home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Turnbull China Bikeride
/
Turnbull China Bikeride - Disc 1.iso
/
ARGONET
/
PD
/
DTP
/
CLEANTXT.SPK
/
!CleanTXT
/
!Help
< prev
next >
Wrap
Text File
|
1997-08-20
|
4KB
|
82 lines
!CleanTXT
Version 0.02ß (20 August 1997)
Needs RISC OS 3.50
!CleanTXT is intended to process text files to make them suitable for import
into word processors.
In its current version, !CleanTXT first removes stray spaces from the beginnings
and ends of lines. It then converts pairs of <new line> characters into single
a <new line> character and converts single <new line> characters into spaces.
To use, drop a text file onto the icon. The control window will open. This
allows additional identifiers for paragraph ends to be setup:
• Lines indented with at least a specified number of spaces will be treated
as new paragraphs. This can be useful for identifying "letter style"
paragraphs, lists, tables, program listings etc.
• Lines starting with a tab can be treated as the beginning of a new
paragraph. This can be useful in the same circumstances as above.
• Lines starting with a digit can be treated as the beginning of a new
paragraph. This can be useful where there are paragraph numbers or
enumerated lists.
• There is a free form input for characters at the beginning of lines (ignoring
initial spaces) that should be treated as commencing a new paragraph. This
can be useful for identifying bulleted lists. By default this looks for
bullet points (•), asterisks (*) or dashes (-).
• There is a free form input for characters at the end of lines (ignoring
trailing spaces) that should be treated as ending a paragraph. This can be
useful where there is no easy identifier of paragraph ends and fulls stops at
the end of sentences or colons introducing lists must be used.
This program has not been through extensive bug testing and so may have
problems or bugs. I should appreciate it if any bugs or additions to the
wishlist could be emailed to me.
Email:
alec.crawford@argonet.co.uk
Wish list:
• Intelligent recognition and formatting of tables (more tricky to implement
and so unlikely to be in the next version). In the short term, this will
be handled by converting a specified number of spaces (or more) into a
tab character.
• Recognition of headings as single lines or lines with "====" or "---"
underlining. The output could be either in the form of RTF of Impression
DDF formats.
• Improved speed. Currently !CleanTXT is not efficiently programmed. In the
short term, speed may be controllable by adjusting (as an option) how
frequently !CleanTXT polls. In addition, it may be helpful to remove some
of the reporting functions from the processing window - let me know if this
would be helpful (I am testing this on a StrongARM).
• Backward compatability. This version requires RISC OS 3.5 (at least) because
it uses "OS_DynamicArea" to allocate itself memory. If there is demand, I
can easily produce a version that predefines space in the program area (and
hence will be useable in RISC OS 3.00 or greater - I believe that Dr Wimp
requires at least RISC OS 3.00) or, if a better programmer (or with more
time to peruse the manuals) wishes to suggest a flexible memory allocation
system from BASIC for earlier OS, I may be able to look into it.
Known bugs:
• !CleanTXT occasionally tags one or two stray characters to the end of the
text
Credits:
Many thanks to Andrew Ayre for Dr Wimp (3.01) without which this program would
not have been programmed. With Dr Wimp, programming was a matter of a few
hours. Thoroughly recommended!
History:
0.01 Version not distributed