idna-test-v2-expected-diffs.txt

# Expected disagreements between strict IDNA2008 (this library) and UTS #46
# Nontransitional, as exercised by the Unicode IdnaTestV2.txt vectors via
# the `unicode-conformance` Cabal-flagged test-suite.
#
# Format: one entry per non-blank, non-comment line.  Fields are separated
# by TAB characters.  The first field is the disposition; the second is
# the source string from IdnaTestV2.txt (with `\uXXXX` escapes preserved
# rather than decoded so the file is greppable); the third (optional)
# field is a short free-form reason.
#
#     <disposition>  <source>  <reason>
#
# Dispositions:
#
#   skip   -- the vector is not run; the library is not expected to
#             agree with the UTS #46 column for this input, and the
#             disagreement isn't worth a finer-grained classification
#             (e.g. UTS #46 applies a mapping we deliberately don't,
#             producing an output we have no equivalent for).
#
#   xfail  -- the vector is run and is expected to disagree with the
#             UTS #46 column.  If the library happens to agree, that's
#             an "unexpected pass" worth investigating (probably means
#             a recent fix or a tightening on the UTS #46 side).
#
# Entries are added as systematic UTS #46-vs-IDNA2008 differences
# surface from running the harness.  Each should cite either the
# §5892 disposition table, the UTS #46 deviation paragraph, or a
# specific RFC 5891/5893 rule.

# (no entries yet -- harness reports raw agreement on every vector)