Add --single-file-directory option: create single directory with all linked files.

This new option copies all files to the directory specified as an argument
to the --single-file-directory option, and also creates dummy shorter
filenames for the files.

This feature was implemented to get around a problem found when zip'ing
the spreadsheet up with the supporting files for users on Windows.  The
Windows users encounter the error 0x80010135 related to some of the ZIP
files going beyond the maximum path name length on windows.  Apparently,
opening ZIP files with long path names just doesn't work on Microsoft
systems.  I've suggested our accountants switch to a Free Software
operating system, but they declined.

1 files changed, 46 insertions, 5 deletions

diff --git a/contrib/non-profit-audit-reports/csv2ods.py b/contrib/non-profit-audit-reports/csv2ods.py
index 8b880648..3a3411ba 100755
--- a/contrib/non-profit-audit-reports/csv2ods.py
+++ b/contrib/non-profit-audit-reports/csv2ods.py
@@ -23,16 +23,23 @@
 import sys, os, os.path, optparse
 import csv
 import ooolib2
+import shutil
+import string
 
 def err(msg):
     print 'error: %s' % msg
     sys.exit(1)
 
-def csv2ods(csvname, odsname, encoding='', verbose = False):
+def csv2ods(csvname, odsname, encoding='', singleFileDirectory=None, verbose = False):
     filesSavedinManifest = {}
 
     if verbose:
         print 'converting from %s to %s' % (csvname, odsname)
+
+    if singleFileDirectory:
+        if not os.path.isdir(os.path.join(os.getcwd(),singleFileDirectory)):
+            os.mkdir(singleFileDirectory)
+
     doc = ooolib2.Calc()
     #  add a pagebreak style
     style = 'pagebreak'
@@ -55,20 +62,51 @@ def csv2ods(csvname, odsname, encoding='', verbose = False):
         if len(fields) > 0:
             for col in range(len(fields)):
                 val = fields[col]
-                if encoding != '':
+                if encoding != '' and val[0:5] != "link:":  # Only utf8 encode if it's not a filename
                     val = unicode(val, 'utf8')
                 if len(val) > 0 and val[0] == '$':
                     doc.set_cell_value(col + 1, row, 'currency', val[1:])
                 else:
                     if (len(val) > 0 and val[0:5] == "link:"):
                         val = val[5:]
-                        linkrel = '../' + val # ../ means remove the name of the *.ods
                         linkname = os.path.basename(val) # name is just the last component
+                        if not singleFileDirectory:
+                            newFile = val
+                        else:
+                            relativeFileWithPath = os.path.basename(val)
+                            fileName, fileExtension = os.path.splitext(relativeFileWithPath)
+                            newFile = fileName[:15]   # 15 is an arbitrary choice.
+                            newFile = newFile + fileExtension
+                            # We'll now test to see if we made this file
+                            # before, and if it matched the same file we
+                            # now want.  If it doesn't, try to make a
+                            # short file name for it.
+                            if filesSavedinManifest.has_key(newFile) and filesSavedinManifest[newFile] != val:
+                                testFile = None
+                                for cc in list(string.letters) + list(string.digits):
+                                    testFile = cc + newFile
+                                    if not filesSavedinManifest.has_key(testFile):
+                                        break
+                                    testFile = None
+                                if not testFile:
+                                    raise Exception("too many similar file names for linkage; giving up")
+                                else:
+                                    newFile = testFile
+                            if not os.path.exists(csvdir + '/' + val):
+                                raise Exception("File" + csvdir + '/' + val + " does not exist in single file directory mode; giving up")
+                            src = os.path.join(csvdir, val)
+                            dest = os.path.join(csvdir, singleFileDirectory, newFile)
+                            shutil.copyfile(src, dest)
+                            shutil.copystat(src, dest)
+                            shutil.copymode(src, dest)
+                            newFile = os.path.join(singleFileDirectory, newFile)
+
+                        linkrel = '../' + newFile # ../ means remove the name of the *.ods
                         doc.set_cell_value(col + 1, row, 'link', (linkrel, linkname))
                         linkpath = csvdir + '/' + val
 
                         if not val in filesSavedinManifest:
-                            filesSavedinManifest[val] = col
+                            filesSavedinManifest[newFile] = val
 
                         if not os.path.exists(linkpath):
                             print "WARNING: link %s DOES NOT EXIST at %s" % (val, linkpath)
@@ -109,7 +147,10 @@ def main():
                       help='ods output filename')
     parser.add_option('-e', '--encoding', action='store', 
                       help='unicode character encoding type')
+    parser.add_option('-d', '--single-file-directory', action='store',
+                      help='directory name to move all files into')
     (options, args) = parser.parse_args()
+
     if len(args) != 0:
         parser.error("not expecting extra args")  
     if not os.path.exists(options.csv):
@@ -122,7 +163,7 @@ def main():
         print 'csv:', options.csv
         print 'ods:', options.ods
         print 'ods:', options.encoding
-    csv2ods(options.csv, options.ods, options.encoding, options.verbose)
+    csv2ods(options.csv, options.ods, options.encoding, options.single_file_directory, options.verbose)
 
 if __name__ == '__main__':
   main()