Engine initialization

You need three things to initialize the engine:

1. data handler
2. cache handler
3. logger

• optional flag which controls if this instance of engine will be marked as default or not.

logger = TextLogger('eos_tq', 'data_folder/logs/eos.log')
data_handler = JsonDataHandler('data_folder/phobos/')
cache_handler = JsonCacheHandler('data_folder/cache/eos.json.bz2', logger)
engine = Eos(data_handler, cache_handler, logger, make_default=True)

Data handler

It will provide access to the data Eos needs. Built-in data handlers:

• [JsonDataHandler] (https://github.com/DarkFenX/Eos/blob/master/data/data_handler/json_data_handler.py), which takes path to folder with [Phobos] (https://github.com/DarkFenX/Phobos) JSON dump as initialization argument.
• [SQLiteDataHandler] (https://github.com/DarkFenX/Eos/blob/master/data/data_handler/sqlite_data_handler.py) which takes path to SQLite database as initialization argument. Data still should be in Phobos-like format.

Of course, you can write your own handler as needed. Interface is defined in abstract base class.

Eos needs access to following data in format of [{field name: field value}, ...] for each table:

• invtypes
  • typeID
  • groupID
  • radius
  • mass
  • volume
  • capacity
• invgroups
  • groupID
  • groupName
  • categoryID
• dgmattribs
  • attributeID
  • maxAttributeID
  • defaultValue
  • highIsGood
  • stackable
• dgmtypeattribs
  • typeID
  • attributeID
  • value
• dgmeffects
  • effectID
  • effectCategory
  • isOffensive
  • isAssistance
  • durationAttributeID
  • dischargeAttributeID
  • rangeAttributeID
  • falloffAttributeID
  • trackingSpeedAttributeID
  • fittingUsageChanceAttributeID
  • preExpression
  • postExpression
  • modifierInfo
• dgmtypeeffects
  • typeID
  • effectID
  • isDefault
• dgmexpressions
  • expressionID
  • operandID
  • arg1
  • arg2
  • expressionValue
  • expressionTypeID
  • expressionGroupID
  • expressionAttributeID
• phbmetadata
  • field_name
  • field_value

Cache handler

Its purpose is provide fast and efficient way to the data Eos needs for calculations. It should take data in the following format ({entity name: set({field name: field value}, ...)}:

• types
  • type_id (integer)
  • group (integer)
  • category (integer)
  • attributes ({integer: float, ...})
  • effects ([integer, ...])
  • default_effect (integer)
• attributes
  • attribute_id (integer)
  • max_attribute (integer)
  • default_value (float)
  • high_is_good (boolean)
  • stackable (boolean)
• effects
  • effect_id (integer)
  • effect_category (integer)
  • is_offensive (boolean)
  • is_assistance (boolean)
  • duration_attribute (integer)
  • discharge_attribute (integer)
  • range_attribute (integer)
  • falloff_attribute (integer)
  • tracking_speed_attribute (integer)
  • fitting_usage_chance_attribute (integer)
  • build_status (integer)
  • modifiers ([integer, ...])
• modifiers
  • modifier_id (integer)
  • state (integer)
  • scope (integer)
  • src_attr (integer)
  • operator (integer)
  • tgt_attr (integer)
  • domain (integer)
  • filter_type (integer)
  • filter_value (integer)

When requested, it should return various objects needed for Eos when it uses handler's corresponding methods.

Built-in cache handlers:

• [JsonCacheHandler] (https://github.com/DarkFenX/Eos/blob/master/data/cache_handler/json_cache_handler.py) - fairly fast handler with medium-to-large memory footprint.

Interface is defined in [BaseCacheHandler] (https://github.com/DarkFenX/Eos/blob/master/data/cache_handler/abc.py).

Logger

As simple as it sounds, provides logging facilities. Following loggers are built-in:

• [TextLogger] (https://github.com/DarkFenX/Eos/blob/master/util/logger/text_logger.py) - write logs into plain text file.

Interface is defined in [BaseLogger] (https://github.com/DarkFenX/Eos/blob/master/util/logger/abc.py).

Default instance flag

If true, all fits created within current python interpreter will be attached to this Eos instance by default; otherwise, you will have to pass Eos instance each time you instantiate a fit.