[sword-svn] r369 - trunk/modules/python

[chrislit at crosswire.org]

Author: chrislit
Date: 2012-08-09 04:38:14 -0700 (Thu, 09 Aug 2012)
New Revision: 369

Modified:
   trunk/modules/python/usfm2osis.py
Log:
added some skeletal code for selection of different file sort algorithms

Modified: trunk/modules/python/usfm2osis.py
===================================================================
--- trunk/modules/python/usfm2osis.py	2012-08-06 16:24:47 UTC (rev 368)
+++ trunk/modules/python/usfm2osis.py	2012-08-09 11:38:14 UTC (rev 369)
@@ -140,6 +140,55 @@
     'ADE':'AddEsth'
     }
 
+canonicalOrder = (
+    # OT
+    'GEN', 'EXO', 'LEV', 'NUM', 'DEU', 'JOS', 'JDG', 'RUT', '1SA', '2SA', '1KI', '2KI', '1CH', '2CH', 'EZR', 'NEH',
+    'EST', 'JOB', 'PSA', 'PRO', 'ECC', 'SNG', 'ISA', 'JER', 'LAM', 'EZK', 'DAN', 'HOS', 'JOL', 'AMO', 'OBA', 'JON',
+    'MIC', 'NAM', 'HAB', 'ZEP', 'HAG', 'ZEC', 'MAL',
+    # DC - Catholic
+    'TOB', 'JDT', 'ESG', 'ADE', 'WIS', 'SIR', 'PSS', 'BAR', 'LJE', 'DAG', 'S3Y', 'SUS', 'BEL', '1MA', '2MA',
+    # DC - Eastern Orthodox
+    '1ES', 'MAN', 'PS2', '3MA', '2ES', '4MA', 
+    # NT
+    'MAT', 'MRK', 'LUK', 'JHN', 'ACT', 'ROM', '1CO', '2CO', 'GAL', 'EPH', 'PHP', 'COL', '1TH', '2TH', '1TI', '2TI',
+    'TIT', 'PHM', 'HEB', 'JAS', '1PE', '2PE', '1JN', '2JN', '3JN', 'JUD', 'REV',
+    # Rahlfs' LXX
+    'ODA', 'ODE',
+    # Esdrae
+    'EZA', '5EZ', '6EZ',
+    # Inconsistency with Esther
+    
+    # Syriac
+    'PS3', '2BA', 'LBA',
+    # Ethiopic
+    'JUB', 'ENO', '1MQ', '2MQ', '3MQ', 'REP', '4BA',
+    # Vulgate
+    'LAO',
+
+    # Additional non-biblical books
+    'XXA', 'XXB', 'XXC', 'XXD', 'XXE', 'XXF', 'XXG',
+
+    # Peripheral books
+    'FRT', 'INT', 'BAK', 'CNC', 'GLO',
+    'TDX', 'NDX', 'OTH'
+    ### Deprecated
+    # Rahlfs
+    'JSA', 'JDB', 'TBS', 'SST', 'DNT', 'BLT',
+    # Esdrae
+    '4ES', '5ES', '6ES',
+
+    # Alternate Psalms
+    'PSB',
+    # Vulgate
+    'PSO', 'PJE',
+    # Armenian
+    'WSI', 'COP', '3CO', 'EUT', 'DOJ',
+    # Apostolic Fathers
+    '1CL', '2CL', 'SHE', 'LBA', 'DID',
+    ###
+    # Proposed replacements <http://lc.bfbs.org.uk/e107_files/downloads/canonicalissuesinparatext.pdf>
+    )
+
 specialBooks = ['FRONT', 'INTRODUCTION', 'BACK', 'CONCORDANCE', 'GLOSSARY', 'INDEX', 'GAZETTEER', 'X-OTHER']
 
 peripherals = {
@@ -936,6 +985,7 @@
     print('  -h, --help       print this usage information')
     print('  -o FILENAME      output filename (default is: <osisWork>.osis.xml)')
     print('  -r               enable relaxed markup processing (for non-standard USFM)')
+    print('  -s mode          set book sorting mode: natural (default), alpha, canonical, none')
     print('  -v               verbose feedback')
     print('')
     print('As an example, if you want to generate the osisWork <Bible.KJV> and your USFM')
@@ -983,6 +1033,7 @@
     encoding = ''
     relaxedConformance = False
     inputFilesIdx = 2 # This marks the point in the sys.argv array, after which all values represent USFM files to be converted.
+    usfmDocList = list()
 
     if '-v' in sys.argv:
         verbose = True
@@ -1016,6 +1067,27 @@
             bookDict = dict(bookDict.items() + addBookDict.items())
             inputFilesIdx += 1
 
+        if '-s' in sys.argv:
+            i = sys.argv.index('-s')+1
+            if len(sys.argv) < i+1:
+                printUsage()
+            if sys.argv[i].startsWith('a'):
+                sortHelper = keynat # TODO: write appropriate helpers
+                print('Sorting book files alphanumerically.')
+            elif sys.argv[i].startsWith('na'):
+                sortHelper = keynat
+                print('Sorting book files naturally.')
+            elif sys.argv[i].startsWith('c'):
+                sortHelper = keynat # TODO: write appropriate helpers
+                print('Sorting book files canonically.')
+            else:
+                sortHelper = usfmDocList.index
+                print('Leaving book files unsorted.')
+        else:
+            sortHelper = keynat
+            print('Sorting book files naturally.')
+
+
         usfmDocList = sys.argv[inputFilesIdx:]
 
         OSISfile = codecs.open(OSISfileName, 'w', 'utf-8')
@@ -1025,7 +1097,7 @@
         # run
         # load up work queue
         work_queue = multiprocessing.Queue()
-        for job in sorted(usfmDocList, key=keynat):
+        for job in sorted(usfmDocList, key=sortHelper):
             work_queue.put(job)
 
         # create a queue to pass to workers to store the results