Changeset 549

Timestamp:

08/23/05 12:48:57

Author:

fumanchu

Message:

1. Initial support for partial GET (Range request header): new _cphttptools.get_ranges() function for use in page handlers and serve_file().
2. New HTTPClientError exception; NotFound? now inherits from it.

Files:

• trunk/cherrypy/_cperror.py (modified) (2 diffs)
• trunk/cherrypy/_cphttptools.py (modified) (5 diffs)
• trunk/cherrypy/test/test_core.py (modified) (3 diffs)

trunk/cherrypy/_cperror.py

@@ -39 +39 @@
-class NotReady(Error):
-    """A request was made before the app server has been started."""
-    pass
-
-class NotFound(Error):
-    """ Happens when a URL couldn't be mapped to any class.method """
-    pass
-
@@ -173 +169 @@
-        else:
-            raise ValueError("The %s status code is unknown." % status)
+
+
+_missing = object()
+
+class HTTPClientError(Error):
+    """Exception raised when the client has made an error in its request."""
+    
+    def __init__(self, status=400, body=_missing):
+        self.status = status = int(status)
+        if status < 400 or status > 499:
+            raise ValueError("status must be between 400 and 499.")
+        
+        import cherrypy
+        cherrypy.response.status = status
+        if body is not _missing:
+            cherrypy.response.body = body
+
+
+class NotFound(HTTPClientError):
+    """ Happens when a URL couldn't be mapped to any class.method """
+    
+    def __init__(self, path):
+        self.args = (path,)
+        HTTPClientError.__init__(self, 404)

trunk/cherrypy/_cphttptools.py

@@ -303 +303 @@
-            finally:
-                applyFilters('onEndResource')
-        except cherrypy.NotFound:
-            cherrypy.response.status = 404
-            handleError(sys.exc_info())
-        except:
+            # This includes HTTPClientError and NotFound
-            handleError(sys.exc_info())
-    
@@ -580 +578 @@
-            cherrypy.response.headerMap['Content-Length'] = len(content)
-        else:
-
+
+
+
+
-    
-    # For some statuses, Internet Explorer 5+ shows "friendly error messages"
@@ -660 +661 @@
-
-
+def get_ranges(content_length):
+    """Return a list of (start, stop) indices from a Range header, or None.
+    
+    Each (start, stop) tuple will be composed of two ints, which are suitable
+    for use in a slicing operation. That is, the header "Range: bytes=3-6",
+    if applied against a Python string, is requesting resource[3:7]. This
+    function will return the list [(3, 7)].
+    """
+    
+    r = cherrypy.request.headerMap.get('Range')
+    if not r:
+        return None
+    
+    result = []
+    bytesunit, byteranges = r.split("=", 1)
+    for brange in byteranges.split(","):
+        start, stop = [x.strip() for x in brange.split("-", 1)]
+        if start:
+            if not stop:
+                stop = content_length - 1
+            start, stop = map(int, (start, stop))
+            if start >= content_length:
+                # From rfc 2616 sec 14.16:
+                # "If the server receives a request (other than one
+                # including an If-Range request-header field) with an
+                # unsatisfiable Range request-header field (that is,
+                # all of whose byte-range-spec values have a first-byte-pos
+                # value greater than the current length of the selected
+                # resource), it SHOULD return a response code of 416
+                # (Requested range not satisfiable)."
+                continue
+            if stop < start:
+                # From rfc 2616 sec 14.16:
+                # "If the server ignores a byte-range-spec because it
+                # is syntactically invalid, the server SHOULD treat
+                # the request as if the invalid Range header field
+                # did not exist. (Normally, this means return a 200
+                # response containing the full entity)."
+                return None
+            result.append((start, stop + 1))
+        else:
+            if not stop:
+                # See rfc quote above.
+                return None
+            # Negative subscript (last N bytes)
+            result.append((content_length - int(stop), content_length))
+    
+    if result == []:
+        cherrypy.response.headerMap['Content-Range'] = "bytes */%s" % content_length
+        b = "Invalid Range (first-byte-pos greater than Content-Length)"
+        raise cherrypy.HTTPClientError(416, b)
+    
+    return result
+
+
-def serve_file(filename):
-    """Set status, headers, and body in order to serve the given file."""
@@ -683 +739 @@
-        ext = ""
-    
+    resp = cherrypy.response
+    
-    contentType = mimetypes.types_map.get(ext, "text/plain")
-cherrypy.response
+resp
-    
-    strModifTime = httpdate(time.gmtime(stat.st_mtime))
@@ -690 +748 @@
-        # Check if if-modified-since date is the same as strModifTime
-        if cherrypy.request.headerMap['If-Modified-Since'] == strModifTime:
-cherrypy.response
-cherrypy.response
+resp
+resp
-            if getattr(cherrypy, "debug", None):
-                cherrypy.log("    Found file (304 Not Modified): %s" % filename, "DEBUG")
-            return
-cherrypy.response
+resp
-    
-    # Set Content-Length and use an iterable (file object)
-    #   this way CP won't load the whole file in memory
-herrypy.response.headerMap['Content-Length']
+_len
-    bodyfile = open(filename, 'rb')
-    cherrypy.response.body = fileGenerator(bodyfile)
-    if getattr(cherrypy, "debug", None):
-        cherrypy.log("    Found file: %s" % filename, "DEBUG")
+    
+    resp.headerMap["Accept-Ranges"] = "bytes"
+    r = get_ranges(c_len)
+    if r:
+        if len(r) == 1:
+            # Return a single-part response.
+            start, stop = r[0]
+            r_len = stop - start
+            resp.status = "206 Partial Content"
+            resp.headerMap['Content-Range'] = ("bytes %s-%s/%s" %
+                                               (start, stop - 1, c_len))
+            resp.headerMap['Content-Length'] = r_len
+            bodyfile.seek(start)
+            resp.body = [bodyfile.read(r_len)]
+        else:
+            # Return a multipart/byteranges response.
+            resp.status = "206 Partial Content"
+            import mimetools
+            boundary = mimetools.choose_boundary()
+            resp.headerMap['Content-Type'] = "multipart/byteranges; boundary=%s" % boundary
+            del resp.headerMap['Content-Length']
+            def fileRanges():
+                for start, stop in r:
+                    yield "--" + boundary
+                    yield "\nContent-type: %s" % contentType
+                    yield ("\nContent-range: bytes %s-%s/%s\n\n"
+                           % (start, stop - 1, c_len))
+                    bodyfile.seek(start)
+                    yield bodyfile.read((stop + 1) - start)
+                    yield "\n"
+                # Final boundary
+                yield "--" + boundary
+            resp.body = fileRanges()
+    else:
+        resp.headerMap['Content-Length'] = c_len
+        resp.body = fileGenerator(bodyfile)
-
-

trunk/cherrypy/test/test_core.py

@@ -30 +30 @@
-
-import cherrypy
+from cherrypy import _cphttptools
-import types
-import os
@@ -170 +171 @@
-        # Since status must start with an int, this should error.
-        cherrypy.response.status = "ZOO OK"
+
+
+class Ranges(Test):
+    
+    def get_ranges(self):
+        return repr(_cphttptools.get_ranges(8))
+    
+    def slice_file(self):
+        path = os.path.join(os.getcwd(), os.path.dirname(__file__))
+        _cphttptools.serve_file(os.path.join(path, "static/index.html"))
+        # Ugly hack but it works
+        return cherrypy.response.body
-
-
@@ -446 +459 @@
-            ignore.pop()
-    
+    def test_Ranges(self):
+        self.getPage("/ranges/get_ranges", [('Range', 'bytes=3-6')])
+        self.assertBody("[(3, 7)]")
+        
+        # Test multiple ranges and a suffix-byte-range-spec, for good measure.
+        self.getPage("/ranges/get_ranges", [('Range', 'bytes=2-4,-1')])
+        self.assertBody("[(2, 5), (7, 8)]")
+        
+        # Get a partial file.
+        self.getPage("/ranges/slice_file", [('Range', 'bytes=2-5')])
+        self.assertStatus("206 Partial Content")
+        self.assertHeader("Content-Type", "text/html")
+        self.assertHeader("Content-Range", "bytes 2-5/14")
+        self.assertBody("llo,")
+        
+        # What happens with overlapping ranges (and out of order, too)?
+        self.getPage("/ranges/slice_file", [('Range', 'bytes=4-6,2-5')])
+        self.assertStatus("206 Partial Content")
+        ct = ""
+        for k, v in self.headers:
+            if k.lower() == "content-type":
+                ct = v
+                break
+        expected_type = "multipart/byteranges; boundary="
+        self.assert_(ct.startswith(expected_type))
+        boundary = ct[len(expected_type):]
+        expected_body = """--%s
+Content-type: text/html
+Content-range: bytes 4-6/14
+
+o, w
+--%s
+Content-type: text/html
+Content-range: bytes 2-5/14
+
+llo, 
+--%s""" % (boundary, boundary, boundary)
+        self.assertBody(expected_body)
+        self.assertNoHeader("Content-Length")
+        
+        # Test "416 Requested Range Not Satisfiable"
+        self.getPage("/ranges/slice_file", [('Range', 'bytes=2300-2900')])
+        self.assertStatus("416 Requested Range Not Satisfiable")
+        self.assertHeader("Content-Range", "bytes */14")
+    
-    def testHeaderCaseSensitivity(self):
-        # Tests that each header only appears once, regardless of case.