diff --git a/plwn/enums.py b/plwn/enums.py
index 246fc3c8290958889f6d55fa5bf0d1098045bc2f..6dfcc70530dfb08d64189a99370872dacae0002e 100644
--- a/plwn/enums.py
+++ b/plwn/enums.py
@@ -104,10 +104,10 @@ class PoS(Enum):
__order__ = 'verb noun adverb adjective ' \
'verb_en noun_en adverb_en adjective_en'
- verb = u'verb'
- noun = u'noun'
- adverb = u'adverb'
- adjective = u'adjective'
+ verb = u'czasownik'
+ noun = u'rzeczownik'
+ adverb = u'przysłówek'
+ adjective = u'przymiotnik'
# English (PWN) PoSes
verb_en = u'verb_en'
@@ -346,8 +346,8 @@ class Domain(Enum):
sys = u'systematyka, klasyfikacja'
- adj = u'PWN: all adjective clusters'
- adv = u'PWN: all adverbs'
+ adj = u'all adjective clusters'
+ adv = u'all adverbs'
mat = u'przymiotniki materiałowe'
diff --git a/plwn/storages/sqlite.py b/plwn/storages/sqlite.py
index 4d918a8444c918b70dd14436b73d4acca0b615af..54f5f329dc582cb504c646ab1cff36075dfc8aae 100644
--- a/plwn/storages/sqlite.py
+++ b/plwn/storages/sqlite.py
@@ -63,195 +63,195 @@ _DB_SCHEMA_SCRIPT = u"""
PRAGMA foreign_keys = ON;
-- Metadata table. Used for version number, currently
-CREATE TABLE "tbl_plwn_meta" (
- "name" VARCHAR(255) NOT NULL ,
- "value" BLOB NULL
+CREATE TABLE tbl_plwn_meta (
+ name VARCHAR(255) NOT NULL ,
+ value BLOB NULL
);
-- Tables for constant values
-CREATE TABLE "tbl_pos" (
- "id" INTEGER NOT NULL ,
- "value" VARCHAR(255) NOT NULL ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_pos (
+ id INTEGER NOT NULL ,
+ value VARCHAR(255) NOT NULL ,
+ PRIMARY KEY (id)
);
-CREATE TABLE "tbl_verbaspect" (
- "id" INTEGER NOT NULL ,
- "value" VARCHAR(255) NOT NULL ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_verbaspect (
+ id INTEGER NOT NULL ,
+ value VARCHAR(255) NOT NULL ,
+ PRIMARY KEY (id)
);
-CREATE TABLE "tbl_emotionmark" (
- "id" INTEGER NOT NULL ,
- "value" VARCHAR(255) NOT NULL ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_emotionmark (
+ id INTEGER NOT NULL ,
+ value VARCHAR(255) NOT NULL ,
+ PRIMARY KEY (id)
);
-CREATE TABLE "tbl_emotionname" (
- "id" INTEGER NOT NULL ,
- "value" VARCHAR(255) NOT NULL ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_emotionname (
+ id INTEGER NOT NULL ,
+ value VARCHAR(255) NOT NULL ,
+ PRIMARY KEY (id)
);
-CREATE TABLE "tbl_emotionvaluation" (
- "id" INTEGER NOT NULL ,
- "value" VARCHAR(255) NOT NULL ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_emotionvaluation (
+ id INTEGER NOT NULL ,
+ value VARCHAR(255) NOT NULL ,
+ PRIMARY KEY (id)
);
-CREATE TABLE "tbl_domain" (
- "id" INTEGER NOT NULL ,
- "value" VARCHAR(255) NOT NULL ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_domain (
+ id INTEGER NOT NULL ,
+ value VARCHAR(255) NOT NULL ,
+ PRIMARY KEY (id)
);
-CREATE UNIQUE INDEX "value" ON "tbl_domain" ("value");
+CREATE UNIQUE INDEX value ON tbl_domain (value);
-- Synset only gets one simple table
-CREATE TABLE "tbl_synset" (
- "id" BLOB NOT NULL ,
- "legacy_id" INTEGER NULL ,
- "definition" TEXT NULL ,
- "isartificial" INTEGER NOT NULL DEFAULT '0' ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_synset (
+ id BLOB NOT NULL ,
+ legacy_id INTEGER NULL ,
+ definition TEXT NULL ,
+ isartificial INTEGER NOT NULL DEFAULT '0' ,
+ PRIMARY KEY (id)
);
-- Lexical units have several tables, since they have several list-like
-- properties. They also need indexes for lookup.
-CREATE TABLE "tbl_lexicalunit" (
- "id" BLOB NOT NULL ,
- "legacy_id" INTEGER NULL ,
- "lemma" VARCHAR(255) NOT NULL ,
- "pos" INTEGER NOT NULL ,
- "variant" INTEGER NOT NULL ,
- "synset" BLOB NOT NULL ,
- "unitindex" INTEGER NOT NULL ,
- "definition" TEXT NULL ,
- "domain" INTEGER NOT NULL ,
- "verbaspect" INTEGER NULL ,
- "isemotional" INTEGER NULL ,
- "emotionmark" INTEGER NULL ,
- "emotionexample1" TEXT NULL ,
- "emotionexample2" TEXT NULL ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_lexicalunit (
+ id BLOB NOT NULL ,
+ legacy_id INTEGER NULL ,
+ lemma VARCHAR(255) NOT NULL ,
+ pos INTEGER NOT NULL ,
+ variant INTEGER NOT NULL ,
+ synset BLOB NOT NULL ,
+ unitindex INTEGER NOT NULL ,
+ definition TEXT NULL ,
+ domain INTEGER NOT NULL ,
+ verbaspect INTEGER NULL ,
+ isemotional INTEGER NULL ,
+ emotionmark INTEGER NULL ,
+ emotionexample1 TEXT NULL ,
+ emotionexample2 TEXT NULL ,
+ PRIMARY KEY (id)
);
-CREATE UNIQUE INDEX "lemma" ON "tbl_lexicalunit" ("lemma", "pos", "variant");
-CREATE INDEX "lex_i_lem_var" ON "tbl_lexicalunit" ("lemma", "variant");
-CREATE INDEX "lex_i_pos" ON "tbl_lexicalunit" ("pos");
-CREATE UNIQUE INDEX "synset" ON "tbl_lexicalunit" ("synset", "unitindex");
+CREATE UNIQUE INDEX lemma ON tbl_lexicalunit (lemma, pos, variant);
+CREATE INDEX lex_i_lem_var ON tbl_lexicalunit (lemma, variant);
+CREATE INDEX lex_i_pos ON tbl_lexicalunit (pos);
+CREATE UNIQUE INDEX synset ON tbl_lexicalunit (synset, unitindex);
-- Tables dependant on lexicalunit
-CREATE TABLE "tbl_senseexample" (
- "unitid" BLOB NOT NULL ,
- "example" TEXT NOT NULL ,
- "source" TEXT NOT NULL
+CREATE TABLE tbl_senseexample (
+ unitid BLOB NOT NULL ,
+ example TEXT NOT NULL ,
+ source TEXT NOT NULL
);
-CREATE INDEX "sen_i" ON "tbl_senseexample" ("unitid");
+CREATE INDEX sen_i ON tbl_senseexample (unitid);
-CREATE TABLE "tbl_externallink" (
- "unitid" BLOB NOT NULL ,
- "link" TEXT NOT NULL
+CREATE TABLE tbl_externallink (
+ unitid BLOB NOT NULL ,
+ link TEXT NOT NULL
);
-CREATE INDEX "link_i" ON "tbl_externallink" ("unitid");
+CREATE INDEX link_i ON tbl_externallink (unitid);
-CREATE TABLE "tbl_usagenote" (
- "unitid" BLOB NOT NULL ,
- "note" TEXT NOT NULL
+CREATE TABLE tbl_usagenote (
+ unitid BLOB NOT NULL ,
+ note TEXT NOT NULL
);
-CREATE INDEX "note_i" ON "tbl_usagenote" ("unitid");
+CREATE INDEX note_i ON tbl_usagenote (unitid);
-CREATE TABLE "tbl_unitemotionname" (
- "unitid" BLOB NOT NULL ,
- "nameid" INTEGER NOT NULL ,
- PRIMARY KEY ("unitid", "nameid")
+CREATE TABLE tbl_unitemotionname (
+ unitid BLOB NOT NULL ,
+ nameid INTEGER NOT NULL ,
+ PRIMARY KEY (unitid, nameid)
);
-CREATE TABLE "tbl_unitemotionvaluation" (
- "unitid" BLOB NOT NULL ,
- "valuationid" INTEGER NOT NULL ,
- PRIMARY KEY ("unitid", "valuationid")
+CREATE TABLE tbl_unitemotionvaluation (
+ unitid BLOB NOT NULL ,
+ valuationid INTEGER NOT NULL ,
+ PRIMARY KEY (unitid, valuationid)
);
-- Relation tables --
-- The for below are used to gather combinations of parent / child relation
-- names.
-CREATE TABLE "tbl_synsetrelationparentpart" (
- "id" INTEGER NOT NULL ,
- "name" VARCHAR(255) NOT NULL ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_synsetrelationparentpart (
+ id INTEGER NOT NULL ,
+ name VARCHAR(255) NOT NULL ,
+ PRIMARY KEY (id)
);
-CREATE TABLE "tbl_synsetrelationchildpart" (
- "id" INTEGER NOT NULL ,
- "name" VARCHAR(255) NOT NULL ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_synsetrelationchildpart (
+ id INTEGER NOT NULL ,
+ name VARCHAR(255) NOT NULL ,
+ PRIMARY KEY (id)
);
-CREATE TABLE "tbl_lexicalrelationparentpart" (
- "id" INTEGER NOT NULL ,
- "name" VARCHAR(255) NOT NULL ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_lexicalrelationparentpart (
+ id INTEGER NOT NULL ,
+ name VARCHAR(255) NOT NULL ,
+ PRIMARY KEY (id)
);
-CREATE TABLE "tbl_lexicalrelationchildpart" (
- "id" INTEGER NOT NULL ,
- "name" VARCHAR(255) NOT NULL ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_lexicalrelationchildpart (
+ id INTEGER NOT NULL ,
+ name VARCHAR(255) NOT NULL ,
+ PRIMARY KEY (id)
);
-CREATE UNIQUE INDEX "name" ON "tbl_lexicalrelationchildpart" ("name");
-CREATE UNIQUE INDEX "parentpart" ON "tbl_lexicalrelationtype" (
- "parentpart",
- "childpart"
+CREATE UNIQUE INDEX name ON tbl_lexicalrelationchildpart (name);
+CREATE UNIQUE INDEX parentpart ON tbl_lexicalrelationtype (
+ parentpart,
+ childpart
);
-- Next, gather these parts into relation types themselves.
-- Parent can't be NULL - the no-parent case will be handled by a special empty
-- string parent. This is so that UNIQUE works correctly.
-CREATE TABLE "tbl_synsetrelationtype" (
- "id" BLOB NOT NULL ,
- "legacy_id" INTEGER NULL ,
- "parentpart" INTEGER NOT NULL ,
- "childpart" INTEGER NOT NULL ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_synsetrelationtype (
+ id BLOB NOT NULL ,
+ legacy_id INTEGER NULL ,
+ parentpart INTEGER NOT NULL ,
+ childpart INTEGER NOT NULL ,
+ PRIMARY KEY (id)
);
-CREATE TABLE "tbl_lexicalrelationtype" (
- "id" BLOB NOT NULL ,
- "legacy_id" INTEGER NULL ,
- "parentpart" INTEGER NOT NULL ,
- "childpart" INTEGER NOT NULL ,
- PRIMARY KEY ("id")
+CREATE TABLE tbl_lexicalrelationtype (
+ id BLOB NOT NULL ,
+ legacy_id INTEGER NULL ,
+ parentpart INTEGER NOT NULL ,
+ childpart INTEGER NOT NULL ,
+ PRIMARY KEY (id)
);
-- The below tables are simply maps of relation aliases to their main IDs.
-- Reverse indexes are needed, too.
-CREATE TABLE "tbl_synsetrelationalias" (
- "name" VARCHAR(255) NOT NULL ,
- "relationid" BLOB NOT NULL ,
- PRIMARY KEY ("name")
+CREATE TABLE tbl_synsetrelationalias (
+ name VARCHAR(255) NOT NULL ,
+ relationid BLOB NOT NULL ,
+ PRIMARY KEY (name)
);
-CREATE INDEX "synsetrelationalias_irev" ON "tbl_synsetrelationalias" (
- "relationid"
+CREATE INDEX synsetrelationalias_irev ON tbl_synsetrelationalias (
+ relationid
);
-CREATE TABLE "tbl_lexicalrelationalias" (
- "name" VARCHAR(255) NOT NULL ,
- "relationid" BLOB NOT NULL ,
- PRIMARY KEY ("name")
+CREATE TABLE tbl_lexicalrelationalias (
+ name VARCHAR(255) NOT NULL ,
+ relationid BLOB NOT NULL ,
+ PRIMARY KEY (name)
);
-CREATE INDEX "lexicalrelationalias_irev" ON "tbl_lexicalrelationalias" (
- "relationid"
+CREATE INDEX lexicalrelationalias_irev ON tbl_lexicalrelationalias (
+ relationid
);
-- Next are finally the relation instances
-CREATE TABLE "tbl_synsetrelation" (
- "source" BLOB NOT NULL ,
- "relationtype" BLOB NOT NULL ,
- "target" BLOB NOT NULL ,
- PRIMARY KEY ("source", "relationtype", "target")
+CREATE TABLE tbl_synsetrelation (
+ source BLOB NOT NULL ,
+ relationtype BLOB NOT NULL ,
+ target BLOB NOT NULL ,
+ PRIMARY KEY (source, relationtype, target)
);
-CREATE TABLE "tbl_lexicalrelation" (
- "source" BLOB NOT NULL ,
- "relationtype" BLOB NOT NULL ,
- "target" BLOB NOT NULL ,
- PRIMARY KEY ("source", "relationtype", "target")
+CREATE TABLE tbl_lexicalrelation (
+ source BLOB NOT NULL ,
+ relationtype BLOB NOT NULL ,
+ target BLOB NOT NULL ,
+ PRIMARY KEY (source, relationtype, target)
);
-- Insert the special empty values for the parent part tables
@@ -284,7 +284,7 @@ _RELINST_TABLES = {
class PLWordNet(bs.PLWordNetBase):
_STORAGE_NAME = 'sqlite3'
- _SCHEMA_VERSION = 4
+ _SCHEMA_VERSION = '4'
@classmethod
def from_reader(cls, reader, dump_to=None):
@@ -366,8 +366,8 @@ class PLWordNet(bs.PLWordNetBase):
with closing(self._db.cursor()) as cur:
cur.execute(
u"""
- SELECT lemma, pos.value, variant, synset
- FROM tbl_lexicalunit
+ SELECT lemma, tbl_pos.value, variant, synset
+ FROM tbl_lexicalunit
JOIN tbl_pos ON tbl_lexicalunit.pos = tbl_pos.id
WHERE tbl_lexicalunit.id = ?
""",
@@ -546,7 +546,8 @@ class PLWordNet(bs.PLWordNetBase):
with closing(self._db.cursor()) as cur:
cur.execute(
u"""
- SELECT tbl_lexicalunit.id, lemma, pos.value, variant, synset
+ SELECT tbl_lexicalunit.id, lemma, tbl_pos.value,
+ variant, synset
FROM tbl_lexicalunit
JOIN tbl_pos ON tbl_lexicalunit.pos = tbl_pos.id
WHERE COALESCE(lemma = :lem, :defval)
@@ -599,7 +600,10 @@ class PLWordNet(bs.PLWordNetBase):
((en.value,) for en in en.EmotionName),
).close()
self._db.executemany(
- u"INSERT OR IGNORE INTO tbl_emotionvaluation (value) VALUES (?)",
+ u"""
+ INSERT OR IGNORE INTO tbl_emotionvaluation (value)
+ VALUES (?)
+ """,
((ev.value,) for ev in en.EmotionValuation),
).close()
self._db.executemany(
@@ -678,6 +682,7 @@ class LexicalUnit(bs.LexicalUnitBase):
self._var = variant
self._synid = synid
# Rest is unitialized
+ self._leg_id = _UNFETCHED
self._syn = _UNFETCHED
self._def = _UNFETCHED
self._usn = _UNFETCHED
@@ -717,6 +722,17 @@ class LexicalUnit(bs.LexicalUnitBase):
def is_english(self):
return self._pos.is_english
+ @property
+ def legacy_id(self):
+ if self._leg_id is _UNFETCHED:
+ with closing(self._db.cursor()) as cur:
+ cur.execute(
+ u"SELECT legacy_id FROM tbl_lexicalunit WHERE id = ?",
+ (self._id,),
+ )
+ self._leg_id = cur.fetchone()[0]
+ return self._leg_id
+
@property
def synset(self):
if self._syn is _UNFETCHED or self._syn() is None:
@@ -934,7 +950,8 @@ class LexicalUnit(bs.LexicalUnitBase):
with closing(self._db.cursor()) as cur:
cur.execute(
u"""
- SELECT tbl_lexicalunit.id, lemma, tbl_pos.value, variant, synset
+ SELECT tbl_lexicalunit.id,
+ lemma, tbl_pos.value, variant, synset
FROM tbl_lexicalrelation
JOIN tbl_lexicalunit ON tbl_lexicalunit.id = target
JOIN tbl_pos ON tbl_lexicalunit.pos = tbl_pos.id
@@ -997,6 +1014,7 @@ class Synset(bs.SynsetBase):
self._units = _UNFETCHED
self._def = _UNFETCHED
+ self._leg_id = _UNFETCHED
self._pos = _UNFETCHED
self._is_polish = _UNFETCHED
@@ -1033,7 +1051,8 @@ class Synset(bs.SynsetBase):
cur.execute(
u"""
SELECT tbl_lexicalunit.id, lemma, tbl_pos.value, variant
- FROM tbl_lexicalunit JOIN tbl_pos ON tbl_lexicalunit.pos = tbl_pos.id
+ FROM tbl_lexicalunit
+ JOIN tbl_pos ON tbl_lexicalunit.pos = tbl_pos.id
WHERE synset = ?
ORDER BY unitindex
""",
@@ -1054,6 +1073,17 @@ class Synset(bs.SynsetBase):
assert self._units
return self._units
+ @property
+ def legacy_id(self):
+ if self._leg_id is _UNFETCHED:
+ with closing(self._db.cursor()) as cur:
+ cur.execute(
+ u"SELECT legacy_id FROM tbl_synset WHERE id = ?",
+ (self._id,),
+ )
+ self._leg_id = cur.fetchone()[0]
+ return self._leg_id
+
@property
def definition(self):
if self._def is _UNFETCHED:
@@ -1313,18 +1343,18 @@ class _DBBuilder(object):
# Synset to lexical units relations also need to be deferred.
self._synid2lexids = coll.defaultdict(list)
# Cache IDs of constant values
- with closing(db.execute(u"SELECT value, id FROM tbl_pos")) as cur:
+ with closing(db.execute(u"SELECT value, id FROM pos")) as cur:
self._posids = dict(cur)
- with closing(db.execute(u"SELECT value, id FROM tbl_verbaspect")) as cur:
+ with closing(db.execute(u"SELECT value, id FROM verbaspect")) as cur:
self._vaids = dict(cur)
- with closing(db.execute(u"SELECT value, id FROM tbl_emotionmark")) as cur:
+ with closing(db.execute(u"SELECT value, id FROM emotionmark")) as cur:
self._emids = dict(cur)
- with closing(db.execute(u"SELECT value, id FROM tbl_emotionname")) as cur:
+ with closing(db.execute(u"SELECT value, id FROM emotionname")) as cur:
self._enids = dict(cur)
- with closing(db.execute(u"SELECT value, id FROM tbl_emotionvaluation")) \
+ with closing(db.execute(u"SELECT value, id FROM emotionvaluation")) \
as cur:
self._evids = dict(cur)
- with closing(db.execute(u"SELECT value, id FROM tbl_domain")) as cur:
+ with closing(db.execute(u"SELECT value, id FROM domain")) as cur:
self._dmids = dict(cur)
def __call__(self, reader):
@@ -1351,7 +1381,7 @@ class _DBBuilder(object):
def _insert_synset(self, syn_node):
self._db.execute(
u"""
- INSERT INTO tbl_synset (id, definition, isartificial)
+ INSERT INTO synset (id, definition, isartificial)
VALUES (?, ?, ?)
""",
(syn_node.id, syn_node.definition, syn_node.is_artificial),
@@ -1402,7 +1432,7 @@ class _DBBuilder(object):
try:
cur.execute(
u"""
- INSERT INTO tbl_lexicalunit (
+ INSERT INTO lexicalunit (
id, lemma, pos, variant,
synset, unitindex,
definition, domain, verbaspect,
@@ -1453,7 +1483,7 @@ class _DBBuilder(object):
cur.executemany(
u"""
- INSERT INTO tbl_senseexample (unitid, example, source)
+ INSERT INTO senseexample (unitid, example, source)
VALUES (?, ?, ?)
""",
(
@@ -1464,14 +1494,14 @@ class _DBBuilder(object):
)
cur.executemany(
u"""
- INSERT INTO tbl_usagenote (unitid, note)
+ INSERT INTO usagenote (unitid, note)
VALUES (?, ?)
""",
((lu_node.id, note) for note in lu_node.usage_notes),
)
cur.executemany(
u"""
- INSERT INTO tbl_externallink (unitid, link)
+ INSERT INTO externallink (unitid, link)
VALUES (?, ?)
""",
((lu_node.id, link)
@@ -1479,7 +1509,7 @@ class _DBBuilder(object):
)
cur.executemany(
u"""
- INSERT INTO tbl_unitemotionname (unitid, nameid)
+ INSERT INTO unitemotionname (unitid, nameid)
VALUES (?, ?)
""",
(
@@ -1489,7 +1519,7 @@ class _DBBuilder(object):
)
cur.executemany(
u"""
- INSERT INTO tbl_unitemotionvaluation (unitid, valuationid)
+ INSERT INTO unitemotionvaluation (unitid, valuationid)
VALUES (?, ?)
""",
(
@@ -1626,10 +1656,10 @@ class _DBBuilder(object):
with closing(self._db.cursor()) as cur:
cur.execute(
u"""
- SELECT tbl_synset.id
- FROM tbl_synset
- LEFT JOIN tbl_lexicalunit ON tbl_synset.id = tbl_lexicalunit.synset
- WHERE tbl_lexicalunit.synset IS NULL
+ SELECT synset.id
+ FROM synset
+ LEFT JOIN lexicalunit ON synset.id = lexicalunit.synset
+ WHERE lexicalunit.synset IS NULL
""",
)
empties = tuple(row[0] for row in cur)
@@ -1642,7 +1672,7 @@ class _DBBuilder(object):
_LOG.warning('Synset %d is empty', synid)
self._db.execute(
- u"DELETE FROM tbl_synset WHERE id IN ({})".format(
+ u"DELETE FROM synset WHERE id IN ({})".format(
u','.join(u'?' * len(empties))
),
empties,