When analyzing, count decoder errors excluding the preserved points
authorPhilip Jägenstedt <philipj@opera.com>
Mon, 16 Apr 2012 09:12:20 +0000 (11:12 +0200)
committerPhilip Jägenstedt <philipj@opera.com>
Mon, 16 Apr 2012 09:12:20 +0000 (11:12 +0200)
commit325edebc5ed13c8fcd63f52518463dbd8f119265
tree028bb6373764f1f283110f4b09f4e6260bbec121
parent0c7ba3845e41e6706a8ed7c719da90a177f7128c
When analyzing, count decoder errors excluding the preserved points

Since Big5-UAO maps some bytes that Big5-HKSCS does not, UAO content
will produce more decoder errors using big5-index.txt. To avoid
excluding such content, count decoder errors after decoding with the
preserve list.

This produces a raw list of 427 pages, previously it was 294.
big5-hkscs-vs-uao.txt
tw-analyze.py