crc/retroforth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

some small work on edit.forth

Browse source 
FossilOrigin-Name: 1dbd50b3cfe48b522714712d040af801c2b6905fee761e1e24580f2bcc1506fc
This commit is contained in:
crc 2019-02-02 04:45:38 +00:00
parent c3c3e1fda2
commit fcbfebb54e
1 changed files with 88 additions and 52 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

140
example/edit.forth
View file

@ -2,25 +2,41 @@
# Hua: a text editor written in RETRO # Hua: a text editor written in RETRO
Hua is a small, functional text editor written in RETRO, using the Hua is a small, functional text editor written in RETRO, using
*RRE* interface. It is line oriented, visual, and tries to be very the *RRE* interface. It is line oriented, visual, and tries to
simple to use. be very simple to use.
First up, several variables and constants that are used through ## Starting
the rest of the code.
Hua is intended to run as a standalone tool. Use a line like:
edit.forth filename
To create a new file:
edit.forth new filename
## A Word of Warning
Hua saves changes as you edit the file. I advise using it along
with a version control system so you can revert changes if or
when needed.
## The Code
Since this runs as a standalone application I use a quick check
to exit if no arguments were passed.
~~~ ~~~
'SourceFile var sys:argc n:zero? [ #0 unix:exit ] if
'CurrentLine var
'LineCount var
'ShowEOL var
'FID var
'CopiedLine d:create #1025 allot
~~~ ~~~
The configuration here is for two items. The number of lines from the If I get here, a filename was provided. So I start by creating
file to show on screen, and the name of the temporary file to use a few variables and constants.
when editing.
The configuration here is for two items. The number of lines
from the file to show on screen, and the name of the temporary
file to use when editing.
~~~ ~~~
#80 'COLS const #80 'COLS const
@ -28,11 +44,28 @@ when editing.
'/tmp/rre.edit 'TEMP-FILE s:const '/tmp/rre.edit 'TEMP-FILE s:const
~~~ ~~~
Get the name of the file to edit. If no file is provided, exit. Next are the variables that I use to track various bits of
state.
~~~
'SourceFile var
'CurrentLine var
'LineCount var
'ShowEOL var
'FID var
'CopiedLine d:create #1025 allot
~~~
Get the name of the file to edit.
~~~ ~~~
sys:argc n:zero? [ #0 unix:exit ] if
#0 sys:argv s:keep !SourceFile #0 sys:argv s:keep !SourceFile
~~~
To create a new file, Hua allows for the use of `new` followed
by the filename. I handle the file creation here.
~~~
@SourceFile 'new s:eq? @SourceFile 'new s:eq?
[ #1 sys:argv s:keep !SourceFile [ #1 sys:argv s:keep !SourceFile
@SourceFile file:A file:open file:close ] if @SourceFile file:A file:open file:close ] if
@ -48,15 +81,15 @@ easier.
I now turn my attention to displaying the file. I am aiming for I now turn my attention to displaying the file. I am aiming for
an interface like: an interface like:
<filename> : <line-count> <filename> : <line-count>
--------------------------------------------------------------- ---------------------------------------------------------------
* 99: * 99:
100: :n:square dup * ; 100: :n:square dup * ;
101: 101:
102: This is the current line 102: This is the current line
103: 103:
--------------------------------------------------------------- ---------------------------------------------------------------
j: down | k: up | ... other helpful text ... j: down | k: up | ... other helpful text ...
The * denotes the currently selected line. The * denotes the currently selected line.
@ -71,11 +104,11 @@ advance to the currently selected line.
@CurrentLine MAX-LINES #2 / - #0 n:max [ @FID file:read-line drop ] times ; @CurrentLine MAX-LINES #2 / - #0 n:max [ @FID file:read-line drop ] times ;
~~~ ~~~
Now for words to format the output. This should all be pretty clear in Now for words to format the output. This should all be pretty
intent. clear in intent.
`clear-display` uses an ANSI/VT100 escape sequence. This might need to `clear-display` uses an ANSI/VT100 escape sequence. This might
be adjusted for your chosen terminal. need to be adjusted for your chosen terminal.
~~~ ~~~
:clear-display (-) :clear-display (-)
@ -89,33 +122,35 @@ This just displays the separator bars.
COLS [ $- c:put ] times nl ; COLS [ $- c:put ] times nl ;
~~~ ~~~
Next, a word to display the header. Currently just the name of the file Next, a word to display the header. Currently just the name of
being edited and the line count. the file being edited and the line count.
~~~ ~~~
:header (-) :header (-)
count-lines @SourceFile '%s_:_%n_lines\n s:format s:put ; count-lines @SourceFile '%s_:_%n_lines\n s:format s:put ;
~~~ ~~~
The `pad` word is used to make sure line numbers are all the same width. The `pad` word is used to make sure line numbers are all the
same width.
~~~ ~~~
:pad (n-n) :pad (n-n)
dup #0 #9 n:between? [ '____ s:put ] if dup #0 #9 n:between? [ '____ s:put ] if
dup #10 #99 n:between? [ '___ s:put ] if dup #10 #99 n:between? [ '___ s:put ] if
dup #100 #999 n:between? [ '__ s:put ] if dup #100 #999 n:between? [ '__ s:put ] if
dup #1000 #9999 n:between? [ '_ s:put ] if ; dup #1000 #9999 n:between? [ '_ s:put ] if ;
~~~ ~~~
A line has a form: A line has a form:
<indicator><number>: <text><eol> <indicator><number>: <text><eol>
The indicator is an asterisk, and visually marks the current line. The indicator is an asterisk, and visually marks the current
line.
EOL is optional. If `ShowEOL` is `TRUE`, it'll display a ~ at the end EOL is optional. If `ShowEOL` is `TRUE`, it'll display a ~ at
of each line. This is useful when looking for trailing whitespace. The the end of each line. This is useful when looking for trailing
indicator can be toggled via the ~ key. whitespace. The indicator can be toggled via the ~ key.
~~~ ~~~
:mark-if-current (n-n) :mark-if-current (n-n)
@ -138,20 +173,21 @@ indicator can be toggled via the ~ key.
---- @FID file:close ; ---- @FID file:close ;
~~~ ~~~
With the code to display the file done, I can proceed on to words for With the code to display the file done, I can proceed to the
handling editing. words for handling editing.
I add a custom combinator, `process-lines` to iterate over the lines in I add a custom combinator, `process-lines` to iterate over the
the file. This takes a quote, and runs it once for each line in the file. lines in the file. This takes a quote, and runs it once for
The quote gets passed two values: a counter and a pointer to the current each line in the file. The quote gets passed two values: a
line in the file. The quote should consume the pointer an increment the counter and a pointer to the current line in the file. The
counter. This also sets up `FID` as a pointer to the temporary file where quote should consume the pointer an increment the counter. This
changes can be written. The combinator will replace the original file also sets up `FID` as a pointer to the temporary file where
after execution completes. changes can be written. The combinator will replace the
original file after execution completes.
Additionally, I define a word named `current?` which returns `TRUE` if Additionally, I define a word named `current?` which returns
the specified line is the current one. This is just to aid in later `TRUE` if the specified line is the current one. This is just
readability. to aid in later readability.
~~~ ~~~
:process-lines (q-) :process-lines (q-)