# URI class bug? **URL:** \ **Category:** Ruby\ **Created:** [June 5, 2007, 8:35am UTC](https://www.ruby-forum.com/t/uri-class-bug/105980 "2007-06-05T08:35:27Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![mog](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/m/82dd89/32.png) [@mog](https://www.ruby-forum.com/u/mog)\ **Post date:** [June 5, 2007, 8:35am UTC](https://www.ruby-forum.com/t/uri-class-bug/105980/1 "2007-06-05T08:35:27Z")
I am using Ruby 1.8.6. I found URI cannot parse URI with “\_” is host. uri = “http://dr\_gabriele.podomatic.com/enclosure/ 2006-08-03T15\_09\_59-07\_00.m4v” URI.parse(uri) Is there any way to work around that? thanks ---
**Author:** ![mog](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/m/82dd89/32.png) [@mog](https://www.ruby-forum.com/u/mog)\ **Post date:** [June 5, 2007, 9:01am UTC](https://www.ruby-forum.com/t/uri-class-bug/105980/2 "2007-06-05T09:01:30Z")
On 6/5/07, Morgan C. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > I am using Ruby 1.8.6. I found URI cannot parse URI with “\_” is host. > > uri = “http://dr\_gabriele.podomatic.com/enclosure/ > 2006-08-03T15\_09\_59-07\_00.m4v” > URI.parse(uri) > > Is there any way to work around that? > thanks It seems underscores are not allowed in host part of an URI. So it’s not a bug. See RFC 2396 (URI), and 1035 (DNS). If you really want it, you can open the class and redefine some of the methods and/or manually edit URI sources. ---
**Author:** ![mog](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/m/82dd89/32.png) [@mog](https://www.ruby-forum.com/u/mog)\ **Post date:** [June 5, 2007, 9:41am UTC](https://www.ruby-forum.com/t/uri-class-bug/105980/3 "2007-06-05T09:41:33Z")
On Jun 5, 3:00 pm, “Jano S.” [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > It seems underscores are not allowed in host part of an URI. So it’s > not a bug. See RFC 2396 (URI), and 1035 (DNS). If you really want it, > you can open the class and redefine some of the methods and/or > manually edit URI sources. In RFC 2396, “\_” is taken as “Unreserved Characters”. Unreserved characters can be escaped without changing the semantics of the URI, but this should not be done unless the URI is being used in a context that does not allow the unescaped character to appear. However, URI.escape doesn’t escape “\_”. require ‘URI’ original\_uri = “http://dr\_gabriele.podomatic.com/enclosure/” uri = URI.escape(original\_uri) puts uri == original\_uri ---
**Author:** ![mog](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/m/82dd89/32.png) [@mog](https://www.ruby-forum.com/u/mog)\ **Post date:** [June 5, 2007, 11:51am UTC](https://www.ruby-forum.com/t/uri-class-bug/105980/4 "2007-06-05T11:51:01Z")
On Jun 5, 5:24 pm, “Jano S.” [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > > > > 2006-08-03T15\_09\_59-07\_00.m4v" > > > > In RFC 2396, “_" is taken as “Unreserved Characters”. > > > > puts uri == original\_uri > > > > alpha = lowalpha | upalpha > > > > There’s no "_” there. YMMV 😉 > > > > Thanks a lot for your help. I am just wandering, internet is a wild world. Wierd non-standard stuff is all around. The non-standard host name is a example. Popular browser can handle these URLs well. Perhaps ruby should be more strong to survive better in such wild world 🙂 ---
**Author:** ![mog](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/m/82dd89/32.png) [@mog](https://www.ruby-forum.com/u/mog)\ **Post date:** [June 5, 2007, 11:24am UTC](https://www.ruby-forum.com/t/uri-class-bug/105980/5 "2007-06-05T11:24:59Z")
On 6/5/07, Morgan C. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > > > thanks > > > used > > > in a context that does not allow the unescaped character to appear. > > However, URI.escape doesn’t escape “\_”. > > require ‘URI’ > original\_uri = “http://dr\_gabriele.podomatic.com/enclosure/” > uri = URI.escape(original\_uri) > puts uri == original\_uri I’m no expert on DNS, this is what I have found in appendix A: host = hostname | IPv4address hostname = _( domainlabel “.” ) toplabel [“.”] domainlabel = alphanum | alphanum ( alphanum | “-” ) alphanum toplabel = alpha | alpha ( alphanum | “-” ) alphanum IPv4address = 1digit “.” 1digit “.” 1_digit “.” 1\*digit alphanum = alpha | digit alpha = lowalpha | upalpha lowalpha = “a” | “b” | “c” | “d” | “e” | “f” | “g” | “h” | “i” | “j” | “k” | “l” | “m” | “n” | “o” | “p” | “q” | “r” | “s” | “t” | “u” | “v” | “w” | “x” | “y” | “z” upalpha = “A” | “B” | “C” | “D” | “E” | “F” | “G” | “H” | “I” | “J” | “K” | “L” | “M” | “N” | “O” | “P” | “Q” | “R” | “S” | “T” | “U” | “V” | “W” | “X” | “Y” | “Z” digit = “0” | “1” | “2” | “3” | “4” | “5” | “6” | “7” | “8” | “9” There’s no “\_” there. YMMV 😉 ---
**Author:** ![mog](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/m/82dd89/32.png) [@mog](https://www.ruby-forum.com/u/mog)\ **Post date:** [June 5, 2007, 1:13pm UTC](https://www.ruby-forum.com/t/uri-class-bug/105980/6 "2007-06-05T13:13:46Z")
Morgan C. wrote: > On Jun 5, 5:24 pm, “Jano S.” [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > > > There’s no “\_” there. YMMV 😉 > > Thanks a lot for your help. > > I am just wandering, internet is a wild world. Wierd non-standard > stuff is all around. The non-standard host name is a example. Popular > browser can handle these URLs well. > Perhaps ruby should be more strong to survive better in such wild > world 🙂 > There was mention a few days ago of bringing the URI class up to a more > recent RFC compliance (3986, I think). Would that help in this > instance? ---
**Author:** ![mog](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/m/82dd89/32.png) [@mog](https://www.ruby-forum.com/u/mog)\ **Post date:** [June 5, 2007, 1:14pm UTC](https://www.ruby-forum.com/t/uri-class-bug/105980/7 "2007-06-05T13:14:47Z")
On 6/5/07, Morgan C. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > > > > > uri = "http://dr\_gabriele.podomatic.com/enclosure/ > > > > > uri = URI.escape(original\_uri) > > > alphanum = alpha | digit > > > > There’s no “\_” there. YMMV 😉 > > Thanks a lot for your help. > > I am just wandering, internet is a wild world. Wierd non-standard > stuff is all around. The non-standard host name is a example. Popular > browser can handle these URLs well. > Perhaps ruby should be more strong to survive better in such wild > world 🙂 If you want to really use underscores, modify lib/1.8/uri/common.rb: add the following after line HOSTNAME=… and comment out (prefix with #) the original HOSTNAME line. ``` ALPHA_ = "a-zA-Z_" ALNUM_ = "#{ALPHA_}\\d" DOMLABEL_ = "(?:[#{ALNUM_}](?:[-#{ALNUM_}]*[#{ALNUM_}])?)" TOPLABEL_ = "(?:[#{ALPHA_}](?:[-#{ALNUM_}]*[#{ALNUM_}])?)" HOSTNAME = "(?:#{DOMLABEL_}\\.)*#{TOPLABEL_}\\.?" ``` J.