From 52dc76c81fffa709fae35af92538723d23ad18d6 Mon Sep 17 00:00:00 2001 From: Fred Drake Date: Mon, 3 Jul 2000 09:56:23 +0000 Subject: Eric S. Raymond : This patch implements relative-path semantics for the "source" facility resembling those of cpp(1), documents the change, and improves the shlex test main to make it easier to test this feature. Along the way, it fixes a name error in the existing docs. [Additional documentation markup changes for consistency by FLD.] --- Doc/lib/libshlex.tex | 85 ++++++++++++++++++++++++++++++---------------------- Lib/shlex.py | 30 +++++++++++++------ 2 files changed, 70 insertions(+), 45 deletions(-) diff --git a/Doc/lib/libshlex.tex b/Doc/lib/libshlex.tex index a070908..f1eea3a 100644 --- a/Doc/lib/libshlex.tex +++ b/Doc/lib/libshlex.tex @@ -13,7 +13,7 @@ simple syntaxes resembling that of the \UNIX{} shell. This will often be useful for writing minilanguages, e.g.\ in run control files for Python applications. -\begin{classdesc}{shlex}{\optional{stream}, \optional{file}} +\begin{classdesc}{shlex}{\optional{stream\optional{, file}}} A \class{shlex} instance or subclass instance is a lexical analyzer object. The initialization argument, if present, specifies where to read characters from. It must be a file- or stream-like object with @@ -35,6 +35,7 @@ equal to \code{sys.stdin}, this second argument defauilts to ``stdin''. A \class{shlex} instance has the following methods: + \begin{methoddesc}{get_token}{} Return a token. If tokens have been stacked using \method{push_token()}, pop a token off the stack. Otherwise, read one @@ -52,35 +53,45 @@ requests. (This is not ordinarily a useful entry point, and is documented here only for the sake of completeness.) \end{methoddesc} -\begin{methoddesc}{openhook}{filename} -When shlex detects a source request (see \member{source} below) -this method is given the following token as argument, and expected to -return a tuple consisting of a filename and an opened stream object. - -Normally, this method just strips any quotes off the argument and -treats it as a filename, calling \code{open()} on it. It is exposed so that -you can use it to implement directory search paths, addition of -file extensions, and other namespace hacks. - +\begin{methoddesc}{sourcehook}{filename} +When \class{shlex} detects a source request (see +\member{source} below) this method is given the following token as +argument, and expected to return a tuple consisting of a filename and +an open file-like object. + +Normally, this method first strips any quotes off the argument. If +the result is an absolute pathname, or there was no previous source +request in effect, or the previous source was a stream +(e.g. \code{sys.stdin}), the result is left alone. Otherwise, if the +result is a relative pathname, the directory part of the name of the +file immediately before it on the source inclusion stack is prepended +(this behavior is like the way the C preprocessor handles +\code{\#include "file.h"}). The result of the manipulations is treated +as a filename, and returned as the first component of the tuple, with +\function{open()} called on it to yield the second component. + +This hook is exposed so that you can use it to implement directory +search paths, addition of file extensions, and other namespace hacks. There is no corresponding `close' hook, but a shlex instance will call -the \code{close()} method of the sourced input stream when it returns EOF. +the \method{close()} method of the sourced input stream when it +returns \EOF. \end{methoddesc} -\begin{methoddesc}{error_leader}{\optional{file}, \optional{line}} +\begin{methoddesc}{error_leader}{\optional{file\optional{, line}}} This method generates an error message leader in the format of a -Unix C compiler error label; the format is '"\%s", line \%d: ', -where the \%s is replaced with the name of the current source file and -the \%d with the current input line number (the optional arguments -can be used to override these). - -This convenience is provided to encourage shlex users to generate -error messages in the standard, parseable format understood by Emacs -and other Unix tools. +\UNIX{} C compiler error label; the format is '"\%s", line \%d: ', +where the \samp{\%s} is replaced with the name of the current source +file and the \samp{\%d} with the current input line number (the +optional arguments can be used to override these). + +This convenience is provided to encourage \module{shlex} users to +generate error messages in the standard, parseable format understood +by Emacs and other \UNIX{} tools. \end{methoddesc} Instances of \class{shlex} subclasses have some public instance -variables which either control lexical analysis or can be used -for debugging: +variables which either control lexical analysis or can be used for +debugging: \begin{memberdesc}{commenters} The string of characters that are recognized as comment beginners. @@ -90,7 +101,7 @@ Includes just \character{\#} by default. \begin{memberdesc}{wordchars} The string of characters that will accumulate into multi-character -tokens. By default, includes all \ASCII{} alphanumerics and +tokens. By default, includes all \ASCII{} alphanumerics and underscore. \end{memberdesc} @@ -114,24 +125,26 @@ be useful to examine this when constructing error messages. \end{memberdesc} \begin{memberdesc}{instream} -The input stream from which this shlex instance is reading characters. +The input stream from which this \class{shlex} instance is reading +characters. \end{memberdesc} \begin{memberdesc}{source} -This member is None by default. If you assign a string to it, that -string will be recognized as a lexical-level inclusion request similar -to the `source' keyword in various shells. That is, the immediately -following token will opened as a filename and input taken from that -stream until EOF, at which point the \code{close()} method of that -stream will be called and the input source will again become the -original input stream. Source requests may be stacked any number of -levels deep. +This member is \code{None} by default. If you assign a string to it, +that string will be recognized as a lexical-level inclusion request +similar to the \samp{source} keyword in various shells. That is, the +immediately following token will opened as a filename and input taken +from that stream until \EOF, at which point the \method{close()} +method of that stream will be called and the input source will again +become the original input stream. Source requests may be stacked any +number of levels deep. \end{memberdesc} \begin{memberdesc}{debug} -If this member is numeric and 1 or more, a shlex instance will print -verbose progress output on its behavior. If you need to use this, -you can read the module source code to learn the details. +If this member is numeric and \code{1} or more, a \class{shlex} +instance will print verbose progress output on its behavior. If you +need to use this, you can read the module source code to learn the +details. \end{memberdesc} Note that any character not declared to be a word character, diff --git a/Lib/shlex.py b/Lib/shlex.py index 575fc35..876dcdf 100644 --- a/Lib/shlex.py +++ b/Lib/shlex.py @@ -3,8 +3,10 @@ # Module and documentation by Eric S. Raymond, 21 Dec 1998 # Input stacking and error message cleanup added by ESR, March 2000 +import os.path import sys + class shlex: "A lexical analyzer class for simple shell-like syntaxes." def __init__(self, instream=None, infile=None): @@ -26,7 +28,8 @@ class shlex: self.filestack = [] self.source = None if self.debug: - print 'shlex: reading from %s, line %d' % (self.instream,self.lineno) + print 'shlex: reading from %s, line %d' \ + % (self.instream, self.lineno) def push_token(self, tok): "Push a token onto the stack popped by the get_token method" @@ -47,7 +50,7 @@ class shlex: # Handle inclusions while raw == self.source: (newfile, newstream) = self.sourcehook(self.read_token()) - self.filestack = [(self.infile,self.instream,self.lineno)] + self.filestack + self.filestack.insert(0, (self.infile, self.instream, self.lineno)) self.infile = newfile self.instream = newstream self.lineno = 1 @@ -63,7 +66,8 @@ class shlex: (self.infile, self.instream, self.lineno) = self.filestack[0] self.filestack = self.filestack[1:] if self.debug: - print 'shlex: popping to %s, line %d' % (self.instream, self.lineno) + print 'shlex: popping to %s, line %d' \ + % (self.instream, self.lineno) self.state = ' ' raw = self.get_token() # Neither inclusion nor EOF @@ -82,7 +86,8 @@ class shlex: if nextchar == '\n': self.lineno = self.lineno + 1 if self.debug >= 3: - print "shlex: in state " + repr(self.state) + " I see character: " + repr(nextchar) + print "shlex: in state", repr(self.state), \ + "I see character:", repr(nextchar) if self.state == None: self.token = ''; # past end of file break @@ -156,6 +161,9 @@ class shlex: "Hook called on a filename to be sourced." if newfile[0] == '"': newfile = newfile[1:-1] + # This implements cpp-like semantics for relative-path inclusion. + if type(self.infile) == type("") and not os.path.isabs(newfile): + newfile = os.path.join(os.path.dirname(self.infile), newfile) return (newfile, open(newfile, "r")) def error_leader(self, infile=None, lineno=None): @@ -166,12 +174,16 @@ class shlex: lineno = self.lineno return "\"%s\", line %d: " % (infile, lineno) -if __name__ == '__main__': - lexer = shlex() +if __name__ == '__main__': + if len(sys.argv) == 1: + lexer = shlex() + else: + file = sys.argv[1] + lexer = shlex(open(file), file) while 1: tt = lexer.get_token() - print "Token: " + repr(tt) - if not tt: + if tt: + print "Token: " + repr(tt) + else: break - -- cgit v0.12