Txt2tags User Guide

Aurelio, Fri Dec 20 15:01:52 2002

About this document

"Hi! I'm the txt2tags manual document.

Here you'll find all available information about the txt2tags text conversion tool.

My updated version can be found at http://txt2tags.sf.net/userguide/

For more informations and recent releases, please visit the txt2tags website.


About txt2tags

This chapter is a txt2tags overview, that will introduce the program purpose and features.

What is it?

Txt2tags is a text formatting and conversion tool.

Txt2tags converts a plain text file with little marks, to any of the supported targets:

Why should I use it?

You'll find txt2tags really useful if you:

And the main purpose motivation is:

Why it is a good choice among other tools?

Txt2tags has a very straight way of growing, following basic concepts. These are the highlights:

Source file readable Txt2tags marks are very simplistic, almost natural.
Target document readable As the source file, the target document is readable also, with indentation and short lines.
Marks consistent Txt2tags marks are unique enough to fit at all kind of documents and don't be confused with the document contents.
Rules consistent As the marks, the rules that applies to them are tied to each other, there are no "exceptions" or "special cases".
Simple structures All the supported formatting is simple, with no extra-options or complicated behaviour modifiers. A mark is just a mark, with no options at all.
Easy to learn With simple marks and source readable, the txt2tags learning curve is user friendly.
Nice examples The sample files included on the package gives real life examples of simple and over-complicated documents written on the txt2tags format.
Valuable Tools The syntax files included on the package (for vim and emacs editors) help you to write documents with no syntax errors.
Three user interfaces There is a Graphical Tk interface that is very user friendly, a Web interface to use it remotely or on the intranet, and a Command Line interface for powerusers and scripting.
Scripting With the full featured comand line mode, an experienced user can automatize tasks and do post-editting on the converted files.
Download and run / Multi-platform Txt2tags is a single Python script. There is no need to compile it or download extra modules. So it runs nicely on *NIX, Linux, Windows and Macintosh machines.
Frequent Updates The program has a mailing list with active users who suggest corrections and improvements. The author himself is an extensive user at home and at work, so the development won't stop briefly.

I have to pay for it?

Absolutely NO!

It's free, GPL, open source, public domain, <put-your-favorite-buzzword-here>.

You can copy, use, modify, sell, release as yours. Software politics/copyright is not one of the author's major concern.

Detailed info about txt2tags

On this section the program features will be seen in a detailed form, solving the doubts you may have about it.

Supported Formatting Structures

The following is a list of all the structures supported by txt2tags.

Supported Targets

It is a common document format which has powerful sgmltools conversion applications. From a single sgml file you can generate html, pdf, ps, info, latex, lyx, rtf and xml documents. The sgml2* tools also does automatic TOC and break sections into subpages (sgml2html).

txt2regex generates SGML files in the linuxdoc system type, ready to be converted with sgml2* tools without any extra catalog files or any SGML annoying requirements.

Everybody knows what HTML is. (hint: internet)

txt2regex generates clean HTML documents, that look pretty and have its source readable. It DOES NOT use CSS, javascript, frames or other futile formatting techniques, that aren't required for simple, techie documents.


I guess you didn't know, but Adobe PageMaker 6.0 has its own tagged language! You can define styles, colortable, beautifiers, and most of all the PageMaker mouse-clicking features are available on its tagged language also. You just need to access "Import tagged text" menu item. Just for the records, it's an <HTML "like"> tag format.

txt2regex generates all the tags and already defines a extensive and working header, setting paragraph styles and formatting. This is the hard part. GOTCHA: No line breaks! A paragraph must be one single line.

Author's note: My entire portuguese regular expression book was written in vi, converted to PageMaker with txt2tags and went to press.

Magic Point is a very handy presentation tool (hint: Microsoft PowerPoint), that uses a tagged language to define all the screens. So you can do complex presentations in vi/emacs/notepad.

txt2tags generates a ready-to-use .mgp file, defining all the necessary headers for fonts and appearence definitions, as long as ISO-8859 accents support.

HOTSPOT 1: txt2tags created .mgp file uses the XFree86 Type1 fonts! So you do not need to carry TrueType fonts files with your presentation.

HOTSPOT 2: the color definitions for fonts are clean, so even on a poor color palette system (as startx -- -bpp 8), the presentation will look pretty!

The key is: convert and use. No adaptation or requirements needed.

UNIX man pages resist over the years. Document formats come and go, and there they are, unbeatable.

There are other tools to generate man documents, but the txt2tags has one advantage: one source, multi targets. so the same man page contents can be converted as HTML page, Magic Point presentation, etc.

You don't know what MoinMoin is? It is a WikiWiki!

Moin syntax is kinda boring when you need to keep {{{'''''adding braces and quotes'''''}}}, so txt2tags comes with the simplified marks and unified solution: one source, multi targets.

TXT is text. The only true formatting type.

Besides txt2tags marks are very intuitive and discrete, you can remove them by converting the file to pure TXT.

The titles are underlined, and the text is basicaly left as is on the source.

Target status for supported structures

structure txt html sgml tex mgp pm6 moin man
headers Y Y Y Y Y N N Y
section title Y Y Y Y Y Y Y Y
paragraphs Y Y Y Y Y Y Y Y
bold - Y Y Y Y Y Y Y
italic - Y Y Y Y Y Y Y
bold-italic - Y Y Y Y Y Y Y
underline - Y - Y Y Y ? -
preformatted - Y Y Y Y Y Y -
preformatted line - Y Y Y Y Y Y Y
preformatted area - Y Y Y Y Y Y Y
quoted area Y Y Y Y Y Y ? N
internet links - Y Y - - - Y -
e-mail links - Y Y - - - Y -
local links - Y Y N - - Y -
named links - Y Y - - - Y -
bulleted list Y Y Y Y Y Y Y Y
numbered list Y Y Y Y Y Y Y N
definition list Y Y ? Y N N N Y
horizontal line Y Y - Y Y N Y -
image - Y Y N Y N Y -
table N Y Y Y N N Y N


  Y   supported
  N   not supported (may be in future releases)
  -   not supported (can't be done on this target)
  ?   not supported (not sure if it can be done or not)

Download & Installation

1. Download & Install Python

First of all, you must download and install the Python interpreter on your system. If you already have it, just skip this step.

Python is one of the nicest programming languages out there, it works on Windows, Linux, UNIX, Macintosh, and others and it can be downloaded from the Python web site. Installation hints are found on the same site.

If you are not sure if you have Python or not, open a console (tty, xterm, MSDOS) and type python. If it is not installed, the system will tell you.

2. Download txt2tags

The official location for txt2tags distribution is on the program homepage, at http://txt2tags.sf.net/src.

All the program files are on the tarball (.tgz file), which can be expanded by most of the compression utilities (including Winzip).

Just get the latest one (more recent date, higher version number). The previous versions remains for historical purposes only.

3. Install txt2tags

As a single Python script, txt2tags needs no installation at all.

The only needed file to use the program is the txt2tags script. The other files of the tarball are documentation, tools and sample files.

The fail-proof way to run txt2tags, is calling Python with it:

  prompt$ python txt2tags

If you want to "install" txt2tags on the system as a stand alone program, just copy (or link) the txt2tags script to a System PATH directory and make sure the system knows how to run it.

Make the script executable (chmod +x txt2tags) and copy it to a $PATH directory (cp txt2tags /usr/bin)

Rename the script adding the .py extension (ren txt2tags txt2tags.py) and copy it to a system PATH directory (copy txt2tags.py C:\WINNT)
Done that, you can create an icon on your desktop for it, if you want to use the program's Graphical Interface.

User Interfaces

Txt2tags has three user interfaces. Now we will take a look at them.

Graphical Tk Interface

Since version 1.0, there is a nice Graphical Interface, that works on Linux, Windows and Mac (and others).

It's pretty simple and easy to use:

And it also has the ability to dump the result file to a window, instead of writing to the disc, so you can do quick testings before save the target file:

Web Interface

The Web Interface is up and running on the internet at http://txt2tags.sf.net/online.php, so you can use and test the program instantly, before download.

One can also put this interface on the local intranet for common use, avoiding to install txt2tags in all machines.

Command Line Interface

For command line powerusers, the --help should be enough:

  usage: txt2tags -t <type> [OPTIONS] file.t2t
         txt2tags -t html -s <split level> -l <lang> file.t2t
    -t, --type       target document type. actually supported:
                     txt, sgml, html, pm6, mgp, moin, man, tex
        --stdout     by default, the output is written to file.<type>
                     with this option, STDOUT is used (no files written)
        --noheaders  suppress header, title and footer information
        --enumtitle  enumerate all title lines as 1, 1.1, 1.1.1, etc
        --maskemail  hide email from spam robots. x@y.z turns to <x (a) y z>
        --toc        add TOC (Table of Contents) to target document
        --toconly    print document TOC and exit
        --gui        invoke Graphical Tk Interface
    -h, --help       print this help information and exit
    -V, --version    print program version and exit
  extra options for HTML target (needs sgml-tools):
        --split      split documents. values: 0, 1, 2 (default 0)
        --lang       document language (default english)


Assuming you have written a file.t2t marked file, let's have some converting fun.

Convert to HTML $ txt2tags -t html file.t2t
The same, using redirection $ txt2tags -t html --stdout file.t2t > file.html
Including Table Of Contents $ txt2tags -t html --toc file.t2t
And also, numbering titles $ txt2tags -t html --toc --enumtitle file.t2t
Contents quick view $ txt2tags --toconly file.t2t
Maybe enumerate them? $ txt2tags --toconly --enumtitle file.t2t
Oneliners from STDIN $ echo -e "\n**bold**" | txt2tags -t html --noheaders -
Testing Mask Email feature $ echo -e "\njohn.wayne@farwest.com" | txt2tags -t txt --maskemail --noheaders -
Post-convert editting $ txt2tags -t html --stdout file.t2t | sed "s/^<BODY .*/<BODY BGCOLOR=green>/" > file.html

The .t2t document Areas

Txt2tags marked files are divided in 3 areas. Each area have its own rules and purpose. They are:

Headers Area
Place for Document Title, Author, Version and Date information. (optional)
Settings Area
Place for general Document Settings and Parser behaviour modifiers. (optional)
Body Area
Place for the Document Content. (required)
As seen on the reminders, the first two Areas are optional, being Body Area the only required one. (Note: The Settings Area was introduced on txt2tags version 1.3)

The areas are delimited by special rules, which will be seen ahead. For now, this is a graphical representation of the areas on a document:

              |            |
              |   HEADERS  |       1. First, the Headers
              |            |
              |  SETTINGS  |       2. Then the Settings
              |            |
              |    BODY    |       3. And finally the Document Body,
              |            |
              |    ...     |          which goes until the end
              |    ...     |

In short, this is how the areas are defined:

Headers First 3 lines of the file, or the first line blank for No Headers.
Settings Begins right after the Header (4th or 2nd line) and ends when the Body Area starts.
Body The first valid text line (not comment or setting) after the Headers Area.

The Headers Area


The Headers Area is the only one that has a fixed position, line oriented. They are located at the first three lines of the source file.

These lines are content-free, with no static information type needed. But the following is recomended for the most documents:

Keep in mind that the first 3 lines of the source document will be the first 3 lines on the target document, separated and with high contrast to the text body (i.e. big letters, bold). If paging is allowed, the headers will be alone and centralized on the first page.

Less (or None) Header lines

Sometimes user wants to specify less then tree lines for headers, giving just document title and/or date information.

Just let the 2nd and/or the 3rd lines empty (blank) and this position will not be placed at the target document. But keep in mind that even blanks, these lines are still part of the headers, so the document body must start after the 3rd line anyway.

The title is the only required header (the first line), but if you leave it blank, you are saying that your document has no headers. So the Body Area will begin right after, on the 2nd line.

This is useful to use together with the command line --noheaders option.

Straight to the point

In short: "Headers are just positions, not contents".

Place one text on the first line, and it will appear on the target's first line. The same for 2nd and 3rd header lines.

The Settings Area


The Settings Area is optional, and an average English writter user should life fine with txt2tags without even know it exists. The primary use of this area is to define settings that affects the program behaviour.

So, how to set something? What's the syntax?

Setting lines are special comment lines, marked by a leading identifier ("!") that makes them different from plain comments. The syntax is just as simple as variable setting, composed by a keyword and a value, separated from each other by the canonical separator colon (":"). Example:

     %! keyword : value

The exclamation mark should be placed together with the comment char ("%!"), no spaces between them. The spaces around keyword and the separator are optional, and both keyword and value are case insensitive (case doesn't matter).

What can i set? Which are the valid keywords?

For now, the only setting that could be done is Encoding. It's needed by non-english writters, who uses accented letters and other locale specific details, so the target document Character Set must be customized (if allowed).

A real life example is:

    %! Encoding: iso-8859-1

To specify the latin charset.

The valid values for the Encoding setting are the same charset names valid for HTML documents, like iso-8859-1 and koi8-r. If you're not sure which encoding you should use, this complete (and long!) list should help.

The LateX target use alias names for encoding. This is not a problem for the user, because txt2tags translate the names internally. Some examples:

txt2tags/HTML > LaTeX
windows-1250 >>> cp1250
windows-1252 >>> cp1252
ibm850 >>> cp850
ibm852 >>> cp852
iso-8859-1 >>> latin1
iso-8859-2 >>> latin2
koi8-r >>> koi8-r

If the value is unknown to txt2tags, it will be passed "as is", allowing user to specify custom encodings.

Some rules about Settings

The Body Area


Well, the body is anything outside Headers and Settings.

The body holds the document contents and all formatting and structures txt2tags can recognize. Inside the body you can also put comments for TODOs and self notes.

You can use the --noheaders command line option to convert only the document body, supressing the headers. This is useful to set your own headers on a separate file, then join the converted body.

Full Example

  My nice doc Title
  Mr. John Doe
  Last Updated: %%date(%c)
  %! Encoding: iso8859-1
  Hi! This is my test document.
  Its content will end here.

Even more detailed info about txt2tags

Marks (RULES)

All marks and syntax used by txt2tags are detailed on a separate RULES file.

The %%date macro

The %%date macro called alone, returns the current date on the ISO yyyymmdd format. Optional formatting can be specified using the %%date(format-string) format.

This format-string is made of plain text plus the formatting directives, which are a percent sign % followed by an identification character.

Following is a list of some common use directives. The full list can be found in http://www.python.org/doc/current/lib/module-time.html.

Directive Description
%a Locale's abbreviated weekday name.
%A Locale's full weekday name.
%b Locale's abbreviated month name.
%B Locale's full month name.
%c Locale's appropriate date and time representation.
%d Day of the month as a decimal number [01,31].
%H Hour (24-hour clock) as a decimal number [00,23].
%I Hour (12-hour clock) as a decimal number [01,12].
%m Month as a decimal number [01,12].
%M Minute as a decimal number [00,59].
%p Locale's equivalent of either AM or PM.
%S Second as a decimal number [00,61]. (1)
%x Locale's appropriate date representation.
%X Locale's appropriate time representation.
%y Year without century as a decimal number [00,99].
%Y Year with century as a decimal number.
%% A literal "%" character.


%%date(format) Results for: 2002, Jan31, 15:00
Last Update: %c Last Update: Thu Jan 31 15:00:00 2002
%Y-%m-%d 2002-01-31
%I:%M %p 03:00 PM
Today is %A, on %B. Today is Thursday, on January.

Txt2tags HISTORY

On July 2001, was launched the first public release of txt2tags (v0.1). But its origins date more than an year before that...

This chapter illustrates in a few words the tool development since its very first draw until the current series.

1999 January: Pre-History

From the author:

"My really first attempts of a text conversion tool began back in 1999, as a very simple and limited Bourne Shell script that convert marked text to an HTML page. Yes, Yet-Another txt2html tool. Everyone Everywhere already must have done one of this... In short, it just recognized simple marks as *bold*, /italic/, _under_, and escape the classic < & > HTML special characters. Not impressive, but hey! I was young ;)"

1999 June: Still Pre-History

The author wants to speak some more:

"Some months passed, and a big Sgml hype arrived at the company I was working (Conectiva). So the txt2html turned into a txt2sgml script. I was really trying to learn about SED* at that moment so txt2sgml was a 110 lines Bourne Shell script with lots of SED code."

* SED: UNIX Stream EDitor - an automatic text editing tool

This improved Sgml version had more supported structures as lists and preformatted text. On the following sample file, you can see the txt2tags marks origins:

  		 * This was a bold line (BOLD line oriented? well...)
  		 - bullet list was very similar to txt2tags list
  		 - but with these -- to begin and close a list
  		 Preformatted text was delimited by the =-- pattern.
  		 The other ------- was just cosmetic.

Still not impressive, but the big step is comming...

2000 August: Not Pre-History anymore

TODO (txt2sgml.sed)

2001 July: Debut of 0.x series (World Release)


2002 September: Debut of 1.x series

This release starts my 1.x series.

More than a year of almost-monthly updates, and the 0.x series provided me a nice set of features, as Command Line and Web interface, TOC handling, numbering titles and lists, STDIN/STDOUT facilities, vim/emacs syntax files and seven supported target formats.

For the incoming 1.x series, I'll try to spread myself out, providing a nice GUI, extensive documentation, mailing list, user base, Unix/Windows/Mac full compatibility and including more targets (as tex, rtf and xhtml).

On this 1.0 release I'm already at full speed ahead, with a new suit (Graphical Tk Interface) and compatibility with Unix/Windows/Mac, handling line breaks and other platform specific issues. Fortunely, now my master can reach Linux, Windows 2000, Cygwin and MacOS 8.6 systems for testing me.

The End. (see source)