From c9c8c71a1204fbd4dc19fe525f62b04b5fd1134f Mon Sep 17 00:00:00 2001
From: Ville Skyttä <ville.skytta@iki.fi>
Date: Sat, 5 Feb 2011 10:03:06 +0200
Subject: Implement subdirectory handling.

Each /^/ in the given URL is replaced by the latest version dir at that level.
---
 cnucnu/helper.py            | 24 ++++++++++++++++++++++++
 cnucnu/package_list.py      |  3 ++-
 cnucnu/tests/helper_test.py |  3 ++-
 3 files changed, 28 insertions(+), 2 deletions(-)

(limited to 'cnucnu')

diff --git a/cnucnu/helper.py b/cnucnu/helper.py
index 8614c44..a3d20d3 100644
--- a/cnucnu/helper.py
+++ b/cnucnu/helper.py
@@ -24,10 +24,34 @@ __docformat__ = "restructuredtext"
 
 #from twisted.internet import reactor
 
+import re
 import pprint as pprint_module
 pp = pprint_module.PrettyPrinter(indent=4)
 pprint = pp.pprint
 
+__html_regex = re.compile(r'\bhref\s*=\s*["\']([^"\'/]+)/["\']', re.I)
+__text_regex = re.compile(r'^d.+\s(\S+)\s*$', re.I|re.M)
+
+def expand_subdirs(url):
+    """ Expand all /^/'s in the given URL with the latest dir at that level """
+    ix = url.find("/^/")
+    while ix != -1:
+        ls = get_html(url[0:ix+1])
+        if not ls:
+            break
+        subdirs = []
+        regex = url.startswith("ftp://") and __text_regex or __html_regex
+        for match in regex.finditer(ls):
+            subdir = match.group(1)
+            if subdir not in (".", ".."):
+                subdirs.append(subdir)
+        if not subdirs:
+            break
+        latest = upstream_max(subdirs)
+        url = "%s/%s/%s" % (url[0:ix], latest, url[ix+len("/^/"):])
+        ix = url.find("/^/", ix + len(latest) + 1)
+    return url
+
 def get_html(url, callback=None, errback=None):
     if url.startswith("ftp://"):
         import urllib
diff --git a/cnucnu/package_list.py b/cnucnu/package_list.py
index de15cdb..182795d 100755
--- a/cnucnu/package_list.py
+++ b/cnucnu/package_list.py
@@ -229,9 +229,10 @@ class Package(object):
 
     def get_html(self):
         if not self._html:
-            from cnucnu.helper import get_html
+            from cnucnu.helper import get_html, expand_subdirs
 
             try:
+                self.__url = expand_subdirs(self.url)
                 html = get_html(self.url)
             # TODO: get_html should raise a generic retrieval error
             except IOError, ioe:
diff --git a/cnucnu/tests/helper_test.py b/cnucnu/tests/helper_test.py
index fddd489..b925a12 100755
--- a/cnucnu/tests/helper_test.py
+++ b/cnucnu/tests/helper_test.py
@@ -22,7 +22,7 @@ import unittest
 import sys
 sys.path.insert(0, '../..')
 
-from cnucnu.helper import upstream_cmp, upstream_max, split_rc, cmp_upstream_repo, get_rc, get_html
+from cnucnu.helper import upstream_cmp, upstream_max, split_rc, cmp_upstream_repo, get_rc, get_html, expand_subdirs
 
 class HelperTest(unittest.TestCase):
 
@@ -130,6 +130,7 @@ class HelperTest(unittest.TestCase):
         http_url = ("http://www.fedoraproject.org")
         res = StringIO.StringIO()
 
+        http_url = expand_subdirs(http_url)
         data1 = get_html(http_url)
 
         callback = [res.write, lambda ignore: reactor.stop()]
-- 
cgit