discuss: Thread: Plain text documents

These two quotes were fro the "History of LDP" thread
> David Lawyer wrote
> I agree but one nice feature about linuxdoc (seldom used anymore) or
> wikis is that they automatically create a table of contents.  If there
> is no such table in a doc, one can as a substitute type in words to
> search the doc but a table of contents is sometimes easier to use to
> find what you are looking for (or for something of interest you were
> not looking for):.
>   
> On 2/4/2016 10:46 AM, David Niklas wrote:
> Many people create their own TOC in txt docs, some of them IMHO are very
> pretty :)
>
> Sincerely, David
I don't understand.  You likely mean that the created TOC's contain more 
that just the section headings and numbering.  But how does one get to a 
topic found in the TOC?  In html one just clicks on it but not for a 
text doc.  If the doc is very short (a screen or two) then no TOC is 
needed (except for smartphones with small screens).

One proposal would be to require (or strongly suggest) a minimal markup 
for text docs submitted to LDP.  They would need to use 4 tags from 
linuxdoc markup and separate paragraphs by blank lines. The tags:
<title>Serial-Howto
<author>David S. Lawyer
<date>Nov. 2011

<sect>Introduction

Optional would the <sect1> tag for subsections.  These submissions would 
be converted by a script to turn them into valid linuxdoc format with a 
table of contents.  It's very easy to create such a script just by using 
the vim editor (in record mode).  To make this valid linuxdoc: Add first 
line: <!doctype linudoc system> and then enclose the rest of the article 
in <article> tags.  It's now valid linuxdoc.  The script would also add 
a <TOC> (Table of Contents).  The doc would now be converted by linuxdoc 
to html format, etc.  So it would be not really a text documents.

The simplest idea would jest be to keep it a .txt file and not convert 
it to linuxdoc.  The problem with the above proposal is that it doesn't 
support lists or tables
||


    |

    .

    |

> On Tue, Feb 09, 2016 at 01:35:43AM -0800, David Lawyer wrote:
>These two quotes were fro the "History of LDP" thread
>> David Lawyer wrote
>> I agree but one nice feature about linuxdoc (seldom used anymore) or
>> wikis is that they automatically create a table of contents.  If there
>> is no such table in a doc, one can as a substitute type in words to
>> search the doc but a table of contents is sometimes easier to use to
>> find what you are looking for (or for something of interest you were
>> not looking for):.
>>   
>> On 2/4/2016 10:46 AM, David Niklas wrote:
>> Many people create their own TOC in txt docs, some of them IMHO are very
>> pretty :)
>>
>> Sincerely, David

>The simplest idea would jest be to keep it a .txt file and not convert it to 
>linuxdoc.  The problem with the above proposal is that it doesn't support 
>lists or tables

Ditto. Keep it simple, or adhere to standards first.

Publish a simple ascii/txt file.

What I usually see done is an HTML file converted backwards to a simple markup 
text.  Headings and sub-headings can be easily found within HTML files versus 
text only files.  So running a script on an HTML file to back-port the file to 
text only, but with simple mark-up headings for table of contents.  Similar to 
what Amazon's Kindlegen performs when converting HTML to Mobi/PRC files, albeit 
proprietary.  I would think something simlar can be easily performed within 
Linux (and likely already can be performed) using a simple script (ie.  
AWK/GAWK) for converting HTML to a simplified text only with some markup 
commands.

-- 
Roger
http://rogerx.freeshell.org/

> > On Tue, Feb 09, 2016 at 01:35:43AM -0800, David Lawyer wrote:
> >The simplest idea would jest be to keep it a .txt file and not convert it to 
> >linuxdoc.  The problem with the above proposal is that it doesn't support 
> >lists or tables
> 
On Tue, Feb 09, 2016 at 02:28:18PM -0500, Roger wrote:
> Ditto. Keep it simple, or adhere to standards first.
> 
> Publish a simple ascii/txt file.
> 
> What I usually see done is an HTML file converted backwards to a simple markup 
> text.  Headings and sub-headings can be easily found within HTML files versus 
> text only files.  So running a script on an HTML file to back-port the file to 
> text only, but with simple mark-up headings for table of contents.  Similar to 
> what Amazon's Kindlegen performs when converting HTML to Mobi/PRC files, albeit 
> proprietary.  I would think something similar can be easily performed within 
> Linux (and likely already can be performed) using a simple script (ie.  
> AWK/GAWK) for converting HTML to a simplified text only with some markup 
> commands.

Both linuxdoc and docbook can output plain text files and Linux has
programs to convert html to text.  Also, most browsers can do this too.
But for LDP, we're talking about cases where the author writes in plain
text.  Does any smart converter exist for converting that to html.  It
would need to detect lists, tables, table of contents etc. all of which
may be in various styles in a text document and convert them to html, etc.
Does such software exist?  Well, there's txt2html which claims it can
convert lists and most tables.  It seems that it actually requires a
little markup since an example shows underlined headings in the text doc.
to be converted.  Perhaps I'll try it out on one of my txt docs. 
			David Lawyer

On Tue, 9 Feb 2016 01:35:43 -0800
David Lawyer ####@####.#### wrote:

> These two quotes were fro the "History of LDP" thread
> > David Lawyer wrote
> > I agree but one nice feature about linuxdoc (seldom used anymore) or
> > wikis is that they automatically create a table of contents.  If there
> > is no such table in a doc, one can as a substitute type in words to
> > search the doc but a table of contents is sometimes easier to use to
> > find what you are looking for (or for something of interest you were
> > not looking for):.
> >   
> > On 2/4/2016 10:46 AM, David Niklas wrote:
> > Many people create their own TOC in txt docs, some of them IMHO are
> > very pretty :)
> >
> > Sincerely, David
> I don't understand.  You likely mean that the created TOC's contain
> more that just the section headings and numbering.  But how does one
> get to a topic found in the TOC?  In html one just clicks on it but not
> for a text doc.  If the doc is very short (a screen or two) then no TOC
> is needed (except for smartphones with small screens).
> 
<snip>

Your going to kick yourself.
1. Open the find dialogue.
2. Type in the section number. If that does not exist the section name
  will do.
3. Hit find.
Viola!

Sincerely, David

On Wednesday, February 17, 2016 06:16:02 PM you wrote:
> Your going to kick yourself.
> 1. Open the find dialogue.
> 2. Type in the section number. If that does not exist the section name
>   will do.
> 3. Hit find.

I'm not going to kick myself ;-)  That is quite a bit more than a simple click 
(even if you use the mouse to copy and paste the section number or name)--
plus, if that section number (or name) appears in multiple locations in the 
document.

True, it is not the end of the world, and I will read plain text documents if 
they (appear to) have the information for which I'm looking.

> Viola!

She has to be the most well known female in the world ;-)