Re: [BUGS]log can not be output when use DataSource

From: Chen Huajun <chenhj(at)cn(dot)fujitsu(dot)com>
To: dmp <danap(at)ttc-cmc(dot)net>
Cc: Dave Cramer <pg(at)fastcrypt(dot)com>, pgsql-jdbc(at)postgresql(dot)org
Subject: Re: [BUGS]log can not be output when use DataSource
Date: 2013-01-29 05:43:18
Message-ID: 51076176.5070800@cn.fujitsu.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-jdbc


> 1. Dave add tasks to get the Manual to the same state as existing properties
> that are implemented. So lacking documentation are:
>
> ApplicationName
> binaryTransfer
> binaryTransferDisable
> binaryTransferEnable
> receiveBufferSize
> sendBufferSize

My work was not based the newest source,
so i checked again ,and found the "readOnly" was alse lacked.

> 2. Chen, No I do not think at this time that ALL properties be implemented
> in BaseDataSource at this time unless you desire to do so. As indicated
> my initial concern was properties that are implemented and have the same
> bug as you have already identified with logLevel.

I agree with you danap.

> 3. The fourth column Chen in your spreadsheet is usefull in identifying two
> properties that are defined in the BaseDataSource but not implemented
> in getURL(). The properties user & password are included in getConnection()
> and therefore do not not need to be in getURL(). The two not implemented:

I advocate user & password also need to be in getURL().
Because getURL() is the key way to affect a connection created by the DataSource.
If a property not need be included in getURL(),it also seems not needed to appear in BaseDataSource.
And just like DriverManager,by which can set user & password via getConnection's arguments or url,
DataSource should same to DriverManager.

Connection db = DriverManager.getConnection(url, username, password);

or

String url = "jdbc:postgresql://localhost/test?user=fred&password=secret&ssl=true";
Connection conn = DriverManager.getConnection(url);

Chen Huajun
(2013/01/29 1:54), dmp wrote:
> Indeed Chen this is very good work. Thank you providing this information.
>
> Dave though it should be the ultimate goal to have the one to one match
> for all supported properties in the BaseDataSource this was not my initial
> intend for asking for the review of BaseDataSource.
>
> Chen you have defined the complete solution, but I was more interested
> in checking the existing BaseDataSource Properties being properly implemented
> because of the bug you found with logLevel. I suspect that other properties
> that are now implemented have the same bug.
>
> I would advocate that fixing the existing class is more important at this
> time then implementing the complete properties set. Of course if this is
> acceptable to you Chen for a course of action it is up to you.
>
> Recommend:
>
> 1. Dave add tasks to get the Manual to the same state as existing properties
> that are implemented. So lacking documentation are:
>
> ApplicationName
> binaryTransfer
> binaryTransferDisable
> binaryTransferEnable
> receiveBufferSize
> sendBufferSize
>
> 2. Chen, No I do not think at this time that ALL properties be implemented
> in BaseDataSource at this time unless you desire to do so. As indicated
> my initial concern was properties that are implemented and have the same
> bug as you have already identified with logLevel.
>
> 3. The fourth column Chen in your spreadsheet is usefull in identifying two
> properties that are defined in the BaseDataSource but not implemented
> in getURL(). The properties user & password are included in getConnection()
> and therefore do not not need to be in getURL(). The two not implemented:
>
> binaryTransferDisable
> binaryTransferEnable
>
> These two should probably be added since that were already defined,
> but not implemented.
>
> danap.
>
>
> Dave Cramer wrote:
>> Chen,
>>
>> This is excellent work.
>>
>> I don't see any reason why all of the documented url parameters should
>> not be available manually, and in both the data source and the connection.
>>
>> Dave
>>
>> Dave Cramer
>>
>> dave.cramer(at)credativ(dot)ca
>> http://www.credativ.ca
>>
>>
>> On Mon, Jan 28, 2013 at 6:28 AM, Chen Huajun <chenhj(at)cn(dot)fujitsu(dot)com
>> <mailto:chenhj(at)cn(dot)fujitsu(dot)com>> wrote:
>>
>> Hi
>>
>> I reviewed the source related to properties.
>> My idea is that the set of supported properties should be the same
>> no matter the means of setting value(by url or by DataSource).
>> So I get all properties by searching keyword ".getProperty(" in
>> pgjdbc source,
>> and then check whether BaseDataSource has correctly deal with all them.
>> As a result i found many inconsistencies as listed in the attachment.
>>
>> I am wondering about if it's needed to add all lacked properties to
>> BaseDataSource?
>>
>>
>> Chen Huajun
>>
>> (2013/01/24 2:35), dmp wrote:
>>
>> Thank you sir for helping in identifying problems with the pgjdbc
>> code and contributing to maintaining this project.
>>
>> danap.
>>
>> Chen Huajun wrote:
>>
>> danap,
>>
>> > Could you and possibly while considering the new possible
>> bug you
>> found with
>> > copying DataSource do a more through review of the
>> BaseDataSource class?
>>
>> OK,I will review the BaseDataSource class first.
>
>
>

In response to

Responses

Browse pgsql-jdbc by date

  From Date Subject
Next Message dmp 2013-01-29 16:54:10 Re: Bug report: NullPointerException from Driver.connect when passed a Properties with non-string values
Previous Message Hari Babu 2013-01-29 05:09:45 Re: JDBC connection test with SSL on PG 9.2.1 server