Mercurial > hg > ocrHandling

annotate copyFiles.py @ 2:90c0df483890 default tip

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
inital3
author dwinter
date Tue, 19 Aug 2014 14:25:29 +0200
parents 5e33fa5a2fdc
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
0
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
1 from _csv import Error
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
2 start ="/Volumes/Folivora/MPG_Archiv/"
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
3 target = "/Volumes/MPGARCHIV/input/"
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
4
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
5
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
6 import os
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
7 import shutil
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
8
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
9 errorf = open("/tmp/error.out","w",encoding="utf-8")
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
10 errorf.write("START")
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
11
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
12 for root, dirs, files in os.walk(start, topdown=False):
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
13 for name in files:
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
14 path = os.path.join(root, name)
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
15 ext = os.path.splitext(path)
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
16 print (name)
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
17 if ext[1] == ".pdf":
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
18
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
19 fld = os.path.split(root)
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
20 print (fld)
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
21 neu = os.path.join(target,fld[1]+"___"+name.replace(" ","_"))
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
22 if not os.path.exists(neu):
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
23 try:
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
24 shutil.copy(path, neu)
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
25 except:
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
26 errorf.write(path+"\n")
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
27 print (path)
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
28 else:
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
29 print("EXISTS:" + path)
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
30
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
31 errorf.close();
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
32
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
33
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
34
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
35 #for name in dirs:
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
36 # print(os.path.join(root, name))
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
37
5e33fa5a2fdc initial
Dirk Wintergruen <dwinter@mpiwg-berlin.mpg.de>
parents:
diff changeset
38