Merge pull request #6 from rlpowell/master

Added an original-camxes-style parser

README.txt

@@ -63,6 +63,17 @@ Partial parsing:
     >>> node.end < len("klama ku ku")
     True
 
+original-camxes-style parsing:
+
+    >>> import camxes_py
+    >>> from camxes_py.transformers import minimal
+    >>> minimal_transformer = minimal.Transformer()
+    >>> text = camxes_py.match("mi la cmen broda fu'ivla li 1 la'o gy english words gy", None, None, minimal_transformer)
+    >>> text
+    ['sentence', [['terms', [['KOhA', 'mi'], ['sumti_6', [['LA', 'la'], ['CMEVLA', 'cmen']]]]], ['bridi_tail_3', [['selbri_3', [['gismu', 'broda'], ['lujvo', "fu'ivla"]]], ['nonabs_terms', [['li_clause', [['LI', 'li'], ['PA', '1']]], ['ZOI_pre', [['ZOI', "la'o"], ['BY', 'gy'], [['zoi_word', 'english '], ['zoi_word', 'words ']], ['BY', 'gy']]]]]]]]]
+    >>>
+
+
 TESTING
 =======
 

TODO

@@ -0,0 +1,3 @@
+- merge any pull requests at
+  https://github.com/teleological/camxes-py/pulls to this repo
+- add a test case in for the minimal transformer

camxes.py

@@ -10,10 +10,10 @@
 
 import camxes_py.parsimonious_ext # expression_nodes
 
-__version__ = "v0.9.0"
+__version__ = "v0.10.0"
 
 PARSERS      = [ 'camxes-ilmen' ]
-TRANSFORMERS = [ 'camxes-json', 'camxes-morphology', 'vlatai', 'node-coverage', 'debug', 'raw' ]
+TRANSFORMERS = [ 'camxes-json', 'camxes-morphology', 'minimal', 'vlatai', 'node-coverage', 'debug', 'raw' ]
 SERIALIZERS  = [ 'json', 'json-pretty', 'json-compact', 'xml' ]
 
 IMPLEMENTATION_RECURSION_LIMIT = {
@@ -102,6 +102,9 @@ def build_transformer(transformer_option, parser):
     elif transformer_option == 'vlatai':
         from camxes_py.transformers import vlatai
         return vlatai.Transformer()
+    elif transformer_option == 'minimal':
+        from camxes_py.transformers import minimal
+        return minimal.Transformer()
     elif transformer_option == 'node-coverage':
         from camxes_py.transformers import node_coverage
         return node_coverage.Transformer(parser)

camxes_py/transformers/minimal.py

@@ -0,0 +1,80 @@
+
+# pylint: disable=I0011, C0111, C0326, no-self-use, unused-argument, invalid-name
+
+import re
+
+from parsimonious.nodes import NodeVisitor
+
+def is_selmaho_expression(name):
+    return re.match(r"^[ABCDFGIJKLMNPRSTUVXYZ]([AEIOUY]([IU]|h[AEIOU])?)?$", name)
+
+class Transformer(object):
+
+    def transform(self, parsed):
+        return Visitor().visit(parsed)
+
+    def default_serializer(self):
+        return lambda x : x.as_json()
+
+class Visitor(NodeVisitor):
+
+    def visit_space_char(self, node, visited_children):
+        return None
+
+    def visit_EOF(self, node, visited_children):
+        return None
+
+    def visit_CMEVLA(self, node, visited_children):
+        return ['CMEVLA', node.text]
+
+    def visit_zoi_word(self, node, visited_children):
+        return ['zoi_word', node.text]
+
+    def visit_gismu_2(self, node, visited_children):
+        return ['gismu', node.text]
+
+    def visit_lujvo(self, node, visited_children):
+        return ['lujvo', node.text]
+
+    def visit_fuhivla(self, node, visited_children):
+        return ["fu'ivla", node.text]
+
+    def generic_visit(self, node, visited_children):
+        # Catch all the cmavo
+        if node.expr_name and is_selmaho_expression(node.expr_name):
+            return [node.expr_name, node.text]
+
+        # Catch all the spaces
+        if node.expr_name == "" and node.text.strip() == "":
+            return None
+
+        # Drop the crap
+        proper_children = list(filter(None, visited_children))
+
+        if len(proper_children) == 0:
+            if len(node.text) > 0:
+                return [node.expr_name, node.text]
+            else:
+                return None
+        else:
+            # Make a tree
+            if len(proper_children) == 1:
+                # Catch the special case where an empty name got used
+                # *and* we're the only parent so we can use our name
+                if len(proper_children[0]) > 1 and proper_children[0][0] == "":
+                    proper_children[0][0] = node.expr_name
+
+                return proper_children[0]
+            else:
+                # Catch empty names that we *can't* replace because
+                # we're not the only parent by just dropping that
+                # layer (this happens with ZOI clauses, for example)
+
+                new_children = []
+                for child in proper_children:
+                    if len(child) > 1 and child[0] == "":
+                        new_children.append(child[1])
+                    else:
+                        new_children.append(child)
+
+                return [node.expr_name, list(new_children)]

setup.py

@@ -15,7 +15,7 @@
 
 setuptools.setup(
     name=PACKAGE_NAME,
-    version="0.9.0",
+    version="0.10.0",
     author="Robin Lee Powell",
     author_email="rlpowell@digitalkingdom.org",
     description="A pure Python implementation of the lojban 'camxes' PEG parser.",

test.py

@@ -55,6 +55,7 @@ def process_spec(input_spec, parser, json_transformer, morph_transformer):
 
     out = morph = None
     try:
+        print("text: " + text)
         parsed = parser.parse(text)
         out = transform_to_serial(parsed, json_transformer)
         morph = transform_to_serial(parsed, morph_transformer)