kdabir / gstorm Goto Github PK

We need to try this out, so that different groovy version is not downloaded because of transitive dependencies. Gstorm is not used in Java projects, so it is assumed some version of groovy is already available in the project / script in which gstorm is included.

Will be awesome if new Gstorm() can detect if a file with database connetion properties in the classpath and use it. For example gstorm.properties with something like:

database=myDatabase
user=myUser
password=myPassword
....

a static method ModelClass.count() that returns the count of rows in table would be a useful addition.

I would expect the where method to return a collection of the Item class, but it returns a collection of GroovyRowResult, which kind of cancels the purpose of ORM

Hi, is there any way to stormify an object but not create a new table? Would be helpful with legacy databases.

Currently on intialization of gstorm, a create table query is executed with not exists clause (CREATE TABLE IF NOT EXISTS...). It's better to give user an option to exclude creation of table.

Try compile time meta-programming instead of runtime meta-programming. Key benefit includes IDE support and compilation failures instead of runtime failures.

So that csv files can be directly opened with gstorm and sql queries can be performed.

It would be nice to have capability to be able to create indexed fields for better performance.
class Person { @Indexed String name }

Either github-pages (maybe using jekyll, or homegrown solution) or Wiki. Only considering markdown as the format in which docs will be written.

It would be great to have doc as part of main source repository, so not github considering wiki for now.

With jekyll md files can be converted to html by github, but need lot of steup and jekyll is more blog oriented than user manual.

Homegrown solution, generated html will also needed to be checked in to gh-pages.

Remove file based versioning. Can use https://github.com/allegro/axion-release-plugin

After upgrading groovy version, running tests gives this error occurs :

Execution failed for task ':compileTestGroovy'.
> Bad <init> method call from inside of a branch
  Exception Details:
    Location:
      gstorm/Gstorm.<init>(Ljava/sql/Connection;)V @91: invokespecial
    Reason:
      Error exists in the bytecode
    Bytecode:
      0x0000000: b800 1a4d 04bd 0004 5903 2c12 5632 122b
      0x0000010: 2bb9 0059 0300 5359 10ff 1202 b800 292a
      0x0000020: 5fab 0000 0000 0071 0000 0004 8065 35b8
      0x0000030: 0000 002b aad3 b1ff 0000 0040 e83e bc00
      0x0000040: 0000 0053 0000 9b75 0000 0068 5f5a 5903
      0x0000050: 3212 2bb8 002f c000 2b5f 57b7 0032 a700
      0x0000060: 3e5f 5a59 0332 b800 38c0 003a 5f57 b700
      0x0000070: 3da7 002b 5f5a 5903 3212 3fb8 002f c000
      0x0000080: 3f5f 57b7 0042 a700 165f 5a57 b700 44a7
      0x0000090: 000d bb00 4659 1248 b700 49bf 57b1
    Stackmap Table:
      full_frame(@76,{UninitializedThis,Object[#63],Object[#77]},{Object[#79],UninitializedThis})
      full_frame(@97,{UninitializedThis,Object[#63],Object[#77]},{Object[#79],UninitializedThis})
      full_frame(@116,{UninitializedThis,Object[#63],Object[#77]},{Object[#79],UninitializedThis})
      full_frame(@137,{UninitializedThis,Object[#63],Object[#77]},{Object[#79],UninitializedThis})
      full_frame(@146,{UninitializedThis,Object[#63],Object[#77]},{Object[#79],UninitializedThis})
      full_frame(@156,{Object[#2],Object[#63],Object[#77]},{Object[#79]})

Happens with groovy 2.3.4+ on JDK 1.8 u11/20. Works on Groovy 2.3.2. So can not bump groovy until this is fixed. Found related issues in both groovy and JDK:

• https://jira.codehaus.org/browse/GROOVY-6951
• https://bugs.openjdk.java.net/browse/JDK-8051012

Will be awesome if you can annotate a model class, and gstorm do all job:

@GstormTable
class Person { 
  String name, project 
}

So, no need of:

def g = new Gstorm()
g.stormify(Person)   

like Double and Float.

Upgrade versions of dependencies to latest

• Gradle
• Gradle Wrapper
• Travis Java versions (build matrix)
• Groovy
• Gradle plugins
• HSQLDB

The default size is not good enough to accommodate all lengths of String.

Can considering adding SQLite DB as it is very commonly used file based db.

In existing tables, where ID field can have different name, add capability to configure name of the field (and perhaps the type)

Currently one test is failing on jre11, possibly because of a Groovy Bug : https://issues.apache.org/jira/browse/GROOVY-8871

gstorm.GstormDataTypesTest > initializationError FAILED
java.lang.IllegalAccessError at GstormDataTypesTest.groovy:-1
73 tests completed, 1 failed

Try to update to groovy 2.5.4 once released to see if that fixes the issue.

In groovy scripts, in grape jcenter is included by default, so user won't have to include a custom repo. in Gradle build, jcenter() repo is also easily configurable.

This would need a change in groupId, something like io.github.kdabir.gstorm.

There is lot of boilerplate code that goes in the beginning of the script to gstormify a model class. Few step to reduce can be to add constructors that automatically use in memory / file db / connection etc.

Also can include hsqldb with gstorm to addition grab is gone. Just be careful with (systemClassLoader = true) stuff.

this way user of library can control logging in a standard way