A few changes, plus treeprint.py.

2 files changed, 82 insertions, 0 deletions

diff --git a/dotXCompose b/dotXCompose
index 9ca14c4..83376be 100644
--- a/dotXCompose
+++ b/dotXCompose
@@ -303,6 +303,8 @@ include "%L"
 <Multi_key> <underscore> <k>           :  "ₖ"  U2096	  # LATIN SUBSCRIPT SMALL LETTER K
 <Multi_key> <underscore> <m>	       :  "ₘ"  U2098	  # LATIN SUBSCRIPT SMALL LETTER M
 <Multi_key> <underscore> <n>	       :  "ₙ"  U2099	  # LATIN SUBSCRIPT SMALL LETTER N
+<Multi_key> <underscore> <minus>       : "₋"   U208B	  # SUBSCRIPT MINUS
+<Multi_key> <underscore> <plus>	       : "₊"   U208A	  # SUBSCRIPT PLUS
 
 
 # Custom additions: Greek letters.  Mapping corresponds to Emacs Greek
@@ -614,6 +616,8 @@ include "%L"
 <Multi_key> <asciicircum> <ampersand> <y>       : "ʸ"   U02B8           # SUPERSCRIPT Y
 #Maybe add: ˀˁ˃˂
 <Multi_key> <asciicircum> <minus>       : "⁻"   U207B           # SUPERSCRIPT MINUS
+<Multi_key> <asciicircum> <plus>	: "⁺"	U207A		# SUPERSCRIPT PLUS
+
 <Multi_key> <asciitilde> <asciitilde>        : "≈"  U2248           # ALMOST EQUAL TO
 <Multi_key> <s> <h>			: "ʃ"	U0283		# LATIN SMALL LETTER ESH
 <Multi_key> <z> <h>			: "ʒ"	U0292		# LATIN SMALL LETTER EZH
@@ -672,6 +676,7 @@ include "%L"
 <Multi_key> <numbersign> <C>              : "𝄡" U0001d121 # MUSICAL SYMBOL C CLEF
 <Multi_key> <numbersign> <o> <slash>	: "♪" U266a	# EIGHTH NOTE
 <Multi_key> <numbersign> <o> <o>	: "♫" U266b	# BEAMED EIGHTH NOTES
+<Multi_key> <numbersign> <percent>	: "♫" U266b	# BEAMED EIGHTH NOTES
 
 
 # Combining accents, for making things you don't have precomposed chars or keystrokes for:
diff --git a/treeprint.py b/treeprint.py
new file mode 100644
index 0000000..ce4f672
--- /dev/null
+++ b/treeprint.py
@@ -0,0 +1,77 @@
+#!/usr/bin/env python
+
+import sys
+import re
+
+"""
+This program slurps in a .XCompose file on standard input (or several
+concatenated together, since it won't follow includes) and outputs the
+compose sequences in an S-expression-like syntax, showing the prefix tree
+of sequences.  This should bring together some of the groups that use a
+prefix-character, like * for the Greek alphabet and # for musical symbols.
+And scatter other related things far and wide.  But it might be fun to look
+at.
+
+Prefix conflicts (in which you have a compose sequence that is the proper
+prefix of another) and exact conflicts (in which you have a compose
+sequence listed two with two different translations) cannot be handled
+gracefully in this notation, and they are not handled gracefully by this
+program.  The tie is not broken in a consistent or predictable fashion,
+etc: this is a case of GIGO.  Deal with it.
+"""
+
+def showdict(data, indent):
+    for (key, value) in data.iteritems():
+        print " "*indent + "("+key,
+        if type(value)==dict:
+            print ""
+            showdict(value, indent+4)
+        else:
+            print "    "+value+")"
+        
+
+listing={}
+
+try:
+    while True:
+        line=sys.stdin.next()
+        # print "((%s))"%line
+        startpos=0
+        name=[]
+        dupsfound=[]
+        while True:
+            m=re.match("\s*<(\w+)>",line[startpos:])
+            if not m:
+                break
+            word=m.group(1)
+            name.append(word)
+            startpos+=m.end()
+        if startpos<=0:
+            continue
+        m=re.match(r'[^"]*"(.+?)"',line)
+        if not m:
+            # shouldn't happen, but just in case
+            val='???'
+            print "couldn't make sense of line: "+line
+        else:
+            val=m.group(1)
+        cur=listing
+        for elt in name[:-1]:
+            if type(cur)==dict:
+                if not cur.has_key(elt):
+                    cur[elt]={}
+                cur=cur[elt]        # This will fail for prefix conflicts
+            else:
+                break           # prefix conflict
+        # Presumably by now we're at the end, pointing to an empty dict.
+        if type(cur)==dict:
+            cur[name[-1]]=val
+        else:
+            # fail.  Prefix conflict.  Let's ignore it.
+            pass
+except StopIteration:
+    print "hit end"
+
+showdict(listing,0)
+
+