Better suggestion with diacritics (#1223)

* Update cdict

* scripts/subst_of_compose.py: Compute substitutions

from compose mappings. They are used when building dictionaries.

* Add substitutions compose data

* Better suggestion with diacritics

This improves the suggestions for words that contain diacritics and
uppercase letters.

This works by stripping diacritics both when building the dictionaries
(using word aliases added in cdict: https://github.com/Julow/cdict/pull/3)
and during lookup. Cdict then takes care of resolving the correct word.

The substitutions are generated using mappings from `fn`, `shift` and
all the `accent_*` modifiers into srcs/compose/substitutions.json
This can be updated easily when more mappings are added.

1 files changed, 23 insertions, 4 deletions

diff --git a/srcs/juloo.keyboard2/ComposeKey.java b/srcs/juloo.keyboard2/ComposeKey.java
index 5be9597..3d97d69 100644
--- a/srcs/juloo.keyboard2/ComposeKey.java
+++ b/srcs/juloo.keyboard2/ComposeKey.java
@@ -22,14 +22,14 @@ public final class ComposeKey
     return null;
   }
 
-  /** Apply the pending compose sequence to char [c]. Returns [null] if no
+  /** Apply a char to the pending compose sequence. Returns [null] if no
       sequence matched. */
-  public static KeyValue apply(int prev, char c)
+  public static KeyValue apply(int state, char c)
   {
     char[] states = ComposeKeyData.states;
     char[] edges = ComposeKeyData.edges;
-    int prev_length = edges[prev];
-    int next = Arrays.binarySearch(states, prev + 1, prev + prev_length, c);
+    int state_length = edges[state];
+    int next = Arrays.binarySearch(states, state + 1, state + state_length, c);
     if (next < 0)
       return null;
     next = edges[next];
@@ -46,6 +46,25 @@ public final class ComposeKey
       return KeyValue.makeCharKey((char)next_header);
   }
 
+  /** Apply char [c] to the pending compose sequence. If the application resolves
+      to a final char state, return it, otherwise return [0]. This is faster
+      than [apply] but do not support sequences of more than 1 char. Used to
+      apply substitutions for word suggestions. */
+  public static char transform_char(int state, char c)
+  {
+    char[] states = ComposeKeyData.states;
+    char[] edges = ComposeKeyData.edges;
+    int state_length = edges[state];
+    int next = Arrays.binarySearch(states, state + 1, state + state_length, c);
+    if (next < 0)
+      return 0;
+    next = edges[next];
+    int next_header = states[next];
+    if (next_header == 0 || next_header == 0xFFFF)
+      return 0;
+    return (char)next_header;
+  }
+
   /** Apply each char of a string to a sequence. Returns [null] if no sequence
       matched. */
   public static KeyValue apply(int prev, String s)