Roberto Andrade created WAGON-419:
-------------------------------------
Summary: HTTP wagons incorrectly encode blank space in URLs
causing resources to be uploaded/looked up with a "+" sign
Key: WAGON-419
URL: https://jira.codehaus.org/browse/WAGON-419
Project: Maven Wagon
Issue Type: Bug
Components: wagon-http, wagon-http-lightweight, wagon-webdav
Affects Versions: 2.6, 2.5, 2.4, 2.3, 2.2, 2.1, 2.0, 1.0
Environment: Apache Maven 3.1.1
(0728685237757ffbf44136acec0402957f723d9a; 2013-09-17 11:22:22-0400)
Maven home: /usr/local/Cellar/maven/3.1.1/libexec
Java version: 1.7.0_51, vendor: Oracle Corporation
Java home: /Library/Java/JavaVirtualMachines/jdk1.7.0_51.jdk/Contents/Home/jre
Default locale: en_US, platform encoding: US-ASCII
OS name: "mac os x", version: "10.9.4", arch: "x86_64", family: "mac"
Reporter: Roberto Andrade
Priority: Minor
The wagons all use
[{{java.net.URLEncoder.encode}}|http://docs.oracle.com/javase/7/docs/api/java/net/URLEncoder.html]
to encode the path components of the URL which uses
{{application/x-www-form-urlencoded}} format instead of the more generic one
defined by RFC2396. As documented by the
[java.net.URL|http://docs.oracle.com/javase/7/docs/api/java/net/URL.html]
class, this should only be used for HTML/form encoding and not generic URL
encoding:
{quote}
The URLEncoder and URLDecoder classes can also be used, but only for HTML form
encoding, which is not the same as the encoding scheme defined in RFC2396.
{quote}
The use of such encoder affects URLs with spaces (i.e.: on the resource path)
by encoding it to {{+}} instead of the traditional {{%20}} which would
ultimately result in an actual space being stored on the server when receiving
a {{PUT}} request. Additionally, the {{application/x-www-form-urlencoded}}
format should only be used if the HTTP request itself has that as a
{{Content-Type}} header value, which in the majority of the cases here is not
the case (the HTTP client will either be sending up the correctly translated
MIME type, a generic {{application/octet-stream}} or some other form of
encoding - i.e.: base64, but the URL encoding is completely independent of
that).
This is probably a known issue given there are {{TODO}} comments on the code
for every one of the wagons.
I'm submitting a PR on Github with a proposed fix for them all.
--
This message was sent by Atlassian JIRA
(v6.1.6#6162)
