Propose refactor

[pherrymason]

Propose refactor

[pherrymason]

I see this as an improvement some complex projects may benefit.
Some questions thouhg:

• A writer that stores info in database, will write data in the same format as in a .po file?
• If I understand correclty Parser will always expect a string as input previously supplied by the reeader.

[newage]

1. Standard reader and writer can work with .po files.
   Extend reader/writer can work with DB. For example: fast search messages, translate with WHERE, create pagination for web UI. (For ExtJS UI). Or maybe use console translate-tool reader and writer.
2. Yes, correctly about Parser

[pherrymason]

While I see your point in extending the class to support database, I'm thinking in the following scenarios:

1. Storing the whole .po file in a row inside a database. I see no point in supporting this by PoParser. While currently you can only parse files, there is a pull request that allows parsing a string. This will allow anyone parse any .po string format from any source.
2. Designing a database schema to read and write msgid efficiently and do it transparent to the developer. While it is a good idea, I think this should belong to a whole new project by itself. The main purpose of PoParse is to be able to read easily a .po file/string and modify it. If msgid's are stored in a completely different way, then this is not the job of PoParse anymore.

I will recheck the pull request as I think there was some caveats still pending and we'll be able to parse any source string.

[newage]

I agree with you. Parser object must contains read and write methods. And read and write only .po files

[pherrymason]

I've improved PoParser and now it has three new methods to parse PO data:

• parseFile( filepath )
• parseString( string )
• parse( Handler )

Handler must be an object of a class implementing the InterfaceHandler so theoretically this could allow to implement a way of reading PO data file from any source.