final String content = extractData.getContent();
        CloseableUtil.closeQuietly(in);
        logger.info(content);
        assertTrue(content.contains("テスト"));
    }

    // TODO tika needs to support pdfbox 2.0
    //    public void test_getTika_pdf() {
    //        final InputStream in = ResourceUtil
    //                .getResourceAsStream("extractor/test.pdf");

          "jat?otamuk???nan&ak?n&ah?es???o&ay?n&a&ganihcawak?simuzi?tak??eba?ikibah?oyot??t&anim?iad?omamihs??uhc??ust&oimuzi?tes????ou&kuf!.&a&d&amay?eos??g&no?ok?usak??hiku?k&awayim?uzii??ma&kan?y&asih?im???rawak?t&a&gon?ka&h?num?t???umo??wa&g&a&kan?nay?t??ias??ko!rih???y&ihsa?usak???e&m&ay?uruk??taruk?us??i&a&nohs?raihcat??goruk?h&cukuf?s&a&gih?hukuy??in???k&a&gako?muzim??iust?o?ustani??m&anim?otihsoynihs?u??r&ogo?u...

# Type of hot thread monitoring (e.g., cpu).
crawler.hotthread.type=cpu
# Metadata fields to exclude from document content.
crawler.metadata.content.excludes=resourceName,X-Parsed-By,Content-Encoding.*,Content-Type.*,X-TIKA.*,X-FESS.*
# Mapping for document metadata names.
crawler.metadata.name.mapping=\
title=title:string\n\
Title=title:string\n\
dc:title=title:string\n\

# html

    /** The key of the configuration. e.g. cpu */
    String CRAWLER_HOTTHREAD_TYPE = "crawler.hotthread.type";

    /** The key of the configuration. e.g. resourceName,X-Parsed-By,Content-Encoding.*,Content-Type.*,X-TIKA.*,X-FESS.* */
    String CRAWLER_METADATA_CONTENT_EXCLUDES = "crawler.metadata.content.excludes";

    /** The key of the configuration. e.g. title=title:string<br>
     * Title=title:string<br>

takazaki.miyazaki.jp
takehara.hiroshima.jp
taketa.oita.jp
taketomi.okinawa.jp
taki.mie.jp
takikawa.hokkaido.jp
takino.hyogo.jp
takinoue.hokkaido.jp
takko.aomori.jp
tako.chiba.jp
taku.saga.jp
talk
tama.tokyo.jp
tamakawa.fukushima.jp
tamaki.mie.jp
tamamura.gunma.jp
tamano.okayama.jp
tamatsukuri.ibaraki.jp
tamayu.shimane.jp
tamba.hyogo.jp
tana.no
tanabe.kyoto.jp

ijayo] dddd [saat] LT",lastDay:"[jana] LT",lastWeek:"[wiki iliyopita] dddd [saat] LT",sameElse:"L"},relativeTime:{future:"%s baadaye",past:"tokea %s",s:"hivi punde",ss:"sekunde %d",m:"dakika moja",mm:"dakika %d",h:"saa limoja",hh:"masaa %d",d:"siku moja",dd:"siku %d",M:"mwezi mmoja",MM:"miezi %d",y:"mwaka mmoja",yy:"miaka %d"},week:{dow:1,doy:7}});var Hr={1:"\u0be7",2:"\u0be8",3:"\u0be9",4:"\u0bea",5:"\u0beb",6:"\u0bec",7:"\u0bed",8:"\u0bee",9:"\u0bef",0:"\u0be6"},jr={"\u0be7":"1","\u0be8":"2","\u0b...

takazaki.miyazaki.jp
takehara.hiroshima.jp
taketa.oita.jp
taketomi.okinawa.jp
taki.mie.jp
takikawa.hokkaido.jp
takino.hyogo.jp
takinoue.hokkaido.jp
takko.aomori.jp
tako.chiba.jp
taku.saga.jp
talk
tama.tokyo.jp
tamakawa.fukushima.jp
tamaki.mie.jp
tamamura.gunma.jp
tamano.okayama.jp
tamatsukuri.ibaraki.jp
tamayu.shimane.jp
tamba.hyogo.jp
tana.no
tanabe.kyoto.jp

kashima.saga.jp
kitagata.saga.jp
kitahata.saga.jp
kiyama.saga.jp
kouhoku.saga.jp
kyuragi.saga.jp
nishiarita.saga.jp
ogi.saga.jp
omachi.saga.jp
ouchi.saga.jp
saga.saga.jp
shiroishi.saga.jp
taku.saga.jp
tara.saga.jp
tosu.saga.jp
yoshinogari.saga.jp
arakawa.saitama.jp
asaka.saitama.jp
chichibu.saitama.jp
fujimi.saitama.jp
fujimino.saitama.jp
fukaya.saitama.jp
hanno.saitama.jp