Source_scanner.xml

<?xml version="1.0" encoding="iso-8859-1"?>
<!DOCTYPE muclient>

<muclient>
<plugin
   name="Source_scanner"
   author="Nick Gammon"
   id="efc3c27a95ba6d8b534e46d5"
   language="Lua"
   purpose="Scans source files for keywords"
   date_written="2012-11-04"
   requires="4.41"
   version="1.1"
   >
<description>
find &lt;phrase&gt; 

eg. 

cat AND dog              --&gt; both words 
cat dog                  --&gt; both words, the "AND" is implied
fish OR bicycle          --&gt; one or the other
cat NOT food             --&gt; one word but not the other
bite NEAR me             --&gt; one near the other (within 10 words)
disk NEAR/3 drive        --&gt; one within 3 words of the other
"trouble brewing"        --&gt; exact phrase
chip*                    --&gt; prefix query, matches chip, chips, chipping etc.
fish NOT (bacon OR eggs) --&gt; brackets can be used to clarify groupings
    
The words AND / OR / NEAR / NOT must be in upper case or they just match those words literally.

find name &lt;wildcard&gt;   --&gt; filter on file name, not contents
    
You can use "f" instead of "filter".

index   --&gt;  scans new directory (choose one) to re-make the database

help    --&gt;  this text
</description>

</plugin>

<aliases>
 
 <alias
   script="filter_files"
   match="^(?:f|find) (.+)$"
   enabled="y"
   regexp="y"
   sequence="100"
  >
  </alias>
  
  <alias
   script="load_text_files"
   match="index"
   enabled="y"
   sequence="100"
  >
  </alias>
  
  <alias
   script="OnHelp"
   match="help"
   enabled="y"
   sequence="100"
  >
  </alias>
    
</aliases>

<!--  Script  -->


<script>
<![CDATA[

-- ----------- CONFIGURATION -----------

-- file types, separate by spaces, commas, semicolons, whatever. 
-- We assume suffixes are alphanumeric
SUFFIXES = "cpp,c,h,xml,lua"

-- default drive for directory picker
DRIVE_TO_SCAN = "C:\\\\"

-- file name for the database (MUSHclient directory)
SOURCE_FILES = GetInfo (66) .. "\\source_files.db"

-- colour to show file names in
HYPERLINK_COLOUR = "dodgerblue"

-- viewer program 
TEXT_VIEWER = "C:\\Program Files\\Crimson Editor\\cedt.exe"

-- or maybe:
-- TEXT_VIEWER = "C:\\Windows\\notepad.exe"

-- view program title (for bringing to the front)
EDITOR_WINDOW_NAME = "Crimson Editor"

-- or maybe:
-- EDITOR_WINDOW_NAME = "Notepad"

-- show this many before suggesting they narrow the search
MAX_RESULTS = 500

-- number of tokens to display around the snippet
SNIPPET_SIZE = -7

-- strip from displayed file names to avoid clutter, like always having "Desktop" or
--  "My Documents" there, which you might not care about.
PREFIX_TO_OMIT = "^%a:\\Documents and Settings\\.-\\"

-- for opening editor and bringing it to the front
-- available here: http://www.gammon.com.au/files/mushclient/lua5.1_extras/windows_utils.zip

assert (package.loadlib ("windows_utils.dll", "luaopen_windows_utils")) ()

require "alphanum"
os.setlocale ("", "time")

-- build suffixes into a table (keyed by suffix) for quick lookup later
local suffixes = {}
for w in string.gmatch (SUFFIXES, "[%a%d]+") do
  suffixes [w:lower ()] = true 
end -- for
  
-- helper function to check for database errors
function dbcheck (code)

 if code ~= sqlite3.OK and       -- no error
    code ~= sqlite3.ROW and      -- completed OK with another row of data
    code ~= sqlite3.DONE then    -- completed OK, no more rows
      local err = db:errmsg()    -- the rollback will change the error message
      db:exec ("ROLLBACK")       -- rollback any transaction to unlock the database
      error (err, 2)             -- show error in caller's context
  end -- if error

end -- dbcheck 

-- find where to start scanning from
function get_text_path ()
  local path = utils.directorypicker ("Directory for source", DRIVE_TO_SCAN)
  if not path then
    return false
  end -- if
  text_drive, text_path = string.match (path, "^(%a:)(.+)$")
  return true
end --  get_text_path
 
-- process one file
function read_file (text_drive, name, stats)
  
  SetStatus ("Processing: " .. name)
  
  -- ignore empty files
  if stats.size == 0 then
    return
  end -- if

  -- look for wanted suffix  
  if not suffixes [string.match (name:lower (), "%.(%a+)$")] then
    return
  end -- if

  count = count + 1

  -- read contents into memory  
  local f = assert (io.open (text_drive .. name, "rb")) 
  local s = f:read ("*a")  
  f:close ()  -- close it
  
  -- quotes have to be doubled
  local fixed_name = string.gsub (text_drive .. name, "'", "''")
  s = string.gsub (s, "%z", "")  -- get rid of 0x00 bytes
  local fixed_contents = string.gsub (s, "'", "''")
   
  -- save contents
  dbcheck (db:exec ( string.format ([[
  INSERT INTO source (name, contents, size, date_written) 
         VALUES ('%s', '%s', %i, %i);
  ]], 
      fixed_name, 
      fixed_contents, 
      stats.size, 
      stats.write_time)))

end -- read_file

-- process one directory (recurses if necessary)
function scan_text_dir (text_drive, text_path, f)
  
  -- find all files in that directory
  local t = assert (utils.readdir (text_drive .. text_path .. "\\*"), 
                    "Could not find " .. text_drive .. text_path  .. "\\*")
    
  for k, v in pairs (t) do
   if not v.hidden and 
      not v.system and
      k:sub (1, 1) ~= "." then
      
      -- recurse to process file or subdirectory
      if v.directory then
        scan_text_dir (text_drive, text_path .. "\\" .. k, f)
      else
        f (text_drive, text_path .. "\\" .. k, v)
      end -- if directory or not
     
   end -- if not hidden or system
  
  end -- for

end -- scan_text_dir 

-- scan entire tree
function load_text_files (name, line, wildcards)

  if not get_text_path () then
    return
  end -- no directory

  db, err, errmsg = sqlite3.open(SOURCE_FILES)
 
  if not db then 
    ColourNote ("white", "red", errmsg .. ": " .. SOURCE_FILES)
    return 
  end

  -- omit the "tokenize" line to get normal tokenization
  -- that line omits underscores as tokens (for C source)
  -- the "X" is there because of SQLITE3 bug:
  -- http://sqlite.1065341.n5.nabble.com/FTS-simple-tokenizer-with-custom-delimeters-td43926.html
  
  dbcheck (db:exec [[
    DROP TABLE IF EXISTS source;
    CREATE VIRTUAL TABLE source USING FTS4(name, contents, size, date_written, 
    tokenize=simple X ' !"#$%&''()*+,-./:;<=>?@[\]^`{|}~'
    );
   ]])
 
  count = 0   -- actual
  start_time = utils.timer ()
  
  -- write database inside a transaction  
  db:exec "BEGIN TRANSACTION;"
  scan_text_dir (text_drive, text_path, read_file)
  db:exec "COMMIT;"
  db:close()  -- close

  end_time = utils.timer ()
  
  SetStatus ("Ready")
  print (string.format ("Loaded %i files in %.1f seconds.", count, end_time - start_time))
 
end -- load_text_files

local snippets = {}
local reset = ANSI (0)
local highlight = ANSI (1) .. ANSI (36)  -- bold and cyan
local unhighlight = ANSI (22) .. ANSI (37)  -- unbold and white

-- here when hyperlink clicked to open a particular file
function display_text_file (name)
  assert (windows_utils.shell_execute (TEXT_VIEWER, '"' .. name .. '"'))
  windows_utils.send_to_front (EDITOR_WINDOW_NAME)
end -- display_text_file

function show_filtered (show_snippet)
  local names = {}
  for k in pairs (filtered) do
    table.insert (names, k)
  end -- for
  
  table.sort(names, alphanum (names))
 
  ColourNote ("yellow", "", "Search term: " .. match)
  print ""
  -- now pull out all entries
   
  for _, name in ipairs (names) do
     Hyperlink ("!!" .. GetPluginID () .. ":display_text_file(" .. name .. ")", 
          (string.gsub (name, PREFIX_TO_OMIT, "")), 
          "Click to open", HYPERLINK_COLOUR, "", 0)
     if name == current then
       print ("  <-- this file")
     else
       print ""
     end -- if
     if show_snippet then
       AnsiNote (reset .. filtered [name])
       print ""
     end -- if
  end -- for each file

  ColourNote ("yellow", "", "Matched " .. #names .. " file(s) (Search term: " .. match .. ")")
  
end -- show_filtered

-- search function ... searches on name or contents
function filter_on (what, field)

  db, err, errmsg = sqlite3.open(SOURCE_FILES)
 
  if not db then 
   ColourNote ("white", "red", errmsg .. ": " .. SOURCE_FILES)
   return 
  end

  local matches = 0
 
  filtered = {}
  local fixed_match = string.gsub (what, "'", "''")
  
  for row in db:nrows(string.format ([[
      SELECT name,
             snippet(source, '%s', '%s', ' ... ', -1, %i) AS snippet 
             FROM source WHERE %s MATCH '%s']], 
      highlight,    -- for snippet
      unhighlight,  -- also for snippet
      SNIPPET_SIZE,
      field,        -- name or contents
      fixed_match)) do 
     filtered [row.name] = row.snippet
     matches = matches + 1
  end -- for

  print (string.rep ("+", 79))
  
  -- show matches  
  if matches == 0 then
    ColourNote ("red", "", "No matches on '" .. match .. "'.")
  elseif matches <= MAX_RESULTS then
    show_filtered (field == "contents")
  else
    ColourNote ("red", "", "Too many to list - try using 'this AND that' to narrow it down")
    ColourNote ("yellow", "", "Matched " .. matches .. " files (Search term: " .. match .. ")")
  end -- if
  
  local source_files = 0
  
  -- show how many files were considered
  for row in db:nrows "SELECT COUNT(*) AS row_count FROM source;" do
    source_files = row.row_count
  end -- for
  
  print (source_files, "files searched.")
  print (string.rep ("-", 79))

  db:close ()
  
end -- filter_on

-- here when they want to filter on name or contents
function filter_files (name, line, wildcards)

  match = wildcards [1]
  
  local a, b = string.match (Trim (match), "(name)%s+(.+)")
  if a and a:lower () == "name" then
    filter_on (b, "name")
  else
    filter_on (match, "contents")
  end
end -- filter_files

]]>

-- show help
function OnHelp ()
  ColourNote ("cyan", "", GetPluginInfo (GetPluginID (), 3))
end

</script>
</muclient>