From: Diane Trout <diane@caltech.edu>
Date: Mon, 9 May 2011 21:54:13 +0000 (-0700)
Subject: Add utility program to help find what exists in georgi's submission dirs
X-Git-Tag: 0.5.2~40^2~1
X-Git-Url: http://woldlab.caltech.edu/gitweb/?p=htsworkflow.git;a=commitdiff_plain;h=eb32c527bcde51ab412170ccc315c8e8bb53619f

Add utility program to help find what exists in georgi's submission dirs
---

diff --git a/extra/ucsc_encode_submission/scan_extension.py b/extra/ucsc_encode_submission/scan_extension.py
new file mode 100644
index 0000000..f3e45a6
--- /dev/null
+++ b/extra/ucsc_encode_submission/scan_extension.py
@@ -0,0 +1,46 @@
+from optparse import OptionParser
+import os
+import sys
+from pprint import pprint
+
+def main(cmdline=None):
+    parser = make_parser()
+    opts, args = parser.parse_args(cmdline)
+
+    extensions = scan(args)
+    #pprint(extensions)
+    print find_common_suffix(extensions)
+
+def make_parser():
+    parser = OptionParser("%prog: directory [directory...]")
+    return parser
+
+def scan(toscan):
+    index = {}
+    for cur_scan_dir in toscan:
+        for path, dirnames, filenames in os.walk(cur_scan_dir):
+            for filename in filenames:
+                next_index = index
+                for c in filename[::-1]:
+                    next_index = next_index.setdefault(c, {})
+    return index
+
+def find_common_suffix(index, tail=[]):
+    if len(tail) > 0 and len(index) > 1:
+        return "".join(tail[::-1])
+
+    results = []
+    for key, choice in index.items():
+        r = find_common_suffix(choice, tail+[key])
+        if r is not None:
+            results.append (r)
+        
+    if len(results) == 0:
+        return None
+    elif len(results) == 1:
+        return results[0]
+    else:
+        return results
+
+if __name__ == "__main__":
+    main()