data types supported by input and output parameters

case "ANY": return NTAny;
case "MAP": return NTMap;
case "NODE": return NTNode;
case "REL": return NTRelationship;
case "RELATIONSHIP": return NTRelationship;
case "EDGE": return NTRelationship;
case "PATH": return NTPath;
case "NUMBER": return NTNumber;
case "LONG": return NTInteger;
case "INT": return NTInteger;
case "INTEGER": return NTInteger;
case "FLOAT": return NTFloat;
case "DOUBLE": return NTFloat;
case "BOOL": return NTBoolean;
case "BOOLEAN": return NTBoolean;
case "DATE": return NTDate;
case "TIME": return NTTime;
case "LOCALTIME": return NTLocalTime;
case "DATETIME": return NTDateTime;
case "LOCALDATETIME": return NTLocalDateTime;
case "DURATION": return NTDuration;
case "POINT": return NTPoint;
case "GEO": return NTGeometry;
case "GEOMETRY": return NTGeometry;
case "STRING": return NTString;
case "TEXT": return NTString;
default: return NTString;

Copy

Check the list of functions and procedures

apoc.custom.list()

Copy

Delete function

apoc.custom.removeFunction(name, type)

Copy

Delete process

apoc.custom.removeProcedure(name)

Copy

Use case 1

• Custom function

CALL apoc.custom.asFunction(
'double',
'RETURN $input*2 as answer',
'long',
[['input','number']]
);
RETURN custom.double(12) AS value;

Copy

Copy

• Custom process

CALL apoc.custom.asProcedure('answer','RETURN $input as answer','read',[['answer','number']],[['input','int','42']])
CALL custom.answer() YIELD answer RETURN answer
CALL custom.answer(13) YIELD answer RETURN answer

Copy

Use case 2

• Functions and procedures need to be used for publishing to other users admin structure • Based on the length of time distance, add up the number of effective shares

WITH 20160630000000 AS endTime,' Shenzhen Investment Holding Co., Ltd ' AS name
//  Filter out the edge of the effective number of shares and find the nearest shareholding relationship based on the length of time distance 
//  First filter out the effective number of shares holdAmontCalc>0 The edge of , Then the number of shares is selected from the side based on the length of time distance detail
MATCH (n:HORGShareHoldV002 {name:name})-[r: holding ]->(m:HORGShareHoldV002) WHERE ANY(detail IN apoc.convert.fromJsonList(r.shareholding_detail) WHERE detail.holdAmontCalc>0) WITH apoc.convert.fromJsonMap(olab.samplingByDate.dis.jsonArray(olab.convert.json(FILTER(detail IN apoc.convert.fromJsonList(r.shareholding_detail) WHERE detail.holdAmontCalc>0)),'defineDate',endTime)) AS detail
//  Total shares 
RETURN SUM(detail.holdAmontCalc) AS holdTotal

Copy

Copy

• Based on the length of time distance, add up the number of effective shares - Defined function

CALL apoc.custom.asFunction(
'sum.hold',
'WITH $endTime AS endTime,$name AS name MATCH (n:HORGShareHoldV002 {name:name})-[r: holding ]->(m:HORGShareHoldV002) WHERE ANY(detail IN apoc.convert.fromJsonList(r.shareholding_detail) WHERE detail.holdAmontCalc>0) WITH apoc.convert.fromJsonMap(olab.samplingByDate.dis.jsonArray(olab.convert.json(FILTER(detail IN apoc.convert.fromJsonList(r.shareholding_detail) WHERE detail.holdAmontCalc>0)),\'defineDate\',endTime)) AS detail RETURN SUM(detail.holdAmontCalc) AS holdTotal',
'LONG',
[['endTime','LONG'],['name','STRING']],
 false,
' Total effective holdings '
);

Copy

Copy

• Use the function to query the total number of effective holdings

RETURN custom.sum.hold(20160630000000,' Shenzhen Investment Holding Co., Ltd ')

Copy

Use case 3

• Functions and procedures need to be used for publishing to other users admin structure • Query the basic information of the company and the actual controller

// 1、 Define company name 
WITH [' Energy Technology Co., Ltd ',' Hkust xunfei ',' COFCO Group Co., Ltd ',' Harvest Fund ',' Wuhan Contemporary Technology Industry Group Co., Ltd '] AS nameList
UNWIND nameList AS companyName
// 2、 Definition GraphQL Inquire about 
// #  Standard name of the company name
// #  company HCODE hcode
// #  Tag information  Tag
// #  Source code SrcCompanyCode source #  Source main table code -code #  Main table name -table
// legal_person_repr #  Legal representative  reg_capital #  The registered capital ( element ) establishment_date #  Date of establishment  biz_scope #  Scope of business  business_major #  Main business 
WITH REPLACE('{\"query\":\"query myConcernedCompany($name: String) { horgByName(name: $name) {name hcode Tag SrcCompanyCode { source code table Hold_Controller { name ratio } } HOrgInfo {legal_person_repr reg_capital establishment_date biz_scope business_major } }}\","variables":{\"name\":\"company-name\"},\"operationName\":\"myConcernedCompany\"}','company-name',companyName) AS query
WITH apoc.convert.fromJsonMap(olab.http.post('http://10.20.13.130/ongdb/graphql',query)) AS result
// 3、 Obtain the company legal person 、 The registered capital 、 Date of establishment 、 Business scope information   also   Get Wande code and Caihui code 
WITH result.data.horgByName[0].hcode AS hcode,result.data.horgByName[0].name AS name,result.data.horgByName[0].Tag AS Tag,result.data.horgByName[0].SrcCompanyCode AS SrcCompanyCode,result.data.horgByName[0].HOrgInfo AS HOrgInfo
//  Analyze the code of Wande and Caihui   as well as   Main business information   Analyze the object of the actual controller 
WITH hcode,name,Tag,FILTER(source IN SrcCompanyCode WHERE source.table='TCR0001')[0].code AS ciahuiCode,FILTER(source IN SrcCompanyCode WHERE source.table='CompIntroduction')[0].code AS windCode,HOrgInfo,FILTER(holdCont IN SrcCompanyCode WHERE holdCont.Hold_Controller<>[])[0].Hold_Controller[0] AS holdController
WITH hcode,name,Tag,REPLACE(ciahuiCode,'api','') AS caihuiCode,windCode,HOrgInfo[0].reg_capital AS reg_capital,HOrgInfo[0].legal_person_repr AS legal_person_repr,HOrgInfo[0].business_major AS business_major,HOrgInfo[0].biz_scope AS biz_scope,HOrgInfo[0].establishment_date AS establishment_date,holdController.name AS holdShareController,holdController.ratio AS holdShareControllerRatio
CALL apoc.case([business_major IS NOT NULL,'RETURN $business_major AS scope',biz_scope IS NOT NULL,'RETURN $biz_scope AS scope'],'',{business_major:business_major,biz_scope:biz_scope}) YIELD value
//  Scope of business   The main body of code   Standard name   Array of tags   Caihui code  WIND Code   The registered capital   legal person   Date of establishment   Real controller   Shareholding ratio of actual controller 
RETURN value.scope AS scope ,hcode,name,Tag,caihuiCode,windCode,reg_capital,legal_person_repr,establishment_date,holdShareController,holdShareControllerRatio

Copy

• Use the process to query the basic information of the company and the actual controller - The build process

CALL apoc.custom.asProcedure(
'org.basicinfo',
'WITH $name AS companyName WITH REPLACE(\'{\\\"query\\\":\\\"query myConcernedCompany($name: String) { horgByName(name: $name) {name hcode Tag SrcCompanyCode { source code table Hold_Controller { name ratio } } HOrgInfo {legal_person_repr reg_capital establishment_date biz_scope business_major } }}\\\","variables":{\\\"name\\\":\\\"company-name\\\"},\\\"operationName\\\":\\\"myConcernedCompany\\\"}\',\'company-name\',companyName) AS query WITH apoc.convert.fromJsonMap(olab.http.post(\'http://10.20.13.130/ongdb/graphql\',query)) AS result WITH result.data.horgByName[0].hcode AS hcode,result.data.horgByName[0].name AS name,result.data.horgByName[0].Tag AS Tag,result.data.horgByName[0].SrcCompanyCode AS SrcCompanyCode,result.data.horgByName[0].HOrgInfo AS HOrgInfo WITH hcode,name,Tag,FILTER(source IN SrcCompanyCode WHERE source.table=\'TCR0001\')[0].code AS ciahuiCode,FILTER(source IN SrcCompanyCode WHERE source.table=\'CompIntroduction\')[0].code AS windCode,HOrgInfo,FILTER(holdCont IN SrcCompanyCode WHERE holdCont.Hold_Controller<>[])[0].Hold_Controller[0] AS holdController WITH hcode,name,Tag,REPLACE(ciahuiCode,\'api\',\'\') AS caihuiCode,windCode,HOrgInfo[0].reg_capital AS reg_capital,HOrgInfo[0].legal_person_repr AS legal_person_repr,HOrgInfo[0].business_major AS business_major,HOrgInfo[0].biz_scope AS biz_scope,HOrgInfo[0].establishment_date AS establishment_date,holdController.name AS holdShareController,holdController.ratio AS holdShareControllerRatio CALL apoc.case([business_major IS NOT NULL,\'RETURN $business_major AS scope\',biz_scope IS NOT NULL,\'RETURN $biz_scope AS scope\'],\'\',{business_major:business_major,biz_scope:biz_scope}) YIELD value RETURN value.scope AS scope ,hcode,name,olab.convert.json(Tag) as Tag,caihuiCode,windCode,reg_capital,legal_person_repr,establishment_date,holdShareController,holdShareControllerRatio',
'READ',
[['scope','STRING'],['hcode','STRING'],['name','STRING'],['Tag','STRING'],['caihuiCode','STRING'],['windCode','STRING'],['reg_capital','STRING'],['legal_person_repr','STRING'],['establishment_date','STRING'],['holdShareController','STRING'],['holdShareControllerRatio','STRING']],
[['name','STRING']],
' Use the process to query the basic information of the company and the return of the actual controller 【 Scope of business   The main body of code   Standard name   Array of tags   Caihui code  WIND Code   The registered capital   legal person   Date of establishment   Real controller   Shareholding ratio of actual controller 】'
);

Copy

• View the defined process

CALL dbms.procedures() YIELD name,signature,description,roles WHERE name CONTAINS 'org.basicinfo' RETURN name,signature,description,roles

Copy

• Use the process to query the basic information of the company and the actual controller

WITH [' Energy Technology Co., Ltd ',' Hkust xunfei ',' COFCO Group Co., Ltd ',' Harvest Fund ',' Wuhan Contemporary Technology Industry Group Co., Ltd '] AS nameList
UNWIND nameList AS companyName
CALL custom.org.basicinfo(companyName) YIELD scope,hcode,name,Tag,caihuiCode,windCode,reg_capital,legal_person_repr,establishment_date,holdShareController,holdShareControllerRatio RETURN scope,hcode,name,apoc.convert.fromJsonList(Tag) AS tag,caihuiCode,windCode,reg_capital,legal_person_repr,establishment_date,holdShareController,holdShareControllerRatio

Copy

Use case 4

• Chinese function name

CALL apoc.custom.asFunction(
' Digital printing function ',
'RETURN $input*2 as answer',
'long',
[['input','number']]
);
RETURN custom. Digital printing function (12) AS value;

Copy

• Chinese process name

CALL apoc.custom.asProcedure(' Digital printing process ','RETURN $input as answer','read',[['answer','number']],[['input','int','42']]);CALL custom. Digital printing process (12) YIELD answer RETURN answer;

Copy

Custom function and procedure storage location

• New properties

KEY::apoc.custom
KEY::apoc.custom.update

Copy

Component download

•apoc-3.5.0.15-all

https://github.com/ongdb-contrib/neo4j-apoc-procedures/releases/tag/3.5.0.15.1

Copy

•ongdb-lab-apoc-3.5.22

https://github.com/ongdb-contrib/ongdb-lab-apoc/releases/tag/3.5.22.x

Copy

•ongdb-enterprise-3.5.22

https://github.com/graphfoundation/ongdb/releases/tag/3.5.22

Copy

References

[1] TOC: ONgDB-APOC Custom functions and procedures [2] Baidu concept explanation : https://baike.baidu.com/item/%E5%AE%8F/2648286?fr=aladdin