isolate podofo to prepare for native parser, autogenerate PDF API during build

experimental/PdfViewer/spec2def.py

 #!/usr/local/bin/python
 # coding: utf-8
 
 import sys
 import re
 
 # TODO(edisonn): put processed part of file in a new file
 # put unprocessed part, in a new file, so we see what we miss
 # keep blank lines, and generate a version without the blank lines
 
@@ skipping 274 matching lines @@
   ret = ret.replace(u'\u2122', '(TM)')
 
 
   # how enable to emit this a python string
   ret = ret.replace('\'', '\\\'')
   ret = ret.replace('\n', '\\n')
 
   
   return ret
   
-def commitRow():
+def commitRow(fspecPy):
   global columnValues
   global emitedDitionaryName
   global table
   global tableToClassName
   
   if columnValues == None:
     return
   
   #print columnValues
   
@@ skipping 80 matching lines @@
     #else:
       #print table
     
     #print tableKey
     #print('\'' + tableKey + '\': [\'' + emitedDitionaryName + '\', \'' + table[len(tableKey) + 1:] + '\'],')
 
     emitedDitionaryName = tableToClassName[tableKey][0]
     comment = fix(tableToClassName[tableKey][1])
     
     if len(tableToClassName[tableKey]) >= 3 and tableToClassName[tableKey][2] != '':
-      print('  pdfspec.addClass(\'' + emitedDitionaryName + '\', \'' + tableToClassName[tableKey][2] + '\', \'' + comment + '\')\\')
+      fspecPy.write('  pdfspec.addClass(\'' + emitedDitionaryName + '\', \'' + tableToClassName[tableKey][2] + '\', \'' + comment + '\')\\\n')
     else:
-      print('  pdfspec.addClass(\'' + emitedDitionaryName + '\', \'Dictionary\', \'' + comment + '\')\\')
+      fspecPy.write('  pdfspec.addClass(\'' + emitedDitionaryName + '\', \'Dictionary\', \'' + comment + '\')\\\n')
 
   if len(tableToClassName[tableKey]) >= 4 and columnValues[0] in tableToClassName[tableKey][3]:
     required = True
 
   if required:
-    print('      .required(\'NULL\')\\')
+    fspecPy.write('      .required(\'NULL\')\\\n')
   else:
-    print('      .optional()\\')
+    fspecPy.write('      .optional()\\\n')
     
-  print('          .field(\'' + columnValues[0] + '\')\\')
-  print('          .name(\'' + columnValues[0] + '\')\\')
-  print('          .type(\'' + columnValues[1] + '\')\\')
-  print('          .comment(\'' + columnValues[2] + '\')\\')
+  fspecPy.write('          .field(\'' + columnValues[0] + '\')\\\n')
+  fspecPy.write('          .name(\'' + columnValues[0] + '\')\\\n')
+  fspecPy.write('          .type(\'' + columnValues[1] + '\')\\\n')
+  fspecPy.write('          .comment(\'' + columnValues[2] + '\')\\\n')
 
   if len(tableToClassName[tableKey]) >= 4 and columnValues[0] in tableToClassName[tableKey][3]:
-    print('          .must(' + tableToClassName[tableKey][3][columnValues[0]] + ')\\')
+    fspecPy.write('          .must(' + tableToClassName[tableKey][3][columnValues[0]] + ')\\\n')
 
-  print('          .done().done()\\')
+  fspecPy.write('          .done().done()\\\n')
   
   
   columnValues = None
   
 def newRow(first, second, third):
   global columnValues 
   columnValues = [first.rstrip(), second.rstrip(), third.rstrip()]
 
 def appendRow(second, third):
   global columnValues
   if second.rstrip() != '':
     columnValues[1] = columnValues[1] + ' ' + second.rstrip()
   if third.rstrip() != '':
     columnValues[2] = columnValues[2] + '\n' + third.rstrip()
 
-def rebaseTable(line):
+def rebaseTable(fspecPy, line):
   global knownTypes
   global columnWidth
   
   line2 = line.replace(',', ' , ')
   
   words = line2.split()
   
   if len(words) < 3:
     return False
 
   i = 1
   while i < len(words) - 1 and words[i] in knownTypes:
     i = i + 1
     
   if words[i].startswith('(Optional') or words[i].startswith('(Required'):
-    commitRow()
+    commitRow(fspecPy)
     
     columnWidth[0] = line.find(words[1])
     
     if words[i].startswith('(Optional'):
       columnWidth[1] = line.find('(Optional') - columnWidth[0] 
     if words[i].startswith('(Required'):
       columnWidth[1] = line.find('(Required') - columnWidth[0] 
     return True
     
   return False
     
     
-def stopTable():
+def stopTable(fspecPy):
   global tableHeaderFound
   global emitedDitionaryName
 
   if not inTable():
     return
   
-  commitRow()
+  commitRow(fspecPy)
   tableHeaderFound = False
   emitedDitionaryName = ''
-  print('      .done()')
-  print
+  fspecPy.write('      .done()\n')
+  fspecPy.write('\n')
     
 
 def killTable():
   return
 
-def processLineCore(line):
+def processLineCore(fspecPy, line):
   global lines
   global tableLine
   global tableRow
   global columnWidth
   global columnValues
   global mustFollowTableHeader
   
-  global fnewspec
+  #global fnewspec
   
   lines = lines + 1
   
   line = unicode(line, 'utf8')
   
   striped = line.rstrip()
   
   words = line.split()
   if len(words) == 0:
-    stopTable()
+    stopTable(fspecPy)
     return False
         
   isTableHeader = re.search('^[\s]*(TABLE [0-9].[0-9][0-9]?)', striped)
   if isTableHeader:
-    stopTable()
+    stopTable(fspecPy)
     tableDescriptionFound(striped)
     mustFollowTableHeader = True
     return False
   
   if mustFollowTableHeader:
     mustFollowTableHeader = False
     if len(words) != 3:
       killTable()
       return False
  
@@ skipping 16 matching lines @@
     third = striped[columnWidth[0] + columnWidth[1] :]
 
     if tableLine == 1:
       if third[0] != '(':
         killTable()
         return False
 
       newRow(first, second, third)
       return True
     
-    if rebaseTable(striped):
+    if rebaseTable(fspecPy, striped):
       first = striped[0 : columnWidth[0]]
       second = striped[columnWidth[0] : columnWidth[0] + columnWidth[1]]
       third = striped[columnWidth[0] + columnWidth[1] :]
     
     first = first.rstrip()
     second = second.rstrip()
     third = third.rstrip()
         
     if first == '' and second == '' and third != '':
       appendRow(second, third)
       return True
       
     if len(first.split()) > 1:
-      stopTable()
+      stopTable(fspecPy)
       return False
 
     if first != '' and first[0] == ' ':
-      stopTable()
+      stopTable(fspecPy)
       return False
 
     if first != '' and second != '' and third == '':
-      stopTable()
+      stopTable(fspecPy)
       return False
 
     if first == '' and second != '' and second[0] != ' ':
       if acceptType(second):
         appendRow(second, third)
         return True
       else:
-        stopTable()
+        stopTable(fspecPy)
         return False
 
     if first != '' and second != '' and third[0] != '(':
       stopTable()
       return False
       
     if first == '' and second != '' and second[0] == ' ':
-      stopTable()
+      stopTable(fspecPy)
       return False
 
     if first != '' and second != '' and third[0] == '(':
-      commitRow()
+      commitRow(fspecPy)
       newRow(first, second, third)
       return True
     
     return False
   return False
   
-def processLine(line):
-  global fnewspec
+def processLine(fspecPy, line):
+  #global fnewspec
   
-  inSpec = processLineCore(line)
+  inSpec = processLineCore(fspecPy, line)
   
   #just return, use the next lines if you wish to rewrite spec
-  #return
+  return
   
   if inSpec:
     #resize colum with types
     line = line[:columnWidth[0] + columnWidth[1]] + (' ' * (60 - columnWidth[1])) + line[columnWidth[0] + columnWidth[1]:]
     line = line[:columnWidth[0]] + (' ' * (40 - columnWidth[0])) + line[columnWidth[0]:]
   
-  fnewspec.write(line)
+  #fnewspec.write(line)
   
 
 def generateDef():
   global lines
-  global fnewspec
+  #global fnewspec
   
-  fnewspec = open('PdfReference-okular-2.txt', 'w')
+  #fnewspec = open('PdfReference-okular-2.txt', 'w')
   
-  print 'import datatypes'
-  print
+  # pdf spec in text format
+  fspecText = open(sys.argv[1], 'r')
+  
+  # pdf spec in python directives 
+  fspecPy = open(sys.argv[2], 'w')
+  
+  fspecPy.write('import datatypes\n')
+  fspecPy.write('\n')
 
-  print 'def buildPdfSpec(pdfspec):'
+  fspecPy.write('def buildPdfSpec(pdfspec):\n')
   
-  for line in sys.stdin:
-    processLine(line)
+  for line in fspecText:
+    processLine(fspecPy, line)
    
   # close last table if it was not closed already 
-  stopTable()
+  stopTable(fspecPy)
   
-  print
+  fspecPy.write('\n')
 
-  print 'def addDictionaryTypesTo(knowTypes):'  
+  fspecPy.write('def addDictionaryTypesTo(knowTypes):\n')  
   for e in tableToClassName:
-    print('  knowTypes[\'' + tableToClassName[e][0] + '\'] = [\'SkPdf' + tableToClassName[e][0] + '*\', \'' + tableToClassName[e][0] + 'FromDictionary\', datatypes.CppNull(), \'ret->podofo()->GetDataType() == ePdfDataType_Dictionary\']')
-  print
-  print
-
+    fspecPy.write('  knowTypes[\'' + tableToClassName[e][0] + '\'] = [\'SkPdf' + tableToClassName[e][0] + '*\', \'SkPdf' + tableToClassName[e][0] + 'FromDictionary\', datatypes.CppNull(), \'ret->podofo()->GetDataType() == PoDoFo::ePdfDataType_Dictionary\', \'A_DICTIONARY\']\n')
+  fspecPy.write('\n')
   
   #print lines
-  fnewspec.close()
+  #fnewspec.close()
 
 if '__main__' == __name__:
   sys.exit(generateDef())