bibtex_to_scrapbox 0.1.0

data/gems/gems/latex-decode-0.2.2/lib/latex/decode/base.rb

@@ -0,0 +1,62 @@
+# -*- coding: utf-8 -*-
+
+module LaTeX
+  module Decode
+
+    class Decoder
+      class << self
+        attr_reader :patterns, :map
+
+        def inherited (base)
+          subclasses << base
+        end
+
+        def subclasses
+          @subclasses ||= []
+        end
+
+        def decode (string)
+          decode!(string.dup)
+        end
+
+        def decode! (string)
+          patterns.each do |pattern|
+            string.gsub!(pattern) { |m| [$2,map[$1],$3].compact.join }
+          end
+          string
+        end
+      end
+    end
+
+    module Base
+
+      module_function
+
+      def normalize (string)
+        string.gsub!(/\\(?:i|j)\b/) { |m| m == '\\i' ? 'ı' : 'ȷ' }
+
+        # \foo\ bar -> \foo{} bar
+        string.gsub!(/(\\[a-zA-Z]+)\\(\s+)/, '\1{}\2')
+
+        # Aaaa\o, -> Aaaa\o{},
+        string.gsub!(/([^{]\\\w)([;,.:%])/, '\1{}\2')
+
+        # \c cb -> \c{cb}
+        string.gsub!(/(\\[^\sij&#\$\{\}_~%])\s+([[:alpha:]]+)\b/i, '\1{\2}')
+
+        # non-breaking spaces
+        string.gsub!(/(\A|[^\\])~/, LaTeX.to_unicode("\\1\u00a0"))
+
+        string
+      end
+
+      def strip_braces (string)
+        string.gsub!(/(^|[^\\])([\{\}]+)/, '\1')
+        string.gsub!(/\\(\{|\})/, '\1')
+        string
+      end
+
+    end
+
+  end
+end

data/gems/gems/latex-decode-0.2.2/lib/latex/decode/compatibility.rb

@@ -0,0 +1,89 @@
+
+if RUBY_VERSION < "1.9"
+  $KCODE = 'U'
+
+  module LaTeX
+    def self.to_unicode(string)
+      string.gsub(/\\?u([\da-f]{4})/i) { |m| [$1.to_i(16)].pack('U') }
+    end
+  end
+
+  def ruby_18; yield; end
+  def ruby_19; false; end
+else
+
+  module LaTeX
+    def self.to_unicode(string)
+      string
+    end
+  end
+
+  def ruby_18; false; end
+  def ruby_19; yield; end
+end
+
+if RUBY_PLATFORM == 'java'
+  require 'java'
+
+  # Use the Java native Unicode normalizer
+  module LaTeX
+    def self.normalize_C(string)
+      java.text.Normalizer.normalize(string, java.text.Normalizer::Form::NFC).to_s
+    end
+  end
+
+else
+  begin
+    require 'unicode'
+
+    # Use the Unicode gem
+    module LaTeX
+      def self.normalize_C(string)
+        Unicode::normalize_C(string)
+      end
+    end
+  rescue LoadError
+    begin
+      require 'active_support/multibyte/chars'
+
+      # Use ActiveSupport's normalizer
+      module LaTeX
+        def self.normalize_C(string)
+          ActiveSupport::Multibyte::Chars.new(string).normalize(:c).to_s
+        end
+      end
+    rescue LoadError
+      fail "Failed to load unicode normalizer: please gem install unicode (or active_support)"
+    end
+  end
+end
+
+module LaTeX
+  begin
+    require 'ritex'
+
+    def self.ritex
+      Ritex::Parser.new(:mathml)
+    end
+
+    def self.to_math_ml(string)
+      ritex.parse string, :nowrap => true, :display => false
+    end
+
+  rescue LoadError
+    begin
+      require 'math_ml'
+
+      def self.to_math_ml(string)
+        MathML::String.mathml_latex_parser.parse(string, false)
+      end
+
+    rescue LoadError
+      # No MathML conversion
+
+      def self.to_math_ml(string)
+        string
+      end
+    end
+  end
+end

data/gems/gems/latex-decode-0.2.2/lib/latex/decode/diacritics.rb

@@ -0,0 +1,46 @@
+# -*- coding: utf-8 -*-
+
+module LaTeX
+  module Decode
+    
+    class Diacritics < Decoder
+      @macros = Hash[*%W{
+        r \u030A
+        H \u030B
+        u \u0306
+        v \u030C
+        G \u030F
+        M \u0322
+        d \u0323
+        c \u0327
+        k \u0328
+        b \u0331
+        B \u0335
+        t \u0361
+      }.map { |s| LaTeX.to_unicode(s) }].freeze
+
+      @map = @macros.merge(Hash[*%w{
+        l  ł
+        L  Ł
+        o  ø
+        O  Ø
+        AA Å
+        aa å
+        AE Æ
+        ae æ
+      }]).freeze
+      
+      @patterns = [
+        ruby_18 {
+          /\\(#{ @macros.keys.map { |k| Regexp.escape(k) }.join('|') })\{([[:alpha:]]?)([[:alpha:]]*)\}/ou          
+        } ||
+        ruby_19 {
+          /\\(#{ @macros.keys.map { |k| Regexp.escape(k) }.join('|') })\{(\p{L}\p{M}*)([[:alpha:]]*)\}/ou
+        },
+        /\\(o|O|l|L|aa|AA|ae|AE)\b/,
+      ].freeze
+      
+    end
+    
+  end
+end

data/gems/gems/latex-decode-0.2.2/lib/latex/decode/greek.rb

@@ -0,0 +1,61 @@
+# -*- coding: utf-8 -*-
+
+module LaTeX
+  module Decode
+
+    class Greek < Decoder
+      @map = Hash[*%w{
+        alpha      α
+        beta       β
+        gamma      γ
+        delta      δ
+        epsilon    ε
+        zeta       ζ
+        eta        η
+        theta      θ
+        iota       ι
+        kappa      κ
+        lambda     λ
+        mu         μ
+        nu         ν
+        xi         ξ
+        rho        ρ
+        sigma      σ
+        tau        τ
+        upsilon    υ
+        phi        φ
+        chi        χ
+        psi        ψ
+        omega      ω
+        Alpha      Α
+        Beta       Β
+        Gamma      Γ
+        Delta      Δ
+        Epsilon    Ε
+        Zeta       Ζ
+        Eta        Η
+        Theta      Θ
+        Iota       Ι
+        Kappa      Κ
+        Lambda     Λ
+        Mu         Μ
+        Nu         Ν
+        Xi         Ξ
+        Rho        Ρ
+        Sigma      Σ
+        Tau        Τ
+        Upsilon    Υ
+        Phi        Φ
+        Chi        Χ
+        Psi        Ψ
+        Omega      Ω
+      }].freeze
+
+      @patterns = [
+        /\\(#{ map.keys.map { |k| Regexp.escape(k) }.join('|') })(?:\{\}|\s+|\b)/ou
+      ].freeze
+
+    end
+
+  end
+end

data/gems/gems/latex-decode-0.2.2/lib/latex/decode/maths.rb

@@ -0,0 +1,21 @@
+module LaTeX
+  module Decode
+
+    class Maths < Decoder
+      @patterns = [
+        /\$([^\$]+)\$/
+      ].freeze
+
+      def self.decode! (string)
+        patterns.each do |pattern|
+          string.gsub!(pattern) do
+            LaTeX.to_math_ml($1)
+          end
+        end
+        string
+      end
+
+    end
+
+  end
+end

data/gems/gems/latex-decode-0.2.2/lib/latex/decode/punctuation.rb

@@ -0,0 +1,59 @@
+# -*- coding: utf-8 -*-
+
+module LaTeX
+  module Decode
+
+    class Punctuation < Decoder
+
+      @macros = Hash[*%w{
+        textendash         –
+        textemdash         —
+        textquoteleft      ‘
+        textquoteright     ’
+        quotesinglbase     ‚
+        textquotedblleft   “
+        textquotedblright  ”
+        quotedblbase       „
+        dag                †
+        ddag               ‡
+        textbullet         •
+        dots               …
+        textperthousand    ‰
+        textpertenthousand ‱
+        guilsinglleft      ‹
+        guilsinglright     ›
+        textreferencemark  ※
+        textinterrobang    ‽
+        textoverline       ‾
+        langle             ⟨
+        rangle             ⟩
+        textasciicircum    ^
+        textbackslash      \\
+        textasciitilde     ~
+        textquotesingle    '
+      }].freeze
+
+      @symbols = Hash[*%w[
+        -     -
+        --    –
+        ---   —
+        ``    “
+        `     ‘
+        ''    ”
+        '     ’
+      ]].freeze
+
+      @map = @macros.merge(@symbols).freeze
+
+      @patterns = [
+        /\\(#{ @macros.keys.map { |k| Regexp.escape(k) }.compact.join('|') })(?:\{\}|\s+|\b|$)/ou,
+        /(-+|`{1,2}|'{1,2})/,
+        /()\\([$%;#_&])(\{\})?/,
+        /()\\(~)\{\}/,
+        /()\\( )/
+      ].freeze
+
+    end
+
+  end
+end

data/gems/gems/latex-decode-0.2.2/lib/latex/decode/symbols.rb

@@ -0,0 +1,225 @@
+# -*- coding: utf-8 -*-
+
+module LaTeX
+  module Decode
+
+    class Symbols < Decoder
+
+      @map = Hash[*%W{
+        textcolonmonetary ₡
+        textlira          ₤
+        textnaira         ₦
+        textwon           ₩
+        textdong          ₫
+        euro              €
+        texteuro          €
+        textnumero        №
+        texttrademark     ™
+        leftarrow         ←
+        uparrow           ↑
+        rightarrow        →
+        downarrow         ↓
+        leftrightarrow    ↔
+        updownarrow       ↕
+        leadsto           ↝
+        rightleftharpoons ⇌
+        Rightarrow        ⇒
+        Leftrightarrow    ⇔
+        forall            ∀
+        complement        ∁
+        partial           ∂
+        exists            ∃
+        nexists           ∄
+        set               ∅
+        Delta             ∆
+        nabla             ∇
+        in                ∈
+        notin             ∉
+        ni                ∋
+        prod              ∏
+        coprod            ∐
+        sum               ∑
+        mp                ∓
+        dotplus           ∔
+        setminus          ∖
+        ast               ∗
+        circ              ∘
+        bullet            ∙
+        surd              √
+        propto            ∝
+        infty             ∞
+        angle             ∠
+        measuredangle     ∡
+        sphericalangle    ∢
+        mid               ∣
+        nmid              ∤
+        parallel          ∥
+        nparallel         ∦
+        wedge             ∧
+        vee               ∨
+        cap               ∩
+        cup               ∪
+        int               ∫
+        iint              ∬
+        iiint             ∭
+        oint              ∮
+        therefore         ∴
+        because           ∵
+        sim               ∼
+        backsim           ∽
+        wr                ≀
+        nsim              ≁
+        simeq             ≃
+        cong              ≅
+        ncong             ≇
+        approx            ≈
+        approxeq          ≊
+        asymp             ≍
+        Bumpeq            ≎
+        bumpeq            ≏
+        doteq             ≐
+        doteqdot          ≑
+        fallingdotseq     ≒
+        risingdotseq      ≓
+        eqcirc            ≖
+        circeq            ≗
+        triangleq         ≜
+        neq               ≠
+        equiv             ≡
+        leq               ≤
+        geq               ≥
+        leqq              ≦
+        geqq              ≧
+        lneqq             ≨
+        gneqq             ≩
+        ll                ≪
+        gg                ≫
+        between           ≬
+        nless             ≮
+        ngtr              ≯
+        nleq              ≰
+        ngeq              ≱
+        lesssim           ≲
+        gtrsim            ≳
+        lessgtr           ≶
+        gtrless           ≷
+        prec              ≺
+        succ              ≻
+        preccurlyeq       ≼
+        succcurlyeq       ≽
+        precsim           ≾
+        succsim           ≿
+        nprec             ⊀
+        nsucc             ⊁
+        subset            ⊂
+        supset            ⊃
+        subseteq          ⊆
+        supseteq          ⊇
+        nsubseteq         ⊈
+        nsupseteq         ⊉
+        subsetneq         ⊊
+        supsetneq         ⊋
+        uplus             ⊎
+        sqsubset          ⊏
+        sqsupset          ⊐
+        sqsubseteq        ⊑
+        sqsupseteq        ⊒
+        sqcap             ⊓
+        sqcup             ⊔
+        oplus             ⊕
+        ominus            ⊖
+        otimes            ⊗
+        oslash            ⊘
+        odot              ⊙
+        circledcirc       ⊚
+        circledast        ⊛
+        circleddash       ⊝
+        boxplus           ⊞
+        boxminus          ⊟
+        boxtimes          ⊠
+        boxdot            ⊡
+        vdash             ⊢
+        dashv             ⊣
+        top               ⊤
+        bot               ⊥
+        Vdash             ⊩
+        Vvdash            ⊪
+        nVdash            ⊮
+        lhd               ⊲
+        rhd               ⊳
+        unlhd             ⊴
+        unrhd             ⊵
+        multimap          ⊸
+        intercal          ⊺
+        veebar            ⊻
+        barwedge          ⊼
+        bigwedge          ⋀
+        bigvee            ⋁
+        bigcap            ⋂
+        bigcup            ⋃
+        diamond           ⋄
+        cdot              ⋅
+        star              ⋆
+        divideontimes     ⋇
+        bowtie            ⋈
+        ltimes            ⋉
+        rtimes            ⋊
+        leftthreetimes    ⋋
+        rightthreetimes   ⋌
+        backsimeq         ⋍
+        curlyvee          ⋎
+        curlywedge        ⋏
+        Subset            ⋐
+        Supset            ⋑
+        Cap               ⋒
+        Cup               ⋓
+        pitchfork         ⋔
+        lessdot           ⋖
+        gtrdot            ⋗
+        lll               ⋘
+        ggg               ⋙
+        lesseqgtr         ⋚
+        gtreqless         ⋛
+        curlyeqprec       ⋞
+        curlyeqsucc       ⋟
+        lnsim             ⋦
+        gnsim             ⋧
+        precnsim          ⋨
+        succnsim          ⋩
+        ntriangleleft     ⋪
+        ntriangleright    ⋫
+        ntrianglelefteq   ⋬
+        ntrianglerighteq  ⋭
+        vdots             ⋮
+        cdots             ⋯
+        ddots             ⋱
+        lceil             ⌈
+        rceil             ⌉
+        lfloor            ⌊
+        rfloor            ⌋
+        Box               □
+        spadesuit         ♠
+        heartsuit         ♡
+        diamondsuit       ♢
+        clubsuit          ♣
+        flat              ♭
+        natural           ♮
+        sharp             ♯
+        tone5             ˥
+        tone4             ˦
+        tone3             ˧
+        tone2             ˨
+        tone1             ˩
+        ss                ß
+        ,                 \u2009
+      }.map { |s| LaTeX.to_unicode(s) }].freeze
+
+
+      @patterns = [
+        /\\(#{ map.keys.map { |k| Regexp.escape(k) }.join('|') })(?:\{\}|\s+|\b)/ou
+      ].freeze
+
+    end
+
+  end
+end