[Epydoc-commits] SF.net SVN: epydoc: [1808] trunk/epydoc/src/epydoc/docwriter/latex_sty.py

[<ed...@us...>]

Revision: 1808
          http://epydoc.svn.sourceforge.net/epydoc/?rev=1808&view=rev
Author:   edloper
Date:     2008-03-05 10:40:33 -0800 (Wed, 05 Mar 2008)

Log Message:
-----------
- Use \raggedright for the subclass list
- Use utf8x encoding, instead of utf8
- \EpydocDottedName now expects its argument to use escaped characters
  (underscore in particular.)  E.g., \EpydocDottedName{foo\_bar}.  It
  no longer uses the latex url package.  This makes it work better with
  the hyperref package.
- Misc changes to show_latex_warnings().

Modified Paths:
--------------
    trunk/epydoc/src/epydoc/docwriter/latex_sty.py

Modified: trunk/epydoc/src/epydoc/docwriter/latex_sty.py
===================================================================
--- trunk/epydoc/src/epydoc/docwriter/latex_sty.py	2008-03-04 02:32:58 UTC (rev 1807)
+++ trunk/epydoc/src/epydoc/docwriter/latex_sty.py	2008-03-05 18:40:33 UTC (rev 1808)
@@ -75,7 +75,6 @@
 \RequirePackage[headings]{fullpage}
 \RequirePackage[usenames]{color}
 \RequirePackage{graphicx}
-\RequirePackage{url}
 
 \@ifclassloaded{memoir}{%
     \RequirePackage[other,notbib]{tocbibind}
@@ -107,6 +106,9 @@
 % ======================================================================
 % General Formatting
 
+% Setting this will make us less likely to get overfull hboxes:
+%\setlength{\emergencystretch}{1in}
+
 % Separate paragraphs by a blank line (do not indent them).  We define
 % a new length variable. \EpydocParskip, for the paragraph-skip length.
 % This needs to be assigned to \parskip here as well as inside several
@@ -566,23 +568,28 @@
 % This section defines the EpydocDottedName command, which is used to
 % display the names of Python objects.
 
-% The \EpydocDottedName command is used to escape dotted names.  In
-% particular, it escapes all characters except '#', '%', and newlines,
-% and allows non-hyphenated wrapping at '.' separator characters.
-% [xx] this generates a warning for names containing _ -- we need
-% to somehow strip that out of the second argument.
-\newcommand\EpydocDottedName[1]{%
-    \texorpdfstring{\protect\Epydoc@DottedName{#1}}{#1}}
+% The \EpydocDottedName command adds a possible break-point after every
+% period in the given string.  It no longer escapes characters such as
+% underscore, since this interfered with the hyperref package; instead,
+% epydoc is responsible for escaping them.  E.g., correct usage for a
+% name with an underscore is \EpydocDottedName{some\_name}.
+\newcommand\EpydocDottedName[1]{
+  \Epydoc@DottedName #1.@}
 
-\DeclareUrlCommand\Epydoc@DottedName{%
-  \urlstyle{rm}%
-  \def\UrlBigBreaks{\do\.}%
-  % We should never see most of these chars in a name, but handle
-  % them if they happen to show up (eg script names)
-  \def\UrlOrds{\do\@\do\#\do\%\do\^\do\&\do\(\do\)\do\-\do\_%
-               \do\|\do\\\do\"\do\/\do\?\do\>\do\<\do\{\do\}%
-               \do\`\do\~}
-}
+% This helper function performs the work of \EpydocDottedName. It
+% scans forward, looking for a period, and putting all text up to the
+% period into #1.  The single character after the period is put in
+% #2.  This function then checks if #2 is '@', indicating that we're
+% done, or some other character, indicating that we should continue.
+% Note that \ifx tests character codes; and that when '@' appears
+% in user code, it gets the character code 'other', but when it
+% appears here, it gets the character code 'letter'.
+\def\Epydoc@DottedName#1.#2{%
+  \ifx#2@\relax #1\else 
+    #1\discretionary{.}{}{.}%
+    \expandafter\expandafter\expandafter\Epydoc@DottedName
+    \expandafter #2\fi%
+  }
 
 """+NIST_DISCLAIMER
 ######################################################################


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.