Skip to content

Commit cd95291

Browse files
committed
Edit alphabetical sort in config
1 parent bd13687 commit cd95291

File tree

4 files changed

+22
-22
lines changed

4 files changed

+22
-22
lines changed

core/src/main/scala/org/dbpedia/extraction/config/dataparser/DateTimeParserConfig.scala

Lines changed: 6 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -12,6 +12,7 @@ object DateTimeParserConfig
1212
"el" -> Map("ιανουάριος"->1,"φεβρουάριος"->2,"μάρτιος"->3,"απρίλιος"->4,"μάϊος"->5,"μάιος"->5,"ιούνιος"->6,"ιούλιος"->7,"αύγουστος"->8,"σεπτέμβριος"->9,"οκτώβριος"->10,"νοέμβριος"->11,"δεκέμβριος"->12,
1313
"ιανουαρίου"->1,"φεβρουαρίου"->2,"μαρτίου"->3,"απριλίου"->4,"μαΐου"->5,"μαίου"->5,"ιουνίου"->6,"ιουλίου"->7,"αυγούστου"->8,"σεπτεμβρίου"->9,"οκτωβρίου"->10,"νοεμβρίου"->11,"δεκεμβρίου"->12),
1414
"en" -> Map("january"->1,"february"->2,"march"->3,"april"->4,"may"->5,"june"->6,"july"->7,"august"->8,"september"->9,"october"->10,"november"->11,"december"->12),
15+
"eo" -> Map("januaro"->1,"februaro"->2,"marto"->3,"aprilo"->4,"majo"->5,"junio"->6,"julio"->7,"aŭgusto"->8,"septembro"->9,"oktobro"->10,"novembro"->11,"decembro"->12),
1516
"es" -> Map("enero"->1,"febrero"->2,"marzo"->3,"abril"->4,"mayo"->5,"junio"->6,"julio"->7,"agosto"->8,"septiembre"->9,"octubre"->10,"noviembre"->11,"diciembre"->12),
1617
"fr" -> Map("janvier"->1,"février"->2,"mars"->3,"avril"->4,"mai"->5,"juin"->6,"juillet"->7,"août"->8,"septembre"->9,"octobre"->10,"novembre"->11,"décembre"->12),
1718
"hr" -> Map("siječanj"->1,"veljača"->2,"ožujak"->3,"travanj"->4,"svibanj"->5,"lipanj"->6,"srpanj"->7,"kolovoz"->8,"rujan"->9,"listopad"->10,"studeni"->11,"prosinac"->12),
@@ -22,8 +23,7 @@ object DateTimeParserConfig
2223
"pt" -> Map("janeiro"->1,"fevereiro"->2,"março"->3,"abril"->4,"maio"->5,"junho"->6,"julho"->7,"agosto"->8,"setembro"->9,"outubro"->10,"novembro"->11,"dezembro"->12,
2324
"jan"->1,"fev"->2,"mar"->3,"abr"->4,"mai"->5,"jun"->6,"jul"->7,"ago"->8,"set"->9,"out"->10,"nov"->11,"dez"->12),
2425
"ru" -> Map("январь"->1,"февраль"->2,"март"->3,"апрель"->4,"май"->5,"июнь"->6,"июль"->7,"август"->8,"сентябрь"->9,"октябрь"->10,"ноябрь"->11,"декабрь"->12,
25-
"янв"->1,"фев"->2,"мар"->3,"апр"->4,"май"->5,"июн"->6,"июл"->7,"авг"->8,"сен"->9,"окт"->10,"ноя"->11,"дек"->12),
26-
"eo" -> Map("januaro"->1,"februaro"->2,"marto"->3,"aprilo"->4,"majo"->5,"junio"->6,"julio"->7,"aŭgusto"->8,"septembro"->9,"oktobro"->10,"novembro"->11,"decembro"->12)
26+
"янв"->1,"фев"->2,"мар"->3,"апр"->4,"май"->5,"июн"->6,"июл"->7,"авг"->8,"сен"->9,"окт"->10,"ноя"->11,"дек"->12)
2727
)
2828

2929
//set of wiki codes for which this parser can be applied
@@ -37,24 +37,24 @@ object DateTimeParserConfig
3737
// Don't change this unless you know how it is done.
3838
"ar" -> Map("ق.م." -> -1, "م." -> 1),
3939
"el" -> Map("ΠΧ"-> -1, "Π\\\\."-> -1, "Π\\"-> -1 , "ΜΧ"-> 1 , "Μ\\\\."-> 1, "Μ\\"-> 1),
40+
"eo" -> Map("a.K." -> -1, "p.K." -> -1),
4041
"es" -> Map("AC"-> -1, "A\\.C\\."-> -1, "DC"-> 1, "D\\.C\\."-> 1, "AD"-> 1, "A\\.D\\."-> 1, "AEC"-> 1, "A\\.E\\.C\\."-> 1 , "EC"-> 1, "E\\.C\\."-> 1),
4142
"it" -> Map("AC"-> -1, "A\\.C\\."-> -1, "DC"-> 1, "D\\.C\\."-> 1, "AD"-> 1, "A\\.D\\."-> 1, "PEV"-> -1, "P\\.E\\.V\\."-> -1, "EV"-> 1, "E\\.V\\." -> 1),
4243
"nl" -> Map("v\\.Chr\\." -> -1, "n\\.C\\."-> 1, "v\\.C\\." -> -1, "n\\.Chr\\."-> 1, "voor Chr\\." -> -1, "na Chr\\."-> 1),
4344
"pt" -> Map("AC"-> -1, "A\\.C\\."-> -1, "DC"-> 1, "D\\.C\\."-> 1, "AD"-> 1, "A\\.D\\."-> 1, "AEC"-> 1, "A\\.E\\.C\\."-> 1 , "EC"-> 1, "E\\.C\\."-> 1),
44-
"fr" -> Map("av\\. J\\.-C\\."-> -1, "ap\\. J\\.-C\\." -> 1),
45-
"eo" -> Map("a.K." -> -1, "p.K." -> -1)
45+
"fr" -> Map("av\\. J\\.-C\\."-> -1, "ap\\. J\\.-C\\." -> 1)
4646
)
4747

4848
//suffixes for 1st, 2nd etc. (maybe add this to infobox extractor RankRegex val)
4949
val cardinalityRegexMap = Map(
5050
"en" -> "st|nd|rd|th",
5151
"el" -> "η|ης",
52+
"eo" -> "-a|-an",
5253
"es" -> "°|\\.°|°\\.",
5354
"it" -> "°|\\.°|°\\.",
5455
"nl" -> "ste|de|e",
5556
"pt" -> "°|\\.°|°\\.",
56-
"fr" -> "er|nd|ème",
57-
"eo" -> "-a|-an"
57+
"fr" -> "er|nd|ème"
5858
)
5959

6060
//specifies for a template name (lower-cased) the property keys of year, month and day

core/src/main/scala/org/dbpedia/extraction/config/dataparser/ParserUtilsConfig.scala

Lines changed: 8 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -46,6 +46,14 @@ object ParserUtilsConfig
4646
"τρισεκατομμυρια" -> 12,
4747
"τετράκις εκατομμύρια" -> 15
4848
),
49+
"eo" -> Map(
50+
"mil" -> 3,
51+
"miliono" -> 6,
52+
"miliardo" -> 9,
53+
"biliono" -> 12,
54+
"biliardo" -> 15,
55+
"triliono" -> 18
56+
),
4957

5058
"nl" -> Map(
5159
"honderd" -> 2,
@@ -92,14 +100,6 @@ object ParserUtilsConfig
92100
"milliard" -> 9,
93101
"billion" -> 12,
94102
"trillion" -> 18
95-
),
96-
"eo" -> Map(
97-
"mil" -> 3,
98-
"miliono" -> 6,
99-
"miliardo" -> 9,
100-
"biliono" -> 12,
101-
"biliardo" -> 15,
102-
"triliono" -> 18
103103
)
104104
)
105105

core/src/main/scala/org/dbpedia/extraction/config/mappings/DisambiguationExtractorConfig.scala

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -11,6 +11,7 @@ object DisambiguationExtractorConfig
1111
"de" -> " (Begriffsklärung)",
1212
"el" -> " (αποσαφήνιση)",
1313
"en" -> " (disambiguation)",
14+
"eo" -> " (apartigilo)",
1415
"es" -> " (desambiguación)",
1516
"eu" -> " (argipena)",
1617
"fr" -> " (homonymie)",
@@ -20,7 +21,6 @@ object DisambiguationExtractorConfig
2021
"nl" -> " (doorverwijspagina)", //TODO make it Set() for multiple “nl” -> " (disambigueren)"
2122
"pl" -> " (ujednoznacznienie)",
2223
"pt" -> " (desambiguação)",
23-
"ru" -> " (значения)",
24-
"eo" -> " (apartigilo)"
24+
"ru" -> " (значения)"
2525
)
2626
}

core/src/main/scala/org/dbpedia/extraction/config/mappings/HomepageExtractorConfig.scala

Lines changed: 6 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -14,6 +14,7 @@ object HomepageExtractorConfig
1414
"de" -> Set("website", "homepage", "webpräsenz", "web", "site", "siteweb", "site web"),/*cleanup*/
1515
"el" -> Set("ιστότοπος", "ιστοσελίδα"),
1616
"en" -> Set("website", "homepage", "web", "site"),
17+
"eo" -> Set("ĉefpaĝo", "retejo"),
1718
"es" -> Set("website", "homepage", "web", "site", "siteweb", "site web", "página", "sitio", "pagina"),/*cleanup*/
1819
"eu" -> Set("webgunea"),
1920
"fr" -> Set("website", "homepage", "web", "site", "siteweb", "site web"),/*cleanup*/
@@ -22,8 +23,7 @@ object HomepageExtractorConfig
2223
"nl" -> Set("website", "homepage", "hoofdpagina", "webpagina", "web", "site"),
2324
"pl" -> Set("web", "strona"),
2425
"pt" -> Set("website", "homepage", "web", "site", "siteweb", "site web", "página", "sitio", "pagina"),/*cleanup*/
25-
"ru" -> Set("сайт"),
26-
"eo" -> Set("ĉefpaĝo", "retejo")
26+
"ru" -> Set("сайт")
2727
)
2828

2929
val supportedLanguages = propertyNamesMap.keySet
@@ -34,6 +34,7 @@ object HomepageExtractorConfig
3434
"de" -> "Weblinks?",
3535
"el" -> "(?:Εξωτερικοί σύνδεσμοι|Εξωτερικές συνδέσεις)",
3636
"en" -> "External links?",
37+
"eo" -> "Eksteraj ligiloj",
3738
"es" -> "(?:Enlaces externos|Enlace externo|Links externos|Link externo)",
3839
"eu" -> "Kanpo loturak?",
3940
"fr" -> "(?:Lien externe|Liens externes|Liens et documents externes)",
@@ -42,8 +43,7 @@ object HomepageExtractorConfig
4243
"nl" -> "(?:Externe links|Externe link)",
4344
"pl" -> "(?:Linki zewnętrzne|Link zewnętrzny)",
4445
"pt" -> "(?:Ligações externas|Ligação externa|Links externos|Link externo)",
45-
"ru" -> "Ссылки",
46-
"eo" -> "Eksteraj ligiloj"
46+
"ru" -> "Ссылки"
4747
)
4848

4949
val officialMap = Map(
@@ -52,6 +52,7 @@ object HomepageExtractorConfig
5252
"de" -> "offizielle",
5353
"el" -> "(?:επίσημος|επίσημη)",
5454
"en" -> "official",
55+
"eo" -> "oficiala",
5556
"es" -> "oficial",
5657
"eu" -> "ofiziala?",
5758
"fr" -> "officiel",
@@ -60,8 +61,7 @@ object HomepageExtractorConfig
6061
"nl" -> "(?:officieel|officiële)",
6162
"pl" -> "oficjalna",
6263
"pt" -> "oficial",
63-
"ru" -> "официальный",
64-
"eo" -> "oficiala"
64+
"ru" -> "официальный"
6565
)
6666

6767
}

0 commit comments

Comments
 (0)