Rewritten a bit to address some criticism in the newsgroup.
diff --git a/Doc/lib/libgzip.tex b/Doc/lib/libgzip.tex
index fd2bc28..f649a4d 100644
--- a/Doc/lib/libgzip.tex
+++ b/Doc/lib/libgzip.tex
@@ -9,39 +9,52 @@
compressing or decompressing the data so it looks like an ordinary
file object.
-\class{GzipFile} objects simulate most of the methods of a file
-object, though it's not possible to use the \method{seek()} and
-\method{tell()} methods to access the file randomly.
-\withsubitem{(class in gzip)}{\ttindex{GzipFile}}
+The module defines the following items:
+\begin{classdesc}{GzipFile}{\optional{filename\optional{, mode\optional{,
+ compresslevel\optional{, fileobj}}}}}
+Constructor for the \class{GzipFile} class, which simulates most of
+the methods of a file object, with the exception of the
+\method{seek()} and \method{tell()} methods. At least one of
+\var{fileobj} and \var{filename} must be given a non-trivial value.
-\begin{funcdesc}{open}{fileobj\optional{, filename\optional{,
- mode\optional{, compresslevel}}}}
- Returns a new \class{GzipFile} object on top of \var{fileobj}, which
- can be a regular file, a \class{StringIO} object, or any object which
- simulates a file.
+The new class instance is based on \var{fileobj}, which can be a
+regular file, a \class{StringIO} object, or any other object which
+simulates a file. It defaults to \code{None}, in which case
+\var{filename} is opened to provide a file object.
- The \program{gzip} file format includes the original filename of the
- uncompressed file; when opening a \class{GzipFile} object for
- writing, it can be set by the \var{filename} argument. The default
- value is an empty string.
+When \var{fileobj} is not \code{None}, the \var{filename} argument is
+only used to be included in the \program{gzip} file header, which may
+includes the original filename of the uncompressed file. It defaults
+to the filename of \var{fileobj}, if discernible; otherwise, it
+defaults to the empty string, and in this case the original filename
+is not included in the header.
- \var{mode} can be either \code{'r'} or \code{'w'} depending on
- whether the file will be read or written. \var{compresslevel} is an
- integer from \code{1} to \code{9} controlling the level of
- compression; \code{1} is fastest and produces the least compression,
- and \code{9} is slowest and produces the most compression. The
- default value of \var{compresslevel} is \code{9}.
+The \var{mode} argument can be either \code{'r'} or \code{'w'},
+depending on whether the file will be read or written. The default is
+the mode of \var{fileobj} if discernible; otherwise, the default is
+\code{'r'}.
- Calling a \class{GzipFile} object's \method{close()} method does not
- close \var{fileobj}, since you might wish to append more material
- after the compressed data. This also allows you to pass a
- \class{StringIO} object opened for writing as \var{fileobj}, and
- retrieve the resulting memory buffer using the \class{StringIO}
- object's \method{getvalue()} method.
+The \var{compresslevel} argument is an integer from \code{1} to
+\code{9} controlling the level of compression; \code{1} is fastest and
+produces the least compression, and \code{9} is slowest and produces
+the most compression. The default is \code{9}.
+
+Calling a \class{GzipFile} object's \method{close()} method does not
+close \var{fileobj}, since you might wish to append more material
+after the compressed data. This also allows you to pass a
+\class{StringIO} object opened for writing as \var{fileobj}, and
+retrieve the resulting memory buffer using the \class{StringIO}
+object's \method{getvalue()} method.
+\end{classdesc}
+
+\begin{funcdesc}{open}{filename\optional{, mode\optional{, compresslevel}}}
+This is a shorthand for \code{GzipFile(\var{filename},}
+\code{\var{mode},} \code{\var{compresslevel})}. The \var{filename}
+argument is required; \var{mode} defaults to \code{'r'} and
+\var{compresslevel} defaults to \code{9}.
\end{funcdesc}
\begin{seealso}
\seemodule{zlib}{the basic data compression module}
\end{seealso}
-