FINERACT-2110: Introduce type-safety API for commands (Note API) #3994

Zeyad2003 · 2024-07-26T05:51:31Z

Description

Describe the changes made and why they were made.

Ignore if these details are present on the associated Apache Fineract JIRA ticket.

Checklist

Please make sure these boxes are checked before submitting your pull request - thanks!

Write the commit message as per https://github.com/apache/fineract/#pull-requests
Acknowledge that we will not review PRs that are not passing the build ("green") - it is your responsibility to get a proposed PR to pass the build, not primarily the project's maintainers.
Create/update unit or integration tests for verifying the changes made.
Follow coding conventions at https://cwiki.apache.org/confluence/display/FINERACT/Coding+Conventions.
Add required Swagger annotation and update API documentation at fineract-provider/src/main/resources/static/legacy-docs/apiLive.htm with details of any API changes
Submission is not a "code dump". (Large changes can be made "in repository" via a branch. Ask on the developer mailing list for guidance, if required.)

FYI our guidelines for code reviews are at https://cwiki.apache.org/confluence/display/FINERACT/Code+Review+Guide.

vidakovic · 2024-07-26T06:14:54Z

...om/v3/commands/data/src/main/java/org/apache/fineract/v3/commands/data/JsonTypedCommand.java

+    private final Long commandId;
+    private final Long resourceId;
+    private final Long subresourceId;
+    private final Long groupId;


Why are these IDs necessary here? This is mimicking exactly the same implementation as the current production one... in other words: if you have a command that has IDs belonging to different entities then it's not really type safe.

removed.
Now we depend on resourceId and subresourceId only for representing IDs

...om/v3/commands/data/src/main/java/org/apache/fineract/v3/commands/data/JsonTypedCommand.java

...dler/src/main/java/org/apache/fineract/v3/commands/handler/NewTypedCommandSourceHandler.java

custom/v3/commands/provider/dependencies.gradle

...ovider/src/main/java/org/apache/fineract/v3/commands/provider/NewCommandHandlerProvider.java

vidakovic · 2024-07-26T06:30:40Z

...ice/src/main/java/org/apache/fineract/v3/commands/service/TypedCommandProcessingService.java

+
+public interface TypedCommandProcessingService {
+
+    <T> CommandProcessingResult executeCommand(CommandTypedWrapper<T> wrapper);


Where is the type safety for the "CommandProcessingResult"?

vidakovic · 2024-07-26T06:32:47Z

custom/v3/note/api/src/main/java/org/apache/fineract/v3/note/api/NotesApiController.java

+import org.springframework.web.bind.annotation.RequestMapping;
+import org.springframework.web.bind.annotation.RestController;
+
+@RestController


You are mixing read and write here... read related stuff is one PR, write related stuff is a separate PR... this cannot be approved, because this code is duplicated in 2 PRs!

A cleaner approach would be to have 1 controller for read and a separate one for write...

I can split the controller into 2 files, but the Commands' PR depends on Queries PR (same structure, share packages, and dependencies ...etc).
If the #3915 got merged then this one can be rebased with it seamlessly (no duplication).

Am I missing something?

.../service/src/main/java/org/apache/fineract/v3/note/service/NoteWritePlatformServiceImpl.java

...ava/org/apache/fineract/commands/service/PortfolioCommandSourceWritePlatformServiceImpl.java

vidakovic · 2024-07-26T06:38:42Z

fineract-provider/src/main/resources/application.properties

@@ -44,6 +44,9 @@ fineract.tenant.read-only-password=${FINERACT_DEFAULT_TENANTDB_RO_PWD:}
 fineract.tenant.read-only-parameters=${FINERACT_DEFAULT_TENANTDB_RO_CONN_PARAMS:}
 fineract.tenant.read-only-name=${FINERACT_DEFAULT_TENANTDB_RO_NAME:}

+# backwards compatibility for 'org.apache.fineract.infrastructure.security.filter.TenantAwareBasicAuthenticationFilter'
+baseUrl=https://localhost:8443${FINERACT_SERVER_SERVLET_CONTEXT_PATH:/fineract-provider}


We need to use a variable for host and port... those are available already with Spring Boot (just don't remember the names). Should be used here instead of hardcoded values.

We need to use a variable for host and port... those are available already with Spring Boot (just don't remember the names). Should be used here instead of hardcoded values.

How to use spring built-in vars?
I tried this but didn't work:

# FINERACT-914 server.forward-headers-strategy=framework server.port=${FINERACT_SERVER_PORT:8443} server.address=${FINERACT_SERVER_ADDRESS:localhost} server.servlet.context-path=${FINERACT_SERVER_SERVLET_CONTEXT_PATH:/fineract-provider} server.compression.enabled=${FINERACT_SERVER_COMPRESSION_ENABLED:true} # backwards compatibility for 'org.apache.fineract.infrastructure.security.filter.TenantAwareBasicAuthenticationFilter baseUrl=https://${server.address}:${server.port}${server.servlet.context-path}

and the checks failed as following:

Run curl -f -k --retry 10 --retry-connrefused --connect-timeout 30 --retry-delay 30 https://localhost:8443/fineract-provider/actuator/health % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0 0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0 curl: (35) error:0A000[12](https://github.com/Zeyad2003/fineract/actions/runs/10129914159/job/28010527803#step:8:13)6:SSL routines::unexpected eof while reading Error: Process completed with exit code 35.

any thoughts?

This approach has passed the checks:

server.forward-headers-strategy=framework server.port=${FINERACT_SERVER_PORT:8443} server.domain=${FINERACT_SERVER_DOMAIN:localhost} server.servlet.context-path=${FINERACT_SERVER_SERVLET_CONTEXT_PATH:/fineract-provider} server.compression.enabled=${FINERACT_SERVER_COMPRESSION_ENABLED:true} # backwards compatibility for 'org.apache.fineract.infrastructure.security.filter.TenantAwareBasicAuthenticationFilter' baseUrl=https://${server.domain}:${server.port}${server.servlet.context-path}

It appears that the server.address property causes some issues. The exact reason is ambiguous to me.

vidakovic · 2024-07-29T13:34:59Z

...v3/commands/data/src/main/java/org/apache/fineract/v3/commands/data/CommandTypedWrapper.java

-        return new CommandWrapper(officeId, groupId, clientId, loanId, savingsId, actionName, entityName, entityId,
-                subentityId, href, json, transactionId, productId, null, creditBureauId, organisationCreditBureauId,
-                jobName, idempotencyKey);
+        return new CommandWrapper(null, null, null, null, null, actionName, entityName, entityId,


No... this whole function is written in a way that goes against all best practices of dependency injection; "ObjectMapper" should be re-used/injected (with all the global configurations)... here we create it with default configurations over and over again... but even without that: this function is acts like copy function... how many times do we use this? My bet is again: once... this programming style is obscuring more things than it saves or helps...

It's just temporary. Not implemented there to remain.
Actually, I'm breaking some rules to get things to work first.
Then it can be improved or even removed.

vidakovic · 2024-07-29T13:36:09Z

custom/v3/commands/data/src/main/java/org/apache/fineract/v3/commands/data/TypedCommand.java

@@ -57,7 +47,7 @@ public JsonCommand getJsonCommand() {
        }

        return new JsonCommand(null, json, null, null, entityName, resourceId,


We have a "CommandWrapper" and a "JsonCommand"... why? Can't we just express what we need in one class?

Yes, I've thought about that, there's no major differences. (just action & entity names).
Thanks for mentioning it!

vidakovic · 2024-07-29T13:39:22Z

custom/v3/note/util/src/main/java/org/apache/fineract/v3/note/util/NoteUtilityMethods.java

-        }
-
-        return resourceDetails.toBuilder().entityName(resourceNameForPermissions).build();
+        String resourceNameForPermissions = switch (type) {


Without even reading more details: very likely the wrong place for ANYTHING permission related! Note: we don't need to squeeze every concept into one class... we are going to run out of space if we create another class... remidner: this belongs to Spring Security (remember when we talked about "Access Desicion Voters"?).

vidakovic · 2024-09-05T12:30:32Z

Please rebase into 1 commit.

vidakovic · 2024-09-05T12:35:39Z

custom/v3/command/domain/src/main/resources/db/custom-changelog/0001_create_command_table.xml

+            <column name="result_status_code" type="INTEGER" remarks="Result status code"/>
+            <column name="created_date" type="TIMESTAMP WITH TIME ZONE" remarks="Creation date"/>
+            <column name="processed_at" type="TIMESTAMP WITH TIME ZONE" remarks="Processing timestamp"/>
+            <column name="request_body" type="TEXT" remarks="Request body"/>


First: keep attributes simple... rename to body (we know that it's from a request... and there will never be another body). Alternative name payload.

Also: type TEXT is not really the way to go... it should be in the database jsonb and Java side com.fasterxml.jackson.databind.JsonNode. That way we can keep the structure and the parsing is done automatically.

Check https://www.baeldung.com/spring-boot-jpa-storing-postgresql-jsonb section 3.

Actually, I checked it, but couldn't resolve the mapping because the ORM needs to know how to map the attribute to a jsonb column, and the way explaied in the tutorial uses hibernate (which doesn't work with eclipselink).

@vidakovic I dont think jsonb would work with mysql or mariadb... isnt it postgres specific?

I reckon we can differentiate json and jsonb based on which engine is used if needed

@adamsaghy

You're right, jsonb is specific to PostgreSQL, but I think it's possible to configure the column as json for MySQL (and MariaDB) and jsonb for PostgreSQL.

The challenge I'm facing is how to map a jsonb attribute using EclipseLink.

vidakovic · 2024-09-05T12:36:28Z

custom/v3/command/domain/src/main/resources/db/custom-changelog/0001_create_command_table.xml

+            </column>
+            <column name="command_processing_status" type="VARCHAR(255)" remarks="Command processing status"/>
+            <column name="command_result" type="TEXT" remarks="Command result"/>
+            <column name="result_status_code" type="INTEGER" remarks="Result status code"/>


Let's name this just status. Simpler and says everything.

vidakovic · 2024-09-05T12:37:11Z

custom/v3/command/domain/src/main/resources/db/custom-changelog/0001_create_command_table.xml

+                <constraints nullable="false" unique="true"/>
+            </column>
+            <column name="command_processing_status" type="VARCHAR(255)" remarks="Command processing status"/>
+            <column name="command_result" type="TEXT" remarks="Command result"/>


Let's name this just result. Simpler and says everything.

Also: type TEXT is not really the way to go... it should be in the database jsonb and Java side com.fasterxml.jackson.databind.JsonNode. That way we can keep the structure and the parsing is done automatically.

Check https://www.baeldung.com/spring-boot-jpa-storing-postgresql-jsonb section 3.

vidakovic · 2024-09-05T12:38:01Z

custom/v3/command/domain/src/main/resources/db/custom-changelog/0001_create_command_table.xml

+            <column name="request_idempotency_key" type="VARCHAR(50)" remarks="Unique request idempotency key">
+                <constraints nullable="false" unique="true"/>
+            </column>
+            <column name="command_processing_status" type="VARCHAR(255)" remarks="Command processing status"/>


Why do we need yet another status... and in a different type? Let's do either integer or string or enum status...

Because they should represent different things.

The integer => Represents the http status code.(200, 201, 403, 404 ...etc)

The enum => Represents the command execution status within the system. (PROCESSED, UNDER_PROCESSING, ERROR ...etc)

vidakovic · 2024-09-05T12:47:43Z

...m/v3/command/mapping/src/main/java/org/apache/fineract/v3/command/mapping/CommandMapper.java

+    }
+
+    @Named("mapRequestBody")
+    default JsonNode mapRequestBody(Object body) {


This should be done by a Spring Converter. See my previous comments above and the Baeldung link. If implemented correctly then this default function is not necessary (Spring Data JPA will automatically take care of it).

vidakovic · 2024-09-05T12:52:47Z

.../note/starter/src/main/java/org/apache/fineract/v3/note/starter/NoteV3AutoConfiguration.java

+@EnableConfigurationProperties({ NoteProperties.class })
+@ComponentScan("org.apache.fineract.v3.note")
+@ConditionalOnProperty(value = "fineract.note.enabled", havingValue = "true")
+public class NoteV3AutoConfiguration {}


Just NoteAutoConfiguration will do.

I tried it but there's a conflict between it and the other configuration in the upstream code.

fineract/fineract-provider/src/main/java/org/apache/fineract/portfolio/note/starter/NoteAutoConfiguration.java

Line 38 in 24c83b5

public class NoteAutoConfiguration {

Here's the error logs:

fineract-fineract-1 | org.springframework.beans.factory.BeanDefinitionStoreException: Failed to process import candidates for configuration class [org.apache.fineract.ServerApplication$Configuration]: Annotation-specified bean name 'noteAutoConfiguration' for bean class [org.apache.fineract.portfolio.note.starter.NoteAutoConfiguration] conflicts with existing, non-compatible bean definition of same name and class [org.apache.fineract.v3.note.starter.NoteAutoConfiguration]

…ajor improvements with the request processing flow. This commit squashes 2 commits (Read & Write) type-safe request **Read Requests** FINERACT-2021: refactor the note GET end-points to eliminate manual serialization with GSON. - Implemented a new custom module for the note API. - Used Spring Web-MVC annotations instead of JAX-RS. - Dispensed the redundant query parameters like "fields, prettyPrint ..etc" - Implemented a new `CommonWebConfiguration` for establishing the security configurations (not complete yet). - Implemented configuration modules (core, starter, configuration) for configuring the note module. - Implemented a new `docker-compose-api-v3.yml` configuration file. **Write Requests** FINERACT-2110: Add basic structure for the new mirror type-safe API (Commands Proof of Concept) - Made a unique command (Request & Response) for each http method. - Define a new way for locating the dedicated handlers. - Add basic structure for the new mirror type-safe API (Commands). - Add a new domain module for storing `Command` entity and add a repository for it. - Re-Implement the logic within the Note Write Service. - Implement a `ControllerAdvice` class for handling the exceptions that's thrown when working with note API.

github-actions · 2024-10-08T00:19:01Z

This pull request seems to be stale. Are you still planning to work on it? We will automatically close it in 30 days.

vidakovic self-requested a review July 26, 2024 06:04

Zeyad2003 force-pushed the FINERACT-2021/refactor_Write_Methods branch 2 times, most recently from 1381b9e to 885a2d8 Compare July 26, 2024 06:23

vidakovic requested changes Jul 26, 2024

View reviewed changes

Zeyad2003 changed the title ~~Introduce type-safety API for commands~~ FINERACT-2110: Introduce type-safety API for commands (Note API) Jul 27, 2024

Zeyad2003 force-pushed the FINERACT-2021/refactor_Write_Methods branch 7 times, most recently from 026a952 to 00e499a Compare July 29, 2024 12:15

vidakovic reviewed Jul 29, 2024

View reviewed changes

Zeyad2003 force-pushed the FINERACT-2021/refactor_Write_Methods branch 12 times, most recently from 0b7d73d to 654d70d Compare August 5, 2024 15:55

Zeyad2003 force-pushed the FINERACT-2021/refactor_Write_Methods branch 3 times, most recently from 831d1e7 to e6c3504 Compare August 9, 2024 07:27

Zeyad2003 force-pushed the FINERACT-2021/refactor_Write_Methods branch from 2514be6 to dd88e8c Compare August 27, 2024 20:39

Zeyad2003 mentioned this pull request Aug 29, 2024

FINERACT-2119: Address jobs by name #4037

Merged

6 tasks

Zeyad2003 force-pushed the FINERACT-2021/refactor_Write_Methods branch 12 times, most recently from 24959a7 to 6d617fe Compare September 1, 2024 07:12

Zeyad2003 requested a review from vidakovic September 1, 2024 07:22

Zeyad2003 force-pushed the FINERACT-2021/refactor_Write_Methods branch 2 times, most recently from df0a5a0 to 004e718 Compare September 1, 2024 23:24

vidakovic reviewed Sep 5, 2024

View reviewed changes

Zeyad2003 mentioned this pull request Sep 6, 2024

FINERACT-2021: refactor the note GET end-points to use Jackson instead of GSON. #3915

Closed

6 tasks

Zeyad2003 force-pushed the FINERACT-2021/refactor_Write_Methods branch from 004e718 to f59cbbb Compare September 7, 2024 04:12

github-actions bot added the stale label Oct 8, 2024

github-actions bot closed this Nov 7, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

FINERACT-2110: Introduce type-safety API for commands (Note API) #3994

FINERACT-2110: Introduce type-safety API for commands (Note API) #3994

Zeyad2003 commented Jul 26, 2024

vidakovic Jul 26, 2024

Zeyad2003 Jul 29, 2024

vidakovic Jul 26, 2024

vidakovic Jul 26, 2024

vidakovic Jul 26, 2024

Zeyad2003 Jul 28, 2024

vidakovic Jul 26, 2024

Zeyad2003 Jul 28, 2024

Zeyad2003 Jul 29, 2024

vidakovic Jul 29, 2024

Zeyad2003 Jul 29, 2024 •

edited

Loading

vidakovic Jul 29, 2024

Zeyad2003 Jul 29, 2024

vidakovic Jul 29, 2024

vidakovic commented Sep 5, 2024

vidakovic Sep 5, 2024

vidakovic Sep 5, 2024 •

edited

Loading

Zeyad2003 Sep 5, 2024

adamsaghy Sep 5, 2024

Zeyad2003 Sep 5, 2024

vidakovic Sep 5, 2024

vidakovic Sep 5, 2024

vidakovic Sep 5, 2024 •

edited

Loading

vidakovic Sep 5, 2024

Zeyad2003 Sep 5, 2024

vidakovic Sep 5, 2024

vidakovic Sep 5, 2024

Zeyad2003 Sep 5, 2024

github-actions bot commented Oct 8, 2024


		public interface TypedCommandProcessingService {

		<T> CommandProcessingResult executeCommand(CommandTypedWrapper<T> wrapper);

		@@ -57,7 +47,7 @@ public JsonCommand getJsonCommand() {
		}

		return new JsonCommand(null, json, null, null, entityName, resourceId,

FINERACT-2110: Introduce type-safety API for commands (Note API) #3994

FINERACT-2110: Introduce type-safety API for commands (Note API) #3994

Conversation

Zeyad2003 commented Jul 26, 2024

Description

Checklist

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Zeyad2003 Jul 29, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

vidakovic commented Sep 5, 2024

Choose a reason for hiding this comment

vidakovic Sep 5, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

vidakovic Sep 5, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

github-actions bot commented Oct 8, 2024

Zeyad2003 Jul 29, 2024 •

edited

Loading

vidakovic Sep 5, 2024 •

edited

Loading

vidakovic Sep 5, 2024 •

edited

Loading