[orca] Add additional support for presentation of mathematical symbols * Add ability to fallback on the nam

[Joanmarie Diggs <joanied src gnome org>]

commit 5962a7ef1fece557b734e5ab616d4c1c87037cd8
Author: Joanmarie Diggs <jdiggs igalia com>
Date:   Fri Jan 9 22:31:02 2015 -0500

    Add additional support for presentation of mathematical symbols
    * Add ability to fallback on the name from unicode data
    * Add fallbacks for arrows, shapes, and additional operators
    * Add ability for the user to add and override symbol names
    
    https://bugzilla.gnome.org/show_bug.cgi?id=741473

 src/orca/mathsymbols.py |   35 +++++++++++++++++++++++++++++++----
 1 files changed, 31 insertions(+), 4 deletions(-)
---
diff --git a/src/orca/mathsymbols.py b/src/orca/mathsymbols.py
index 219e7d6..b6bfe7d 100644
--- a/src/orca/mathsymbols.py
+++ b/src/orca/mathsymbols.py
@@ -26,9 +26,12 @@ __copyright__ = "Copyright (c) 2014 Igalia, S.L."
 __license__   = "LGPL"
 
 import re
+import unicodedata
 
 from .orca_i18n import C_
 
+fallbackOnUnicodeData = True
+
 SPEAK_NEVER = 1
 SPEAK_ALWAYS = 2
 SPEAK_FOR_CHARS = 3
@@ -36,7 +39,18 @@ speakStyle = SPEAK_ALWAYS
 
 _all = {}
 _alnum = {}
+_arrows = {}
 _operators = {}
+_shapes = {}
+
+# Note that the following are to help us identify what is likely a math symbol
+# (as opposed to one serving the function of an image in "This way up.")
+_arrows.update(dict.fromkeys(map(chr, range(0x2190, 0x2200))))
+_arrows.update(dict.fromkeys(map(chr, range(0x2750, 0x2800))))
+_arrows.update(dict.fromkeys(map(chr, range(0x2b30, 0x2b50))))
+_operators.update(dict.fromkeys(map(chr, range(0x2220, 0x2300))))
+_operators.update(dict.fromkeys(map(chr, range(0x2a00, 0x2b00))))
+_shapes.update(dict.fromkeys(map(chr, range(0x25a0, 0x2600))))
 
 # Unicode has a huge number of individual symbols to include styles, such as
 # bold, italic, double-struck, etc. These are so far not supported by speech
@@ -2003,7 +2017,9 @@ _operators['\u22fe'] = C_('math symbol', 'small contains with overbar')
 _operators['\u22ff'] = C_('math symbol', 'z notation bag membership')
 
 _all.update(_alnum)
+_all.update(_arrows)
 _all.update(_operators)
+_all.update(_shapes)
 _RE = re.compile('[%s]' % ''.join(list(_all.keys())), re.UNICODE)
 
 def _getStyleString(symbol):
@@ -2039,11 +2055,20 @@ def _getStyleString(symbol):
 
     return "%s"
 
+def updateSymbols(symbolDict):
+    global _all
+    _all.update(symbolDict)
+
 def _getSpokenName(symbol, includeStyle):
-    name = _all.get(symbol)
-    if not name:
+    if not symbol in _all:
         return ""
 
+    name = _all.get(symbol)
+    if not name and fallbackOnUnicodeData:
+        name = unicodedata.name(symbol).lower()
+        updateSymbols({symbol: name})
+        return name
+
     if includeStyle and symbol in _alnum:
         name = _getStyleString(symbol) % name
 
@@ -2054,8 +2079,10 @@ def getCharacterName(symbol):
 
 def adjustForSpeech(string):
     chars = set(re.findall(_RE, string))
+    includeStyle = speakStyle == SPEAK_ALWAYS
     for char in chars:
-        name = " %s " % _getSpokenName(char, speakStyle == SPEAK_ALWAYS)
-        string = re.sub(char, name, string)
+        name = _getSpokenName(char, includeStyle)
+        if name:
+            string = re.sub(char, " %s " % name, string)
 
     return string