Models

`models` ¶

Defines kotobase's database SQLAlchemy 2.0 typed ORM schema

Schema

Preserves every field from the upstream sources
Includes the JMdict part of speech, register, field, dialect, sense information and priority tags + furigana segmentation from jmdict_furigana
Inclues the JMNedict kanji, kana, and translation information + furigana segmentation from jmnedict_furigana
Includes the full KanjiDic2 reading, meaning and reference set, radical decompositions from KRADFILE, stroke orders from KanjiVG, and pronunciation audios from kanjialive (separate database, attached when present)
Includes japanese example sentence / translation pairs and audio provenance from Tatoeba
Includes grammar, kanji, and vocabulary information extracted from the Tanos JLPT Lists

Data Format

List shaped, read-only values such as tag code lists, cross references and furigana segments are stored in JSON columns using the SQLite json1 extension
Anything that is searched or joined is normalized into its own table
Full text search is provided by FTS5 virtual tables that the Build Pipeline creates at build time, so they are not declared here

Versioning

The schema is versioned by db_meta['schema_version']
Bump SCHEMA_VERSION whenever the table layout changes so that a stale database can be detected

`SCHEMA_VERSION = 1` `module-attribute` ¶

Layout version stored in db_meta and checked by the read layer

`Audio` ¶

Bases: Base

A pronunciation audio clip together with its provenance

Optional Audio Pack

The data column is filled only in the optional audio pack
In the core database, a row may instead carry a url that points at a remote clip, for example a Tatoeba recording

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`kind`	`str`	What the clip pronounces, such as `kanji_word`, `kanji_reading` or `sentence`
`key`	`str`	Lookup key for the clip, such as a kanji, a word or a sentence identifier
`reading`	`str \| None`	The reading the clip pronounces when relevant
`fmt`	`str \| None`	Audio container or codec such as `ogg` or `mp3`
`sample_rate`	`int \| None`	Sample rate of the clip in hertz
`data`	`bytes \| None`	Raw audio bytes when bundled in the audio pack
`url`	`str \| None`	Remote location of the clip when it is not bundled
`source`	`str`	Name of the upstream source, such as `kanjialive`
`license`	`str \| None`	License identifier for the clip
`attribution`	`str \| None`	Required attribution text or link

`Base` ¶

Bases: DeclarativeBase

Declarative base class shared by every kotobase ORM model

`DbMeta` ¶

Bases: Base

A build metadata key and value pair

Contains

Schema Version
Build Date
Build Duration
Database Size
Version / Date Of Each Data Source

Attributes:

Name	Type	Description
`key`	`str`	Primary key, the metadata name
`value`	`str \| None`	The metadata value, stored as text

`Furigana` ¶

Bases: Base

Furigana segmentation for a spelling and reading pair

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`text`	`str`	The written spelling, usually containing kanji
`reading`	`str`	The full kana reading of the spelling
`segments`	`list[dict]`	The JmdictFurigana segmentation, a list of `{"ruby": ..., "rt": ...}` records that align spans of the spelling with their readings, where the pair of text and reading is unique

`JMDictEntry` ¶

Bases: Base

A single JMdict dictionary entry

Represents one ent_seq record from JMdict, which is the root of a Japanese to English entry. Its surface forms, readings and senses are attached through relationships

Attributes:

Name	Type	Description
`id`	`int`	Primary key, the JMdict `ent_seq` sequence number
`is_common`	`bool`	True when any form of the entry carries a priority marker that classifies it as common
`freq_rank`	`int \| None`	Frequency rank where a lower value is more frequent, or None when the entry has no priority information
`kanji`	`list[JMDictKanji]`	Ordered kanji surface forms of the entry
`kana`	`list[JMDictKana]`	Ordered kana readings of the entry
`senses`	`list[JMDictSense]`	Ordered senses, each holding its glosses

`JMDictGloss` ¶

Bases: Base

A single gloss (translation) belonging to a JMdict sense

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`sense_id`	`int`	Foreign key to `JMDictSense`
`position`	`int`	Zero based order of the gloss within the sense
`lang`	`str`	ISO 639 language code of the gloss, defaults to `eng`
`text`	`str`	The translated text
`gender`	`str \| None`	Grammatical gender of the gloss when given
`gtype`	`str \| None`	Gloss type such as `lit`, `fig`, `expl` or `tm`
`sense`	`JMDictSense`	The owning sense

`JMDictKana` ¶

Bases: Base

A kana (reading) form of a JMdict entry

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`entry_id`	`int`	Foreign key to JMDictEntry
`position`	`int`	Zero based order of the reading within the entry
`text`	`str`	The kana reading
`is_common`	`bool`	True when this reading carries a common priority marker
`no_kanji`	`bool`	True when the reading is not a reading of any kanji form
`restrictions`	`list[str]`	Kanji forms this reading is restricted to, empty when the reading applies to all kanji forms
`info`	`list[str]`	Reading information tag codes such as `ik` or `ok`
`priority`	`list[str]`	Priority code list such as `news1` or `ichi1`
`entry`	`JMDictEntry`	The owning entry

`JMDictKanji` ¶

Bases: Base

A kanji (written) surface form of a JMdict entry

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`entry_id`	`int`	Foreign key to JMDictEntry
`position`	`int`	Zero based order of the form within the entry
`text`	`str`	The kanji spelling
`is_common`	`bool`	True when this form carries a common priority marker
`info`	`list[str]`	Spelling information tag codes such as `iK` or `ateji`
`priority`	`list[str]`	Priority code list such as `news1` or `ichi1`
`entry`	`JMDictEntry`	The owning entry

`JMDictSense` ¶

Bases: Base

A sense (one meaning) of a JMdict entry

A sense groups one or more glosses that share the same part of speech and usage information

Misc Tags

The misc list carries register and slang markers such as sl (slang), net-sl (internet slang), col (colloquial) and vulg (vulgar)

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`entry_id`	`int`	Foreign key to `JMDictEntry`
`position`	`int`	Zero based order of the sense within the entry
`pos`	`list[str]`	Part of speech tag codes such as `n` or `v5r`
`field`	`list[str]`	Field of application tag codes such as `comp` or `med`
`misc`	`list[str]`	Miscellaneous register tag codes, see the note above
`dialect`	`list[str]`	Dialect tag codes such as `ksb` for the Kansai dialect
`info`	`list[str]`	Free text sense information notes
`xref`	`list[str]`	Cross reference targets to related entries
`antonym`	`list[str]`	Antonym references for the sense
`applies_to_kanji`	`list[str]`	Kanji forms the sense is restricted to, empty when it applies to all kanji forms
`applies_to_kana`	`list[str]`	Kana forms the sense is restricted to, empty when it applies to all kana forms
`lsource`	`list[dict]`	Source language records, each holding the language, text, type and a wasei flag for loanwords
`entry`	`JMDictEntry`	The owning entry
`glosses`	`list[JMDictGloss]`	Ordered glosses belonging to the sense

`JMnedictEntry` ¶

Bases: Base

A JMnedict proper name entry

Attributes:

Name	Type	Description
`id`	`int`	Primary key, the JMnedict sequence number
`kanji`	`list[JMnedictKanji]`	Kanji forms of the name
`kana`	`list[JMnedictKana]`	Kana forms of the name
`translations`	`list[JMnedictTranslation]`	Ordered translation blocks

`JMnedictGloss` ¶

Bases: Base

A single translated name belonging to a JMnedict translation block

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`translation_id`	`int`	Foreign key to `JMnedictTranslation`
`position`	`int`	Zero based order of the gloss within the block
`lang`	`str`	ISO 639 language code of the gloss, defaults to `eng`
`text`	`str`	The translated name text
`translation`	`JMnedictTranslation`	The owning translation block

`JMnedictKana` ¶

Bases: Base

A kana form of a JMnedict entry

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`entry_id`	`int`	Foreign key to `JMnedictEntry`
`position`	`int`	Zero based order of the form within the entry
`text`	`str`	The kana reading of the name
`entry`	`JMnedictEntry`	The owning entry

`JMnedictKanji` ¶

Bases: Base

A kanji form of a JMnedict entry

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`entry_id`	`int`	Foreign key to `JMnedictEntry`
`position`	`int`	Zero based order of the form within the entry
`text`	`str`	The kanji spelling of the name
`entry`	`JMnedictEntry`	The owning entry

`JMnedictTranslation` ¶

Bases: Base

A translation block of a JMnedict entry

Each block records the kind of name and holds one or more translated glosses

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`entry_id`	`int`	Foreign key to JMnedictEntry
`position`	`int`	Zero based order of the block within the entry
`name_type`	`list[str]`	Name type tag codes such as `place`, `surname`, `given` or `company`
`xref`	`list[str]`	Cross reference targets to related entries
`entry`	`JMnedictEntry`	The owning entry
`glosses`	`list[JMnedictGloss]`	Ordered translated names in this block

`JlptGrammar` ¶

Bases: Base

A Tanos JLPT grammar point

Note

The formation and examples columns are kept for forward compatibility. The current Tanos data does not populate them, so they are normally empty

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`level`	`int`	JLPT level from 1 (hardest) to 5 (easiest)
`grammar`	`str`	The grammar point itself
`formation`	`str \| None`	How the grammar point is formed when known
`examples`	`list[str]`	Example sentences for the grammar point

`JlptKanji` ¶

Bases: Base

A Tanos JLPT kanji item

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`level`	`int`	JLPT level from 1 (hardest) to 5 (easiest)
`kanji`	`str`	The kanji character
`on_yomi`	`str \| None`	On readings, space separated
`kun_yomi`	`str \| None`	Kun readings, space separated
`meaning`	`str \| None`	The English meaning, with senses comma separated

`JlptVocab` ¶

Bases: Base

A Tanos JLPT vocabulary item

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`level`	`int`	JLPT level from 1 (hardest) to 5 (easiest)
`word`	`str \| None`	The headword, written with kanji when one exists and falling back to the kana reading otherwise
`reading`	`str \| None`	The kana reading of the headword
`meaning`	`str \| None`	The English meaning, with senses comma separated

`Kanji` ¶

Bases: Base

A KanjiDic2 character and its scalar attributes

The repeating attributes of a character such as readings, meanings and references live in dedicated child tables that are reachable through the relationships below

Attributes:

Name	Type	Description
`literal`	`str`	Primary key, the kanji character itself
`grade`	`int \| None`	School grade in which the kanji is taught
`stroke_count`	`int \| None`	Accepted stroke count
`freq`	`int \| None`	Newspaper frequency rank where a lower value is more frequent
`jlpt_old`	`int \| None`	Pre 2010 four level JLPT class from KanjiDic2
`rad_classical`	`int \| None`	Classical (Kangxi) radical number
`rad_nelson`	`int \| None`	Nelson radical number when it differs
`stroke_miscounts`	`list[int]`	Alternative miscount stroke values
`readings`	`list[KanjiReading]`	On, kun and foreign readings
`meanings`	`list[KanjiMeaning]`	Meanings keyed by language
`nanori`	`list[KanjiNanori]`	Name only readings
`dic_refs`	`list[KanjiDicRef]`	External dictionary references
`query_codes`	`list[KanjiQueryCode]`	Lookup codes such as SKIP
`variants`	`list[KanjiVariant]`	Variant form references
`codepoints`	`list[KanjiCodepoint]`	Character encoding codepoints
`strokes`	`KanjiStrokes \| None`	KanjiVG stroke order data when present

`KanjiCodepoint` ¶

Bases: Base

A character encoding codepoint of a kanji

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`literal`	`str`	Foreign key to `Kanji`
`type`	`str`	Codepoint type such as `ucs` or `jis208`
`value`	`str`	The codepoint value in that encoding
`kanji`	`Kanji`	The owning kanji

`KanjiDicRef` ¶

Bases: Base

An external dictionary reference for a kanji

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`literal`	`str`	Foreign key to `Kanji`
`type`	`str`	Reference type such as `nelson_c`, `heisig` or `moro`
`value`	`str`	The reference value within that dictionary
`extra`	`dict \| None`	Extra metadata, for example volume and page for Morohashi references
`kanji`	`Kanji`	The owning kanji

`KanjiMeaning` ¶

Bases: Base

A meaning of a kanji in a given language

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`literal`	`str`	Foreign key to `Kanji`
`lang`	`str`	ISO 639 language code of the meaning, defaults to `en`
`value`	`str`	The meaning text
`position`	`int`	Zero based order of the meaning for its language
`kanji`	`Kanji`	The owning kanji

`KanjiNanori` ¶

Bases: Base

A nanori (name only) reading of a kanji

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`literal`	`str`	Foreign key to `Kanji`
`value`	`str`	The nanori reading text
`position`	`int`	Zero based order of the nanori for the kanji
`kanji`	`Kanji`	The owning kanji

`KanjiQueryCode` ¶

Bases: Base

A lookup query code for a kanji

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`literal`	`str`	Foreign key to `Kanji`
`type`	`str`	Code type such as `skip`, `four_corner` or `deroo`
`value`	`str`	The code value
`skip_misclass`	`str \| None`	SKIP misclassification kind when present
`kanji`	`Kanji`	The owning kanji

`KanjiRadical` ¶

Bases: Base

A kanji to radical decomposition edge, taken from KRADFILE

Each row records that a kanji contains a given radical component. The pair of kanji and radical is unique

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`literal`	`str`	The kanji that contains the radical
`radical`	`str`	The radical component contained in the kanji

`KanjiReading` ¶

Bases: Base

A reading of a kanji

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`literal`	`str`	Foreign key to `Kanji`
`type`	`str`	Reading type such as `ja_on`, `ja_kun`, `pinyin` or `korean_r`
`value`	`str`	The reading text
`position`	`int`	Zero based order of the reading for its type
`kanji`	`Kanji`	The owning kanji

`KanjiStrokes` ¶

Bases: Base

KanjiVG stroke order data for a kanji

Licensing

Provenance and licensing for KanjiVG are recorded once in db_meta rather than on every row to keep the table small

Attributes:

Name	Type	Description
`literal`	`str`	Primary key and foreign key to `Kanji`
`stroke_count`	`int \| None`	Number of strokes in the diagram
`svg`	`str`	Serialized KanjiVG `<kanji>` group markup with the stroke paths, from which a consumer can render stroke order
`kanji`	`Kanji`	The owning kanji

`KanjiVariant` ¶

Bases: Base

A variant form reference for a kanji

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`literal`	`str`	Foreign key to `Kanji`
`type`	`str`	Encoding of the variant value such as `jis208`
`value`	`str`	The variant reference value
`kanji`	`Kanji`	The owning kanji

`Radical` ¶

Bases: Base

A search radical and its stroke count, taken from RADKFILE

Attributes:

Name	Type	Description
`radical`	`str`	Primary key, the radical character
`stroke_count`	`int \| None`	Number of strokes in the radical

`Sentence` ¶

Bases: Base

A Tatoeba sentence in a single language

A row is either a Japanese sentence or an English translation of one. The lang column distinguishes them

Attributes:

Name	Type	Description
`id`	`int`	Primary key, the Tatoeba sentence identifier
`lang`	`str`	ISO 639 language code of the sentence
`text`	`str`	The sentence text

`SentenceLink` ¶

Bases: Base

A translation link from a Japanese sentence to another sentence

Attributes:

Name	Type	Description
`id`	`int`	Primary key row identifier
`source_id`	`int`	Foreign key to the Japanese `Sentence`
`target_id`	`int`	Foreign key to the translated `Sentence`

`Tag` ¶

Bases: Base

An entity tag code and its human readable description

Populated from the JMdict and JMnedict <!ENTITY> definitions so that codes such as sl (slang) or ksb (Kansai dialect) can be expanded to text

Attributes:

Name	Type	Description
`code`	`str`	Primary key part, the tag code as it appears in the source
`category`	`str`	Primary key part, the tag family such as `pos`, `misc`, `field`, `dialect` or `name_type`
`description`	`str`	Human readable description of the tag

Models

models ¶

SCHEMA_VERSION = 1 module-attribute ¶

Audio ¶

Base ¶

DbMeta ¶

Furigana ¶

JMDictEntry ¶

JMDictGloss ¶

JMDictKana ¶

JMDictKanji ¶

JMDictSense ¶

JMnedictEntry ¶

JMnedictGloss ¶

JMnedictKana ¶

JMnedictKanji ¶

JMnedictTranslation ¶

JlptGrammar ¶

JlptKanji ¶

JlptVocab ¶

Kanji ¶

KanjiCodepoint ¶

KanjiDicRef ¶

KanjiMeaning ¶

KanjiNanori ¶

KanjiQueryCode ¶

KanjiRadical ¶

KanjiReading ¶

KanjiStrokes ¶

KanjiVariant ¶

Radical ¶

Sentence ¶

SentenceLink ¶

Tag ¶

`models` ¶

`SCHEMA_VERSION = 1` `module-attribute` ¶

`Audio` ¶

`Base` ¶

`DbMeta` ¶

`Furigana` ¶

`JMDictEntry` ¶

`JMDictGloss` ¶

`JMDictKana` ¶

`JMDictKanji` ¶

`JMDictSense` ¶

`JMnedictEntry` ¶

`JMnedictGloss` ¶

`JMnedictKana` ¶

`JMnedictKanji` ¶

`JMnedictTranslation` ¶

`JlptGrammar` ¶

`JlptKanji` ¶

`JlptVocab` ¶

`Kanji` ¶

`KanjiCodepoint` ¶

`KanjiDicRef` ¶

`KanjiMeaning` ¶

`KanjiNanori` ¶

`KanjiQueryCode` ¶

`KanjiRadical` ¶

`KanjiReading` ¶

`KanjiStrokes` ¶

`KanjiVariant` ¶

`Radical` ¶

`Sentence` ¶

`SentenceLink` ¶

`Tag` ¶