ajout du module base.url

lib/pyulib/src/ulib/all/__init__.py

@@ -20,7 +20,7 @@ except ImportError:
                'procs', 'paths', 'htmlentities', 'dates', 'times', 'words',
                'tmpfiles', 'lines', 'files', 'config', 'editor', 'pager',
                'getopt', 'optparse', 'args', 'control',
-               'functions', 'json', 'web',
+               'functions', 'json', 'web', 'url',
                'flock',
                'password',
                )

lib/pyulib/src/ulib/base/url.py

@@ -0,0 +1,49 @@
+# -*- coding: utf-8 mode: python -*- vim:sw=4:sts=4:et:ai:si:sta:fenc=utf-8
+import i_need_py23
+
+"""Des fonction pour télécharger des resources
+"""
+
+__all__ = ('Downloader',
+           )
+
+import os, re, urlparse, urllib, shutil
+from os import path
+
+from base import make_prop
+
+class Downloader(object):
+    _starturl, starturl = make_prop('_starturl', None)[:2]
+    _referer, referer = make_prop('_referer', None)[:2]
+
+    def __init__(self, starturl=None):
+        if starturl is not None: self._starturl = starturl
+
+    _headers, headers = make_prop('_headers', None)[:2]
+    def reget(self, url, output=None, content_type=None):
+        baseurl = self._referer
+        if baseurl is None: baseurl = self._starturl
+        if baseurl is not None: url = urlparse.urljoin(baseurl, url)
+
+        filename, headers = urllib.urlretrieve(url)
+        ct = headers.get('content-type', None)
+        if content_type is not None and (ct is None or (ct != content_type and not ct.startswith(content_type))):
+            os.unlink(filename)
+            return None
+
+        if output is None:
+            output = filename
+        else:
+            shutil.copy2(filename, output)
+            os.unlink(filename)
+
+        self._referer = url
+        return output
+
+    def get(self, url, output=None, content_type=None):
+        if output is not None and path.exists(output):
+            return output
+        return self.reget(url, output, content_type)
+
+    def cleanup(self):
+        urllib.urlcleanup()