Newer
Older
# download and extraction target dir
#base-dir= moved to $extraction-framework/core/src/main/resources/universal.properties
# Source file. If source file name ends with .gz or .bz2, it is unzipped on the fly.
# Must exist in the directory xxwiki/yyyymmdd and have the prefix xxwiki-yyyymmdd-
# where xx is the wiki code and yyyymmdd is the dump date.
# default:
#source=# moved to $extraction-framework/core/src/main/resources/universal.properties
# use only directories that contain a 'download-complete' file? Default is false.
require-download-complete=false
# List of languages or article count ranges, e.g. 'en,de,fr' or '10000-20000' or '10000-', or '@mappings'
languages=@mappings
# extractor class names starting with "." are prefixed by "org.dbpedia.extraction.mappings"
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
#extractors.ar=.MappingExtractor,.TopicalConceptsExtractor
#
#extractors.be=.MappingExtractor
#
#extractors.bg=.MappingExtractor
#
#extractors.bn=.MappingExtractor
#
#extractors.ca=.MappingExtractor,.DisambiguationExtractor,.HomepageExtractor,.TopicalConceptsExtractor
#
#extractors.ced=.MappingExtractor
#
#extractors.commons=.MappingExtractor,.ContributorExtractor,.TemplateParameterExtractor,.FileTypeExtractor,.GalleryExtractor,.ImageAnnotationExtractor,.CommonsKMLExtractor,.DBpediaResourceExtractor
#
#extractors.cs=.MappingExtractor
#
#extractors.cy=.MappingExtractor
#
#extractors.da=.MappingExtractor
#
#extractors.de=.MappingExtractor,.DisambiguationExtractor,.HomepageExtractor,.ImageExtractorNew,.PersondataExtractor,.PndExtractor,.CommonsResourceExtractor
#
#extractors.el=.MappingExtractor,.DisambiguationExtractor,.HomepageExtractor,.ImageExtractorNew,.TopicalConceptsExtractor
#
#extractors.en=.MappingExtractor,.CitationExtractor,.DisambiguationExtractor,.GenderExtractor,.HomepageExtractor,.ImageExtractorNew,.PersondataExtractor,.PndExtractor,.TopicalConceptsExtractor,.AnchorTextExtractor,.CommonsResourceExtractor
#
#extractors.eo=.MappingExtractor
#
#extractors.es=.MappingExtractor,.DisambiguationExtractor,.HomepageExtractor,.ImageExtractorNew,.TopicalConceptsExtractor,.CommonsResourceExtractor
#
#extractors.et=.MappingExtractor
#
#extractors.eu=.MappingExtractor,.DisambiguationExtractor,.HomepageExtractor,.ImageExtractorNew,.TopicalConceptsExtractor
#
#extractors.fa=.MappingExtractor
#
#extractors.fi=.MappingExtractor
#
#extractors.fr=.MappingExtractor,.DisambiguationExtractor,.HomepageExtractor,.ImageExtractorNew,.PndExtractor,.TopicalConceptsExtractor,.fr.PopulationExtractor,.CommonsResourceExtractor
#
#extractors.ga=.MappingExtractor,.HomepageExtractor
#
#extractors.gl=.MappingExtractor
#
#extractors.hi=.MappingExtractor
#
#extractors.hr=.MappingExtractor
#
#extractors.hu=.MappingExtractor
#
#extractors.id=.MappingExtractor
#
#extractors.it=.MappingExtractor,.DisambiguationExtractor,.HomepageExtractor,.ImageExtractorNew,.TopicalConceptsExtractor
#
#extractors.ja=.MappingExtractor,.DisambiguationExtractor,.HomepageExtractor,.ImageExtractorNew,.TopicalConceptsExtractor,.CommonsResourceExtractor
#
#extractors.ko=.MappingExtractor,.DisambiguationExtractor
#
#extractors.lt=.MappingExtractor
#
#extractors.lv=.MappingExtractor
#
#extractors.nl=.MappingExtractor,.DisambiguationExtractor,.ImageExtractorNew,.CommonsResourceExtractor
#
#extractors.mk=.MappingExtractor
#
#extractors.mt=.MappingExtractor
#
#extractors.pl=.MappingExtractor,.DisambiguationExtractor,.HomepageExtractor,.ImageExtractorNew
#
#extractors.pt=.MappingExtractor,.DisambiguationExtractor,.HomepageExtractor,.ImageExtractorNew,.TopicalConceptsExtractor,.CommonsResourceExtractor
#
#extractors.ru=.MappingExtractor,.DisambiguationExtractor,.HomepageExtractor,.ImageExtractorNew,.TopicalConceptsExtractor
#
#extractors.sk=.MappingExtractor
#
#extractors.sl=.MappingExtractor
#
#extractors.sr=.MappingExtractor
#
#extractors.tr=.MappingExtractor
#
#extractors.ur=.MappingExtractor
#
#extractors.vi=.MappingExtractor
#
#extractors.war=.MappingExtractor
#only the raw extractor here: all other wikidata extractors are executed in an separate extraction for wikidata (see: extraction.wikidata.properties)
#extractors.wikidata=.WikidataSameAsExtractor,.WikidataRawExtractor
#extractors.zh=.MappingExtractor
# If we need to Exclude Non-Free Images in this Extraction, set this to true
copyrightCheck=false