jtok-user.jtok.en.en_punct.xml Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of marytts-runtime Show documentation
The newest version!
<?xml version="1.0" encoding="UTF-8"?>
<!-- This contains the punctuation description for English. -->
<PUNCTUATION_DESCRIPTION>

  <!-- This defines the specific punctuation elements. The regexp -->
  <!-- attribute is the regular expression used to match -->
  <!-- the punctuation. -->
  <!-- ATTENTION! Because the regex package we use always -->
  <!-- stops at the first disjunct that could be matched, the -->
  <!-- order of the definition with a class is critical! So make -->
  <!-- sure that punctuations with the potentially longer match -->
  <!-- stand BEFORE others, e.g. use the order `abc|ab|a' -->
  <!-- instead of `a|ab|abc' -->
  <DEFINITIONS>  
    
    <!-- possible terminal punctuation -->
    <PERIOD regexp="\." class="TERM_PUNCT_P"/>

    <!-- terminal punctuation -->
    <MULTI_PERIOD regexp="\.{2,}" class="TERM_PUNCT"/>
    <EXCLAM regexp="!+" class="TERM_PUNCT"/>
    <QUEST regexp="\?+" class="TERM_PUNCT"/>

    <!-- opening punctuation -->
    <OINF regexp="&lt;&lt;+" class="OPEN_PUNCT"/>    
    <RSQUOTE_L regexp="``?" class="OPEN_PUNCT"/>
    <OGUILLEMET regexp="«" class="OPEN_PUNCT"/>
    <ODQUO1 regexp="“" class="OPEN_PUNCT"/> <!-- Unicode U+201C -->
    <ODQUO2 regexp="„" class="OPEN_PUNCT"/> <!-- Unicode U+201F -->
    <ODQUO3 regexp="‟" class="OPEN_PUNCT"/> <!-- Unicode U+201E -->
    <OSQUO1 regexp="‘" class="OPEN_PUNCT"/> <!-- Unicode U+2018 -->
    <OSQUO2 regexp="‚" class="OPEN_PUNCT"/> <!-- Unicode U+201A -->
    <OSQUO3 regexp="‛" class="OPEN_PUNCT"/> <!-- Unicode U+201B -->
    <OSAQUO regexp="‹" class="OPEN_PUNCT"/> <!-- Unicode U+2039 -->
    <!-- opening brackets -->
    <OPAR regexp="\(" class="OPEN_BRACKET"/>
    <OPAR2 regexp="\{" class="OPEN_BRACKET"/>
    <OCROCHE regexp="\[" class="OPEN_BRACKET"/>

    <!-- closing punctuation -->
    <CSUP regexp="&gt;&gt;+" class="CLOSE_PUNCT"/>
    <CGUILLEMET regexp="»" class="CLOSE_PUNCT"/>
    <CDQUO regexp="”" class="CLOSE_PUNCT"/> <!-- Unicode U+201D -->
    <CSQUO regexp="’" class="CLOSE_PUNCT"/> <!-- Unicode U+2019 -->
    <CSAQUO regexp="›" class="CLOSE_PUNCT"/> <!-- Unicode U+203A -->
    <!-- closing brackets -->
    <CPAR regexp="\)" class="CLOSE_BRACKET"/>
    <CPAR2 regexp="\}" class="CLOSE_BRACKET"/>
    <CCROCHE regexp="\]" class="CLOSE_BRACKET"/>

    <!-- ambigue opening/closing punctuation -->
    <OCQUOTE regexp="\&quot;" class="OPENCLOSE_PUNCT"/>
    <RSQUOTE_R regexp="&apos;&apos;?" class="OPENCLOSE_PUNCT"/>

    <!-- this is a matcher for the special case when two internal -->
    <!-- punctuations don't cause a split; it's used for -->
    <!-- enclitics after abbreviations like in L.A.'s -->
    <!-- and abbreviations in composita. -->
    <SPECIAL_INT regexp="\.'|\.-" class="PUNCTUATION"/>

    <!-- other punctuation -->
    <LT regexp="&lt;" class="PUNCTUATION"/>
    <GT regexp="&gt;" class="PUNCTUATION"/>
    <EQ regexp="=" class="PUNCTUATION"/>
    <OCHYPHEN regexp="--+" class="PUNCTUATION"/>
    <FIGDASH regexp="‒" class="PUNCTUATION"/> <!-- Unicode U+2012 -->
    <ENDASH regexp="–" class="PUNCTUATION"/> <!-- Unicode U+2013 -->
    <EMDASH regexp="—" class="PUNCTUATION"/> <!-- Unicode U+2014 -->
    <HBAR regexp="―" class="PUNCTUATION"/> <!-- Unicode U+2015 -->
    <BULLET regexp="•|■|□" class="PUNCTUATION"/>
    <PERCENT regexp="%" class="PUNCTUATION"/>
    <TM regexp="©|®|™" class="PUNCTUATION"/>
    <SECTION regexp="§" class="PUNCTUATION"/>

    <!-- only match a hyphen when it's not followed by a digit -->
    <!-- because it's part of a signed digit then, like in -1 -->
    <HYPHEN regexp="-(?!\d)" class="PUNCTUATION"/>
    <SCOLON regexp=";" class="PUNCTUATION"/>
    <COLON regexp=":" class="PUNCTUATION"/>
    <!-- only match a comma when it's not followed by a digit -->
    <!-- because it's then digit internal, like in 3,5 -->
    <COMMA regexp=",(?!\d)" class="PUNCTUATION"/>
    <SLASH regexp="\/" class="PUNCTUATION"/>
    <BSLASH regexp="\\" class="PUNCTUATION"/>
    <PIPE regexp="\|" class="PUNCTUATION"/>
    <STAR regexp="\*" class="PUNCTUATION"/>
    <TILDE regexp="\~" class="PUNCTUATION"/>
    <AMP regexp="\&amp;" class="PUNCTUATION"/>

    <NBL regexp="@" class="PUNCTUATION"/>

  </DEFINITIONS>

  <!-- These are the mandatory rules defined as -->
  <!-- regular expressions with the definitions above. -->
  <!-- Rule names may not be altered! -->
  <!-- The default behavior is to treat the punctuation -->
  <!-- defined above as breaking punctuation, that is -->
  <!-- this punctuations are always used as separator -->
  <!-- and never as part of clitics or abbreviation. -->
  <RULES>

    <!-- The all-rule is used for basic splitting of strings -->
    <!-- into punctuation and non-punctuation parts. -->
    <!-- It's basically a disjunction of the punctuations -->
    <!-- defined above. -->
    <!-- ATTENTION! Because the regex package we use always -->
    <!-- stops at the first disjunct that could be matched, the -->
    <!-- order of the disjuncts in the rule is critical! So make -->
    <!-- sure that punctuations with the potentially longer match -->
    <!-- stand BEFORE others, e.g. use the order `abc|ab|a' -->
    <!-- instead of `a|ab|abc' -->
    <ALL_PUNCT_RULE>
      (<SPECIAL_INT/>)(?=\w)|<MULTI_PERIOD/>|<PERIOD/>|<OINF/>|<LT/>|<CSUP/>|<GT/>|<EQ/>|<OCHYPHEN/>|<HYPHEN/>|<RSQUOTE_L/>|<RSQUOTE_R/>|<OGUILLEMET/>|<CGUILLEMET/>|<CDQUO/>|<CSAQUO/>|<CSQUO/>|<ODQUO1/>|<ODQUO2/>|<ODQUO3/>|<OSAQUO/>|<OSQUO1/>|<OSQUO2/>|<OSQUO3/>|<OPAR/>|<CPAR/>|<OPAR2/>|<CPAR2/>|<OCROCHE/>|<CCROCHE/>|<OCQUOTE/>|<EXCLAM/>|<QUEST/>|<SCOLON/>|<COLON/>|<COMMA/>|<SLASH/>|<BSLASH/>|<PIPE/>|<STAR/>|<TILDE/>|<AMP/>|<NBL/>|<FIGDASH/>|<ENDASH/>|<EMDASH/>|<HBAR/>|<BULLET/>|<PERCENT/>|<TM/>|<SECTION/>
    </ALL_PUNCT_RULE>

    <!-- punctuation which can mark clitics; this rule is used -->
    <!-- to check if closer clitics examination is needed; -->
    <!-- using '.' here has the effect that ALL tokens are -->
    <!-- checked for special prefixes and suffixes. -->
    <CLITIC_PUNCT_RULE>
      .
    </CLITIC_PUNCT_RULE>

    <!-- punctuation which can be found in compounds -->
    <INTERNAL_PUNCT_RULE>
      <SPECIAL_INT/>|<RSQUOTE_R/>|<HYPHEN/>|<PERIOD/>|<AMP/>|<SLASH/>|<COLON/>|<NBL/>
    </INTERNAL_PUNCT_RULE>

    <!-- non-breaking punctuation on the left side of a token; -->
    <!-- use the beginning-of-input matcher \A to easily -->
    <!-- identify this punctuation -->
    <NON_BREAKING_LEFT_PUNCT_RULE class="PUNCTUATION">
      \A(<NBL/>)
    </NON_BREAKING_LEFT_PUNCT_RULE>

    <!-- non-breaking punctuation on the right side of a token; -->
    <!-- use the end-of-input matcher \Z to easily -->
    <!-- identify this punctuation -->
    <NON_BREAKING_RIGHT_PUNCT_RULE class="TERM_PUNCT_P">
      <PERIOD/>\Z
    </NON_BREAKING_RIGHT_PUNCT_RULE>

    <!-- punctuation which can be found only within sentences -->
    <INTERNAL_TU_PUNCT_RULE>
      <COMMA/>|<SCOLON/>|<COLON/>
    </INTERNAL_TU_PUNCT_RULE>

  </RULES>
    
</PUNCTUATION_DESCRIPTION>