infod 0.0.3.1 → 0.0.3.2

data/infod/infod.rb

@@ -4,13 +4,12 @@ lambda
 mime
 404
 500
+grep
 audio
 blog
 cal
 code
-css
 csv
-du
 edit
 facets
 feed
@@ -20,7 +19,6 @@ fs
 GET
 glob
 graph
-grep
 groonga
 HEAD
 histogram
@@ -28,7 +26,6 @@ html
 HTTP
 image
 index
-json
 kv
 ls
 mail
@@ -36,13 +33,11 @@ man
 microblog
 names
 page
-PATCH
 POST
 postscript
 rdf
 ruby
 schema
-search
 sh
 text
 threads

data/infod/kv.rb

@@ -21,8 +21,8 @@ class E
       if t.e             # old triple exists?
         t.deleteNode     # remove triple
         indexEdit p,o,'' # unindex
-      end
-      self[p,oO] unless oO.empty? # add triple
+      end                # add
+      self[p,oO] unless oO.class==String && oO.empty? # 3rd arg is new value, empty-string -> nil
     else
       unless t.e          # triple exists?
         indexEdit p,o,nil # index triple
@@ -37,10 +37,12 @@ class E
     end
   end
 
-  def triplrFsStore
-    properties.map{|p|
-      self[p].map{|o|
-        yield uri, p.uri, o}}
+  def triplrDoc &f
+    docBase.glob('#*').map{|s| s.triplrResource &f}
+  end
+
+  def triplrResource
+    properties.map{|p|self[p].map{|o| yield uri, p.uri, o}}
   end
 
   def deletePredicate p

data/infod/lambda.rb

@@ -22,6 +22,8 @@ class E
   F={}
   Watch={}
 
+  NullView = -> d,e {}
+
   def self.dev
     Watch.each{|f,ts|
       if ts < File.mtime(f)

data/infod/ls.rb

@@ -16,6 +16,8 @@ class E
   F['view/'+MIMEtype+'inode/directory'] = F['view/dir']
 
   fn 'view/ls',->i,e{
+    e.q['sort'] ||= 'stat:mtime'
+    e.q['reverse'] ||= true
     dir = e['uri'].E
     path = dir.pathSegment
     up = (!path || path.uri == '/') ? '/' : dir.parent.url
@@ -30,20 +32,24 @@ class E
      {_: :a, class: :up, href: up+'?view=ls', c: '↑'},
      {class: :ls, c: (Fn 'view/table',i,e)},'<br clear=all>',
      {_: :a, class: :down, href: e['uri'].E.url.t, c: '&darr;'}]}
-  
-  # user-patchable default-handler
-  fn '/GET',->e,r{
-    x = 'index.html'
-    i = [e,e.pathSegment].compact.map{|e|e.as x}.find &:e
-    if i
-      if e.uri[-1] == '/' # inside dir?
-        i.env(r).getFile  # show index
-      else                # descend to indexed dir
-        [301, {Location: e.uri.t}, []]
-      end
-    else
-      # default handler
-      e.response
-    end}
 
+  fn 'protograph/du',->d,q,m{
+    d.pathSegment.do{|path|
+    GREP_DIRS.find{|p|path.uri.match p}.do{|ok|
+      e = [d,path].compact.find &:e
+      q['view'] ||= 'table'
+      q['sort'] = Stat+'size'
+      q['reverse'] = true
+      m[e.uri] = e if e
+      rand.to_s.h}}}
+
+  fn 'graph/du',->e,_,m{
+    `du -a #{m.values[0].sh}`.each_line{|l|
+      s,p = l.chomp.split /\t/ # size, path
+      p = p.unpathFs           # path -> URI
+      m[p.uri] = {'uri' => p.uri,
+        Posix+'util#du' => E[p.uri+'?graph=du#du'],
+            Stat+'size' => [s.to_i]}}
+    m }
+  
 end

data/infod/mail.rb

@@ -6,9 +6,19 @@ class E
   rescue LoadError => e
   end
 
-  MessagePath = ->id{
-    h = id.h # hash identifier
-    '/msg/' + h[0..1] + '/' + h[2] + '/' + id}
+  MessagePath = ->id{ h = id.h # hash identifier
+    '/msg/' + h[0..2] + '/' + id}
+
+  GREP_DIRS.push /^\/m\/[^\/]+\// # allow grep within a single address
+
+  F['/m/GET'] = -> e,r{
+    if m = e.pathSegment.uri.match(/^\/m\/([^\/]+)$/)
+      r.q['set'] = 'depth'
+      r.q['view'] ||= 'threads'
+       e.response
+    else
+      false      
+    end}
 
   def triplrTmail &f
     (TMail::Mail.load node).do{|m|      # load
@@ -24,49 +34,48 @@ class E
       yield e, Date, m.date.iso8601 if m.date
       yield e, Title, m.subject.to_utf8
       yield e, Creator, E[creator]
-      yield e, SIOC+'has_discussion', E[e+'?graph=thread']
-
-      yield creator, SIOC+'name', m.friendly_from.to_utf8
+      yield e, SIOC+'has_discussion', E[e+'?graph=thread&view=timegraph#discussion']
+      yield creator, Name, m.friendly_from.to_utf8
       yield creator, DC+'identifier', E['mailto:'+from]
-            posts = '/m/'+from+'#posts'
-      yield creator, SIOC+'creator_of', E[posts]
-      yield posts, Type, E[LDP+'Container']
-      yield posts, LDP+'firstPage', E['/index/sioc:has_creator/'+CGI.escape(creator)]
-      yield e, SIOC+'reply_to', E[URI.escape "mailto:#{m.header['x-original-to']||from}?References=<#{e}>&In-Reply-To=<#{e}>&Subject=#{m.subject.to_utf8}"]
+      yield e, SIOC+'reply_to',
+      E[URI.escape("mailto:#{m.header['x-original-to']||from}?References=<#{id}>&In-Reply-To=<#{id}>&Subject=#{m.subject.to_utf8}&")+'#reply']
 
       %w{to cc bcc}.map{|to|
         m.send(to).do{|to| to.map{|to|
             to = to.to_utf8
-            r = '/m/'+to+'#'+to
-            yield e, To, E[r]
-            yield r, SIOC+'container_of', E['/index/sioc:addressed_to/'+CGI.escape(r)]}}}
+            yield e, To, E['/m/'+to+'#'+to]
+          }}}
 
       %w{in_reply_to references}.map{|ref|
         m.send(ref).do{|refs| refs.map{|r|
           yield e, SIOC+'reply_of', E[MessagePath[r[1..-2]]]}}}
+      m.in_reply_to.do{|refs| refs.map{|r|yield e, SIOC+'has_parent', E[MessagePath[r[1..-2]]]}}
 
-      # RDF:HTML with self-contained minimal styling
+      # RDF:HTML message-body
       yield e, Content,
       H([{_: :pre, class: :mail, style: 'white-space: pre-wrap',
            c: m.concat_message(e.E,0,&f).gsub(/^\s*(&gt;)(&gt;|\s)*\n/,"").lines.to_a.map{|l| # skip quoted empty-lines
-             l.match(/(^\s*(&gt;|On[^\n]+(said|wrote))[^\n]*)\n/) ? {_: :span, class: :q, c: l} : l # wrap quoted lines
-           }},
-         {_: :style, c: "pre.mail .q {background-color:#00f;color:#fff}\npre.mail a{background-color:#ef3}\npre.mail img {max-width:100%}"}])}
+             l.match(/(^\s*(&gt;|On[^\n]+(said|wrote))[^\n]*)\n/) ? {_: :span, class: :q, depth: l.scan(/(&gt;)/).size, c: l} : l # quotes
+           }},(H.css '/css/mail',true)])}
   rescue Exception => e
     puts e
   end
 
   def triplrMailMessage &f
-    insertDocs :triplrTmail, @r['SERVER_NAME'], [To,SIOC+'has_creator',SIOC+'reply_of'], &f
+    # indexing function, called on previously-unseen doc-graphs
+    ix = ->doc, graph, host {
+      graph.map{|u,r|
+        a = [] # addresses
+        r[Creator].do{|c|a.concat c}
+        r[To].do{|t|a.concat t}
+        r[Date].do{|t|
+          st = '/'+t[0].gsub('-','/').sub('T','.').sub(/\+.*/,'.'+u.h[0..1]+'.e')
+          a.map{|rel|
+            doc.ln E[rel.uri.split('#')[0]+st]}}}}
+    addDocs :triplrTmail, @r['SERVER_NAME'], [SIOC+'reply_of'], ix, &f
   end
-=begin
- there's another mail library called Mail, as of v2.5.4 takes 50x as long as tmail (apt-get install ruby-tmail)
-HEAD 200 http://m/m/2013/12/01/?nocache=&triplr=triplrMail curl/7.33.0  5.4003388
-HEAD 200 http://m/m/2013/12/01/?nocache=&triplr=triplrTmail curl/7.33.0  0.1198720
-
-almost a copy of above works but identifiers are not wrapped in <> - with caching it might be fast enough..
 
-=end
+  F['view/'+MIMEtype+'message/rfc822'] = NullView # hide containing file in default render
 
 end
 
@@ -94,7 +103,7 @@ module TMail
       else # just a part
         unicode_body.do{|b|
           if content_type && content_type.match(/html/)
-           (b.split /<body[^>]*>/)[-1].split(/<\/body>/)[0]
+            E::F['cleanHTML'][b]
           else
             b.hrefs true
           end}

data/infod/man.rb

@@ -16,13 +16,13 @@ class E
     if !name || name.empty? || name.match(/\//)
       if section
         # enumerate section children
-        (H [H.css('/css/man'),{_: :style, c: "a {background-color: #{E.cs}}"},
+     body = H [H.css('/css/man'),{_: :style, c: "a {background-color: #{E.cs}}"},
             Pathname(manPath+'/man'+section).c.map{|p|
               n = p.basename.to_s.sub /\.[0-9][a-z]*\...$/,''
             }.group_by{|e|e[0].match(/[a-zA-Z]/) ? e[0].downcase : '0-9'}.sort.map{|g,m|
               [{_: :h3, c: g},
-               m.map{|n|[{_: :a, href: '/man/'+section+'/'+n, c: n },' ']}]}
-           ]).hR
+               m.map{|n|[{_: :a, href: '/man/'+section+'/'+n, c: n },' ']}]}]
+        [200, {'Content-Type'=>'text/html; charset=utf-8'}, [body]]
       else
         e.response
       end

data/infod/microblog.rb

@@ -26,7 +26,7 @@ class E
 
   def tw g
     no.readlines.shuffle.each_slice(22){|s|
-      E['https://twitter.com/search/realtime?q='+s.map{|u|'from:'+u.chomp}.intersperse('+OR+').join].insertDocs :triplrTweets, g}
+      E['https://twitter.com/search/realtime?q='+s.map{|u|'from:'+u.chomp}.intersperse('+OR+').join].addDocs :triplrTweets, g, nil, FeedArchiver}
   end
 
   def triplrTweets
@@ -36,7 +36,7 @@ class E
       yield s, Type, E[SIOCt+'MicroblogPost']
       yield s, Type, E[SIOC+'Post']
       yield s, Creator, E(base+'/'+t.css('.username b')[0].inner_text)
-      yield s, SIOC+'name',t.css('.fullname')[0].inner_text
+      yield s, Name,t.css('.fullname')[0].inner_text
       yield s, Atom+"/link/image", E(t.css('.avatar')[0].attr('src'))
       yield s, Date, Time.at(t.css('[data-time]')[0].attr('data-time').to_i).iso8601
       content = t.css('.tweet-text')[0]
@@ -56,7 +56,7 @@ class E
     [r[Date][0].match(/T([0-9:]{5})/).do{|m|m[1]},
      {_: :span, :class => :nick, c: {_: :a, href: r[Atom+'/link/alternate'].do{|a|a[0].uri}||r.url,
             c: [r[Atom+"/link/image"].do{|p| {_: :img, src: p[0].uri, style: "#{rand(2).zero? ? 'left' : 'right'}: 0"}},
-                {_: :span, c: r[SIOC+'name']||r[Creator]||'#'}]}},' ',
+                {_: :span, c: r[Name]||r[Creator]||'#'}]}},' ',
         {_: :span, :class => :tw, # skip redundant title fields
        c: [((r[Title].to_s == r[Content].to_s || r.uri.match(/twitter/)) && '' ||
             {_: :a, :class => :title, href: r.url, c: r[Title]}), # skip quoted mail-lines & abbreviate

data/infod/mime.rb

@@ -79,6 +79,7 @@ class E
     nfo: 'text/nfo',
     nt:  'text/ntriples',
     ntriples:  'text/ntriples',
+    org: 'application/org',
     owl: 'application/rdf+xml',
     pdf: 'application/pdf',
     pl:  'application/perl',

data/infod/names.rb

@@ -1,9 +1,10 @@
-%w{base64 cgi shellwords}.each{|r|require(r)}
+%w{cgi shellwords}.each{|r|require(r)}
 
 class E
 
   attr_reader :uri
   alias_method :url, :uri
+  alias_method :maybeURI, :uri
 
   def env r=nil
     r ? (@r = r
@@ -58,7 +59,7 @@ class E
   end
   alias_method :dir, :dirname
 
-  # add hostname to URI if missing
+  # add hostname to URI (if missing)
   def hostURL e
     host = 'http://'+e['SERVER_NAME']
     if uri.index('/') == 0 
@@ -70,15 +71,12 @@ class E
 
   # pointer to local data about global URI
   def localURL e
-    # path
-    if uri.index('/') == 0 # already a local path
-      uri
-    # host match
+    if uri.index('/') == 0
+      uri             # already a local path
     elsif e && uri.index('http://'+e['SERVER_NAME']+'/') == 0 
-      pathSegment.uri
-    # non-local
+      pathSegment.uri # host match, unchanged local path
     else
-      URIURL + (CGI.escape uri)
+      '/' + uri       # URI -> local path
     end
   end
 
@@ -87,7 +85,6 @@ class E
     m && m[2] && m[2].E || nil
   end
 
-  # URI extension :: E -> string
   def ext
     File.extname(uri).tail||''
   end
@@ -124,39 +121,25 @@ class E
   alias_method :+, :appendURI
   alias_method :as, :appendSlashURI
 
-  def path?
-    uri.path?
-  end
-
   def shortPath
     @shortPath ||=
-      (if path?
-         if uri.match /^\//
-           uri
-         else
-           '/' + uri.shorten
-         end
+      (if uri.match /^\//
+         uri
        else
-         '/E/' + uri.h.dive[0..5] + (Base64.urlsafe_encode64 uri)
+         '/' + uri.shorten
        end)
   end
 
-  # URI -> path
   def path
-    @path ||=
-      (if path?
-         if uri.match /^\//
-           uri
-         else
-           '/' + uri
-         end
-       else
-         '/E/' + uri.h.dive[0..5] + (Base64.urlsafe_encode64 uri)
-       end)
+    if uri.match /^\//
+      uri
+    else
+      '/' + uri
+    end
   end
 
   def u
-    # data-storage path for resource
+    # path for data about this resource
     @u ||= E (f ? dirname + '/.' + (File.basename path) : path.t + '._')
   end
 
@@ -170,56 +153,22 @@ class E
     d.force_encoding('UTF-8').sh
   end
 
-  # literals to URIs
-  # currently used for iso8601 dates mapping to paths, so date-range queries (depth-first subtrees) can be done w/ dir/fs tools
-  # could also use as a "trie" for autocomplete + sorted-strings
   def E.literal o
-    E['/'].literal o
+    ''.E.literal o
   end
-  
-  Literal={}
-   [Purl+'dc/elements/1.1/date',
-    Date,DC+'created',DC+'modified',
-   ].map{|f|Literal[f]=true}
 
   def literal o
-
-    # already a URI
     return o if o.class == E
-
-    # blob for non-strings
-    return literalBlob o unless o.class == String
-
-    # whitelisted predicateURIs to paths
-    return literalURI o if (Literal[uri] || o.size<=88) && !o.match(/\//)
-
-    # string matches URI format
-    return E o if o.match %r{\A[a-z]+://[^\s]+\Z}
-
-    # blob
-    literalBlob o
-
-  end
-
-  # pathname for short literals
-  def literalURI o
-    E "/l/"+o.gsub(/[\.:\-T+]/,'/')+'/'+o if Literal[uri] && o
-  end
- 
-  def literalBlobURI o
-    if o.class == String
-      E "/E/blob/"+o.h.dive
-    else
-      E "/E/json/"+[o].to_json.h.dive
-    end
-  end
-
-  def literalBlob o
-    u = literalBlobURI o
+    u = (if o.class == String
+           E "/E/blob/"+o.h.dive
+         else
+           E "/E/json/"+[o].to_json.h.dive
+         end)
     u.w o, !o.class == String unless u.f
+    u
   end
 
-  # spaceship
+  # spaceship comparison-operator
   def <=> c
     to_s <=> c.to_s
   end
@@ -232,7 +181,7 @@ class E
    {'uri' => uri}
   end
 
-  # implementation-specific internal pathnames not on the web
+  # internal pathnames not on the web (cached representations, index databases)
   F['/E/GET'] = F[E404]
 
 end
@@ -242,6 +191,7 @@ class Hash
     self["uri"]||""
   end
   alias_method :url, :uri
+  alias_method :maybeURI, :uri
   def label
     self[E::Label] || uri.label
   end
@@ -274,7 +224,7 @@ class String
      self )
   end
 
-  # shrink URI to qname/CURIE/prefix'd identifier
+  # shrink URI to qname/CURIE/prefixed identifier
   def shorten
     E::Prefix.map{|p,f|
       return p + ':' + self[f.size..-1]  if (index f) == 0
@@ -298,16 +248,10 @@ class String
     if m = (match /^\/([a-z]+:)\/+(.*)/)
       (m[1] + '//' + m[2]).E
 
-    # CURIE
+    # prefix-shortened URI
     elsif m = (match /^\/([^\/:]+:[^\/]+)/)
       m[1].expand.E
 
-    # opaque URI w/ optional extension
-    elsif match /^\/E\/..\//
-      self[9..-1].match(/([^.]+)(.*)/).do{|c|
-        (Base64.urlsafe_decode64 c[1]) + c[2]
-    }.E
-
     # String literal
     elsif match /^\/E\/blob/
       self.E.r
@@ -316,10 +260,6 @@ class String
     elsif match /^\/E\/json/
       self.E.r true
 
-    # literal in basename
-    elsif match /^\/l\//
-      File.basename self
-
     # plain path
     else
       self.E
@@ -331,10 +271,6 @@ class String
     E.new self
   end
 
-  def path?
-    (match /^(\.|\/|https?:\/)/) && true || false
-  end
-
   def frag
     split(/#/).pop()
   end