markov chain generation is pretty decent

examples/word_count/markov

@@ -0,0 +1,34 @@
+#!/usr/bin/env ruby
+
+require '../../couchrest'
+
+cr = CouchRest.new("http://localhost:5984")
+db = cr.database('word-count-example')
+
+word = ARGV[0]
+words = [word]
+wprobs = {}
+
+while word
+  $stdout.print ' ' if words.length > 1
+  $stdout.print word
+  $stdout.flush
+  
+  wprobs[word] ||= db.view('markov/chain-reduce', :startkey => [word,nil], :endkey => [word,{}],:group_level => 2)
+
+  # puts
+  # puts "search #{word} #{wprobs[word]['rows'].length}"
+  # wprobs[word]['rows'].sort_by{|r|r['value']}.each{|r|puts [r['value'],r['key']].inspect}
+
+  rows = wprobs[word]['rows'].select{|r|(r['key'][1]!='')}.sort_by{|r|r['value']}
+  row = rows[(-1*[rows.length,5].min)..-1].sort_by{rand}[0]
+  word = row ? row['key'][1] : nil
+  words << word
+end
+
+$stdout.print '.'
+$stdout.flush
+puts
+
+# `say #{words.join(' ')}`
+

examples/word_count/views/books/chunked-map.js

@@ -0,0 +1,3 @@
+function(doc) {
+  doc.title && doc.chunk && emit([doc.title, doc.chunk],null);
+}

examples/word_count/views/books/united-map.js

@@ -0,0 +1 @@
+function(doc){if(doc.text && doc.text.match(/united/))  emit([doc.title, doc.chunk],null)}

examples/word_count/views/markov/chain-map.js

@@ -0,0 +1,6 @@
+function(doc){
+  var words = doc.text.split(/\W/).filter(function(w) {return w.length > 0}).map(function(w){return w.toLowerCase()});
+  for (var i = 0, l = words.length; i < l; i++) {
+    emit(words.slice(i,4),doc.title);
+  }
+}

examples/word_count/views/markov/chain-reduce.js

@@ -0,0 +1,7 @@
+function(key,vs,c){
+  if (c) {
+    return sum(vs);
+  } else {
+    return vs.length;
+  }
+}

examples/word_count/views/word_count/count-map.js

@@ -0,0 +1,6 @@
+function(doc){
+  var words = doc.text.split(/\W/).map(function(w){return w.toLowerCase()});
+  words.forEach(function(word){
+    if (word.length > 0) emit([word,doc.title],1);
+  });
+}

examples/word_count/views/word_count/count-reduce.js

@@ -0,0 +1,3 @@
+function(key,combine){
+  return sum(combine);
+}

examples/word_count/word_count.rb

@@ -2,16 +2,18 @@ require File.dirname(__FILE__) + '/../../couchrest'
 
 couch = CouchRest.new("http://localhost:5984")
 db = couch.database('word-count-example')
-# db.delete! rescue nil
-# db = couch.create_db('word-count-example')
+db.delete! rescue nil
+db = couch.create_db('word-count-example')
 
 %w{america.txt da-vinci.txt outline-of-science.txt ulysses.txt}.each do |book|
+# %w{}.each do |book|
   title = book.split('.')[0]
   puts title
   File.open(File.join(File.dirname(__FILE__),book),'r') do |file|
     lines = []
     chunk = 0
     while line = file.gets
+      puts chunk
       lines << line
       if lines.length > 10
         db.save({

script/push_views.rb

@@ -1,60 +0,0 @@
-$: << File.expand_path(File.dirname(__FILE__)) + '/..'
-require 'lib/parse'
-require 'couchrest/couchrest'
-require 'vendor/jsmin/lib/jsmin'
-# require 'yaml'
-# connect to couchdb
-
-cr = CouchRest.new("http://localhost:5984")
-cr.create_db('grabbit-import') rescue nil
-db = cr.database('grabbit-import')
-
-# create views from files
-views = {}
-viewfiles = Dir.glob(File.join(File.expand_path(File.dirname(__FILE__)),"..","views","**","*.js"))
-
-libfiles = viewfiles.select{|f|/lib\.js/.match(f)}
-
-all = (viewfiles-libfiles).collect do |file|
-  filename = /(\w.*)-(\w.*)\.js/.match file.split('/').pop
-  filename.to_a + [file]
-end
-
-@libfuncs = open(libfiles[0]).read
-
-def readjs(file)
-  st = open(file).read
-  st.sub!(/\/\/include-lib/,@libfuncs)
-  JSMin.minify(st)
-end 
-
-all.group_by do |f|
-  f[3].split('/')[-2]
-end.each do |design,ps|
-  views[design] ||= {}
-  puts "design #{design}"
-  ps.group_by do |f|
-    f[1]
-  end.each do |view,parts|
-    puts "view #{view}"
-    views[design]["#{view}-reduce"] ||= {}
-    parts.each do |p|
-      puts "part #{p.inspect}"
-      views[design]["#{view}-reduce"][p[2]] = readjs(p[3])
-    end
-    views[design]["#{view}-map"] = {:map => views[design]["#{view}-reduce"]['map']}
-    views[design].delete("#{view}-reduce") unless views[design]["#{view}-reduce"]['reduce']
-  end  
-end
-
-views.each do |design,viewfuncs|
-  begin
-    view = db.get("_design/#{design}")
-    db.delete(view)
-  rescue
-  end
-  db.save({
-    "_id" => "_design/#{design}",
-    :views => viewfuncs
-  })
-end

views/mentions/inbound-map.js

@@ -0,0 +1,2 @@
+
+function(doc){if(doc.mp3s){for(var i=0,m;m=doc.mp3s[i];i++){emit(m.href,doc.fetch.url);}}}

views/mentions/inbound-reduce.js

@@ -0,0 +1,3 @@
+
+function(hrefs,ss){log(ss)
+return ss[0];}

views/mentions/mp3links-map.js

@@ -0,0 +1,2 @@
+
+function(doc){var fetchurl=doc.fetch&&doc.fetch.url;if(!fetchurl)return;doc.entries&&doc.entries.forEach(function(e){e.mp3s&&e.mp3s.forEach(function(mp3){mp3.href&&emit(mp3.href,fetchurl);});});doc.playlist&&doc.playlist.track&&doc.playlist.track.forEach(function(t){t.location&&t.location.forEach(function(url){emit(url,fetchurl);});});doc.mp3s&&doc.mp3s.forEach(function(mp3){mp3.href&&emit(mp3.href,fetchurl);});}

views/mentions/mp3links-reduce.js

@@ -0,0 +1,2 @@
+
+function(ks,vs){log({keys:ks});log({values:vs});return 1;};

views/metadata/albums-map.js

@@ -0,0 +1,2 @@
+
+function(doc){doc.playlist&&doc.playlist.track&&doc.playlist.track.forEach(function(t){emit([t.creator||null,t.title||null],t.album||null);});};

views/metadata/track-map.js

@@ -0,0 +1,2 @@
+
+function(doc){doc.playlist&&doc.playlist.track&&doc.playlist.track.forEach(function(t){if(t.creator||t.title){if(t.location){t.location.forEach(function(url){emit([t.creator||null,t.title||null],url);});}else{emit([t.creator||null,t.title||null],null);}}});};

views/metadata/track-reduce.js

@@ -0,0 +1,2 @@
+
+function(ks,vs,c){if(c){return null;}else{log(ks[0][0][0]);return ks[0][0][0];}};

views/test/maponly-map.js

@@ -0,0 +1,2 @@
+
+function(doc){emit(null,doc);}

views/test/reducehaving-map.js

@@ -0,0 +1,585 @@
+/*
+
+ JS Beautifier
+---------------
+  $Date: 2008-05-26 06:34:52 +0300 (Mon, 26 May 2008) $
+  $Revision: 55 $
+
+
+  Written by Einars "elfz" Lielmanis, <elfz@laacz.lv> 
+      http://elfz.laacz.lv/beautify/
+
+  Originally converted to javascript by Vital, <vital76@gmail.com> 
+      http://my.opera.com/Vital/blog/2007/11/21/javascript-beautify-on-javascript-translated
+
+
+  You are free to use this in any way you want, in case you find this useful or working for you.
+
+  Usage:
+    js_beautify(js_source_text);
+
+*/
+
+
+function js_beautify(js_source_text, indent_size, indent_character)
+{
+
+    var input, output, token_text, last_type, last_text, last_word, current_mode, modes, indent_level, indent_string;
+    var whitespace, wordchar, punct, parser_pos, line_starters, in_case;
+    var prefix, token_type, do_block_just_closed, var_line, var_line_tainted;
+
+
+    function trim_output()
+    {
+        while (output.length && (output[output.length - 1] === ' ' || output[output.length - 1] === indent_string)) {
+            output.pop();
+        }
+    }
+
+    function print_newline(ignore_repeated)
+    {
+        ignore_repeated = typeof ignore_repeated === 'undefined' ? true: ignore_repeated;
+        
+        trim_output();
+
+        if (!output.length) {
+            return; // no newline on start of file
+        }
+
+        if (output[output.length - 1] !== "\n" || !ignore_repeated) {
+            output.push("\n");
+        }
+        for (var i = 0; i < indent_level; i++) {
+            output.push(indent_string);
+        }
+    }
+
+
+
+    function print_space()
+    {
+        var last_output = output.length ? output[output.length - 1] : ' ';
+        if (last_output !== ' ' && last_output !== '\n' && last_output !== indent_string) { // prevent occassional duplicate space
+            output.push(' ');
+        }
+    }
+
+
+    function print_token()
+    {
+        output.push(token_text);
+    }
+
+    function indent()
+    {
+        indent_level++;
+    }
+
+
+    function unindent()
+    {
+        if (indent_level) {
+            indent_level--;
+        }
+    }
+
+
+    function remove_indent()
+    {
+        if (output.length && output[output.length - 1] === indent_string) {
+            output.pop();
+        }
+    }
+
+
+    function set_mode(mode)
+    {
+        modes.push(current_mode);
+        current_mode = mode;
+    }
+
+
+    function restore_mode()
+    {
+        do_block_just_closed = current_mode === 'DO_BLOCK';
+        current_mode = modes.pop();
+    }
+
+
+    function in_array(what, arr)
+    {
+        for (var i = 0; i < arr.length; i++)
+        {
+            if (arr[i] === what) {
+                return true;
+            }
+        }
+        return false;
+    }
+
+
+
+    function get_next_token()
+    {
+        var n_newlines = 0;
+        var c = '';
+
+        do {
+            if (parser_pos >= input.length) {
+                return ['', 'TK_EOF'];
+            }
+            c = input.charAt(parser_pos);
+
+            parser_pos += 1;
+            if (c === "\n") {
+                n_newlines += 1;
+            }
+        }
+        while (in_array(c, whitespace));
+
+        if (n_newlines > 1) {
+            for (var i = 0; i < 2; i++) {
+                print_newline(i === 0);
+            }
+        }
+        var wanted_newline = (n_newlines === 1);
+
+
+        if (in_array(c, wordchar)) {
+            if (parser_pos < input.length) {
+                while (in_array(input.charAt(parser_pos), wordchar)) {
+                    c += input.charAt(parser_pos);
+                    parser_pos += 1;
+                    if (parser_pos === input.length) {
+                        break;
+                    }
+                }
+            }
+
+            // small and surprisingly unugly hack for 1E-10 representation
+            if (parser_pos !== input.length && c.match(/^[0-9]+[Ee]$/) && input.charAt(parser_pos) === '-') {
+                parser_pos += 1;
+
+                var t = get_next_token(parser_pos);
+                c += '-' + t[0];
+                return [c, 'TK_WORD'];
+            }
+
+            if (c === 'in') { // hack for 'in' operator
+                return [c, 'TK_OPERATOR'];
+            }
+            return [c, 'TK_WORD'];
+        }
+        
+        if (c === '(' || c === '[') {
+            return [c, 'TK_START_EXPR'];
+        }
+
+        if (c === ')' || c === ']') {
+            return [c, 'TK_END_EXPR'];
+        }
+
+        if (c === '{') {
+            return [c, 'TK_START_BLOCK'];
+        }
+
+        if (c === '}') {
+            return [c, 'TK_END_BLOCK'];
+        }
+
+        if (c === ';') {
+            return [c, 'TK_END_COMMAND'];
+        }
+
+        if (c === '/') {
+            var comment = '';
+            // peek for comment /* ... */
+            if (input.charAt(parser_pos) === '*') {
+                parser_pos += 1;
+                if (parser_pos < input.length) {
+                    while (! (input.charAt(parser_pos) === '*' && input.charAt(parser_pos + 1) && input.charAt(parser_pos + 1) === '/') && parser_pos < input.length) {
+                        comment += input.charAt(parser_pos);
+                        parser_pos += 1;
+                        if (parser_pos >= input.length) {
+                            break;
+                        }
+                    }
+                }
+                parser_pos += 2;
+                return ['/*' + comment + '*/', 'TK_BLOCK_COMMENT'];
+            }
+            // peek for comment // ...
+            if (input.charAt(parser_pos) === '/') {
+                comment = c;
+                while (input.charAt(parser_pos) !== "\x0d" && input.charAt(parser_pos) !== "\x0a") {
+                    comment += input.charAt(parser_pos);
+                    parser_pos += 1;
+                    if (parser_pos >= input.length) {
+                        break;
+                    }
+                }
+                parser_pos += 1;
+                if (wanted_newline) {
+                    print_newline();
+                }
+                return [comment, 'TK_COMMENT'];
+            }
+
+        }
+
+        if (c === "'" || // string
+        c === '"' || // string
+        (c === '/' &&
+        ((last_type === 'TK_WORD' && last_text === 'return') || (last_type === 'TK_START_EXPR' || last_type === 'TK_END_BLOCK' || last_type === 'TK_OPERATOR' || last_type === 'TK_EOF' || last_type === 'TK_END_COMMAND')))) { // regexp
+            var sep = c;
+            var esc = false;
+            c = '';
+
+            if (parser_pos < input.length) {
+
+                while (esc || input.charAt(parser_pos) !== sep) {
+                    c += input.charAt(parser_pos);
+                    if (!esc) {
+                        esc = input.charAt(parser_pos) === '\\';
+                    } else {
+                        esc = false;
+                    }
+                    parser_pos += 1;
+                    if (parser_pos >= input.length) {
+                        break;
+                    }
+                }
+
+            }
+
+            parser_pos += 1;
+            if (last_type === 'TK_END_COMMAND') {
+                print_newline();
+            }
+            return [sep + c + sep, 'TK_STRING'];
+        }
+
+        if (in_array(c, punct)) {
+            while (parser_pos < input.length && in_array(c + input.charAt(parser_pos), punct)) {
+                c += input.charAt(parser_pos);
+                parser_pos += 1;
+                if (parser_pos >= input.length) {
+                    break;
+                }
+            }
+            return [c, 'TK_OPERATOR'];
+        }
+
+        return [c, 'TK_UNKNOWN'];
+    }
+
+
+    //----------------------------------
+
+    indent_character = indent_character || ' ';
+    indent_size = indent_size || 4;
+
+    indent_string = '';
+    while (indent_size--) {
+        indent_string += indent_character;
+    }
+
+    input = js_source_text;
+
+    last_word = ''; // last 'TK_WORD' passed
+    last_type = 'TK_START_EXPR'; // last token type
+    last_text = ''; // last token text
+    output = [];
+
+    do_block_just_closed = false;
+    var_line = false;
+    var_line_tainted = false;
+
+    whitespace = "\n\r\t ".split('');
+    wordchar = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789_$'.split('');
+    punct = '+ - * / % & ++ -- = += -= *= /= %= == === != !== > < >= <= >> << >>> >>>= >>= <<= && &= | || ! !! , : ? ^ ^= |='.split(' ');
+
+    // words which should always start on new line.
+    line_starters = 'continue,try,throw,return,var,if,switch,case,default,for,while,break,function'.split(',');
+
+    // states showing if we are currently in expression (i.e. "if" case) - 'EXPRESSION', or in usual block (like, procedure), 'BLOCK'.
+    // some formatting depends on that.
+    current_mode = 'BLOCK';
+    modes = [current_mode];
+
+    indent_level = 0;
+    parser_pos = 0; // parser position
+    in_case = false; // flag for parser that case/default has been processed, and next colon needs special attention
+    while (true) {
+        var t = get_next_token(parser_pos);
+        token_text = t[0];
+        token_type = t[1];
+        if (token_type === 'TK_EOF') {
+            break;
+        }
+
+        switch (token_type) {
+
+        case 'TK_START_EXPR':
+            var_line = false;
+            set_mode('EXPRESSION');
+            if (last_type === 'TK_END_EXPR' || last_type === 'TK_START_EXPR') {
+                // do nothing on (( and )( and ][ and ]( ..
+            } else if (last_type !== 'TK_WORD' && last_type !== 'TK_OPERATOR') {
+                print_space();
+            } else if (in_array(last_word, line_starters) && last_word !== 'function') {
+                print_space();
+            }
+            print_token();
+            break;
+
+        case 'TK_END_EXPR':
+            print_token();
+            restore_mode();
+            break;
+
+        case 'TK_START_BLOCK':
+            
+            if (last_word === 'do') {
+                set_mode('DO_BLOCK');
+            } else {
+                set_mode('BLOCK');
+            }
+            if (last_type !== 'TK_OPERATOR' && last_type !== 'TK_START_EXPR') {
+                if (last_type === 'TK_START_BLOCK') {
+                    print_newline();
+                } else {
+                    print_space();
+                }
+            }
+            print_token();
+            indent();
+            break;
+
+        case 'TK_END_BLOCK':
+            if (last_type === 'TK_START_BLOCK') {
+                // nothing
+                trim_output();
+                unindent();
+            } else {
+                unindent();
+                print_newline();
+            }
+            print_token();
+            restore_mode();
+            break;
+
+        case 'TK_WORD':
+
+            if (do_block_just_closed) {
+                print_space();
+                print_token();
+                print_space();
+                break;
+            }
+
+            if (token_text === 'case' || token_text === 'default') {
+                if (last_text === ':') {
+                    // switch cases following one another
+                    remove_indent();
+                } else {
+                    // case statement starts in the same line where switch
+                    unindent();
+                    print_newline();
+                    indent();
+                }
+                print_token();
+                in_case = true;
+                break;
+            }
+
+
+            prefix = 'NONE';
+            if (last_type === 'TK_END_BLOCK') {
+                if (!in_array(token_text.toLowerCase(), ['else', 'catch', 'finally'])) {
+                    prefix = 'NEWLINE';
+                } else {
+                    prefix = 'SPACE';
+                    print_space();
+                }
+            } else if (last_type === 'TK_END_COMMAND' && (current_mode === 'BLOCK' || current_mode === 'DO_BLOCK')) {
+                prefix = 'NEWLINE';
+            } else if (last_type === 'TK_END_COMMAND' && current_mode === 'EXPRESSION') {
+                prefix = 'SPACE';
+            } else if (last_type === 'TK_WORD') {
+                prefix = 'SPACE';
+            } else if (last_type === 'TK_START_BLOCK') {
+                prefix = 'NEWLINE';
+            } else if (last_type === 'TK_END_EXPR') {
+                print_space();
+                prefix = 'NEWLINE';
+            }
+
+            if (last_type !== 'TK_END_BLOCK' && in_array(token_text.toLowerCase(), ['else', 'catch', 'finally'])) {
+                print_newline();
+            } else if (in_array(token_text, line_starters) || prefix === 'NEWLINE') {
+
+                if (last_text === 'else') {
+                    // no need to force newline on else break
+                    print_space();
+                } else if ((last_type === 'TK_START_EXPR' || last_text === '=') && token_text === 'function') {
+                    // no need to force newline on 'function': (function
+                    // DONOTHING
+                } else if (last_type === 'TK_WORD' && (last_text === 'return' || last_text === 'throw')) {
+                    // no newline between 'return nnn'
+                    print_space();
+                } else if (last_type !== 'TK_END_EXPR') {
+                    if ((last_type !== 'TK_START_EXPR' || token_text !== 'var') && last_text !== ':') {
+                        // no need to force newline on 'var': for (var x = 0...)
+                        if (token_text === 'if' && last_type === 'TK_WORD' && last_word === 'else') {
+                            // no newline for } else if {
+                            print_space();
+                        } else {
+                            print_newline();
+                        }
+                    }
+                }
+            } else if (prefix === 'SPACE') {
+                print_space();
+            }
+            print_token();
+            last_word = token_text;
+
+            if (token_text === 'var') {
+                var_line = true;
+                var_line_tainted = false;
+            }
+
+            break;
+
+        case 'TK_END_COMMAND':
+
+            print_token();
+            var_line = false;
+            break;
+
+        case 'TK_STRING':
+
+            if (last_type === 'TK_START_BLOCK' || last_type === 'TK_END_BLOCK') {
+                print_newline();
+            } else if (last_type === 'TK_WORD') {
+                print_space();
+            }
+            print_token();
+            break;
+
+        case 'TK_OPERATOR':
+
+            var start_delim = true;
+            var end_delim = true;
+            if (var_line && token_text !== ',') {
+                var_line_tainted = true;
+                if (token_text === ':') {
+                    var_line = false;
+                }
+            }
+
+            if (token_text === ':' && in_case) {
+                print_token(); // colon really asks for separate treatment
+                print_newline();
+                break;
+            }
+
+            in_case = false;
+
+            if (token_text === ',') {
+                if (var_line) {
+                    if (var_line_tainted) {
+                        print_token();
+                        print_newline();
+                        var_line_tainted = false;
+                    } else {
+                        print_token();
+                        print_space();
+                    }
+                } else if (last_type === 'TK_END_BLOCK') {
+                    print_token();
+                    print_newline();
+                } else {
+                    if (current_mode === 'BLOCK') {
+                        print_token();
+                        print_newline();
+                    } else {
+                        // EXPR od DO_BLOCK
+                        print_token();
+                        print_space();
+                    }
+                }
+                break;
+            } else if (token_text === '--' || token_text === '++') { // unary operators special case
+                if (last_text === ';') {
+                    // space for (;; ++i)
+                    start_delim = true;
+                    end_delim = false;
+                } else {
+                    start_delim = false;
+                    end_delim = false;
+                }
+            } else if (token_text === '!' && last_type === 'TK_START_EXPR') {
+                // special case handling: if (!a)
+                start_delim = false;
+                end_delim = false;
+            } else if (last_type === 'TK_OPERATOR') {
+                start_delim = false;
+                end_delim = false;
+            } else if (last_type === 'TK_END_EXPR') {
+                start_delim = true;
+                end_delim = true;
+            } else if (token_text === '.') {
+                // decimal digits or object.property
+                start_delim = false;
+                end_delim = false;
+
+            } else if (token_text === ':') {
+                // zz: xx
+                // can't differentiate ternary op, so for now it's a ? b: c; without space before colon
+                if (last_text.match(/^\d+$/)) {
+                    // a little help for ternary a ? 1 : 0;
+                    start_delim = true;
+                } else {
+                    start_delim = false;
+                }
+            }
+            if (start_delim) {
+                print_space();
+            }
+
+            print_token();
+
+            if (end_delim) {
+                print_space();
+            }
+            break;
+
+        case 'TK_BLOCK_COMMENT':
+
+            print_newline();
+            print_token();
+            print_newline();
+            break;
+
+        case 'TK_COMMENT':
+
+            // print_newline();
+            print_space();
+            print_token();
+            print_newline();
+            break;
+
+        case 'TK_UNKNOWN':
+            print_token();
+            break;
+        }
+
+        last_type = token_type;
+        last_text = token_text;
+    }
+
+    return output.join('');
+
+}

views/test/reducehaving-reduce.js

@@ -0,0 +1,2 @@
+
+function(ks,vs){return vs.length;}