OpenEuphoria

2. Re: AI fun

• Posted by andreasWagner 2 days ago
• 78 views

Didn't work between here either 

-- notepad_extract.ex 
-- 
-- Usage: 
--   eui notepad_extract.ex output.txt file1.html file2.html file3.html ... 
-- 
-- For each HTML file: 
--   - Extracts title between <anotepadtitle>...</anotepadtitle> 
--   - Extracts content between <anotepadcontent>...</anotepadcontent> 
--   - Writes to output as: 
-- 
--       Title text 
--       ========== 
--       Content text 
-- 
--   - Blank line between notes 
 
include std/io.e 
include std/text.e 
include std/sequence.e 
include std/filesys.e 
 
function read_file_as_text(sequence filename) 
    integer fh = open(filename, "r") 
    if fh = -1 then 
        printf(STDERR, "Could not open file: %s\n", {filename}) 
        return "" 
    end if 
 
    sequence data = "" 
    object line 
    while 1 do 
        line = gets(fh) 
        if atom(line) then 
            exit 
        end if 
        data &= line 
    end while 
 
    close(fh) 
    return data 
end function 
 
function extract_between(sequence text, sequence start_tag, sequence end_tag) 
    integer start_pos = match(start_tag, text) 
    if start_pos = 0 then 
        return "" 
    end if 
 
    start_pos += length(start_tag) 
 
    integer end_pos = match(end_tag, text[start_pos..$]) 
    if end_pos = 0 then 
        return "" 
    end if 
 
    end_pos = start_pos + end_pos - 2 
 
    return text[start_pos..end_pos] 
end function 
 
procedure write_note(integer out_fh, sequence title, sequence content) 
    -- Trim whitespace 
    title   = trim(title) 
    content = trim(content) 
 
    if length(title) = 0 and length(content) = 0 then 
        return 
    end if 
 
    -- Title 
    puts(out_fh, title & "\n") 
 
    -- Underline with '=' same length as title 
    sequence underline = repeat('=', length(title)) 
    puts(out_fh, underline & "\n") 
 
    -- Content 
    puts(out_fh, content & "\n\n") 
end procedure 
 
procedure main() 
    sequence cmd = command_line() 
	sequence HTML_list = {} 
 
	integer out_name, out_fh 
 
/*	AI generated code 

-- cmd[1] = interpreter, cmd[2] = script name 
    if length(cmd) < 4 then 
        puts(STDERR, 
            "Usage:\n" & 
            "  eui notepad_extract.ex output.txt file1.html file2.html ...\n") 
        abort(1) 
    end if 
 
    sequence out_name = cmd[3] 
    integer out_fh = open(out_name, "w") 
    if out_fh = -1 then 
        printf(STDERR, "Could not open output file: %s\n", {out_name}) 
        abort(1) 
    end if 
*/ 
 
	--with no command line, just going to get a list of all HTML files 
	HTML_list = dir(current_dir() & "\\*.html") 
 
    out_fh = open("ANotepad_all.txt", "w") 
    if out_fh = -1 then 
        printf(STDERR, "Could not open output file: %s\n", {out_name}) 
        abort(1) 
    end if 
 
    -- Process each HTML file 
    for i = 1 to length(HTML_list) do 
        sequence fname = HTML_list[i][D_NAME] 
        printf(STDOUT, "Processing %s...\n", {fname}) 
 
        sequence html = read_file_as_text(fname) 
        if length(html) = 0 then 
            continue 
        end if 
 
        -- Handle both correct and slightly malformed closing tags for safety 
        sequence title = extract_between( 
            html, 
            "<anotepadtitle>", 
            "</anotepadtitle>" 
        ) 
 
        sequence content = extract_between( 
            html, 
            "<anotepadcontent>", 
            "</anotepadcontent>" 
        ) 
 
        -- If the user s HTML has a typo like </anotepadcontent)>, try that too 
        if length(content) = 0 then 
            content = extract_between( 
                html, 
                "<anotepadcontent>", 
                "</anotepadcontent)>" 
            ) 
        end if 
 
        write_note(out_fh, title, content) 
    end for 
 
    close(out_fh) 
    printf(STDOUT, "Done. Notes written to %s\n", {out_name}) 
end procedure 
 
main() 
 
 
/*This was the prompt 

write me a euphoria program, that I can feed many htmles to, tat will analyse the content and produce a  
single text file. The analysis should produce a title and content for each note. The title should be  
underlined by "=". The html file contains a title and a content for each note. The title is delineated within  
<anotepadtitle></anotepadtitle> and the content is delineated between <anotepadcontent></anotepadcontent) 
 
*/ 

4. Re: AI fun

• Posted by andreasWagner 2 days ago
• 55 views

I've had this problem before. That's why I always send text from the web browser through MS-Edit, convert it to US-ASCII, and then check at least twice to see if I see any strange characters. Yes, I also save and reload twice.

Of course, I did the same with your text.

Hope that helps.