Change Details

e.g. `[[media:Foo.pdf#page=3]] ` should link to third page (assuming pdf client supports that). Currently the fragment identifier is stripped [If in legacy fragment mode, that would also break the link, but that's a separate issue] Anyone wanting to fix this, you probably need to: * Adjust Linker::makeMediaLinkFile so that it takes the fragment from $title (e.g. See Title::getFragment or Title::getFragmentForUrl) (but keeps taking the rest of the url from $file). You do not want to add the fragment in the broken case. * Since this is not linking to a wikipage (or even html), an argument could be made that you shouldn't use Title::getFragmentForUrl(), but instead just use Title::getFragment(), as for other media types, fragments should basically be opaque blobs that could be anything (See https://en.wikipedia.org/wiki/Fragment_identifier#Examples ). * After fixing this, you should add additional parser tests in tests/parser/parserTests.txt

e.g. `[[media:Foo.pdf#page=3]] ` should link to third page (assuming pdf client supports that). Currently the fragment identifier is stripped [If in legacy fragment mode, that would also break the link, but that's a separate issue] Anyone wanting to fix this, you probably need to: * Adjust [[ https://doc.wikimedia.org/mediawiki-core/master/php/classLinker.html#a0b05ecbb5397cc87a30346c83962c3db | Linker::makeMediaLinkFile ]] so that it takes the fragment from $title (e.g. See [[ https://doc.wikimedia.org/mediawiki-core/1.28.2/php/classTitle.html#a343de80fa8b360db8ce95d91943d877c | Title::getFragment ]] or [[ https://doc.wikimedia.org/mediawiki-core/1.28.2/php/classTitle.html#a6c41a1d0ce02439aa61318466d3a6cb9 | Title::getFragmentForUrl ]]) (but keeps taking the rest of the url from $file). You do not want to add the fragment in the broken case. * Since this is not linking to a wikipage (or even html), an argument could be made that you shouldn't use Title::getFragmentForUrl(), but instead just use Title::getFragment(), as for other media types, fragments should basically be opaque blobs that could be anything (See https://en.wikipedia.org/wiki/Fragment_identifier#Examples ). * After fixing this, you should add additional parser tests in tests/parser/parserTests.txt ([[ https://www.mediawiki.org/wiki/Parser_tests | more details ]])

e.g. `[[media:Foo.pdf#page=3]] ` should link to third page (assuming pdf client supports that). Currently the fragment identifier is stripped [If in legacy fragment mode, that would also break the link, but that's a separate issue] Anyone wanting to fix this, you probably need to: * Adjust Linker::makeMediaLinkFile[[ https://doc.wikimedia.org/mediawiki-core/master/php/classLinker.html#a0b05ecbb5397cc87a30346c83962c3db | Linker::makeMediaLinkFile ]] so that it takes the fragment from $title (e.g. See Title::getFragment or Title::getFragmentForUrl[[ https://doc.wikimedia.org/mediawiki-core/1.28.2/php/classTitle.html#a343de80fa8b360db8ce95d91943d877c | Title::getFragment ]] or [[ https://doc.wikimedia.org/mediawiki-core/1.28.2/php/classTitle.html#a6c41a1d0ce02439aa61318466d3a6cb9 | Title::getFragmentForUrl ]]) (but keeps taking the rest of the url from $file). You do not want to add the fragment in the broken case. * Since this is not linking to a wikipage (or even html), an argument could be made that you shouldn't use Title::getFragmentForUrl(), but instead just use Title::getFragment(), as for other media types, fragments should basically be opaque blobs that could be anything (See https://en.wikipedia.org/wiki/Fragment_identifier#Examples ). * After fixing this, you should add additional parser tests in tests/parser/parserTests.txt ([[ https://www.mediawiki.org/wiki/Parser_tests | more details ]])