Always convert ODT to MD using LC_ALL=C.UTF-8 for well known result.

[text-madewithcc.git] / fixup-migrate-gettext.rb

diff --git a/fixup-migrate-gettext.rb b/fixup-migrate-gettext.rb
index eb255daa359f0d40e2ccafe3b73c7db7fcb705e0..b2d9c7ca28c4f6326bda2080e3a3528e666eae79 100755 (executable)
--- a/fixup-migrate-gettext.rb
+++ b/fixup-migrate-gettext.rb
@@ -1,12 +1,14 @@
 #!/usr/bin/ruby
+# coding: utf-8
 class PoStr
   Sections = %w(Headers Source Translated)
-  attr_accessor :status, :src, :dest, :type, :location
+  attr_accessor :status, :src, :srcclean, :dest, :prev, :prevclean, :type, :location
   def initialize(text)
     section = 0
     @status = ''
     @src = []
     @dest = []
+    @prev = []
 
     text.split(/\n/).each do |lin|
       # Header
@@ -19,7 +21,16 @@ class PoStr
           @status = 'fuzzy' if type == ',' and field == 'fuzzy'
           @type = value if type == '.' and field == 'type'
           @location = {:file => field, :line => value} if type == ':'
-        elsif lin =~ /^msgid (".*")/
+          if type == '|'
+            #print lin, "\n"
+            if lin =~ /^\#\| msgid "(.*)"$/
+              #print "first '%s'\n" % $1
+              @prev << $1
+            elsif lin =~ /^\#\| "(.*)"$/
+              @prev << $1
+            end
+          end
+        elsif lin =~ /^msgid "(.*)"/
           section += 1
           @src << $1
         else
@@ -28,9 +39,9 @@ class PoStr
 
       # Source text
       elsif section == 1
-        if lin =~ /^".*"$/
-          @src << lin
-        elsif lin =~ /^msgstr (".*")/
+        if lin =~ /^"(.*)"$/
+          @src << $1
+        elsif lin =~ /^msgstr "(.*)"/
           section += 1
           @dest << $1
         else
@@ -39,8 +50,8 @@ class PoStr
 
       # Translated text
       else
-        if lin =~ /^".*"$/
-          @dest << lin
+        if lin =~ /^"(.*)"$/
+          @dest << $1
         else
           boom(section, lin, text)
         end
@@ -56,9 +67,129 @@ class PoStr
   def fuzzy?
     @status == 'fuzzy'
   end
+
+  def srcstr()
+    return @src.join("")
+  end
+
+  def deststr()
+    return @dest.join("")
+  end
+
+  def prevstr()
+    return @prev.join("")
+  end
+
+  def output()
+    # No use outputting empty translations
+    if '' == srcstr() && '' == deststr()
+      return
+    end
+    if '' != @type
+      print "#. type:%s\n" % [@type]
+    end
+    if @location
+      print "#: %s:%s\n" % [@location[:file], @location[:line]]
+    end
+    if fuzzy?
+      print "#, fuzzy\n"
+      if '' != prevstr()
+        @prev.each_with_index do |lin, idx|
+          if idx == 0
+            print "#| msgid \"%s\"\n" % lin
+          else
+            print "#| \"%s\"\n" % lin
+          end
+        end
+      end
+    end
+    @src.each_with_index do |lin, idx|
+      if idx == 0
+        print "msgid \"%s\"\n" % lin
+      else
+        print "\"%s\"\n" % lin
+      end
+    end
+    @dest.each_with_index do |lin, idx|
+      if idx == 0
+        print "msgstr \"%s\"\n" % lin
+      else
+        print "\"%s\"\n" % lin
+      end
+    end
+    print "\n"
+  end
 end
 
 raise ArgumentError, 'Source file not specified' if ARGV.size != 1
 file = ARGV[0]
 strings = File.open(file,'r').read.split(/\n\n/)[1..-1].map {|str| PoStr.new(str)}
-print strings
+
+#print strings
+
+c = 0
+strings.each do |entry|
+  # Ignore whitespace changes between prev and src
+  entry.prevclean = entry.prevstr().gsub(/ +/, ' ')
+  entry.srcclean = entry.srcstr().gsub(/ +/, ' ')
+    
+  if entry.fuzzy?
+    if entry.prevclean == entry.srcclean
+      entry.status = ''
+      #print "# clear fuzzy (space)\n"
+      c = c + 1
+      entry.output
+      next
+    end
+
+    # Rewrite title strings, which lost '#' at the front
+    if entry.prevclean.gsub(/^#+ +/, '') == entry.srcclean
+      # FIXME Not safe to clear fuzzy flag, might have bogus translation
+      entry.status = ''
+      if entry.dest[0].gsub!(/^#+ +/, '') ||
+         ('' == entry.dest[0] && entry.dest[1].gsub!(/^#+ +/, ''))
+        #print "cleared fuzzy (title)\n"
+        c = c + 1
+      end
+    end
+
+    # Rewrite footnotes to use "<placeholder type=\"footnote\" id=\"0\"/>"
+    if entry.srcclean =~ /<placeholder type=\\"footnote\\" id=\\"0\\"\/>/
+      #print "found footnote\n"
+      p = entry.prevclean
+      p.sub!(/([a-z]\.["”]?)(\d+)(\s)/, "\\1<placeholder type=\\\"footnote\\\" id=\\\"0\\\"\/>\\3")
+      p.sub!(/([a-z]\.["”]?)(\d+)$/, "\\1<placeholder type=\\\"footnote\\\" id=\\\"0\\\"\/>")
+      #print p, "\n"
+      #print entry.src, "\n"
+      if p == entry.srcclean
+        replaced = false
+        entry.dest.each do |part|
+          if part.sub!(/([a-z]\.["”»]?)(\d+)(\s)/, "\\1<placeholder type=\\\"footnote\\\" id=\\\"0\\\"\/>\\3")
+            replaced = true
+          end
+        end
+        entry.dest.each do |part|
+          if part.sub!(/([a-z]\.["”»]?)(\d+)$/, "\\1<placeholder type=\\\"footnote\\\" id=\\\"0\\\"\/>")
+            replaced = true
+          end
+        end
+        if replaced
+          #print "cleared fuzzy (footnote)\n"
+          entry.status = ''
+          c = c + 1
+        end
+      end
+    end
+  end
+
+  # Copy image references and single URLs unchanged, even for non-fuzzies
+  if entry.prevclean =~ /!\[\]\(Pictures\// ||
+     entry.srcclean =~ /^<ulink url=\\\".+\\\"\/>$/
+    entry.dest = entry.src.clone()
+    entry.status = ''
+    c = c + 1
+  end
+
+  entry.output
+end
+print "# Cleared %d fuzzy strings\n" % [c]