[plug] text processing

Mon Dec 9 20:53:27 WST 2002

The problem was an extra extraneous dollar sign.  From the command line;

[alan at laptop alan]$ mkdir test
[alan at laptop alan]$ cd test
[alan at laptop test]$ mkdir doc 
[alan at laptop test]$ mkdir html
[alan at laptop test]$ touch doc/one.doc
[alan at laptop test]$ touch doc/two.doc
[alan at laptop test]$ for file in $(ls ./doc);do
> file_out=${file%doc}html
> cp ./doc/$file ./html/$file_out    
> done
[alan at laptop test]$ ls -l html
total 0
-rw-rw-r--    1 alan     alan            0 Dec  9 20:50 one.html
-rw-rw-r--    1 alan     alan            0 Dec  9 20:50 two.html

BTW, by convention,  the variables should be in upper case, ie, $FILE.

HTH

Alan

On Mon, 2002-12-09 at 20:42, Alan Graham wrote:
> Sorry bout that.  There were two things wrong with my solution;
> extraneous $ signs, 
> 
> for $file in $(ls ../docs/*doc)
> 
> should be
> for file in $(ls ../docs/*doc)
> 
> and bash doesn't support 
> 
> file_out=${$file%doc}html
> 
> which is valid in korn shell.  Sorry.  I'll have a quick look to see if
> I can find out how to do it in bash.
> 
> Alan
> 
> On Mon, 2002-12-09 at 17:39, Sol wrote:
> > 
> > 
> > ----------  Forwarded Message  ----------
> > 
> > Subject: Re: [plug] text processing
> > Date: Mon, 09 Dec 2002 17:38:18 +0800
> > From: Sol <sol at autonomon.net>
> > To: plug at plug.linux.org.au
> > 
> > <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
> > <html>
> > <head>
> >   <title></title>
> > </head>
> > <body>
> > Thanks for your replies Ryan and Graham,<br>
> > <br>
> > I've tried both out with no luck yet. Here's a what I did and what went
> >  wrong:<br>
> >  -----------------------------------------------------------------<br> #First
> >  with Ryan's tip<br>
> > <br>
> > #!/bin/sh<br>
> > <br>
> > DOC_DIR=/home/sol/bswa_publications/DOC/<br>
> > HTML_DIR=/home/sol/bswa_publications/HTML/<br>
> > <br>
> > for FILE in `ls ${DOC_DIR}`<br>
> > do<br>
> >     wvText  ${FILE} | txt2html >
> >  ${HTML_DIR}${FILE}<br> done<br>
> > <br>
> > ## What I got here was a bunch of files in ../HTML/ with exactly the same
> > names as in ../DOC/. They contained the bare minimum amount of HTML elements
> > and none of the text like this:<br>
> > <br>
> > <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN"><br>
> > <HTML><br>
> > <HEAD><br>
> > <TITLE></TITLE><br>
> > <META NAME="generator" CONTENT="txt2html v1.28"><br>
> > </HEAD><br>
> > <BODY><br>
> > <P><br>
> > Usage: /usr/bin/wvText &lt;word document&gt; &lt;text output
> > file&gt;<br>
> > <br>
> > </BODY><br>
> > </HTML><br>
> > <br>
> > I've tried doing "wvText < ${FILE}" which gives me a different error but
> > passed no files so I guess that's a backward step.<br>
> > -----------------------------------------------------------------------<br>
> > <br>
> > -----------------------------------------------------------------------<br>
> > #Now Graham's<br>
> > <br>
> > #!/bin/ksh<br>
> > <br>
> > for $file in $(/home/sol/bswa_publications/DOC/*doc)<br>
> > do<br>
> >     $file_out=${$file%doc}html<br>
> >     wvText < /home/sol/bswa_publications/DOC.$file |
> >  txt2html > /home/sol/bswa_publications/$file_out<br> done<br>
> > <br>
> > ##I got the error<br>
> > DOCtoHTML.ksh: `$file': not a valid identifier<br>
> > <br>
> > I don't know any ksh at all so I don't even know which $file is causing the
> > problem.<br>
> > -----------------------------------------------------------<br>
> > It's interesting to see different solutions for the same simple problem.
> > Ta.<br>
> > <br>
> > sol<br>
> > <br>
> > <br>
> > <br>
> > <br>
> > Graham, Alan A. wrote:<br>
> > <blockquote type="cite"
> >  cite="mid41D499B16A81D4118D2500805F0DD05C084E5AE0 at perm03.woodside.com.au">
> >   <pre wrap="">!#/bin/ksh
> > # I can't test this cos I'm on an NT box :-(
> > # I know korn shell, but I understand bash can run ksh scripts
> > #
> > 
> > for $file in $(ls ../docs/*doc)
> > do
> > 	$file_out=${$file%doc}html
> > 	wvText < ../docs.$file | txt2html > ../html/$file_out
> > done
> > 
> > I can't remember if the output from straight ls looks like ../doc/one.doc or
> > just one.doc.  The code assumes one.doc.  And I say again, I can't test this
> > goes I don't have access to a real OS at this site.
> > 
> > Alan
> > 
> > 
> >   </pre>
> >   <blockquote type="cite">
> >     <pre wrap="">-----Original Message-----
> > From:	Sol [<a class="moz-txt-link-abbreviated"
> >  href="mailto:SMTP:sol at autonomon.net">SMTP:sol at autonomon.net</a>]
> >  Sent:	Monday, 9 December 2002 16:24
> > To:	<a class="moz-txt-link-abbreviated"
> >  href="mailto:plug at plug.linux.org.au">plug at plug.linux.org.au</a>
> >  Subject:	[plug] text processing
> > 
> > Hi PLUG,
> > 
> > I have a bunch of M$ Word files that in a directory that I want to clean
> > up
> > and output to another empty directory as HTML. I've been doing everything
> > the
> > slow way using command line tools, but I'm sure that it can all be done
> > with
> > a single command. I'm using wvText and txt2html. I want to pipe all the
> > files
> > in the directory in order through wvText into the empty directory and then
> > 
> > pass everything in that directory through txt2html.
> > 
> > If I've got these files: one.doc, two.doc and three.doc in ../docs/ and
> > want
> > them to end up as HTML in ../html/ as one.html, two.html and three.html,
> > how
> > can I do this with a single command?
> > 
> > Thanks,
> > sol
> > 
> > 
> > 
> > --
> > This email was received from the Internet.  If this email is unsolicited,
> > non-business related, inappropriate or spam, please forward it to
> > <a class="moz-txt-link-abbreviated"
> >  href="mailto:spamfilter at woodside.com.au">spamfilter at woodside.com.au</a>
> >  </pre>
> >   </blockquote>
> >   <pre wrap=""><!---->
> >   </pre>
> > </blockquote>
> > <br>
> > <br>
> > </body>
> > </html>
> > 
> > -------------------------------------------------------
> > 
> 
> 