Improve test coverage for tools.

* Replace generic Exception with TemplateParserException.
* Add tests to cover many of the uncovered lines in
  tools/lib/template_parser.py.
* Add an exclusion line to the naïve pattern for checking for missing
  tuples in format strings, to keep the linter happy.

tools/lib/template_parser.py

@@ -4,6 +4,10 @@ from typing import Callable, Optional
 from six.moves import range
 import re
 
+class TemplateParserException(Exception):
+    # TODO: Have callers pass in line numbers.
+    pass
+
 class TokenizerState(object):
     def __init__(self):
         # type: () -> None
@@ -128,7 +132,7 @@ def validate(fn=None, text=None, check_indent=True):
 
     def no_start_tag(token):
         # type: (Token) -> None
-        raise Exception('''
+        raise TemplateParserException('''
             No start tag
             fn: %s
             end tag:
@@ -167,7 +171,7 @@ def validate(fn=None, text=None, check_indent=True):
                 if end_col != start_col:
                     problem = 'Bad indentation.'
             if problem:
-                raise Exception('''
+                raise TemplateParserException('''
                     fn: %s
                     %s
                     start:
@@ -238,7 +242,7 @@ def get_handlebars_tag(text, i):
     while end < len(text) -1 and text[end] != '}':
         end += 1
     if text[end] != '}' or text[end+1] != '}':
-        raise Exception('Tag missing }}')
+        raise TemplateParserException('Tag missing }}')
     s = text[i:end+2]
     return s
 
@@ -248,7 +252,7 @@ def get_django_tag(text, i):
     while end < len(text) -1 and text[end] != '%':
         end += 1
     if text[end] != '%' or text[end+1] != '}':
-        raise Exception('Tag missing %}')
+        raise TemplateParserException('Tag missing %}')
     s = text[i:end+2]
     return s
 
@@ -261,7 +265,7 @@ def get_html_tag(text, i):
             quote_count += 1
         end += 1
     if end == len(text) or text[end] != '>':
-        raise Exception('Tag missing >')
+        raise TemplateParserException('Tag missing >')
     s = text[i:end+1]
     return s
 

tools/lint-all

@@ -248,6 +248,9 @@ def build_custom_checkers(by_lang):
         # rare; if we find any, they can be added to the exclude list for
         # this rule.
         {'pattern': '% [a-zA-Z0-9_.]*\)?$',
+         'exclude_line': set([
+             ('tools/tests/test_template_parser.py', '{% foo'),
+         ]),
          'description': 'Used % comprehension without a tuple'},
         # To avoid json_error(_variable) and json_error(_(variable))
         {'pattern': '\Wjson_error\(_\(?\w+\)',

tools/tests/test_template_parser.py

@@ -6,6 +6,8 @@ import unittest
 
 try:
     from tools.lib.template_parser import (
+        TemplateParserException,
+        get_tag_info,
         html_tag_tree,
         is_django_block_tag,
         tokenize,
@@ -16,6 +18,12 @@ except ImportError:
     sys.exit(1)
 
 class ParserTest(unittest.TestCase):
+    def _assert_validate_error(self, error, fn=None, text=None, check_indent=True):
+        # See https://github.com/python/typeshed/issues/372
+        # for why we have to ingore types here.
+        with self.assertRaisesRegexp(TemplateParserException, error): # type: ignore
+            validate(fn=fn, text=text, check_indent=check_indent)
+
     def test_is_django_block_tag(self):
         # type: () -> None
         self.assertTrue(is_django_block_tag('block'))
@@ -54,6 +62,71 @@ class ParserTest(unittest.TestCase):
             '''
         validate(text=my_html)
 
+    def test_validate_no_start_tag(self):
+        # type: () -> None
+        my_html = '''
+            foo</p>
+        '''
+        self._assert_validate_error('No start tag', text=my_html)
+
+    def test_validate_mismatched_tag(self):
+        # type: () -> None
+        my_html = '''
+            <b>foo</i>
+        '''
+        self._assert_validate_error('Mismatched tag.', text=my_html)
+
+    def test_validate_bad_indentation(self):
+        # type: () -> None
+        my_html = '''
+            <p>
+                foo
+                </p>
+        '''
+        self._assert_validate_error('Bad indentation.', text=my_html, check_indent=True)
+
+    def test_validate_incomplete_handlebars_tag_1(self):
+        # type: () -> None
+        my_html = '''
+            {{# foo
+        '''
+        self._assert_validate_error('Tag missing }}', text=my_html)
+
+    def test_validate_incomplete_handlebars_tag_2(self):
+        # type: () -> None
+        my_html = '''
+            {{# foo }
+        '''
+        self._assert_validate_error('Tag missing }}', text=my_html)
+
+    def test_validate_incomplete_django_tag_1(self):
+        # type: () -> None
+        my_html = '''
+            {% foo
+        '''
+        self._assert_validate_error('Tag missing %}', text=my_html)
+
+    def test_validate_incomplete_django_tag_2(self):
+        # type: () -> None
+        my_html = '''
+            {% foo %
+        '''
+        self._assert_validate_error('Tag missing %}', text=my_html)
+
+    def test_validate_incomplete_html_tag_1(self):
+        # type: () -> None
+        my_html = '''
+            <b
+        '''
+        self._assert_validate_error('Tag missing >', text=my_html)
+
+    def test_validate_incomplete_html_tag_2(self):
+        # type: () -> None
+        my_html = '''
+            <a href="
+        '''
+        self._assert_validate_error('Tag missing >', text=my_html)
+
     def test_code_blocks(self):
         # type: () -> None
 
@@ -74,10 +147,7 @@ class ParserTest(unittest.TestCase):
             <code>x =
             5</code>
             '''
-        # See https://github.com/python/typeshed/issues/372
+        self._assert_validate_error('Code tag is split across two lines.', text=my_html)
-        # for why we have to ingore types here.
-        with self.assertRaisesRegexp(Exception, 'split across two lines'): # type: ignore
-            validate(text=my_html)
 
     def test_anchor_blocks(self):
         # type: () -> None
@@ -145,6 +215,19 @@ class ParserTest(unittest.TestCase):
         self.assertEqual(token.kind, 'django_end')
         self.assertEqual(token.tag, 'if')
 
+    def test_get_tag_info(self):
+        html = '''
+            <p id="test" class="test1 test2">foo</p>
+        '''
+
+        start_tag, end_tag = tokenize(html)
+
+        start_tag_info = get_tag_info(start_tag)
+        end_tag_info = get_tag_info(end_tag)
+
+        self.assertEqual(start_tag_info.text(), 'p.test1.test2#test')
+        self.assertEqual(end_tag_info.text(), 'p')
+
     def test_html_tag_tree(self):
         # type: () -> None
         html = '''