Thanks Ryan and Yves. I guess I should look at the modules and Yves should look at core. I will try to have the modules fixed by our next meeting. From:
xliff@lists.oasis-open.org [mailto:
xliff@lists.oasis-open.org] On Behalf Of Ryan King Sent: Tuesday, December 08, 2015 9:41 AM To: Yves Savourel; XLIFF Main List Subject: RE: [xliff] Implementing XLFF 2.0 test suite files Hi Yves thanks for the reply. · .coreinvalidad_nonisolatedecwithstartref.xlf - startRef should be added to it. YS> I think the error is legit for this one: it’s why the file is invalid, no? Since the name is bad_nonisolatedec *with* startref.xlf it seems like there * should * be a startRef. It is then invalid because there is an id with startRef (also it should be id=”2” to avoid duplicates across siblings): e.g. <ec id=”2” startRef="1"/>. Otherwise, this file is identical to bad_NonIsolatedEc *Without* StartRef.xlf which doesn’t make sense. · .modulesinvalidad-fs_subfs-not-allowed-w-o-fs.xlf o <note id="n3" fs:fs="paragraph"> should be <note id="n3" fs:fs="p"> o <note id="n4" fs:subFs="noFS"> should be <note id="n4" fs:subFs="no,FS"> YS> It looks like both are legit error because the files are invalid example, but that file and Bad-fs_fs-not-valid-HTML.xlf are exactly the same: each one should have only one error. The name of the test seems to suggest that it is testing whether a subFS is allowed without a fs and this comment supports that: <!-- 5.3.5.2 subFs violation, If the attribute subFs is used, the attribute fs MUST be specified as well. -->. However, this comment seems to support ill-formedness <!-- 5.3.5.1 fs violation, paragraph not a valid value-->. So I think you are right, there should be two files, however, the subFS test is not about ill-formedness, so this particular file should be fixed as specified above and Bad-fs_fs-not-valid-HTML.xlf should serve as the ill-formed file. Thanks, Ryan From:
xliff@lists.oasis-open.org [ mailto:
xliff@lists.oasis-open.org ] On Behalf Of Yves Savourel Sent: Tuesday, December 8, 2015 4:49 AM To: XLIFF Main List <
xliff@lists.oasis-open.org > Subject: RE: [xliff] Implementing XLFF 2.0 test suite files Hi Ryan, I’ve spot checked several files for the core and modules supported by Okapi and I agree: sometimes expected error mask unexpected one. We should be sure the invalid files have only one error. In a few cases I think the errors are legit because that is the error exemplified by the invalid file. See below for a few notes. (I didn’t look at all files). On a related topic: Ryan, what is your opinion on this file:
https://lists.oasis-open.org/archives/xliff/201512/msg00011/test-file.xlf is it valid or not? Thanks, -ys From:
xliff@lists.oasis-open.org [ mailto:
xliff@lists.oasis-open.org ] On Behalf Of Ryan King Sent: Monday, December 7, 2015 5:44 PM To: '
xliff@lists.oasis-open.org' <
xliff@lists.oasis-open.org > Subject: [xliff] Implementing XLFF 2.0 test suite files Hi TC folks, In creating the XLIFF 2.0 OM that we’ve pushed up to GitHub (
https://github.com/Microsoft/XLIFF2-Object-Model ) we found some issues with some of the test suite files (
https://tools.oasis-open.org/version-control/browse/wsvn/xliff/trunk/xliff-20/test-suite/core/valid/#_trunk_xliff-20_test-suite_core_valid_ ) when implementing validation tests. Apologies if these are known issues. I haven’t been able to attend calls or follow the list as closely as I would have liked over the past couple of months. The following files contain duplicated ids across siblings, which depending on how tests are implemented, may mask the actual issue being tested: · .coreinvalidad_invalidextensionattributeonpc.xlf · .coreinvalidad_nonisolatedecwithoutstartref.xlf · .coreinvalidad_nonisolatedecwithstartref.xlf · .coreinvalidad_unknowndatarefendvalue.xlf · .coreinvalidad_unknowndatarefstartvalue.xlf · .coreinvalidad_unknowndatarefvalue.xlf · .modulesinvalidad-res_resourceitem-not-unique.xlf · .modulesvalidgood-res_source-has-no-content-and-href.xlf The following contain a <res:target> language that doesn’t match document language, which depending on how tests are implemented, may mask the actual issue being tested: · .modulesinvalidad-res_resourceitem-not-unique.xlf · .modulesinvalidad-res_resourceitemref-not-unique.xlf · .modulesinvalidad-res_source-has-content-and-href.xlf · .modulesinvalidad-res_source-xml_lang-not-same-as-xliff.xlf · .modulesvalidgood-res_resourceitem-unique.xlf · .modulesvalidgood-res_resourceitemref-unique.xlf · .modulesvalidgood-res_source-has-no-content-and-href.xlf · .modulesvalidgood-res_source-xml_lang-same-as-xliff.xlf The following have a <source> that should be from another namespace, so it should be pre-fixed, e.g. <res:source …><abc:source>Hi</abc:source></res:source>: · .modulesinvalidad-res_source-has-content-and-href.xlf - · .modulesinvalidad-res_source-xml_lang-not-same-as-xliff.xlf · .modulesvalidgood-res_source-xml_lang-same-as-xliff.xlf There are two “orphan” reourceItemRef that don’t point to anything: <res:resourceItemRef ref="r1"/> AND <res:resourceItemRef ref="r2" id="rr1"/> in the following: · .modulesinvalidad-res_resourceitem-not-unique.xlf · .modulesvalidgood-res_source-has-no-content-and-href.xlf · .modulesvalidgood-res_source-xml_lang-same-as-xliff.xlf The following contain more than one ChangeTrack Module: · .modulesinvalidad-ctr_appliesto-not-using-ref-to-resolvableid.xlf · .modulesinvalidad-ctr_property-not-content-or-valid-attribute-ref.xlf · .modulesinvalidad-ctr_ref-not-pointed-to-resolvableid.xlf · .modulesinvalidad-ctr_revisions-not-using-ref-to-resolvableid.xlf · .modulesvalidgood-ctr_appliesto-using-ref-to-resolvableid.xlf · .modulesvalidgood-ctr_property-content-or-valid-attribute-ref.xlf · .modulesvalidgood-ctr_ref-pointed-to-resolvableid.xlf · .modulesvalidgood-ctr_revisions-using-ref-to-resolvableid.xlf Other issues: · .coreinvalidad_nonisolatedecwithstartref.xlf - startRef should be added to it. YS> I think the error is legit for this one: it’s why the file is invalid, no? · .corevalidsample1.xlf - <gls:translation ref="#m2" source="myTermbase"> should be <gls:translation ref="t-#m2" source="myTermbase"> · .modulesinvalidad-fs_subfs-not-allowed-w-o-fs.xlf o <note id="n3" fs:fs="paragraph"> should be <note id="n3" fs:fs="p"> o <note id="n4" fs:subFs="noFS"> should be <note id="n4" fs:subFs="no,FS"> YS> It looks like both are legit error because the files are invalid example, but that file and Bad-fs_fs-not-valid-HTML.xlf are exactly the same: each one should have only one error. · .modulesinvalidad-slr_equivstorage-ec-not-isolated.xlf - In order for the <ec> to not be isolated, it should be in the same unit and not have an id. Thank you and please let me know if we misunderstood any of these issues! Thanks, Ryan