This simplifies the code quite a bit.
- GERMAN_UMLAUTS are now handled through a key-value attribute.
The dictionary generator does not need to know about it any more.
- FRENCH_LIGATURES are deprecated as we handle them with shortcuts now.
- CONTAINS_BIGRAMS is deprecated. Bigram processing is always applied
regardless of this flag.
Bug: 11281748
Change-Id: If567e52e245a9342adc7f3104a0f7d8d782df8c1
In new format, each bigram entry has flags (1 byte), a terminal id (3 byte),
a time-stamp (4 byte), a counter (1 byte) and a level (1 byte).
Bug: 10920255
Bug: 10920165
Change-Id: I0f7fc125a6178e6d25a07e8462afc41a7f57e3e1
When there are too many bigrams, we stop reading the file,
so the file pointer is in an inconsistent place. This means we
have no idea what's going to happen next. It's better to crash
right away.
Change-Id: Id3b7b78cbe4fda3493b3c9c46758763e1ab5f6a3
