Format flag issue

.travis.yml

@@ -4,6 +4,7 @@ python:
  - "3.4"
  - "3.5"
  - "3.6"
+ - "3.7"
  - "pypy"
  #- "pypy3"  temporarily disable until updated to Python 3.3
 addons:

README.rst

@@ -220,6 +220,23 @@ This is an example for Ruby project:
     jdk_switcher use oraclejdk8
     pip install html5validator
 
+Integration with GitLab CI
+--------------------------------
+
+There is a docker image available to be used with GitLab CI or stand alone.
+`Docker image <https://hub.docker.com/r/cyb3rjak3/html5validator>`_,
+`Docker image repo <https://github.com/Cyb3r-Jak3/html5validator-docker>`_.
+
+Example for html test `(Full) <https://gitlab.com/Cyb3r-Jak3/Portfolio-Website/blob/master/.gitlab-ci.yml>`_:
+
+.. code-block:: yaml
+
+    html_test:
+      stage: html_test
+      image: cyb3rjak3/html5validator:latest
+      script:
+        - html5validator --root public/ --also-check-css --format text
+
 
 Technical Notes
 ---------------

html5validator/tests/test_simple.py

@@ -105,10 +105,43 @@ def test_stack_size():
                             '-lll']) == 0
 
 
+def test_valid_format_flags():
+    assert subprocess.call(['html5validator',
+                            '--root={}/valid/'.format(HTML_TEST_FILES),
+                            '--format', 'text']) == 0
+    assert subprocess.call(['html5validator',
+                            '--root={}/valid/'.format(HTML_TEST_FILES),
+                            '--format', 'gnu']) == 0
+    assert subprocess.call(['html5validator',
+                            '--root={}/valid/'.format(HTML_TEST_FILES),
+                            '--format', 'json']) == 0
+    assert subprocess.call(['html5validator',
+                            '--root={}/valid/'.format(HTML_TEST_FILES),
+                            '--format', 'xml']) == 0
+
+
+def test_invalid_format_flags():
+    assert subprocess.call(['html5validator',
+                            '--root={}/invalid/'.format(HTML_TEST_FILES),
+                            '--format', 'text']) == 3
+    assert subprocess.call(['html5validator',
+                            '--root={}/invalid/'.format(HTML_TEST_FILES),
+                            '--format', 'gnu']) == 1
+    assert subprocess.call(['html5validator',
+                            '--root={}/invalid/'.format(HTML_TEST_FILES),
+                            '--format', 'json']) == 1
+    assert subprocess.call(['html5validator',
+                            '--root={}/invalid/'.format(HTML_TEST_FILES),
+                            '--format', 'xml']) == 8
+
+
 if __name__ == '__main__':
     test_valid()
     test_invalid()
     test_return_value()
     test_angularjs()
     test_multiple_ignoreres()
     test_ignore_and_ignorere()
+    test_stack_size()
+    test_valid_format_flags()
+    test_invalid_format_flags()

html5validator/validator.py

@@ -13,7 +13,20 @@
 
 LOGGER = logging.getLogger(__name__)
 
-DEFAULT_IGNORE_RE = ['Picked up _JAVA_OPTIONS:.*']
+DEFAULT_IGNORE_RE = [
+    r'\APicked up _JAVA_OPTIONS:.*',
+    r'\ADocument checking completed. No errors found.*',
+]
+
+DEFAULT_IGNORE = [
+    '{"messages":[]}'
+]
+
+DEFAULT_IGNORE_XML = [
+    '</messages>',
+    '<?xml version=\'1.0\' encoding=\'utf-8\'?>',
+    '<messages xmlns="http://n.validator.nu/messages/">'
+]
 
 
 class JavaNotFoundException(Exception):
@@ -43,6 +56,9 @@ def __init__(self,
         # add default ignore_re
         self.ignore_re += DEFAULT_IGNORE_RE
 
+        # add default ignore
+        self.ignore += DEFAULT_IGNORE
+
         # process fancy quotes in ignore
         self.ignore = [self._normalize_string(s) for s in self.ignore]
         self.ignore_re = [self._normalize_string(s) for s in self.ignore_re]
@@ -144,14 +160,23 @@ def validate(self, files):
 
         e = stderr.splitlines()
 
+        # Removes any empty items in the list
+        e = list(filter(None, e))
+
+        # Prevents removal of xml tags if there are errors
+        if self.format == "xml" and len(e) < 4:
+            self.ignore = DEFAULT_IGNORE_XML
+
+        LOGGER.debug(e)
+
         for i in self.ignore:
             e = [l for l in e if i not in l]
         for i in self.ignore_re:
             regex = re.compile(i)
             e = [l for l in e if not regex.search(l)]
 
-        if stderr:
-            LOGGER.error(stderr)
+        if e:
+            LOGGER.error(e)
         else:
             LOGGER.info('All good.')
         return len(e)