Integrate castep_outputs parser

[ladinesa]

• benchmark wrt to exisiting TextParser
• write interface

[oerc0122]

Hi, I'm the main developer of castep_outputs and since the project is still in an early stage, I just wanted to say a few things:

• This is a personal project in my own time and while I try to get issues sorted ASAP there may be times where I have other projects taking priority. It is not endorsed by the Castep Developers Group beyond casual discussions (which are positive, but still) and there is not an agreement to maintain it, though I do work closely with several of them and the possibility has been mentioned.
• A little caveat emptor. While the code has tests for the formats I expect to parse based on the test files castep uses, it may not be complete for all combinations of operations and a couple that I know are missing are open issues. The system tests I run on the castep_outputs repo itself are definitely not anywhere near complete. If you have any issues please report them and I'll be more than happy to fix them when I get the time.
• The current structure of the resultant "tree" of dictionaries is not final. I'm actually looking for feedback on changing the structure to improve how it can be best used. If there's anything you feel should be in a different order, please open an issue/discussion and I will see how that might be done.
• I'm currently in the process of restructuring the code as the project has grown, breaking it down into modules and packages. I'm trying to preserve the main external interfaces, but be aware. I can notify you before that changes if you'd like.
• Documentation is a little thin on the ground and some of the features which partially exist are only documented in my head (.castep parsing filters are the main thing).
• It does not currently handle any of the binary castep outputs, only the human readable, though there is another project castepxbin which does, looking to open discussions soon to see about the possibility of adding these features.

[ladinesa]

Hi, I'm the main developer of castep_outputs and since the project is still in an early stage, I just wanted to say a few things:

• This is a personal project in my own time and while I try to get issues sorted ASAP there may be times where I have other projects taking priority. It is not endorsed by the Castep Developers Group beyond casual discussions (which are positive, but still) and there is not an agreement to maintain it, though I do work closely with several of them and the possibility has been mentioned.
• A little caveat emptor. While the code has tests for the formats I expect to parse based on the test files castep uses, it may not be complete for all combinations of operations and a couple that I know are missing are open issues. The system tests I run on the castep_outputs repo itself are definitely not anywhere near complete. If you have any issues please report them and I'll be more than happy to fix them when I get the time.
• The current structure of the resultant "tree" of dictionaries is not final. I'm actually looking for feedback on changing the structure to improve how it can be best used. If there's anything you feel should be in a different order, please open an issue/discussion and I will see how that might be done.
• I'm currently in the process of restructuring the code as the project has grown, breaking it down into modules and packages. I'm trying to preserve the main external interfaces, but be aware. I can notify you before that changes if you'd like.
• Documentation is a little thin on the ground and some of the features which partially exist are only documented in my head (.castep parsing filters are the main thing).
• It does not currently handle any of the binary castep outputs, only the human readable, though there is another project castepxbin which does, looking to open discussions soon to see about the possibility of adding these features.

Thanks for the clarification. I am still on the process of testing it and going through the data structure to see how it matches our schema. I would then wait until the data structure is finalised but it would be great if this works out as the standard castep parser.