Regexp.xml

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook XML V4.5//EN"
"../../../docbook-xml-4.5/docbookx.dtd">
<chapter id="chapter.regexp">
  <title>Regular expressions<indexterm class="singular">
      <primary>Regular expressions</primary>

      <seealso>Segmentation</seealso>

      <seealso>Searching</seealso>
    </indexterm></title>

  <para>The regular expressions (or regex for short) used in searches and
  segmentation rules are those supported by Java. Should you need more
  specific information, consult the <ulink
  url="http://download.oracle.com/javase/1.6.0/docs/api/java/util/regex/Pattern.html">Java
  Regex documentation</ulink>. See additional references and examples
  below.</para>

  <note>
    <para>This chapter is intended for advanced users, who need to define
    their own variants of segmentation rules or devise more complex and
    powerful key search items.</para>
  </note>

  <table>
    <title>Regex - Flags</title>

    <tgroup align="left" cols="2" rowsep="1">
      <colspec align="left" colnum="1"/>

      <thead>
        <row>
          <entry align="left">The construct</entry>

          <entry align="left">... matches the following</entry>
        </row>
      </thead>

      <tbody>
        <row>
          <entry>(?i)</entry>

          <entry>Enables case-insensitive matching (by default, the pattern is
          case-sensitive).</entry>
        </row>
      </tbody>
    </tgroup>
  </table>

  <table>
    <title>Regex - Character</title>

    <tgroup align="left" cols="2" rowsep="1">
      <colspec align="left" colnum="1"/>

      <thead>
        <row>
          <entry align="left">The construct</entry>

          <entry align="left">... matches the following</entry>
        </row>
      </thead>

      <tbody>
        <row>
          <entry>x</entry>

          <entry>The character x, except the following...</entry>
        </row>

        <row>
          <entry>\uhhhh</entry>

          <entry>The character with hexadecimal value 0xhhhh</entry>
        </row>

        <row>
          <entry>\t</entry>

          <entry>The tab character ('\u0009')</entry>
        </row>

        <row>
          <entry>\n</entry>

          <entry>The newline (line feed) character ('\u000A')</entry>
        </row>

        <row>
          <entry>\r</entry>

          <entry>The carriage-return character ('\u000D')</entry>
        </row>

        <row>
          <entry>\f</entry>

          <entry>The form-feed character ('\u000C')</entry>
        </row>

        <row>
          <entry>\a</entry>

          <entry>The alert (bell) character ('\u0007')</entry>
        </row>

        <row>
          <entry>\e</entry>

          <entry>The escape character ('\u001B')</entry>
        </row>

        <row>
          <entry>\cx</entry>

          <entry>The control character corresponding to x</entry>
        </row>

        <row>
          <entry>\0n</entry>

          <entry>The character with octal value 0n (0 &lt;= n &lt;= 7)</entry>
        </row>

        <row>
          <entry>\0nn</entry>

          <entry>The character with octal value 0nn (0 &lt;= n &lt;=
          7)</entry>
        </row>

        <row>
          <entry>\0mnn</entry>

          <entry>The character with octal value 0mnn (0 &lt;= m &lt;= 3, 0
          &lt;= n &lt;= 7)</entry>
        </row>

        <row>
          <entry>\xhh</entry>

          <entry>The character with hexadecimal value 0xhh</entry>
        </row>
      </tbody>
    </tgroup>
  </table>

  <table>
    <title>Regex - Quotation</title>

    <tgroup align="left" cols="2" rowsep="1">
      <colspec align="left" colnum="1"/>

      <thead>
        <row>
          <entry align="left">The construct</entry>

          <entry align="left">...matches the following</entry>
        </row>
      </thead>

      <tbody>
        <row>
          <entry>\</entry>

          <entry>Nothing, but quotes the following character. This is required
          if you would like to enter any of the meta characters
          !$()*+.&lt;&gt;?[\]^{|} to match as themselves.</entry>
        </row>

        <row>
          <entry>\\</entry>

          <entry>For example, this is the backslash character</entry>
        </row>

        <row>
          <entry>\Q</entry>

          <entry>Nothing, but quotes all characters until \E</entry>
        </row>

        <row>
          <entry>\E</entry>

          <entry>Nothing, but ends quoting started by \Q</entry>
        </row>
      </tbody>
    </tgroup>
  </table>

  <table>
    <title>Regex - Classes for Unicode blocks and categories</title>

    <tgroup align="left" cols="2" rowsep="1">
      <colspec align="left" colnum="1"/>

      <thead>
        <row>
          <entry align="left">The construct</entry>

          <entry align="left">...matches the following</entry>
        </row>
      </thead>

      <tbody>
        <row>
          <entry>\p{InGreek}</entry>

          <entry>A character in the Greek block (simple <ulink
          url="http://download.oracle.com/javase/1.6.0/docs/api/java/util/regex/Pattern.html#ubc">
          block</ulink>)</entry>
        </row>

        <row>
          <entry>\p{Lu}</entry>

          <entry>An uppercase letter (simple <ulink
          url="http://download.oracle.com/javase/1.6.0/docs/api/java/util/regex/Pattern.html#ubc">
          category</ulink>)</entry>
        </row>

        <row>
          <entry>\p{Sc}</entry>

          <entry>A currency symbol</entry>
        </row>

        <row>
          <entry>\P{InGreek}</entry>

          <entry>Any character except one in the Greek block
          (negation)</entry>
        </row>

        <row>
          <entry>[\p{L}&amp;&amp;[^\p{Lu}]]</entry>

          <entry>Any letter except an uppercase letter (subtraction)</entry>
        </row>
      </tbody>
    </tgroup>
  </table>

  <table>
    <title>Regex - Character classes</title>

    <tgroup align="left" cols="2" rowsep="1">
      <colspec align="left" colnum="1"/>

      <thead>
        <row>
          <entry align="left">The construct</entry>

          <entry align="left">...matches the following</entry>
        </row>
      </thead>

      <tbody>
        <row>
          <entry>[abc]</entry>

          <entry>a, b, or c (simple class)</entry>
        </row>

        <row>
          <entry>[^abc]</entry>

          <entry>Any character except a, b, or c (negation)</entry>
        </row>

        <row>
          <entry>[a-zA-Z]</entry>

          <entry>a through z or A through Z, inclusive (range)</entry>
        </row>
      </tbody>
    </tgroup>
  </table>

  <table>
    <title>Regex - Predefined character classes</title>

    <tgroup align="left" cols="2" rowsep="1">
      <colspec align="left" colnum="1"/>

      <thead>
        <row>
          <entry align="left">The construct</entry>

          <entry align="left">...matches the following</entry>
        </row>
      </thead>

      <tbody>
        <row>
          <entry>.</entry>

          <entry>Any character (except for line terminators)</entry>
        </row>

        <row>
          <entry>\d</entry>

          <entry>A digit: [0-9]</entry>
        </row>

        <row>
          <entry>\D</entry>

          <entry>A non-digit: [^0-9]</entry>
        </row>

        <row>
          <entry>\s</entry>

          <entry>A whitespace character: [ \t\n\x0B\f\r]</entry>
        </row>

        <row>
          <entry>\S</entry>

          <entry>A non-whitespace character: [^\s]</entry>
        </row>

        <row>
          <entry>\w</entry>

          <entry>A word character: [a-zA-Z_0-9]</entry>
        </row>

        <row>
          <entry>\W</entry>

          <entry>A non-word character: [^\w]</entry>
        </row>
      </tbody>
    </tgroup>
  </table>

  <table>
    <title>Regex - Boundary matchers</title>

    <tgroup align="left" cols="2" rowsep="1">
      <colspec align="left" colnum="1"/>

      <thead>
        <row>
          <entry align="left">The construct</entry>

          <entry align="left">...matches the following</entry>
        </row>
      </thead>

      <tbody>
        <row>
          <entry>^</entry>

          <entry>The beginning of a line</entry>
        </row>

        <row>
          <entry>$</entry>

          <entry>The end of a line</entry>
        </row>

        <row>
          <entry>\b</entry>

          <entry>A word boundary</entry>
        </row>

        <row>
          <entry>\B</entry>

          <entry>A non-word boundary</entry>
        </row>
      </tbody>
    </tgroup>
  </table>

  <table>
    <title>Regex - Greedy quantifiers</title>

    <tgroup align="left" cols="2" rowsep="1">
      <colspec align="left" colnum="1"/>

      <thead>
        <row>
          <entry align="left">The construct</entry>

          <entry align="left">...matches the following</entry>
        </row>
      </thead>

      <tbody>
        <row>
          <entry>X<emphasis>?</emphasis></entry>

          <entry>X, once or not at all</entry>
        </row>

        <row>
          <entry>X<emphasis>*</emphasis></entry>

          <entry>X, zero or more times</entry>
        </row>

        <row>
          <entry>X<emphasis>+</emphasis></entry>

          <entry>X, one or more times</entry>
        </row>
      </tbody>
    </tgroup>
  </table>

  <note>
    <para>greedy quantifiers will match as much as they can. For example,
    <emphasis>a+?</emphasis> will match the aaa in
    <emphasis>aaabbb</emphasis></para>
  </note>

  <table>
    <title>Regex - Reluctant (non-greedy) quantifiers</title>

    <tgroup align="left" cols="2" rowsep="1">
      <colspec align="left" colnum="1"/>

      <thead>
        <row>
          <entry align="left">The construct</entry>

          <entry align="left">...matches the following</entry>
        </row>
      </thead>

      <tbody>
        <row>
          <entry>X??</entry>

          <entry>X, once or not at all</entry>
        </row>

        <row>
          <entry>X*?</entry>

          <entry>X, zero or more times</entry>
        </row>

        <row>
          <entry>X+?</entry>

          <entry>X, one or more times</entry>
        </row>
      </tbody>
    </tgroup>
  </table>

  <note>
    <para>non-greedy quantifiers will match as little as they can. For
    example, <emphasis>a+?</emphasis> will match the first
    <emphasis>a</emphasis> in <emphasis>aaabbb</emphasis></para>
  </note>

  <table>
    <title>Regex - Logical operators</title>

    <tgroup align="left" cols="2" rowsep="1">
      <colspec align="left" colnum="1"/>

      <thead>
        <row>
          <entry align="left">The construct</entry>

          <entry align="left">...matches the following</entry>
        </row>
      </thead>

      <tbody>
        <row>
          <entry>XY</entry>

          <entry>X followed by Y</entry>
        </row>

        <row>
          <entry>X|Y</entry>

          <entry>Either X or Y</entry>
        </row>

        <row>
          <entry>(XY)</entry>

          <entry>XY as a single group</entry>
        </row>
      </tbody>
    </tgroup>
  </table>

  <section id="regex.tools.and.examples.of.use">
    <title><indexterm class="singular">
        <primary>Regular expressions</primary>

        <secondary>Tools</secondary>
      </indexterm>Regex tools and examples of use<indexterm class="singular">
        <primary>Regular expressions</primary>

        <secondary>Examples of use</secondary>
      </indexterm></title>

    <para>A number of interactive tools are available to develop and test
    regular expressions. They generally follow much the same pattern (for an
    example from the Regular Expression Tester see below): the regular
    expression (top entry) analyzes the search text (Text box in the middle) ,
    yielding the hits, shown in the result Text box.</para>

    <figure id="regerx.tester">
      <title>Regex Tester</title>

      <mediaobject>
        <imageobject role="html">
          <imagedata fileref="images/RegexTester.png"/>
        </imageobject>

        <imageobject role="fo">
          <imagedata fileref="images/RegexTester.png" width="80%"/>
        </imageobject>
      </mediaobject>
    </figure>

    <para>See <ulink url="http://weitz.de/regex-coach/">The Regex
    Coach</ulink> for Windows,Linux, FreeBSD versions of a stand-alone tool.
    This is much the same as the above example.</para>

    <para>A nice collection of useful regex cases can be found in
    <application>OmegaT</application> itself (see Options &gt; Segmentation).
    The following list includes expressions you may find useful when searching
    through the translation memory:</para>

    <table>
      <title>Regex - Examples of regular expressions in translations</title>

      <tgroup align="left" cols="2" rowsep="1">
        <colspec align="left" colnum="1"/>

        <thead>
          <row>
            <entry>Regular expression</entry>

            <entry>Finds the following:</entry>
          </row>
        </thead>

        <tbody>
          <row>
            <entry>(\b\w+\b)\s\1\b</entry>

            <entry>double words</entry>
          </row>

          <row>
            <entry>[\.,]\s*[\.,]+</entry>

            <entry>comma or a period, followed by spaces and yet another comma
            or period</entry>
          </row>

          <row>
            <entry>\. \s+$</entry>

            <entry>extra spaces after the period at the end of the
            line</entry>
          </row>

          <row>
            <entry>\s+a\s+[aeiou]</entry>

            <entry>English: words, beginning with vowels, should generally be
            preceded by "an", not "a"</entry>
          </row>

          <row>
            <entry>\s+an\s+[^aeiou]</entry>

            <entry>English: the same check as above, but concerning consonants
            ("a", not "an")</entry>
          </row>

          <row>
            <entry>\s{2,}</entry>

            <entry>more than one space</entry>
          </row>

          <row>
            <entry>\.[A-Z]</entry>

            <entry>Period, followed by an upper-case letter - possibly a space
            is missing between the period and the start of a new
            sentence?</entry>
          </row>

          <row>
            <entry>\bis\b</entry>

            <entry>search for “is”, not “this” or “isn't” etc.</entry>
          </row>
        </tbody>
      </tgroup>
    </table>
  </section>
</chapter>