Add unit tests; fix scanner bug

earley.py → earley/earley.py

@@ -125,7 +125,7 @@ def scan(S, k, state, words, do_state_print=False):
 	# If we've either gone 'over the deep end(k is out of range)', or we're in range,
 	# but the word is not a part of speech, then we can be reasonably certain that
 	# the input is not within the language
-	if words[k] in parts_of_speech and (nxt_elem_scanner == words[k] or nxt_elem_scanner == 'number'):
+	if words[k] in parts_of_speech:
 		n_lax = swap_around_dot(state, dot)
 
 		nitem = (state[0], n_lax, state[2])
@@ -204,7 +204,7 @@ def earley_parse(words, grammar, do_state_print=False):
 				if not is_finished(state, dot):
 					nxt_elem = get_next_element(state, dot)
 					if is_nonterminal(nxt_elem): # predict
-						added = predict(S, k, nxt_elem, words, grammar, do_state_print = do_state_print)
+						added = predict(S, k, nxt_elem, words, grammar, do_state_print = False)
 					else: # scanner(terminal)
 						added = scan(S, k, state, words, do_state_print = do_state_print)
 				else: # we should be finished with a particular production
@@ -245,16 +245,8 @@ def process_grammar(grammar):
 
 
 def main():
-	pgrammar = process_grammar(load_grammar(".\\grammar.txt"))
-
-	# TBD: Move to unit tests
+	pgrammar = process_grammar(load_grammar("..\\grammars\\grammar.txt"))
 	print(earley_parse("2+3*4", pgrammar)) # True
-	print(earley_parse("1", pgrammar)) # True
-	print(earley_parse("1+", pgrammar)) # False
-	print(earley_parse("1+2", pgrammar)) # True
-	print(earley_parse("1+2*", pgrammar)) # False
-	print(earley_parse("1+2*3", pgrammar)) # True
-	print(earley_parse("1+2*3/", pgrammar)) # False
 
 if __name__ == "__main__":
 	main()

tests/test_earley.py

@@ -0,0 +1,87 @@
+'''test_earley.py'''
+import sys
+import pytest
+
+sys.path.insert(0, '..')
+
+from earley.earley import load_grammar, process_grammar, earley_parse
+
+
+@pytest.fixture
+def grammar_raw():
+	gramm = load_grammar('..\\grammars\\grammar.txt')
+	return gramm
+
+@pytest.fixture
+def grammar(grammar_raw):
+	gram_dict = process_grammar(grammar_raw)
+	return gram_dict
+
+
+def test_process_grammar_pass(grammar):
+	assert(grammar=={'P': 'S', 'S': 'S + M | M', 'M': 'M * T | T', 'T': 'number'})
+
+def test_single_number_pass(grammar):
+	res = earley_parse("1", grammar)
+
+	assert(res)
+
+def test_full_example_pass(grammar):
+	res = earley_parse("2+3*4", grammar)
+
+	assert(res)
+
+def test_all_integers_pass(grammar):
+	res = earley_parse("0+1+2+3+4+5+6+7+8+9", grammar)
+
+	assert(res)
+
+def test_addition_pass_1(grammar):
+	res = earley_parse("1+2", grammar)
+
+	assert(res)
+
+def test_addition_pass_2(grammar):
+	res = earley_parse("1+2+3", grammar)
+
+	assert(res)
+
+def test_addition_mul_pass(grammar):
+	res = earley_parse("1+2*3", grammar)
+
+	assert(res)
+
+def test_fail_1(grammar):
+	res = earley_parse("1+", grammar)
+
+	assert res == False
+
+def test_fail_2(grammar):
+	res = earley_parse("1+2*", grammar)
+
+	assert res == False
+
+def test_fail_3(grammar):
+	res = earley_parse("1+2*3/", grammar)
+
+	assert res == False
+
+def test_sub_fail_1(grammar):
+	res = earley_parse("1-2", grammar)
+
+	assert res == False
+
+def test_sub_fail_2(grammar):
+	res = earley_parse("1-", grammar)
+
+	assert res == False
+
+def test_div_fail_1(grammar):
+	res = earley_parse("1/2", grammar)
+
+	assert res == False
+
+def test_div_fail_2(grammar):
+	res = earley_parse("1/", grammar)
+
+	assert res == False