Match
Processor that matches objects and data types based on given configuration.
Classes:
|
Config for processing string-type tag values before matching with |
Matches source items/collections to given result(s). |
- class musify.processors.match.CleanTagConfig(tag, remove=<factory>, split=<factory>, preprocess=<function CleanTagConfig.<lambda>>)
Bases:
PrettyPrinter
Config for processing string-type tag values before matching with
ItemMatcher
Attributes:
The name of the tag to clean.
A set of string values to remove from this tag.
A set of string values for which the cleaner will slice the tag on and remove anything that comes after.
Methods:
A function to apply before the remove and split values are applied.
as_dict
()Return a dictionary representation of the key attributes of this object.
-
remove:
set
[str
] A set of string values to remove from this tag.
-
split:
set
[str
] A set of string values for which the cleaner will slice the tag on and remove anything that comes after.
- preprocess()
A function to apply before the remove and split values are applied.
- as_dict()
Return a dictionary representation of the key attributes of this object.
- Return type:
dict
[str
,Any
]
- The results of this function are used to produce the following:
A JSON representation of the object when calling
json()
The string representation of the object when calling str() on the object
The representation of the object when calling repr() on the object
-
remove:
- class musify.processors.match.ItemMatcher
Bases:
Processor
Matches source items/collections to given result(s).
Attributes:
A set of words to search for in tag values that identify the item as being a karaoke item.
A difference in years of this value gives a score of 0 for the
match_year()
algorithm.Apply these remove settings to all tags when processing tags as per
clean_tags()
method.Apply these split settings to all tags when processing tags as per
clean_tags()
method.A list of configurations in the form of
CleanTagConfig
to apply for each tag type.A set of words to check for when applying name score reduction logic.
The factor to apply to a name score when a word from
reduce_name_score_on
is found in the result but not in the sourceMusifyObject
.The
MusifyLogger
for this objectMethods:
log
(messages[, pad])Log lists of
messages
in a uniform aligned format with a givenpad
character.clean_tags
(source)Clean tags on the input item and assign to its
clean_tags
attribute.match_not_karaoke
(source, result)Checks if a result is not a karaoke item that is either 0 when item is karaoke or 1 when not karaoke.
match_name
(source, result)Match on names and return a score between 0-1.
match_artist
(source, result)Match on artists and return a score between 0-1.
match_album
(source, result)Match on album and return a score between 0-1.
match_length
(source, result)Match on length and return a score between 0-1.
match_year
(source, result)Match on year and return a score between 0-1.
match
(source, results[, min_score, ...])Perform match algorithm for a given item and its results.
as_dict
()Return a dictionary representation of the key attributes of this object.
- karaoke_tags = {'backing', 'instrumental', 'karaoke'}
A set of words to search for in tag values that identify the item as being a karaoke item.
- year_range = 10
A difference in years of this value gives a score of 0 for the
match_year()
algorithm. See thematch_year()
method for more information.
- clean_tags_remove_all = {'&', 'a', 'and', 'the'}
Apply these remove settings to all tags when processing tags as per
clean_tags()
method. See alsoCleanTagConfig
for more info on this configuration.
- clean_tags_split_all = {}
Apply these split settings to all tags when processing tags as per
clean_tags()
method. See alsoCleanTagConfig
for more info on this configuration.
- clean_tags_config = (CleanTagConfig(tag=<TagFields.TITLE: 65>, remove={'part'}, split={'featuring', '/', 'feat.', 'ft.'}, preprocess=<function CleanTagConfig.<lambda>>), CleanTagConfig(tag=<TagFields.ARTIST: 32>, remove=set(), split={'vs', 'featuring', 'feat.', 'ft.'}, preprocess=<function CleanTagConfig.<lambda>>), CleanTagConfig(tag=<TagFields.ALBUM: 30>, remove={'ep'}, split=set(), preprocess=<function ItemMatcher.<lambda>>))
A list of configurations in the form of
CleanTagConfig
to apply for each tag type. See alsoCleanTagConfig
for more info.
- reduce_name_score_on = {'acoustic', 'demo', 'live'}
A set of words to check for when applying name score reduction logic.
If a word from this list is present in the name of the result to score against but not in the source
MusifyObject
, apply thereduce_name_score_factor
to reduce its score. This set is always combined with thekaraoke_tags
.
- reduce_name_score_factor = 0.5
The factor to apply to a name score when a word from
reduce_name_score_on
is found in the result but not in the sourceMusifyObject
.
-
logger:
MusifyLogger
The
MusifyLogger
for this object
- log(messages, pad=' ')
Log lists of
messages
in a uniform aligned format with a givenpad
character.Convenience function for ensuring consistent log format for results of operations of this class and any other classes which use this class.
- Return type:
None
- clean_tags(source)
Clean tags on the input item and assign to its
clean_tags
attribute. Used for better matching/searching. Clean by removing words, and only taking phrases before a certain word e.g. ‘featuring’, ‘part’. Cleaning config for string-type tags is set in_clean_tags_config
.- Parameters:
source (
MusifyObject
) – The base object with tags to clean.- Return type:
None
- match_not_karaoke(source, result)
Checks if a result is not a karaoke item that is either 0 when item is karaoke or 1 when not karaoke.
- Return type:
int
- match_name(source, result)
Match on names and return a score between 0-1. Score=0 when either value is None.
- Return type:
float
- match_artist(source, result)
Match on artists and return a score between 0-1. Score=0 when either value is None. When many artists are present, a scale factor is applied to the score of matches on subsequent artists. i.e. match on artist 1 is scaled by 1, match on artist 2 is scaled by 1/2, match on artist 3 is scaled by 1/3 etc.
- Return type:
float
- match_album(source, result)
Match on album and return a score between 0-1. Score=0 when either value is None.
- Return type:
float
- match_length(source, result)
Match on length and return a score between 0-1. Score=0 when either value is None.
- Return type:
float
- match_year(source, result)
Match on year and return a score between 0-1. Score=0 when either value is None. Matches within
year_range
years on a 0-1 scale where 1 is the exact same year and 0 is a difference in year greater thatyear_range
. User may modify this max range via theyear_range
class attribute.- Return type:
float
- match(source, results, min_score=0.1, max_score=0.8, match_on=frozenset({TagFields.FILENAME, TagFields.TYPE, TagFields.SIZE, TagFields.CHANNELS, TagFields.SAMPLE_RATE, TagFields.BIT_RATE, TagFields.DATE_MODIFIED, TagFields.DATE_ADDED, TagFields.LAST_PLAYED, TagFields.PLAY_COUNT, TagFields.LENGTH, TagFields.ALBUM, TagFields.ALBUM_ARTIST, TagFields.ARTIST, TagFields.YEAR, TagFields.COMPOSER, TagFields.COMMENTS, TagFields.CONDUCTOR, TagFields.DISC_NUMBER, TagFields.DISC_TOTAL, TagFields.GENRES, TagFields.TITLE, TagFields.PUBLISHER, TagFields.RATING, TagFields.BPM, TagFields.TRACK_NUMBER, TagFields.TRACK_TOTAL, TagFields.EXT, TagFields.PATH, TagFields.FOLDER, TagFields.BIT_DEPTH, TagFields.DATE, TagFields.MONTH, TagFields.DAY, TagFields.KEY, TagFields.COMPILATION, TagFields.IMAGES, TagFields.DATE_CREATED, TagFields.DESCRIPTION, TagFields.URI, TagFields.USER_ID, TagFields.USER_NAME, TagFields.OWNER_ID, TagFields.OWNER_NAME, TagFields.FOLLOWERS, TagFields.NAME}), allow_karaoke=False)
Perform match algorithm for a given item and its results.
- Parameters:
source (
TypeVar
(T
, bound=MusifyObject
)) – Source item to compare against and find a match for.results (
Iterable
[TypeVar
(T
, bound=MusifyObject
)]) – Results for comparisons.min_score (
float
(default:0.1
)) – Only return the result as a match if the score is above this value. Value will be limited to between 0.01 and 1.0.max_score (
float
(default:0.8
)) – Stop matching once this score has been reached. Value will be limited to between 0.01 and 1.0.match_on (
GenericAlias
[TagField
] (default:frozenset({<TagFields.FILENAME: 3>, <TagFields.DATE: 900>, <TagFields.MONTH: 901>, <TagFields.DAY: 902>, <TagFields.KEY: 903>, <TagFields.COMPILATION: 904>, <TagFields.IMAGES: 905>, <TagFields.SIZE: 7>, <TagFields.TYPE: 4>, <TagFields.CHANNELS: 8>, <TagFields.BIT_RATE: 10>, <TagFields.SAMPLE_RATE: 9>, <TagFields.DATE_MODIFIED: 11>, <TagFields.LENGTH: 16>, <TagFields.DATE_ADDED: 12>, <TagFields.LAST_PLAYED: 13>, <TagFields.PLAY_COUNT: 14>, <TagFields.DATE_CREATED: 921>, <TagFields.ALBUM: 30>, <TagFields.ALBUM_ARTIST: 31>, <TagFields.ARTIST: 32>, <TagFields.YEAR: 35>, <TagFields.DESCRIPTION: 931>, <TagFields.COMPOSER: 43>, <TagFields.COMMENTS: 44>, <TagFields.CONDUCTOR: 45>, <TagFields.URI: 941>, <TagFields.USER_ID: 942>, <TagFields.USER_NAME: 943>, <TagFields.OWNER_ID: 944>, <TagFields.OWNER_NAME: 945>, <TagFields.FOLDER: 179>, <TagFields.DISC_NUMBER: 52>, <TagFields.FOLLOWERS: 946>, <TagFields.DISC_TOTAL: 54>, <TagFields.BIT_DEPTH: 183>, <TagFields.GENRES: 59>, <TagFields.TITLE: 65>, <TagFields.PUBLISHER: 73>, <TagFields.RATING: 75>, <TagFields.BPM: 85>, <TagFields.TRACK_NUMBER: 86>, <TagFields.TRACK_TOTAL: 87>, <TagFields.EXT: 100>, <TagFields.NAME: 1000>, <TagFields.PATH: 106>})
)) – List of tags to match on. Currently only the following fields are supported:title
,artist
,album
,year
,length
.allow_karaoke (
bool
(default:False
)) – When True, items determined to be karaoke are allowed when matching added items. Skip karaoke results otherwise. Karaoke items are identified using thekaraoke_tags
attribute.
- Returns:
Optional
[TypeVar
(T
, bound=MusifyObject
)] – T. The item that matched best if found, None if no item matched conditions.
- as_dict()
Return a dictionary representation of the key attributes of this object.
- Return type:
dict
[str
,Any
]
- The results of this function are used to produce the following:
A JSON representation of the object when calling
json()
The string representation of the object when calling str() on the object
The representation of the object when calling repr() on the object