Modul:Wikidata utilities: Unterschied zwischen den Versionen

Aus skandinavien-wiki.net
(+getSitelink)
K (89 Versionen von wikivoyage:Modul:Wikidata_utilities importiert)
 
(46 dazwischenliegende Versionen von 2 Benutzern werden nicht angezeigt)
Zeile 1: Zeile 1:
local fw = {}
-- Wikidata convenience utilities


function fw.getEntity( id )
-- documentation
local WikidataUtilities = {
suite  = 'WikidataUtilities',
serial = '2022-09-04',
item  = 65439025
}
 
-- i18n
local wd = {
version  = 'P348',
retrieved = 'P813',
Gregorian = 'Q12138', -- calendar models
prolepticGregorian = 'Q1985727'
}
 
-- module variable and administration
local wu = {
moduleInterface = WikidataUtilities
}
 
-- table storing property ids used
local catTable = {
P0 = ''
}
 
local function isSet( arg )
return arg and arg ~= ''
end
 
function wu.getEntity( id )
local wrongQualifier = false
local wrongQualifier = false
local entity = nil
local entity = nil
if not id or id == '' then
if not isSet( id ) then
return '', entity, wrongQualifier
return '', entity, wrongQualifier
end
end
Zeile 21: Zeile 50:
end
end


function fw.getLabel( entity, lang )
function wu.getEntityId( id )
local isString = type( entity ) == 'string'
local wrongQualifier = false
if not entity or ( isString and entity == '' ) then
local entity = nil
if not isSet( id ) then
id = ''
elseif mw.wikibase.isValidEntityId( id ) and mw.wikibase.entityExists( id ) then
-- expensive function call
-- redirect ids marked false, too
entity = id
else
id = ''
wrongQualifier = true
end
 
return id, entity, wrongQualifier
end
 
function wu.getLabel( entity, lang, noFallback )
if not isSet( entity ) then
return nil
return nil
end
end
if isString then -- entity is id
local tp = type( entity )
if lang and lang ~= '' then
if tp == 'string' and mw.wikibase.isValidEntityId( entity ) then
return mw.wikibase.getLabelByLang( entity, lang )
return isSet( lang ) and mw.wikibase.getLabelByLang( entity, lang )
else
or ( not noFallback and mw.wikibase.getLabel( entity ) )
return mw.wikibase.getLabel( entity )
elseif tp == 'table' and entity.labels then
end
return isSet( lang ) and entity:getLabel( lang )
else -- entity is table
or ( not noFallback and entity:getLabel() )
if lang and lang ~= '' then
else
return entity:getLabel( lang )
return nil
else
return entity:getLabel()
end
end
end
end
end


function fw.getSitelink( entity, globalSiteId )
function wu.getSitelink( entity, globalSiteId )
local isString = type( entity ) == 'string'
if not isSet( entity ) then
if not entity or ( isString and entity == '' ) then
return nil
return nil
end
end
if isString then -- entity is id
local tp = type( entity )
if globalSiteId and globalSiteId ~= '' then
if tp == 'string' and mw.wikibase.isValidEntityId( entity ) then
return mw.wikibase.getSitelink( entity, globalSiteId )
return mw.wikibase.getSitelink( entity, globalSiteId )
else
elseif tp == 'table' and entity.labels then
return mw.wikibase.getSitelink( entity )
return entity:getSitelink( globalSiteId )
end
else
else -- entity is table
return nil
if globalSiteId and globalSiteId ~= '' then
return entity:getSitelink( globalSiteId )
else
return entity:getSitelink( )
end
end
end
end
end


local function getFirstValue( statements )
function wu.getSitelinkTable( entity, globalSiteId )
if #statements == 0 then
if not isSet( entity ) or not isSet( globalSiteId ) then
return nil
elseif type( entity ) == 'string' then -- entity is id
entity = mw.wikibase.getEntity( entity )
end
if entity and entity.sitelinks then
return entity.sitelinks[ globalSiteId ]
else
return nil
return nil
end
end
end


for i = 1, #statements, 1 do
-- convert from url to Q id
if statements[i].mainsnak.snaktype == 'value' then
function wu.getUnit( unit )
return statements[i].mainsnak.datavalue.value
if isSet( unit ) and type( unit ) == 'string' then
end
return unit:gsub( 'https?://www.wikidata.org/entity/', '' )
else
return ''
end
end
end


return nil
-- is calendar model a Gregorian calendar?
function wu.isGregorian( t )
return type( t ) == 'string' or t.calendarmodel == wd.Gregorian
or t.calendarmodel == wd.prolepticGregorian
end
end


local function getNValues( statements, count )
-- extract date from time
local ar = {}
function wu.getDateFromTime( t )
if count > #statements then count = #statements end
local model = wd.prolepticGregorian
if ( #statements == 0 ) or ( count <= 0 ) then
if type( t ) == 'table' then
return ar
model = t.calendarmodel
t = t.time
end
end
 
t = t:gsub( '^+', '' ):gsub( 'T.+Z$', '' ):gsub( '-00$', '' )
local i = 0
:gsub( '-00$', '' )
repeat
return t, model
i = i + 1
if statements[i].mainsnak.snaktype == 'value' then
table.insert( ar, statements[i].mainsnak.datavalue.value )
end
until ( i >= #statements ) or ( #ar >= count )
 
return ar
end
end


function fw.getBestStatements( entity, p )
function wu.getBestStatements( entity, p )
if type( entity ) == 'string' then
local tp = type( entity )
if tp == 'string' and mw.wikibase.isValidEntityId( entity ) then
return mw.wikibase.getBestStatements( entity, p )
return mw.wikibase.getBestStatements( entity, p )
elseif tp == 'table' and entity.labels then
return entity:getBestStatements( p )
else
else
return entity:getBestStatements( p )
return {}
end
end
end
end


function fw.getStatements( entity, p, count )
function wu.getStatements( entity, p, count )
local ar = {}
local ar = {}
if ( not entity ) or ( entity == '' ) then
if not ( isSet( entity ) and isSet( p ) ) then
return ar
return ar
end
end


local statements = fw.getBestStatements( entity, p )
local statements = wu.getBestStatements( entity, p )
 
count = math.min( count or #statements, #statements )
count = count or #statements
if count <= 0 then
if count > #statements then count = #statements end
if ( #statements == 0 ) or ( count <= 0 ) then
return ar
return ar
end
end
Zeile 118: Zeile 165:
repeat
repeat
i = i + 1
i = i + 1
if statements[i].mainsnak.snaktype == 'value' then
if statements[ i ].mainsnak.snaktype == 'value' then
table.insert( ar, statements[i] )
if statements[ i ].mainsnak.datatype == 'quantity' then
statements[ i ].mainsnak.datavalue.value.amount =
statements[ i ].mainsnak.datavalue.value.amount:gsub( '^+', '' )
statements[ i ].mainsnak.datavalue.value.unit = wu.getUnit(
statements[ i ].mainsnak.datavalue.value.unit )
end
table.insert( ar, statements[ i ] )
end
end
until ( i >= #statements ) or ( #ar >= count )
until i >= #statements or #ar >= count


return ar
return ar
end
end


function fw.getValue( entity, p, catArray )
function wu.getValue( entity, p )
local value = ''
local statements = wu.getStatements( entity, p, 1 )
if entity and entity ~= '' and p and p ~= '' then
if #statements > 0 then
value = getFirstValue( fw.getBestStatements( entity, p ) )
catTable[ p ] = ''
if value and catArray then
return statements[ 1 ].mainsnak.datavalue.value
catArray[ p ] = ''
end
value = value or ''
end
if catArray then
return value, catArray
else
else
return value
return ''
end
end
end
end


function fw.getId( entity, p, catArray )
function wu.getId( entity, p )
local value = ''
local value = ''
if entity and entity ~= '' and p and p ~= '' then
local statements = wu.getStatements( entity, p, 1 )
value = getFirstValue( fw.getBestStatements( entity, p ) )
if #statements > 0 then
if value then
value = statements[ 1 ].mainsnak.datavalue.value
if catArray then catArray[ p ] = '' end
value = value.id or ''
value = value.id
if value ~= '' then
else
catTable[ p ] = ''
value = ''
end
end
end
end
if catArray then
return value
return value, catArray
else
return value
end
end
end


function fw.getValues( entity, p, count, catArray )
function wu.getValues( entity, p, count )
local values = '', statements
local statements = wu.getStatements( entity, p, count )
if entity and entity ~= '' and p and p ~= '' then
if #statements > 0 then
statements = fw.getBestStatements( entity, p )
catTable[ p ] = ''
values = getNValues( statements, count or #statements )
for i = 1, #statements, 1 do
if catArray and #values > 0 then catArray[ p ] = '' end
statements[ i ] = statements[ i ].mainsnak.datavalue.value
end
end
if catArray then
return values, catArray
else
return values
end
end
return statements
end
end


function fw.getValuesByLang( entity, p, count, lang, catArray )
function wu.getValuesByLang( entity, p, count, lang )
local ar = '', statements, i, value
local ar = {}
if entity and entity ~= '' and p and p ~= '' then
local statements = wu.getStatements( entity, p )
statements = fw.getBestStatements( entity, p )
if #statements > 0 then
ar = {}
local value
count = count or #statements
for i = 1, #statements, 1 do
if #statements > 0 and count > 0 then
value = statements[ i ].mainsnak.datavalue.value
i = 0
if value.language and lang == value.language then
repeat
table.insert( ar, value.text )
i = i + 1
end
if statements[ i ].mainsnak.snaktype == 'value' then
if count and #ar >= count then
value = statements[ i ].mainsnak.datavalue.value
break
if value.language and lang == value.language then
end
table.insert( ar, statements[ i ].mainsnak.datavalue.value.text )
end
end
until ( i >= #statements ) or ( #ar >= count )
end
end
if catArray and #ar > 0 then catArray[ p ] = '' end
end
end
if catArray then
if #ar > 0 then
return ar, catArray
catTable[ p ] = ''
else
return ar
end
end
end
return ar
end


function fw.getValuesWithQualifierIds( entity, p, qualifierP, defaultId, catArray )
-- get values array for monolingual text
local result = {}, statements, value, id, i, j
function wu.getMonolingualValues( entity, p )
if entity and entity ~= '' and p and p ~= '' and qualifierP and qualifierP ~= '' then
local result = {}
statements = fw.getStatements( entity, p, nil )
local statements = wu.getStatements( entity, p, nil )
if #statements > 0 then
if #statements > 0 and statements[ 1 ].mainsnak.datatype == 'monolingualtext' then
-- defaultId is used if a qualifier is missing
local hyphen, lng, value
if not defaultId or defaultId == '' or type( defaultId ) ~= 'string' then
catTable[ p ] = ''
defaultId = 'unknown'
for i = 1, #statements, 1 do
value = statements[ i ].mainsnak.datavalue.value
lng = value.language
hyphen = lng:find( '-' )
if hyphen then
lng = lng:sub( 1, hyphen - 1 )
end
end
 
if not result[ lng ] then
if catArray then catArray[ p ] = '' end
result[ lng ] = value.text
for i = 1, #statements, 1 do
value = statements[ i ].mainsnak.datavalue.value
id = defaultId
if statements[ i ].qualifiers and statements[ i ].qualifiers[ qualifierP ]
and ( #statements[ i ].qualifiers[ qualifierP ] > 0 ) then
for j = 1, #statements[ i ].qualifiers[ qualifierP ], 1 do
if statements[ i ].qualifiers[ qualifierP ][ j ].snaktype == 'value' then
id = statements[ i ].qualifiers[ qualifierP ][ j ].datavalue.value.id
break
end
end
end
result[ id ] = value
end
end
end
end
end
end
if catArray then
return result
return result, catArray
end
else
 
function wu.getValuesByQualifier( entity, p, qualifierP, defaultId )
local result = {}
if not isSet( qualifierP ) then
return result
return result
elseif type( defaultId ) ~= 'string' or defaultId == '' then
defaultId = 'unknown'
end
end
end


-- get values array for monolingual text
local statements = wu.getStatements( entity, p, nil )
function fw.getValuesWithLanguages( entity, p, catArray )
if #statements > 0 then
local result = {}, statements, hyphen, i, lng, value
catTable[ p ] = ''
if entity and entity ~= '' and p and p ~= '' then
local id, statement, value
statements = fw.getStatements( entity, p, nil )
for i = 1, #statements do
if #statements > 0 and statements[ 1 ].mainsnak.datatype == 'monolingualtext' then
statement = statements[ i ]
if catArray then catArray[ p ] = '' end
-- defaultId is used if a qualifier is missing
for i = 1, #statements, 1 do
id = defaultId
value = statements[i].mainsnak.datavalue.value
value = statement.mainsnak.datavalue.value
lng = value.language
if statement.qualifiers and statement.qualifiers[ qualifierP ] then
hyphen = lng:find( '-' )
for _, qualifier in ipairs( statement.qualifiers[ qualifierP ] ) do
if hyphen then
if qualifier.snaktype == 'value' then
lng = lng:sub( 1, hyphen - 1 )
id = qualifier.datavalue.value.id
if id then
catTable[ qualifierP ] = ''
break
end
end
end
end
if not result[ lng ] then result[ lng ] = value.text end
end
end
result[ id ] = value
end
end
end
end
if catArray then
return result
return result, catArray
else
return result
end
end
end


local function getValueFromDatavalue( datavalue )
local function analyzeDatavalue( datavalue, labelFct, ... )
local v = datavalue.value
local v = datavalue.value
local t = datavalue.type
local t = datavalue.type
if type( v ) == 'table' then
if type( v ) == 'table' then
-- items which can be reduced to a string
-- items which can be reduced to a string
if t == 'wikibase-entityid' then v = v.id
if t == 'wikibase-entityid' then
elseif t == 'time' then v = v.time
v = v.id
if type( labelFct ) == 'function' then
v = labelFct( v, ... )
end
elseif t == 'quantity' then
v.amount = v.amount:gsub( '^+', '' )
if tonumber( v.amount ) == 0 then
v.amount = '0'
end
if v.unit == '1' then
v = tonumber( v.amount ) or 1
else
v.unit = wu.getUnit( v.unit )
end
elseif t == 'time' then
v.calendarmodel = wu.getUnit( v.calendarmodel )
if wu.isGregorian( v ) then
v = v.time
end
end
end
end
end
Zeile 273: Zeile 321:


-- The following function is an experimental one, not for extensive use
-- The following function is an experimental one, not for extensive use
function fw.getValuesWithQualifiers( entity, p, qualifiers, count )
-- for qualifiers, references
local result = {}
--  { item1, item2, ... } : using named qualifiers/references
local statements = fw.getStatements( entity, p, count )
--  {} : using no qualifiers/references
if #statements == 0 then return result end
--  nil : using all qualifiers/references
function wu.getValuesWithQualifiers( entity, p, values, qualifiers, references,
count, labelFct, ... )
local array, qual
local function toQualifierTable( tab, key, qualTab, labelFct, ... )
local v
if not tab[ key ] then
tab[ key ] = {}
end
for i = 1, #qualTab do
qual = qualTab[ i ]
if qual.snaktype == 'value' then
v, tab[ key .. '-type' ] =
analyzeDatavalue( qual.datavalue, labelFct, ... )
table.insert( tab[ key ], v )
end
end
if #tab[ key ] == 0 then
tab[ key ] = nil
tab[ key .. '-type' ] = nil
else
catTable[ key ] = ''
end
end
local function hasValue( tab, val )
for i = 1, #tab do
if tab[ i ] == val then
return true
end
end
return false
end
 
local results = {}
local statements = wu.getStatements( entity, p, count )
if #statements == 0 then
return results
end
local i, v
if type( values ) == 'table' and #values > 0 then
for i = #statements, 1, -1 do
v = statements[ i ].mainsnak.datavalue.value
if type( v ) ~= 'string' then
v = v.id
end
if not isSet( v ) or not hasValue( values, v ) then
table.remove( statements, i )
end
end
if #statements == 0 then
return results
end
end
catTable[ p ] = ''


if qualifiers and ( type( qualifiers ) == 'string' ) then
if type( qualifiers ) == 'string' then
qualifiers = { qualifiers }
qualifiers = { qualifiers }
end
if type( references ) == 'string' then
references = { references }
end
end


local array, key, value, i, j
local key, reference, statement
for i = 1, #statements, 1 do
for i = 1, #statements do
array = { value = statements[i].mainsnak.datavalue.value,
statement = statements[ i ]
['value-type'] = statements[i].mainsnak.datavalue.type }
array = { value = analyzeDatavalue( statement.mainsnak.datavalue, labelFct, ... ),
if statements[i].qualifiers then
[ 'value-type' ] = statement.mainsnak.datavalue.type }
 
if statement.qualifiers then
if not qualifiers then -- all qualifier properties
if not qualifiers then -- all qualifier properties
for key, value in pairs( statements[i].qualifiers ) do
for key, qualTab in pairs( statement.qualifiers ) do
if #value > 0 then
toQualifierTable( array, key, qualTab, labelFct, ... )
for j = 1, #value, 1 do
end
if value[ j ].snaktype == 'value' then
else -- table of selected qualifier properties
array[ key ], array[ key .. '-type' ] =
for j = 1, #qualifiers do
getValueFromDatavalue( value[ j ].datavalue )
key = qualifiers[ j ]
break
if statement.qualifiers[ key ] then
end
toQualifierTable( array, key, statement.qualifiers[ key ], labelFct, ... )
end
end
end
end
end
else -- table of selected qualifier properties
end
for key, value in pairs( qualifiers ) do
end
if statements[i].qualifiers[ value ] and
 
( #statements[i].qualifiers[ value ] > 0 ) then
array.references = {}
for j = 1, #statements[i].qualifiers[ value ], 1 do
if statement.references then
if statements[i].qualifiers[ value ][ j ].snaktype == 'value' then
for k = 1, #statement.references do
array[ value ], array[ value .. '-type' ] =
reference = statement.references[ k ]
getValueFromgetValueFromDatavalue( statements[i].qualifiers[ value ][ j ].datavalue )
if reference and reference.snaks then
break
table.insert( array.references, {} )
if not references then -- all references
for key, refTab in pairs( reference.snaks ) do
toQualifierTable( array.references[ #array.references ],
key, refTab )
end
else -- table of selected references
for j = 1, #references do
key = references[ j ]
if reference.snaks[ key ] then
toQualifierTable( array.references[ #array.references ],
key, reference.snaks[ key ] )
end
end
end
end
Zeile 314: Zeile 430:
end
end
end
end
table.insert( result, array )
 
table.insert( results, array )
end
end
return result
 
-- clustering statements with identical value
local helper = {}
local sort1 = 0
local mult = false
local result
for i = 1, #results do
result = results[ i ]
if helper[ result.value ] then
helper[ result.value ].sort2 = helper[ result.value ].sort2 + 1
mult = true
else
sort1 = sort1 + 1
helper[ result.value ] = { sort1 = sort1, sort2 = 1 }
end
result.sort1 = helper[ result.value ].sort1
result.sort2 = helper[ result.value ].sort2
end
if sort1 > 1 and mult and #results > 2 then
table.sort( results,
function( a, b )
return a.sort1 < b.sort1 or
( a.sort1 == b.sort1 and a.sort2 < b.sort2 )
end
)
end
 
return results
end
end


function fw.typeSearch( p31, list, limit, catArray )
-- get lastEdit from reference retrieve date
-- p31: array of Wikidata values
function wu.getLastedit( lastEdit, statements )
-- list: indexed array of q id - types relations
local isBoolean = type( lastEdit ) == 'boolean'
-- limit: maximum levels to analyse
if isBoolean and lastEdit == false then
if not list or not p31 or #p31 == 0 then
return lastEdit
return 'error', catArray
end
end
 
local le = ''
local function compareIds( ar )
for _, statement in ipairs( statements ) do
local i, t
if statement.references then
for i = 1, #ar, 1 do
for _, reference in ipairs( statement.references ) do
t = list[ ar[ i ].id ]
if reference[ wd.retrieved ] then
if t then
for _, retrieved in ipairs( reference[ wd.retrieved ] ) do
return t
retrieved = wu.getDateFromTime( retrieved )
if retrieved > le then
le = retrieved
end
end
end
end
end
end
end
return nil
end
end
local aType, i, id, ids, j
if isBoolean then
return ( le ~= '' ) and le or lastEdit
else
return ( le > lastEdit ) and le or lastEdit
end
end


aType = compareIds( p31 ) -- check p31 ids first, maybe step 2 is not nessary
function wu.getAliases( entity, lang )
if aType then
if type( entity ) == 'string' then -- is Q id
return aType, catArray
entity = mw.wikibase.getEntity( entity )
end
end
if not lang then
lang = mw.getContentLanguage():getCode()
end
local aliases = {}
if entity and entity.aliases and entity.aliases[ lang ] then
for _, alias in ipairs( entity.aliases[ lang ] ) do
table.insert( aliases, alias.value )
end
end
return aliases
end


-- now functions becomes expensive because of multiple fw.getValues calls
-- maintenance utilities
for i = 1, #p31, 1 do -- step 2: analyse P279 chains of first ids
function wu.addProperty( p )
id = p31[ i ].id -- start id
catTable[ p ] = ''
j = 0
end
repeat
ids, catArray = fw.getValues( id, 'P279', nil, catArray )
if #ids > 0 then
id = ids[ 1 ].id
aType = compareIds( ids )
if aType then
return aType, catArray
end
end
j = j + 1
until j >= limit or #ids == 0
end


return 'error', catArray
function wu.removeProperty( p )
catTable[ p ] = nil
end
end


function fw.getCategories( catArray, formatStr )
function wu.getCategories( formatStr )
result = ''
local result = ''
if not catArray then return result end
 
if not formatStr or formatStr == '' then
if not isSet( formatStr ) then
formatStr = '[[Category:%s]]'
formatStr = '[[Category:%s]]'
end
end
for key, value in pairs( catArray ) do
 
catTable.P0 = nil
for key, value in pairs( catTable ) do
result = result .. string.format( formatStr, key )
result = result .. string.format( formatStr, key )
end
end
Zeile 376: Zeile 530:
end
end


return fw
return wu

Aktuelle Version vom 27. Januar 2023, 08:40 Uhr

Die Dokumentation für dieses Modul kann unter Modul:Wikidata utilities/doc erstellt werden

-- Wikidata convenience utilities

-- documentation
local WikidataUtilities = {
	suite  = 'WikidataUtilities',
	serial = '2022-09-04',
	item   = 65439025
}

-- i18n
local wd = {
	version   = 'P348',
	retrieved = 'P813',
	Gregorian = 'Q12138', -- calendar models
	prolepticGregorian = 'Q1985727'
}

-- module variable and administration
local wu = {
	moduleInterface = WikidataUtilities
}

-- table storing property ids used
local catTable = {
	P0 = ''
}

local function isSet( arg )
	return arg and arg ~= ''
end

function wu.getEntity( id )
	local wrongQualifier = false
	local entity = nil
	
	if not isSet( id ) then
		return '', entity, wrongQualifier
	end
	if mw.wikibase.isValidEntityId( id ) then
		-- expensive function call
		-- redirect ids marked false, too
		entity = mw.wikibase.getEntity( id )
	end
	if not entity then
		id = ''
		wrongQualifier = true
	end

	return id, entity, wrongQualifier
end

function wu.getEntityId( id )
	local wrongQualifier = false
	local entity = nil
	
	if not isSet( id ) then
		id = ''
	elseif mw.wikibase.isValidEntityId( id ) and mw.wikibase.entityExists( id ) then
		-- expensive function call
		-- redirect ids marked false, too
		entity = id
	else
		id = ''
		wrongQualifier = true
	end

	return id, entity, wrongQualifier
end

function wu.getLabel( entity, lang, noFallback )
	if not isSet( entity ) then
		return nil
	end
	local tp = type( entity )
	if tp == 'string' and mw.wikibase.isValidEntityId( entity ) then
		return isSet( lang ) and mw.wikibase.getLabelByLang( entity, lang )
			or ( not noFallback and mw.wikibase.getLabel( entity ) )
	elseif tp == 'table' and entity.labels then
		return isSet( lang ) and entity:getLabel( lang )
			or ( not noFallback and entity:getLabel() )
	else
		return nil
	end
end

function wu.getSitelink( entity, globalSiteId )
	if not isSet( entity ) then
		return nil
	end
	local tp = type( entity )
	if tp == 'string' and mw.wikibase.isValidEntityId( entity ) then
		return mw.wikibase.getSitelink( entity, globalSiteId )
	elseif tp == 'table' and entity.labels then
		return entity:getSitelink( globalSiteId )
	else
		return nil
	end
end

function wu.getSitelinkTable( entity, globalSiteId )
	if not isSet( entity ) or not isSet( globalSiteId ) then
		return nil
	elseif type( entity ) == 'string' then -- entity is id
		entity = mw.wikibase.getEntity( entity )
	end
	if entity and entity.sitelinks then
		return entity.sitelinks[ globalSiteId ]
	else
		return nil
	end
end


-- convert from url to Q id
function wu.getUnit( unit )
	if isSet( unit ) and type( unit ) == 'string' then
		return unit:gsub( 'https?://www.wikidata.org/entity/', '' )
	else
		return ''
	end
end

-- is calendar model a Gregorian calendar?
function wu.isGregorian( t )
	return type( t ) == 'string' or t.calendarmodel == wd.Gregorian
		or t.calendarmodel == wd.prolepticGregorian
end

-- extract date from time
function wu.getDateFromTime( t )
	local model = wd.prolepticGregorian
	if type( t ) == 'table' then
		model = t.calendarmodel
		t = t.time
	end
	t = t:gsub( '^+', '' ):gsub( 'T.+Z$', '' ):gsub( '-00$', '' )
		:gsub( '-00$', '' )
	return t, model
end

function wu.getBestStatements( entity, p )
	local tp = type( entity )
	if tp == 'string' and mw.wikibase.isValidEntityId( entity ) then
		return mw.wikibase.getBestStatements( entity, p )
	elseif tp == 'table' and entity.labels then
		return entity:getBestStatements( p )
	else
		return {}
	end
end

function wu.getStatements( entity, p, count )
	local ar = {}
	if not ( isSet( entity ) and isSet( p ) ) then
		return ar
	end

	local statements = wu.getBestStatements( entity, p )
	count = math.min( count or #statements, #statements )
	if count <= 0 then
		return ar
	end

	local i = 0
	repeat
		i = i + 1
		if statements[ i ].mainsnak.snaktype == 'value' then
			if statements[ i ].mainsnak.datatype == 'quantity' then
				statements[ i ].mainsnak.datavalue.value.amount =
					statements[ i ].mainsnak.datavalue.value.amount:gsub( '^+', '' )
				statements[ i ].mainsnak.datavalue.value.unit = wu.getUnit(
					statements[ i ].mainsnak.datavalue.value.unit )
			end
			table.insert( ar, statements[ i ] )
		end
	until i >= #statements or #ar >= count

	return ar
end

function wu.getValue( entity, p )
	local statements = wu.getStatements( entity, p, 1 )
	if #statements > 0 then
		catTable[ p ] = ''
		return statements[ 1 ].mainsnak.datavalue.value
	else
		return ''
	end
end

function wu.getId( entity, p )
	local value = ''
	local statements = wu.getStatements( entity, p, 1 )
	if #statements > 0 then
		value = statements[ 1 ].mainsnak.datavalue.value
		value = value.id or ''
		if value ~= '' then
			catTable[ p ] = ''
		end
	end
	return value
end

function wu.getValues( entity, p, count )
	local statements = wu.getStatements( entity, p, count )
	if #statements > 0 then
		catTable[ p ] = ''
		for i = 1, #statements, 1 do
			statements[ i ] = statements[ i ].mainsnak.datavalue.value
		end
	end
	return statements
end

function wu.getValuesByLang( entity, p, count, lang )
	local ar = {}
	local statements = wu.getStatements( entity, p )
	if #statements > 0 then
		local value
		for i = 1, #statements, 1 do
			value = statements[ i ].mainsnak.datavalue.value
			if value.language and lang == value.language then
				table.insert( ar, value.text )
			end
			if count and #ar >= count then
				break
			end
		end
	end
	if #ar > 0 then
		catTable[ p ] = ''
	end
	return ar
end	

-- get values array for monolingual text
function wu.getMonolingualValues( entity, p )
	local result = {}
	local statements = wu.getStatements( entity, p, nil )
	if #statements > 0 and statements[ 1 ].mainsnak.datatype == 'monolingualtext' then
		local hyphen, lng, value
		catTable[ p ] = ''
		for i = 1, #statements, 1 do
			value = statements[ i ].mainsnak.datavalue.value
			lng = value.language
			hyphen = lng:find( '-' )
			if hyphen then
				lng = lng:sub( 1, hyphen - 1 )
			end
			if not result[ lng ] then
				result[ lng ] = value.text
			end
		end
	end
	return result
end

function wu.getValuesByQualifier( entity, p, qualifierP, defaultId )
	local result = {}
	if not isSet( qualifierP ) then
		return result
	elseif type( defaultId ) ~= 'string' or defaultId == '' then
		defaultId = 'unknown'
	end

	local statements = wu.getStatements( entity, p, nil )
	if #statements > 0 then
		catTable[ p ] = ''
		local id, statement, value
		for i = 1, #statements do
			statement = statements[ i ]
			-- defaultId is used if a qualifier is missing
			id = defaultId
			value = statement.mainsnak.datavalue.value
			if statement.qualifiers and statement.qualifiers[ qualifierP ] then
				for _, qualifier in ipairs( statement.qualifiers[ qualifierP ] ) do
					if qualifier.snaktype == 'value' then
						id = qualifier.datavalue.value.id
						if id then
							catTable[ qualifierP ] = ''
							break
						end
					end
				end
			end
			result[ id ] = value
		end
	end
	return result
end

local function analyzeDatavalue( datavalue, labelFct, ... )
	local v = datavalue.value
	local t = datavalue.type
	if type( v ) == 'table' then
		-- items which can be reduced to a string
		if t == 'wikibase-entityid' then
			v = v.id
			if type( labelFct ) == 'function' then
				v = labelFct( v, ... )
			end
		elseif t == 'quantity' then
			v.amount = v.amount:gsub( '^+', '' )
			if tonumber( v.amount ) == 0 then
				v.amount = '0'
			end
			if v.unit == '1' then
				v = tonumber( v.amount ) or 1
			else
				v.unit = wu.getUnit( v.unit )
			end
		elseif t == 'time' then
			v.calendarmodel = wu.getUnit( v.calendarmodel )
			if wu.isGregorian( v ) then
				v = v.time
			end
		end
	end
	return v, t
end

-- The following function is an experimental one, not for extensive use
-- for qualifiers, references
--  { item1, item2, ... } : using named qualifiers/references
--  {} : using no qualifiers/references
--  nil : using all qualifiers/references
function wu.getValuesWithQualifiers( entity, p, values, qualifiers, references,
	count, labelFct, ... )
	local array, qual
	local function toQualifierTable( tab, key, qualTab, labelFct, ... )
		local v
		if not tab[ key ] then
			tab[ key ] = {}
		end
		for i = 1, #qualTab do
			qual = qualTab[ i ]
			if qual.snaktype == 'value' then
				v, tab[ key .. '-type' ] =
					analyzeDatavalue( qual.datavalue, labelFct, ... )
				table.insert( tab[ key ], v )
			end
		end
		if #tab[ key ] == 0 then
			tab[ key ] = nil
			tab[ key .. '-type' ] = nil
		else
			catTable[ key ] = ''
		end
	end
	local function hasValue( tab, val )
		for i = 1, #tab do
			if tab[ i ] == val then
				return true
			end
		end
		return false
	end

	local results = {}
	local statements = wu.getStatements( entity, p, count )
	if #statements == 0 then
		return results
	end
	local i, v
	if type( values ) == 'table' and #values > 0 then
		for i = #statements, 1, -1 do
			v = statements[ i ].mainsnak.datavalue.value
			if type( v ) ~= 'string' then
				v = v.id
			end
			if not isSet( v ) or not hasValue( values, v ) then
				table.remove( statements, i )
			end
		end
		if #statements == 0 then
			return results
		end
	end
	catTable[ p ] = ''

	if type( qualifiers ) == 'string' then
		qualifiers = { qualifiers }
	end
	if type( references ) == 'string' then
		references = { references }
	end

	local key, reference, statement
	for i = 1, #statements do
		statement = statements[ i ]
		array = { value = analyzeDatavalue( statement.mainsnak.datavalue, labelFct, ... ),
			[ 'value-type' ] = statement.mainsnak.datavalue.type }

		if statement.qualifiers then
			if not qualifiers then -- all qualifier properties
				for key, qualTab in pairs( statement.qualifiers ) do
					toQualifierTable( array, key, qualTab, labelFct, ... )
				end
			else -- table of selected qualifier properties
				for j = 1, #qualifiers do
					key = qualifiers[ j ]
					if statement.qualifiers[ key ] then
						toQualifierTable( array, key, statement.qualifiers[ key ], labelFct, ... )
					end
				end
			end
		end

		array.references = {}
		if statement.references then
			for k = 1, #statement.references do
				reference = statement.references[ k ]
				if reference and reference.snaks then
					table.insert( array.references, {} )
					if not references then -- all references
						for key, refTab in pairs( reference.snaks ) do
							toQualifierTable( array.references[ #array.references ],
								key, refTab )
						end
					else -- table of selected references
						for j = 1, #references do
							key = references[ j ]
							if reference.snaks[ key ] then
								toQualifierTable( array.references[ #array.references ],
									key, reference.snaks[ key ] )
							end
						end
					end
				end
			end
		end

		table.insert( results, array )
	end

	-- clustering statements with identical value
	local helper = {}
	local sort1 = 0
	local mult = false
	local result
	for i = 1, #results do
		result = results[ i ]
		if helper[ result.value ] then
			helper[ result.value ].sort2 = helper[ result.value ].sort2 + 1
			mult = true
		else
			sort1 = sort1 + 1
			helper[ result.value ] = { sort1 = sort1, sort2 = 1 }
		end
		result.sort1 = helper[ result.value ].sort1
		result.sort2 = helper[ result.value ].sort2
	end
	if sort1 > 1 and mult and #results > 2 then
		table.sort( results,
			function( a, b )
				return a.sort1 < b.sort1 or 
					( a.sort1 == b.sort1 and a.sort2 < b.sort2 )
			end
		)
	end

	return results
end

-- get lastEdit from reference retrieve date
function wu.getLastedit( lastEdit, statements )
	local isBoolean = type( lastEdit ) == 'boolean'
	if isBoolean and lastEdit == false then
		return lastEdit
	end
	local le = ''
	for _, statement in ipairs( statements ) do
		if statement.references then
			for _, reference in ipairs( statement.references ) do
				if reference[ wd.retrieved ] then
					for _, retrieved in ipairs( reference[ wd.retrieved ] ) do
						retrieved = wu.getDateFromTime( retrieved )
						if retrieved > le then
							le = retrieved
						end
					end
				end
			end
		end
	end
	if isBoolean then
		return ( le ~= '' ) and le or lastEdit
	else
		return ( le > lastEdit ) and le or lastEdit
	end
end

function wu.getAliases( entity, lang )
	if type( entity ) == 'string' then -- is Q id
		entity = mw.wikibase.getEntity( entity )
	end
	if not lang then
		lang = mw.getContentLanguage():getCode()
	end
	local aliases = {}
	if entity and entity.aliases and entity.aliases[ lang ] then
		for _, alias in ipairs( entity.aliases[ lang ] ) do
			table.insert( aliases, alias.value )
		end
	end
	return aliases
end

-- maintenance utilities
function wu.addProperty( p )
	catTable[ p ] = ''
end

function wu.removeProperty( p )
	catTable[ p ] = nil
end

function wu.getCategories( formatStr )
	local result = ''

	if not isSet( formatStr ) then
		formatStr = '[[Category:%s]]'
	end

	catTable.P0 = nil
	for key, value in pairs( catTable ) do
		result = result .. string.format( formatStr, key )
	end
	return result
end

return wu