summaryrefslogtreecommitdiff
path: root/Master/texmf/doc/xindy/style-tutorial-2.html
diff options
context:
space:
mode:
Diffstat (limited to 'Master/texmf/doc/xindy/style-tutorial-2.html')
-rw-r--r--Master/texmf/doc/xindy/style-tutorial-2.html361
1 files changed, 0 insertions, 361 deletions
diff --git a/Master/texmf/doc/xindy/style-tutorial-2.html b/Master/texmf/doc/xindy/style-tutorial-2.html
deleted file mode 100644
index b5313e59d72..00000000000
--- a/Master/texmf/doc/xindy/style-tutorial-2.html
+++ /dev/null
@@ -1,361 +0,0 @@
-<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN">
-<HTML>
-<HEAD>
- <META NAME="GENERATOR" CONTENT="SGML-Tools 1.0.9">
- <TITLE>XINDY Style File Tutorial: A Basic Introduction</TITLE>
- <LINK HREF="style-tutorial-3.html" REL=next>
- <LINK HREF="style-tutorial-1.html" REL=previous>
- <LINK HREF="style-tutorial.html#toc2" REL=contents>
-</HEAD>
-<BODY>
-<A HREF="style-tutorial-3.html">Next</A>
-<A HREF="style-tutorial-1.html">Previous</A>
-<A HREF="style-tutorial.html#toc2">Contents</A>
-<HR>
-<H2><A NAME="s2">2. A Basic Introduction</A></H2>
-
-<P>This section incrementally introduces the most important aspects of
-the system. After reading this chapter you should be able to specify
-style files for
-about 80% of the commonly used indexes. The examples are demonstrated
-with a TeX markup so one can easily typeset the results <SF>xindy</SF>
-produces. You need LaTeX and the ISO-Latin enhancements that come
-with the <CODE>inputenc</CODE> package to run the following examples. Also the
-<SF>xindy</SF> system must already be installed.
-<P>
-But this tutorial doesn't reflect real life any more. The concepts are
-introduced below to explain them, but actual usage is probably
-different. In particular, you should not expect to specify sort rules
-by hand; usually one uses the language modules for that. Same way,
-xindy standard modules provide lots of markup functionality that you
-need for your documents, and can be used as a starting point.
-Nevertheless, let's continue with the explanation of <sf>xindy</sf>'s
-style file language.
-<P>
-<H2><A NAME="ss2.1">2.1 Running <SF>xindy</SF></A>
-</H2>
-
-<P>Create a new directory somewhere and copy some files from the
-distribution directory <CODE>doc/style-tutorial/</CODE> by typing
-<P>
-<BLOCKQUOTE><CODE>
-<PRE>
-eg$ mkdir tutorial
-eg$ cd tutorial
-eg$ cp &lt;distrib-dir>/doc/style-tutorial/*.raw .
-eg$ cp &lt;distrib-dir>/doc/style-tutorial/*.tex .
-</PRE>
-</CODE></BLOCKQUOTE>
-<P>with <CODE>distrib-dir</CODE> replaced by the actual location. Now create a
-file <CODE>style1.xdy</CODE> with the following content:
-<P>
-<BLOCKQUOTE><CODE>
-<PRE>
-;; This is a first example using `xindy'.
-
-(define-location-class "page-numbers" ("arabic-numbers"))
-(define-attributes (("definition" "usage")))
-</PRE>
-</CODE></BLOCKQUOTE>
-<P>Now run <SF>xindy</SF> by typing
-<P>
-<BLOCKQUOTE><CODE>
-<PRE>
-eg$ xindy -t ex1.xlg -M style1 -I xindy ex1.raw
-</PRE>
-</CODE></BLOCKQUOTE>
-<P>You should see something like
-<P>
-<BLOCKQUOTE><CODE>
-<PRE>
-Opening logfile "ex1.xlg" (done)
-Reading indexstyle...
-Loading module "style1.xdy"...
-Finished loading module "style1.xdy".
-Finished reading indexstyle.
-Finalizing indexstyle... (done)
-
-Reading raw-index "ex1.raw"...
-Finished reading raw-index.
-
-Processing index... [10%] [20%] [30%] [40%] [50%] [60%] [70%] [80%] [90%] [100%]
-Finished processing index.
-
-Writing markup... [10%] [20%] [30%] [40%] [50%] [60%] [70%] [80%] [90%] [100%]
-Markup written into file "ex1.ind".
-</PRE>
-</CODE></BLOCKQUOTE>
-<P><SF>xindy</SF> has now successfully compiled the index <CODE>ex1.raw</CODE> using
-your index style <CODE>style1.xdy</CODE>. The result is now stored in file
-<CODE>ex1.ind</CODE>. You can view this file but currently it only contains an
-unreadable mix of data.
-<P>But now let's come back to our index style. The syntax of the command
-is in a Lisp-like form with lots of braces, looking a little bit
-weird, but you'll soon get used to it. What is the meaning of the two
-commands we specified? The first command informed <SF>xindy</SF> that we
-like to process page numbers. We do this by defining a new
-<EM>location class</EM> named <CODE>page-numbers</CODE>. The page numbers consist
-of <CODE>arabic-numbers</CODE> as we might expect but this is not necessarily
-true---imagine your page numbers consisted of roman numerals instead.
-When reading the <EM>raw index</EM> contained in file <CODE>ex1.raw</CODE>
-<SF>xindy</SF> checks all locations if they match all known location
-classes. Since in our example the only known location class is the
-class of page numbers which are written using arabic digits, all
-locations will be checked if they are correct page numbers.
-<P>The second command tells <SF>xindy</SF> that we use two types of
-attributes for location references. Most often the locations in an
-index denote different meanings. For example, in mathematical texts
-one will distinguish the <EM>definition</EM> of a mathematical term from
-its <EM>usage</EM>. Sometimes these are typeset using different font
-shapes such as <EM>italic</EM> or font series such as <B>boldface</B>. Each
-location has an associated attribute which, if it is unspecified,
-defaults to the attribute <CODE>default</CODE>. With this command you have
-made these attributes known to the system, which makes it possible to
-assign different markup to these attributes later on.
-<P>
-<P>
-<H2><A NAME="ss2.2">2.2 Adding some Markup</A>
-</H2>
-
-<P>Until now you haven't seen something exciting, so its time to specify
-some markup. Add the following lines to our index style:
-<P>
-<BLOCKQUOTE><CODE>
-<PRE>
-(markup-index :open "~n\begin{theindex}~n"
- :close "~n\end{theindex}~n"
- :tree)
-
-(markup-locref :class "page-numbers" :attr "definition"
- :open "{\bf " :close "}")
-
-(markup-locclass-list :open "\quad{}")
-(markup-locref-list :sep ", ")
-</PRE>
-</CODE></BLOCKQUOTE>
-<P>Now run <SF>xindy</SF> again and afterwards LaTeX:
-<P>
-<BLOCKQUOTE><CODE>
-<PRE>
-eg$ xindy -t ex1.xlg -M style1 -I xindy ex1.raw
-eg$ latex ex1.tex
-</PRE>
-</CODE></BLOCKQUOTE>
-<P>You can view <CODE>ex1.dvi</CODE> with your prefered viewer (maybe <CODE>xdvi</CODE>
-or something else) to get a first impression of your results. Maybe
-your are not satisfied (for sure you aren't), because it still looks
-very confusing. What did the above rules tell <SF>xindy</SF>? When you
-view the file <CODE>ex1.ind</CODE> which is the result <SF>xindy</SF> generates,
-you'll recognize some of the <EM>markup tags</EM> you specified. The
-following is an excerpt of this file:
-<P>
-<BLOCKQUOTE><CODE>
-<PRE>
-\begin{theindex}
- academia\quad{}{\bf 1}acafetado\quad{}{\bf 2}acalmar\quad{}{\bf 4}
- açafrão\quad{}{\bf 3}indexflat\quad{}1hierarchical\quad{}2
- veryhierarchical\quad{}3impressive\quad{}4saber\quad{}{\bf 7}
- sabor\quad{}{\bf 8}sabão\quad{}{\bf 6}sábado\quad{}{\bf 5}
-\end{theindex}
-</PRE>
-</CODE></BLOCKQUOTE>
-<P>First of all you'll see that the file starts with the string
-<CODE>\begin{theindex}</CODE> and ends with <CODE>\end{theindex}</CODE>.
-Additionally some locations are correctly enclosed into a TeX macro
-that typesets them in shape boldface, whereas others aren't. The
-boldface ones are all those locations from the raw index that have the
-attribute <CODE>definition</CODE>.
-<P>The <CODE>:open</CODE> and <CODE>:close</CODE> keyword arguments each take a string as
-argument. The first one is written to the file when <EM>opening</EM> an
-enviroment, whereas the latter one <EM>closes</EM> an environment. What we
-have specified is the markup for the whole index (which is actually
-printed only once) and the markup for all locations of class
-<CODE>page-numbers</CODE> which own the attribute <CODE>definition</CODE>. Here we
-have cleanly separated the structured markup from the visual one,
-allowing an easy redefinition if we decide, for example, to markup the
-<CODE>definition</CODE>-locations in italics instead of boldface.
-<P>Some words on <EM>keyword arguments</EM> and <EM>switches</EM>. Keyword
-arguments such as <CODE>:open</CODE> or <CODE>:close</CODE> always take exactly one
-argument which must be positioned right after the keyword separated by
-a whitespace (a blank or a tab-stop). Switches don't take any
-arguments. For example, <CODE>:tree</CODE> in the command <CODE>markup-index</CODE> is
-a switch and thus it does not take an argument. We will use this
-terminology throughout the rest of this document.
-<P>The third command caused <SF>xindy</SF> to insert a horizontal space
-between the keyword and the locations (the TeX command
-<CODE>\quad{}</CODE> simply inserts a specific horizontal space). The last
-command caused <SF>xindy</SF> to separate all location references from
-each other with a comma followed by a blank, independently of any
-location class.
-<P>As you already may have observed, the tilde sign (<CODE>~</CODE>) serves
-as a <EM>quoting character</EM>.
-<P>We continue specifying markup to get a printable result by adding more
-markup:
-<P>
-<BLOCKQUOTE><CODE>
-<PRE>
-(markup-indexentry :open "~n \item " :depth 0)
-(markup-indexentry :open "~n \subitem " :depth 1)
-(markup-indexentry :open "~n \subsubitem " :depth 2)
-</PRE>
-</CODE></BLOCKQUOTE>
-<P>This assigns different markup for the different hierarchy layers of
-the indexentries. Our index is hierarchically organized. It contains
-items which themselves contain more sub-items which also might contain
-sub-sub-items. Each layer is started by a different markup which is
-correctly assigned with the <CODE>:depth</CODE> keyword argument. The layers
-are numbered by their <EM>depth</EM> starting from zero.
-<P>Now run <SF>xindy</SF> and TeX again and enjoy your first index. It's
-easy, isn't it! The output <CODE>ex1.ind</CODE> looks like the following:
-<P>
-<BLOCKQUOTE><CODE>
-<PRE>
-\begin{theindex}
-
- \item academia\quad{}{\bf 1}
- \item acafetado\quad{}{\bf 2}
- \item acalmar\quad{}{\bf 4}
- \item açafrão\quad{}{\bf 3}
- \item index
- \subitem flat\quad{}1
- \subitem hierarchical\quad{}2
- \subitem very
- \subsubitem hierarchical\quad{}3
- \subsubitem impressive\quad{}4
- \item saber\quad{}{\bf 7}
- \item sabor\quad{}{\bf 8}
- \item sabão\quad{}{\bf 6}
- \item sábado\quad{}{\bf 5}
-
-\end{theindex}
-</PRE>
-</CODE></BLOCKQUOTE>
-<P>Hmm, as you might have seen there are several problems that need
-further investigation. The index contains some Portuguese words that
-are printed correctly but should appear at other positions inside the
-index. For instance, the word <EM>sábado</EM> should appear before the
-word <EM>saber</EM> since <EM>á</EM> must be sorted as if it were simply an
-<EM>a</EM>. The reason why these words are not sorted correctly is
-simple---the accentuated letters have codes beyond position 128 in the
-ISO-Latin alphabet. Sorting based on these codes yields this incorrect
-order.
-<P>What to do? We can define for each of the words containing these
-special characters an explicit <EM>print key</EM>. The print key describes
-the printed representation of the keyword whereas the <EM>key</EM> or
-<EM>main key</EM> is used for sorting and merging. A very tedious task
-which is not a very clever solution since in a non-english language
-many many words contain these special cases. We follow the way
-<SF>xindy</SF> offers: <EM>keyword-mappings</EM>.
-<P>
-<H2><A NAME="ss2.3">2.3 Keyword Mappings</A>
-</H2>
-
-<P>What are keyword mappings for? A good question. I'll try to give some
-answers.
-<P>
-<UL>
-<LI> <EM>Merging of differently written words</EM>. Some text formatting
-systems allow different writings for the same word. For example, TeX
-can be used with ISO-Latin characters as well as with its own
-character sequences. If a document is composed from smaller ones
-possibly written by different authors using different forms of writing
-the index entries, the same terms may happen to be written differently
-and consequently we need a mechanism to identify them as equal.
-</LI>
-<LI> <EM>Specifying the sort order</EM>. As outlined in the previous
-section it is really difficult and error-prone to specify the sort key
-for each keyword explicitly. Sometimes the sort order is even
-different for the type of the document, as it happens in German, where
-two different types of sortings exist, one for everyday indexes and
-one for dictionaries. The sort order actually defines the position of
-arbitrary language-specific letters inside of an index.
-</LI>
-</UL>
-<P>A detailed elaboration of these ideas can be found in the paper <EM>An
-International Version of MakeIndex</EM> by Joachim Schrod [3].
-It describes the ideas that led to modifications on one of the
-ancestors of the <SF>xindy</SF> system: <CODE>makeindex</CODE> [4].
-<P>The keyword mappings are as follows. The <EM>merge key</EM> is generated
-from the <EM>main key</EM> with the so called <EM>merge mapping</EM>. The
-merge mapping can be specified with the command <CODE>merge-rule</CODE>. The
-<EM>sort key</EM> is derived from the merge key using the <EM>sort
-mapping</EM> specified with the <CODE>sort-rule</CODE> command. The following
-scheme shows this mapping process:
-<P>
-<FIGURE>
-<EPS FILE="mappings.eps">
-<IMG SRC="mappings.gif">
-</FIGURE>
-<P>
-<P>We will use this command now to define a suitable sort mapping that
-fits our needs:
-<P>
-<BLOCKQUOTE><CODE>
-<PRE>
-(sort-rule "à" "a")
-(sort-rule "á" "a")
-(sort-rule "ã" "a")
-(sort-rule "è" "e")
-(sort-rule "é" "e")
-(sort-rule "ç" "c")
-</PRE>
-</CODE></BLOCKQUOTE>
-<P>These rules define mappings from a keyword to a <EM>normalized</EM>
-version. In the logfile <CODE>ex1.xlg</CODE> these transformations are logged so
-that one can see how these mappings are performed. In this example we
-do not need any <CODE>merge-rule</CODE> but we will see applications in
-further examples.
-<P>Running <SF>xindy</SF> and TeXing the result now places the indexentries
-at the right positions.
-<P>In reality, such sort rules tend to be much more complex, due to
-the idiosynchrasies of sorting natural languages. Sort rules for
-many languages are available as part of the xindy distribution, as
-so-called language modules. We specify the language with the option
-<code>-L</code>, e.g., in our example we could have used
-<BLOCKQUOTE><CODE>
-<PRE>
-eg$ xindy -t ex1.xlg -M style1 -L portuguese -I xindy ex1.raw
-</PRE>
-</CODE></BLOCKQUOTE>
-<P>If we use one of the available language modules, sort rules in
-user-written xindy styles are ignored.
-<P>The result is now quite satisfying if the index entries weren't
-clumped together that much. We usually want the different index
-entries beginning with the same letter be optically separated from the
-ofhers. This improves readability and there must be a way to
-accomplish this---the <EM>letter groups</EM>.
-<P>
-<H2><A NAME="ss2.4">2.4 Letter Groups</A>
-</H2>
-
-<P>To group indexentries we must define what indexentries form a group.
-The clustering is done by matching the keywords' prefixes (taken from
-the <EM>sort key</EM>) with a user-defined table of prefixes and define
-appropriate markup that separates the groups from each other. Here it
-goes.
-<P>
-<BLOCKQUOTE><CODE>
-<PRE>
-(define-letter-groups
- ("a" "b" "c" "d" "e" "f" "g" "h" "i" "j" "k" "l" "m"
- "n" "o" "p" "q" "r" "s" "t" "u" "v" "w" "x" "y" "z"))
-
-(markup-letter-group-list :sep "~n\indexspace")
-</PRE>
-</CODE></BLOCKQUOTE>
-<P>This defines the given list of letter groups. When forming the letter
-groups, each letter group is checked if it matches a prefix of the
-indexentries' sort key. The longest match assigns the index entry to
-this letter group. If no match was possible the index entry is put into
-group <CODE>default</CODE>.
-<P>The result now looks much better than before. You have now learned the
-basic features that you need to specify everyday indexes. In the next
-chapter we'll continue to make you an expert in indexing.
-<P>
-<P>
-<HR>
-<A HREF="style-tutorial-3.html">Next</A>
-<A HREF="style-tutorial-1.html">Previous</A>
-<A HREF="style-tutorial.html#toc2">Contents</A>
-</BODY>
-</HTML>