{"thread":{"id":"57513","subject":"[PATCH 1/2] comment: fix typo","startedAt":"2022-03-07T03:45:34Z","lastAt":"2022-03-08T15:59:02Z","messageCount":17,"participants":["Jason Yundt","Ævar Arnfjörð Bjarmason","brian m. carlson","Junio C Hamano"],"isPatch":true,"patchVersion":1,"patchTotal":2},"messages":[{"id":"450506","messageId":"20220307033723.175553-2-jason@jasonyundt.email","threadId":"57513","inReplyTo":"20220307033723.175553-1-jason@jasonyundt.email","subject":"[PATCH 1/2] comment: fix typo","fromName":"Jason Yundt","fromEmail":"jason@jasonyundt.email","sentAt":"2022-03-07T03:37:22Z","receivedAt":"2022-03-07T03:45:34Z","isPatch":true,"sender":{"key":"jason@jasonyundt.email","avatar":"https://avatars.githubusercontent.com/u/5579359?v=4"},"body":"Signed-off-by: Jason Yundt \n---\n t/t9502-gitweb-standalone-parse-output.sh | 2 +-\n 1 file changed, 1 insertion(+), 1 deletion(-)\n\ndiff --git a/t/t9502-gitweb-standalone-parse-output.sh b/t/t9502-gitweb-standalone-parse-output.sh\nindex 3167473b30..e7363511dd 100755\n--- a/t/t9502-gitweb-standalone-parse-output.sh\n+++ b/t/t9502-gitweb-standalone-parse-output.sh\n@@ -34,7 +34,7 @@ EOF\n #\n # This will check that gitweb HTTP header contains proposed filename\n # as with '.tar' suffix added, and that generated tarfile\n-# (gitweb message body) has as prefix for al files in tarfile\n+# (gitweb message body) has as prefix for all files in tarfile\n #\n # default to \n check_snapshot () {\n-- \n2.35.1\n\n"},{"id":"450507","messageId":"20220307033723.175553-1-jason@jasonyundt.email","threadId":"57513","inReplyTo":null,"subject":"[PATCH 0/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"Jason Yundt","fromEmail":"jason@jasonyundt.email","sentAt":"2022-03-07T03:37:21Z","receivedAt":"2022-03-07T03:45:36Z","isPatch":true,"sender":{"key":"jason@jasonyundt.email","avatar":"https://avatars.githubusercontent.com/u/5579359?v=4"},"body":"See the second commit's message for more details.\n\nJason Yundt (2):\n comment: fix typo\n gitweb: remove invalid http-equiv=\"content-type\"\n\n gitweb/gitweb.perl | 4 +---\n t/t9502-gitweb-standalone-parse-output.sh | 15 ++++++++++++++-\n 2 files changed, 15 insertions(+), 4 deletions(-)\n\n-- \n2.35.1\n\n"},{"id":"450508","messageId":"20220307033723.175553-3-jason@jasonyundt.email","threadId":"57513","inReplyTo":"20220307033723.175553-1-jason@jasonyundt.email","subject":"[PATCH 2/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"Jason Yundt","fromEmail":"jason@jasonyundt.email","sentAt":"2022-03-07T03:37:23Z","receivedAt":"2022-03-07T03:46:00Z","isPatch":true,"sender":{"key":"jason@jasonyundt.email","avatar":"https://avatars.githubusercontent.com/u/5579359?v=4"},"body":"Before this change, gitweb would generate pages which included:\n\n\t\n\nA meta element with http-equiv=\"content-type\" is said to be in the\n\"Encoding declaration state\". According to the HTML Standard,\n\n\tThe Encoding declaration state may be used in HTML documents,\n\tbut elements with an http-equiv attribute in that state must not\n\tbe used in XML documents.\n\n\tSource: \n\nThis change removes that meta element since gitweb always generates XML\ndocuments.\n\nSigned-off-by: Jason Yundt \n---\n gitweb/gitweb.perl | 4 +---\n t/t9502-gitweb-standalone-parse-output.sh | 13 +++++++++++++\n 2 files changed, 14 insertions(+), 3 deletions(-)\n\ndiff --git a/gitweb/gitweb.perl b/gitweb/gitweb.perl\nindex fbd1c20a23..606b50104c 100755\n--- a/gitweb/gitweb.perl\n+++ b/gitweb/gitweb.perl\n@@ -4213,8 +4213,7 @@ sub git_header_html {\n \tmy %opts = @_;\n \n \tmy $title = get_page_title();\n-\tmy $content_type = get_content_type_html();\n-\tprint $cgi->header(-type=>$content_type, -charset => 'utf-8',\n+\tprint $cgi->header(-type=>get_content_type_html(), -charset => 'utf-8',\n \t -status=> $status, -expires => $expires)\n \t\tunless ($opts{'-no_http_header'});\n \tmy $mod_perl_version = $ENV{'MOD_PERL'} ? \" $ENV{'MOD_PERL'}\" : '';\n@@ -4225,7 +4224,6 @@ sub git_header_html {\n \n \n \n-\n \n \n $title\ndiff --git a/t/t9502-gitweb-standalone-parse-output.sh b/t/t9502-gitweb-standalone-parse-output.sh\nindex e7363511dd..25165edacc 100755\n--- a/t/t9502-gitweb-standalone-parse-output.sh\n+++ b/t/t9502-gitweb-standalone-parse-output.sh\n@@ -207,4 +207,17 @@ test_expect_success 'xss checks' '\n \txss \"\" \"$TAG+\"\n '\n \n+no_http_equiv_content_type() {\n+\tgitweb_run \"$@\" &&\n+\t! grep -Ei \"http-equiv=['\\\"]?content-type\" gitweb.body\n+}\n+\n+# See: \n+test_expect_success 'no http-equiv=\"content-type\" in XHTML' '\n+\tno_http_equiv_content_type &&\n+\tno_http_equiv_content_type \"p=.git\" &&\n+\tno_http_equiv_content_type \"p=.git;a=log\" &&\n+\tno_http_equiv_content_type \"p=.git;a=tree\"\n+'\n+\n test_done\n-- \n2.35.1\n\n"},{"id":"450537","messageId":"220307.861qze0wv9.gmgdl@evledraar.gmail.com","threadId":"57513","inReplyTo":"20220307033723.175553-3-jason@jasonyundt.email","subject":"Re: [PATCH 2/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"Ævar Arnfjörð Bjarmason","fromEmail":"avarab@gmail.com","sentAt":"2022-03-07T12:23:49Z","receivedAt":"2022-03-07T12:29:22Z","isPatch":true,"sender":{"key":"avarab@gmail.com","avatar":"https://avatars.githubusercontent.com/u/45301?v=4"},"body":"\nOn Sun, Mar 06 2022, Jason Yundt wrote:\n\n> Before this change, gitweb would generate pages which included:\n>\n> \t\n>\n> A meta element with http-equiv=\"content-type\" is said to be in the\n> \"Encoding declaration state\". According to the HTML Standard,\n>\n> \tThe Encoding declaration state may be used in HTML documents,\n> \tbut elements with an http-equiv attribute in that state must not\n> \tbe used in XML documents.\n>\n> \tSource: \n>\n> This change removes that meta element since gitweb always generates XML\n> documents.\n>\n> Signed-off-by: Jason Yundt \n> ---\n> gitweb/gitweb.perl | 4 +---\n> t/t9502-gitweb-standalone-parse-output.sh | 13 +++++++++++++\n> 2 files changed, 14 insertions(+), 3 deletions(-)\n>\n> diff --git a/gitweb/gitweb.perl b/gitweb/gitweb.perl\n> index fbd1c20a23..606b50104c 100755\n> --- a/gitweb/gitweb.perl\n> +++ b/gitweb/gitweb.perl\n> @@ -4213,8 +4213,7 @@ sub git_header_html {\n> \tmy %opts = @_;\n> \n> \tmy $title = get_page_title();\n> -\tmy $content_type = get_content_type_html();\n> -\tprint $cgi->header(-type=>$content_type, -charset => 'utf-8',\n> +\tprint $cgi->header(-type=>get_content_type_html(), -charset => 'utf-8',\n\nI think it would be better to just skip this hunk, no behavior will\nchange if it's left in.\n\n> \t -status=> $status, -expires => $expires)\n> \t\tunless ($opts{'-no_http_header'});\n> \tmy $mod_perl_version = $ENV{'MOD_PERL'} ? \" $ENV{'MOD_PERL'}\" : '';\n> @@ -4225,7 +4224,6 @@ sub git_header_html {\n> \n> \n> \n> -\n\n..with this being the only behavior change (yeah the variable will now\nbe used only in one place, but that's fine)\n\nI'm not sure I understand this change really. The result in always XML,\nso application/xhtml+xml is redundant, text/html, or both?\n\nBut aside from that: I have seen browsers get the lack of encoding=\"\"\n\"wrong\" with data at rest, don't some still default to ISO-8859-1?\n\nSo won't this result in badly decoded data if you save the web page &\nview it locally?\n\n> \n> \n> $title\n> diff --git a/t/t9502-gitweb-standalone-parse-output.sh b/t/t9502-gitweb-standalone-parse-output.sh\n> index e7363511dd..25165edacc 100755\n> --- a/t/t9502-gitweb-standalone-parse-output.sh\n> +++ b/t/t9502-gitweb-standalone-parse-output.sh\n> @@ -207,4 +207,17 @@ test_expect_success 'xss checks' '\n> \txss \"\" \"$TAG+\"\n> '\n> \n> +no_http_equiv_content_type() {\n> +\tgitweb_run \"$@\" &&\n> +\t! grep -Ei \"http-equiv=['\\\"]?content-type\" gitweb.body\n\nNit: Should we skip the \"-i\" here since we're testing our own output,\nand not http standards in general (i.e. we don't have to worry about the\ncase of http-equiv?)\n"},{"id":"450669","messageId":"109813056.nniJfEyVGO@jason-desktop-linux","threadId":"57513","inReplyTo":"220307.861qze0wv9.gmgdl@evledraar.gmail.com","subject":"Re: [PATCH 2/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"Jason Yundt","fromEmail":"jason@jasonyundt.email","sentAt":"2022-03-07T22:49:48Z","receivedAt":"2022-03-07T22:49:54Z","isPatch":true,"sender":{"key":"jason@jasonyundt.email","avatar":"https://avatars.githubusercontent.com/u/5579359?v=4"},"body":"On Monday, March 7, 2022 7:23:49 AM EST Ævar Arnfjörð Bjarmason wrote:\n> I'm not sure I understand this change really. The result in always XML,\n> so application/xhtml+xml is redundant, text/html, or both?\n\nTo be honest, using an http-equiv=\"content-type\" in XHTML is confusing. When \nyou do use one, your goal shouldn’t really be to specify the document’s MIME \ntype. After all, the first three lines of each page say\n\n\t\n\t\n\t\n\nThose lines are more than enough to determine that something is using XHTML \nand UTF-8. Instead, the idea is to help out a parser that is incorrectly \nparsing the document as HTML (instead of as XHTML). Historical W3C documents \n(that were applicable when http-equiv=\"content-type\" was allowed in XHTML) [1]\n[2][3] indicate that http-equiv=\"content-type\" should be used like this:\n\n\t\n\nIn other words, to use http-equiv=\"content-type\" properly in XHTML, you had to \nlie about the document’s type. The fact that this is confusing is probably \npart of why WHATWG disallowed it in the HTML Standard.\n\n> But aside from that: I have seen browsers get the lack of encoding=\"\"\n> \"wrong\" with data at rest, don't some still default to ISO-8859-1?\n> \n> So won't this result in badly decoded data if you save the web page &\n> view it locally?\n\nI tested this idea in ungoogled-chromium, Firefox and Pale Moon. Other than \nPale Moon in one specific circumstance, they all used UTF-8 as the encoding. \nPale Moon used windows-1252, but only when the file ended with .html. When the \nfile ended with .xhtml, Pale Moon used UTF-8. That being said, we don’t have to \nuse an http-equiv=\"content-type\" to fix the problem. Instead, we can use a \n which is allowed by the HTML Standard [4].\n\n[1]: \n[2]: \n[3]: \n\n[4]: \n\n\n\n"},{"id":"450671","messageId":"YiaUP3b3PJWZx+AY@camp.crustytoothpaste.net","threadId":"57513","inReplyTo":"20220307033723.175553-3-jason@jasonyundt.email","subject":"Re: [PATCH 2/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"brian m. carlson","fromEmail":"sandals@crustytoothpaste.net","sentAt":"2022-03-07T23:24:47Z","receivedAt":"2022-03-07T23:24:53Z","isPatch":true,"sender":{"key":"sandals@crustytoothpaste.net","avatar":"https://avatars.githubusercontent.com/u/497054?v=4"},"body":"On 2022-03-07 at 03:37:23, Jason Yundt wrote:\n> Before this change, gitweb would generate pages which included:\n> \n> \t\n> \n> A meta element with http-equiv=\"content-type\" is said to be in the\n> \"Encoding declaration state\". According to the HTML Standard,\n> \n> \tThe Encoding declaration state may be used in HTML documents,\n> \tbut elements with an http-equiv attribute in that state must not\n> \tbe used in XML documents.\n> \n> \tSource: \n> \n> This change removes that meta element since gitweb always generates XML\n> documents.\n\nThis change seems fine. We do specify this in the HTTP header,\nincluding the character set, which is what matters, so this should work\nin every browser, and the http-equiv is unneeded.\n\nI also don't think we need a meta header here, since we have an XML\ndeclaration, and that's controlling in this situation. This isn't\nregular HTML and we don't declare it as such, so using a meta header to\ncontrol this isn't correct: the XML declaration should be used instead\nin the event a user downloads this to a local disk and processes it\noutside the context of an HTTP request.\n\nSince we control the HTTP headers, I'd actually argue that your test\nmight well reject all http-equiv headers since they could be done much\nbetter with actual HTTP headers (and would therefore work with\nnon-browser clients), but I don't think that's worth a reroll, nor do I\nthink a test is even needed here (but bonus points for adding one).\n\nSo I think this looks good as is. Thanks for the patch.\n-- \nbrian m. carlson (he/him or they/them)\nToronto, Ontario, CA\n"},{"id":"450692","messageId":"20220308010711.61817-1-jason@jasonyundt.email","threadId":"57513","inReplyTo":"20220307033723.175553-1-jason@jasonyundt.email","subject":"[PATCH v2 0/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"Jason Yundt","fromEmail":"jason@jasonyundt.email","sentAt":"2022-03-08T01:07:09Z","receivedAt":"2022-03-08T01:08:34Z","isPatch":true,"sender":{"key":"jason@jasonyundt.email","avatar":"https://avatars.githubusercontent.com/u/5579359?v=4"},"body":"See the second commit's message for more details. Compared to the first\nversion of this patch, this one\n\n- keeps an extra variable,\n- replaces the http-equiv=\"content-type\" tag with a charset= one, and\n- removes the -i flag from grep.\n\nJason Yundt (2):\n comment: fix typo\n gitweb: remove invalid http-equiv=\"content-type\"\n\n gitweb/gitweb.perl | 2 +-\n t/t9502-gitweb-standalone-parse-output.sh | 18 +++++++++++++++++-\n 2 files changed, 18 insertions(+), 2 deletions(-)\n\n-- \n2.35.1\n\n"},{"id":"450693","messageId":"20220308010711.61817-3-jason@jasonyundt.email","threadId":"57513","inReplyTo":"20220307033723.175553-1-jason@jasonyundt.email","subject":"[PATCH v2 2/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"Jason Yundt","fromEmail":"jason@jasonyundt.email","sentAt":"2022-03-08T01:07:11Z","receivedAt":"2022-03-08T01:08:36Z","isPatch":true,"sender":{"key":"jason@jasonyundt.email","avatar":"https://avatars.githubusercontent.com/u/5579359?v=4"},"body":"Before this change, gitweb would generate pages which included:\n\n\t\n\nA meta element with http-equiv=\"content-type\" is said to be in the\n\"Encoding declaration state\". According to the HTML Standard,\n\n\tThe Encoding declaration state may be used in HTML documents,\n\tbut elements with an http-equiv attribute in that state must not\n\tbe used in XML documents.\n\n\tSource: \n\nGitweb always generates XML documents, so its use of\nhttp-equiv=\"content-type\" was invalid. This change replaces that tag\nwith\n\n\t\n\nwhich is equivalent [1] and allowed in XML documents [2].\n\n[1]: \n[2]: \n\nSigned-off-by: Jason Yundt \n---\n gitweb/gitweb.perl | 2 +-\n t/t9502-gitweb-standalone-parse-output.sh | 16 ++++++++++++++++\n 2 files changed, 17 insertions(+), 1 deletion(-)\n\ndiff --git a/gitweb/gitweb.perl b/gitweb/gitweb.perl\nindex fbd1c20a23..59457c1004 100755\n--- a/gitweb/gitweb.perl\n+++ b/gitweb/gitweb.perl\n@@ -4225,7 +4225,7 @@ sub git_header_html {\n \n \n \n-\n+\n \n \n $title\ndiff --git a/t/t9502-gitweb-standalone-parse-output.sh b/t/t9502-gitweb-standalone-parse-output.sh\nindex e7363511dd..0b06e2d6b0 100755\n--- a/t/t9502-gitweb-standalone-parse-output.sh\n+++ b/t/t9502-gitweb-standalone-parse-output.sh\n@@ -207,4 +207,20 @@ test_expect_success 'xss checks' '\n \txss \"\" \"$TAG+\"\n '\n \n+check_encoding_meta_element() {\n+\tgitweb_run \"$@\" &&\n+\t! grep -E \"http-equiv=['\\\"]?content-type\" gitweb.body &&\n+\tgrep -F '' gitweb.body\n+}\n+\n+# One of those can be used in XHTML, the other one can't. See:\n+# \n+# \n+test_expect_success 'no http-equiv=\"content-type\", yes charset=\"utf-8\"' '\n+\tcheck_encoding_meta_element &&\n+\tcheck_encoding_meta_element \"p=.git\" &&\n+\tcheck_encoding_meta_element \"p=.git;a=log\" &&\n+\tcheck_encoding_meta_element \"p=.git;a=tree\"\n+'\n+\n test_done\n-- \n2.35.1\n\n"},{"id":"450694","messageId":"20220308010711.61817-2-jason@jasonyundt.email","threadId":"57513","inReplyTo":"20220307033723.175553-1-jason@jasonyundt.email","subject":"[PATCH v2 1/2] comment: fix typo","fromName":"Jason Yundt","fromEmail":"jason@jasonyundt.email","sentAt":"2022-03-08T01:07:10Z","receivedAt":"2022-03-08T01:08:56Z","isPatch":true,"sender":{"key":"jason@jasonyundt.email","avatar":"https://avatars.githubusercontent.com/u/5579359?v=4"},"body":"Signed-off-by: Jason Yundt \n---\n t/t9502-gitweb-standalone-parse-output.sh | 2 +-\n 1 file changed, 1 insertion(+), 1 deletion(-)\n\ndiff --git a/t/t9502-gitweb-standalone-parse-output.sh b/t/t9502-gitweb-standalone-parse-output.sh\nindex 3167473b30..e7363511dd 100755\n--- a/t/t9502-gitweb-standalone-parse-output.sh\n+++ b/t/t9502-gitweb-standalone-parse-output.sh\n@@ -34,7 +34,7 @@ EOF\n #\n # This will check that gitweb HTTP header contains proposed filename\n # as with '.tar' suffix added, and that generated tarfile\n-# (gitweb message body) has as prefix for al files in tarfile\n+# (gitweb message body) has as prefix for all files in tarfile\n #\n # default to \n check_snapshot () {\n-- \n2.35.1\n\n"},{"id":"450695","messageId":"Yia2Y2zoOjvVdJ9O@camp.crustytoothpaste.net","threadId":"57513","inReplyTo":"20220308010711.61817-3-jason@jasonyundt.email","subject":"Re: [PATCH v2 2/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"brian m. carlson","fromEmail":"sandals@crustytoothpaste.net","sentAt":"2022-03-08T01:50:27Z","receivedAt":"2022-03-08T01:50:34Z","isPatch":true,"sender":{"key":"sandals@crustytoothpaste.net","avatar":"https://avatars.githubusercontent.com/u/497054?v=4"},"body":"On 2022-03-08 at 01:07:11, Jason Yundt wrote:\n> diff --git a/gitweb/gitweb.perl b/gitweb/gitweb.perl\n> index fbd1c20a23..59457c1004 100755\n> --- a/gitweb/gitweb.perl\n> +++ b/gitweb/gitweb.perl\n> @@ -4225,7 +4225,7 @@ sub git_header_html {\n> \n> \n> \n> -\n> +\n\nI don't actually think this is an improvement. I don't think it's\nnecessary, considering we have an XML declaration and the HTTP header,\nboth of which already say it's UTF-8 and will take precedence over this.\n-- \nbrian m. carlson (he/him or they/them)\nToronto, Ontario, CA\n"},{"id":"450697","messageId":"xmqqbkyhp4wv.fsf@gitster.g","threadId":"57513","inReplyTo":"20220308010711.61817-1-jason@jasonyundt.email","subject":"Re: [PATCH v2 0/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2022-03-08T02:13:52Z","receivedAt":"2022-03-08T02:13:57Z","isPatch":true,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Jason Yundt writes:\n\n> - keeps an extra variable,\n\nI am not sure if this is an improvement. The original had two\nplaces that used $content_type, but after getting rid of one, there\nis only one place that needed the value, which can be used in place;\nand it was quite clear that was what was going on in the previous\niteration.\n\nAbout the thing, it seems that brian already commented on it.\n\nThanks.\n\n\n"},{"id":"450725","messageId":"4386099.LvFx2qVVIh@jason-desktop-linux","threadId":"57513","inReplyTo":"xmqqbkyhp4wv.fsf@gitster.g","subject":"Re: [PATCH v2 0/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"Jason Yundt","fromEmail":"jason@jasonyundt.email","sentAt":"2022-03-08T12:26:51Z","receivedAt":"2022-03-08T12:27:27Z","isPatch":true,"sender":{"key":"jason@jasonyundt.email","avatar":"https://avatars.githubusercontent.com/u/5579359?v=4"},"body":"On Monday, March 7, 2022 9:13:52 PM EST Junio C Hamano wrote:\n> About the thing, it seems that brian already commented on it.\n\nThanks for mentioning that. I now see Brian’s comments on the archive. My mail \nserver was blocking him (via zen.spamhaus.org), but I’ve added his server to \nthe allowlist.\n\n\n"},{"id":"450726","messageId":"220308.86cziwy595.gmgdl@evledraar.gmail.com","threadId":"57513","inReplyTo":"Yia2Y2zoOjvVdJ9O@camp.crustytoothpaste.net","subject":"Re: [PATCH v2 2/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"Ævar Arnfjörð Bjarmason","fromEmail":"avarab@gmail.com","sentAt":"2022-03-08T12:44:35Z","receivedAt":"2022-03-08T12:54:04Z","isPatch":true,"sender":{"key":"avarab@gmail.com","avatar":"https://avatars.githubusercontent.com/u/45301?v=4"},"body":"\nOn Tue, Mar 08 2022, brian m. carlson wrote:\n\n> [[PGP Signed Part:Undecided]]\n> On 2022-03-08 at 01:07:11, Jason Yundt wrote:\n>> diff --git a/gitweb/gitweb.perl b/gitweb/gitweb.perl\n>> index fbd1c20a23..59457c1004 100755\n>> --- a/gitweb/gitweb.perl\n>> +++ b/gitweb/gitweb.perl\n>> @@ -4225,7 +4225,7 @@ sub git_header_html {\n>> \n>> \n>> \n>> -\n>> +\n>\n> I don't actually think this is an improvement. I don't think it's\n> necessary, considering we have an XML declaration and the HTTP header,\n> both of which already say it's UTF-8 and will take precedence over this.\n\nAgeed. I was a bit surprised per Jason's\nhttps://lore.kernel.org/git/109813056.nniJfEyVGO@jason-desktop-linux/\nthat the removal wasn't kept.\n\nI.e. he was replying to a question of mine asking whether we didn't need\nthis data at rest, e.g if you save the page. I didn't notice the \"http('HTTP_ACCEPT') &&\n\t \t $cgi->http('HTTP_ACCEPT') =~ m/(,|;|\\s|^)application\\/xhtml\\+xml(,|;|\\s|$)/ &&\n\t \t $cgi->Accept('application/xhtml+xml') != 0) {\n\t-\t\treturn 'application/xhtml+xml';\n\t+\t\treturn get_content_type_html();\n\t \t} else {\n\t \t\treturn 'text/html';\n\t \t}\n\t@@ -4214,6 +4220,7 @@ sub git_header_html {\n\t \n\t \tmy $title = get_page_title();\n\t \tmy $content_type = get_content_type_html();\n\t+\tmy $content_type_xml = get_content_type_html();\n\t \tprint $cgi->header(-type=>$content_type, -charset => 'utf-8',\n\t \t -status=> $status, -expires => $expires)\n\t \t\tunless ($opts{'-no_http_header'});\n\t@@ -4225,7 +4232,7 @@ sub git_header_html {\n\t \n\t \n\t \n\t-\n\t+\n\t \n\t \n\t $title\n\nOf course we might then *also* decide that in this\ncase isn't needed at all, but isn't that a seperate change?\n\nAnd won't conforming browsers treat application/xhtml+xml differently\nwhen the page is saved? A long time ago (Idid some web development)\nusing it would enable pedantic strictness in browsers, i.e. unclosed\ntags etc. would be a hard error, but I can't reproduce that locally in\neither Firefox or Chrome now (with just the gitweb output as-is with\nthat http-equiv tweaked).\n\nSo maybe it does nothing, or maybe it's just those browser...\n"},{"id":"450750","messageId":"4175455.ejJDZkT8p0@jason-desktop-linux","threadId":"57513","inReplyTo":"220308.86cziwy595.gmgdl@evledraar.gmail.com","subject":"Re: [PATCH v2 2/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"Jason Yundt","fromEmail":"jason@jasonyundt.email","sentAt":"2022-03-08T14:54:21Z","receivedAt":"2022-03-08T14:54:25Z","isPatch":true,"sender":{"key":"jason@jasonyundt.email","avatar":"https://avatars.githubusercontent.com/u/5579359?v=4"},"body":"On Tuesday, March 8, 2022 7:44:35 AM EST Ævar Arnfjörð Bjarmason wrote:\n> Maybe I still don't understand this, but the commit message seems to me\n> be conflating whether we send the *right* http-equiv with whether we\n> send it at all,\n\nThe intent behind the commit message is to say that is never correct in XHTML.\n\n> i.e. if the problem is that XML documents shouldn't be\n> text/html isn't this correct?:\n> \t\n> \tdiff --git a/gitweb/gitweb.perl b/gitweb/gitweb.perl\n> \tindex fbd1c20a232..c1c5af0b197 100755\n> \t--- a/gitweb/gitweb.perl\n> \t+++ b/gitweb/gitweb.perl\n> \t@@ -4049,7 +4049,13 @@ sub get_page_title {\n> \t \treturn $title;\n> \t }\n> \t \n> \t+sub get_content_type_xml {\n> \t+\treturn 'application/xhtml+xml';\n> \t+}\n> \t+\n> \t sub get_content_type_html {\n> \t+\tmy ($want_xml) = @_;\n> \t+\n> \t \t# require explicit support from the UA if we are to send the page as\n> \t \t# 'application/xhtml+xml', otherwise send it as plain old 'text/html'.\n> \t \t# we have to do this because MSIE sometimes globs '*/*', pretending to\n> \t@@ -4057,7 +4063,7 @@ sub get_content_type_html {\n> \t \tif (defined $cgi->http('HTTP_ACCEPT') &&\n> \t \t $cgi->http('HTTP_ACCEPT') =~ m/(,|;|\\s|^)application\\/xhtml\\+xml(,|;|\\s|$)/ &&\n> \t \t $cgi->Accept('application/xhtml+xml') != 0) {\n> \t-\t\treturn 'application/xhtml+xml';\n> \t+\t\treturn get_content_type_html();\n\nI’m guessing that you meant to call get_content_type_xml() here.\n\n> \t \t} else {\n> \t \t\treturn 'text/html';\n> \t \t}\n> \t@@ -4214,6 +4220,7 @@ sub git_header_html {\n> \t \n> \t \tmy $title = get_page_title();\n> \t \tmy $content_type = get_content_type_html();\n> \t+\tmy $content_type_xml = get_content_type_html();\n\nI’m also guessing that you meant to call get_content_type_xml() here.\n\n> \t \tprint $cgi->header(-type=>$content_type, -charset => 'utf-8',\n> \t \t -status=> $status, -expires => $expires)\n> \t \t\tunless ($opts{'-no_http_header'});\n> \t@@ -4225,7 +4232,7 @@ sub git_header_html {\n> \t \n> \t \n> \t \n> \t-\n> \t+\n> \t \n> \t \n> \t $title\n\nWith those assumptions in mind, I don’t think that your code is correct if\nthe problem is that XML documents shouldn't be text/html. Here’s why:\n\n1. XML documents shouldn’t contain http-equiv=\"content-type\" [1].\n2. When a meta’s http-equiv attribute equals content-type, then its content\n attribute should equal “the literal string \"text/html;\", optionally\n followed by any number of ASCII whitespace, followed by the literal\n string \"charset=utf-8\".” [1]\n\n[1]: \n\n\n\n"},{"id":"450752","messageId":"20220308155612.105957-3-jason@jasonyundt.email","threadId":"57513","inReplyTo":"20220308155612.105957-1-jason@jasonyundt.email","subject":"[PATCH v3 2/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"Jason Yundt","fromEmail":"jason@jasonyundt.email","sentAt":"2022-03-08T15:56:12Z","receivedAt":"2022-03-08T15:58:24Z","isPatch":true,"sender":{"key":"jason@jasonyundt.email","avatar":"https://avatars.githubusercontent.com/u/5579359?v=4"},"body":"Before this change, gitweb would generate pages which included:\n\n\t\n\nWhen a meta's http-equiv equals \"content-type\", the http-equiv is said\nto be in the \"Encoding declaration state\". According to the HTML\nStandard,\n\n\tThe Encoding declaration state may be used in HTML documents,\n\tbut elements with an http-equiv attribute in that state must not\n\tbe used in XML documents.\n\n\tSource: \n\nThis change removes that meta element since gitweb always generates XML\ndocuments.\n\nSigned-off-by: Jason Yundt \n---\n gitweb/gitweb.perl | 4 +---\n t/t9502-gitweb-standalone-parse-output.sh | 13 +++++++++++++\n 2 files changed, 14 insertions(+), 3 deletions(-)\n\ndiff --git a/gitweb/gitweb.perl b/gitweb/gitweb.perl\nindex fbd1c20a23..606b50104c 100755\n--- a/gitweb/gitweb.perl\n+++ b/gitweb/gitweb.perl\n@@ -4213,8 +4213,7 @@ sub git_header_html {\n \tmy %opts = @_;\n \n \tmy $title = get_page_title();\n-\tmy $content_type = get_content_type_html();\n-\tprint $cgi->header(-type=>$content_type, -charset => 'utf-8',\n+\tprint $cgi->header(-type=>get_content_type_html(), -charset => 'utf-8',\n \t -status=> $status, -expires => $expires)\n \t\tunless ($opts{'-no_http_header'});\n \tmy $mod_perl_version = $ENV{'MOD_PERL'} ? \" $ENV{'MOD_PERL'}\" : '';\n@@ -4225,7 +4224,6 @@ sub git_header_html {\n \n \n \n-\n \n \n $title\ndiff --git a/t/t9502-gitweb-standalone-parse-output.sh b/t/t9502-gitweb-standalone-parse-output.sh\nindex e7363511dd..8cb582f0e6 100755\n--- a/t/t9502-gitweb-standalone-parse-output.sh\n+++ b/t/t9502-gitweb-standalone-parse-output.sh\n@@ -207,4 +207,17 @@ test_expect_success 'xss checks' '\n \txss \"\" \"$TAG+\"\n '\n \n+no_http_equiv_content_type() {\n+\tgitweb_run \"$@\" &&\n+\t! grep -E \"http-equiv=['\\\"]?content-type\" gitweb.body\n+}\n+\n+# See: \n+test_expect_success 'no http-equiv=\"content-type\" in XHTML' '\n+\tno_http_equiv_content_type &&\n+\tno_http_equiv_content_type \"p=.git\" &&\n+\tno_http_equiv_content_type \"p=.git;a=log\" &&\n+\tno_http_equiv_content_type \"p=.git;a=tree\"\n+'\n+\n test_done\n-- \n2.35.1\n\n"},{"id":"450753","messageId":"20220308155612.105957-1-jason@jasonyundt.email","threadId":"57513","inReplyTo":"20220308010711.61817-1-jason@jasonyundt.email","subject":"[PATCH v3 0/2] gitweb: remove invalid http-equiv=\"content-type\"","fromName":"Jason Yundt","fromEmail":"jason@jasonyundt.email","sentAt":"2022-03-08T15:56:10Z","receivedAt":"2022-03-08T15:59:01Z","isPatch":true,"sender":{"key":"jason@jasonyundt.email","avatar":"https://avatars.githubusercontent.com/u/5579359?v=4"},"body":"See the second commit's message for more details. Compared to the second\nversion of this patch, this one\n\n- removes the extra variable again,\n- doesn't include a and\n- corrects a technical error in the second commit’s message.\n\nJason Yundt (2):\n comment: fix typo\n gitweb: remove invalid http-equiv=\"content-type\"\n\n gitweb/gitweb.perl | 4 +---\n t/t9502-gitweb-standalone-parse-output.sh | 15 ++++++++++++++-\n 2 files changed, 15 insertions(+), 4 deletions(-)\n\n-- \n2.35.1\n\n"},{"id":"450754","messageId":"20220308155612.105957-2-jason@jasonyundt.email","threadId":"57513","inReplyTo":"20220308155612.105957-1-jason@jasonyundt.email","subject":"[PATCH v3 1/2] comment: fix typo","fromName":"Jason Yundt","fromEmail":"jason@jasonyundt.email","sentAt":"2022-03-08T15:56:11Z","receivedAt":"2022-03-08T15:59:02Z","isPatch":true,"sender":{"key":"jason@jasonyundt.email","avatar":"https://avatars.githubusercontent.com/u/5579359?v=4"},"body":"Signed-off-by: Jason Yundt \n---\n t/t9502-gitweb-standalone-parse-output.sh | 2 +-\n 1 file changed, 1 insertion(+), 1 deletion(-)\n\ndiff --git a/t/t9502-gitweb-standalone-parse-output.sh b/t/t9502-gitweb-standalone-parse-output.sh\nindex 3167473b30..e7363511dd 100755\n--- a/t/t9502-gitweb-standalone-parse-output.sh\n+++ b/t/t9502-gitweb-standalone-parse-output.sh\n@@ -34,7 +34,7 @@ EOF\n #\n # This will check that gitweb HTTP header contains proposed filename\n # as with '.tar' suffix added, and that generated tarfile\n-# (gitweb message body) has as prefix for al files in tarfile\n+# (gitweb message body) has as prefix for all files in tarfile\n #\n # default to \n check_snapshot () {\n-- \n2.35.1\n\n"}]}