yamllint version 0.4.0

For instance:

    yamllint .
    yamllint file.yml ../my-other-dir

.travis.yml

@@ -0,0 +1,18 @@
+---
+language: python
+python:
+  - 2.6
+  - 2.7
+  - 3.3
+  - 3.4
+  - 3.5
+  - nightly
+install:
+  - pip install pyyaml flake8 coveralls
+  - pip install .
+script:
+  - flake8 .
+  - yamllint $(git ls-files '*.yml')
+  - coverage run --source=yamllint setup.py test
+after_success:
+  coveralls

Makefile

@@ -1,3 +0,0 @@
-.PHONY: tests
-tests:
-	python -m unittest discover

README.md

@@ -1 +1,70 @@
 # yamllint
+
+A linter for YAML files.
+
+[![Build Status](https://travis-ci.org/adrienverge/yamllint.svg?branch=master)](https://travis-ci.org/adrienverge/yamllint)
+[![Coverage Status](https://coveralls.io/repos/adrienverge/yamllint/badge.svg?branch=master&service=github)](https://coveralls.io/github/adrienverge/yamllint?branch=master)
+
+Compatible with Python 2 & 3.
+
+## Usage
+
+```sh
+yamllint my_file.yml my_other_file.yaml ...
+```
+
+```sh
+yamllint .
+```
+
+```sh
+yamllint -c ~/myconfig file.yml
+```
+
+```sh
+# To output a format parsable (by editors like Vim, emacs, etc.)
+yamllint -f parsable file.yml
+```
+
+## Installation
+
+```sh
+pip install yamllint
+```
+
+## Configuration
+
+There is no documentation yet, so here is what you need to know: you can
+override some rules, disable them or pass them in *warning* (instead of
+*error*). Have a look at the content of `yamllint/conf/default.yml` and create
+your own configuration file.
+
+It could look like this:
+
+```yaml
+# This is my first, very own configuration file for yamllint!
+# It extends the default conf by adjusting some options.
+
+extends: default
+
+rules:
+  # 80 should be enough, but don't fail if a line is longer
+  line-length:
+    max: 80
+    level: warning
+
+  # accept both
+  #    key:
+  #      - item
+  # and
+  #    key:
+  #    - item
+  indentation:
+    indent-sequences: whatever
+
+  # don't bother me with this rule
+  comments-indentation: disable
+```
+
+Tip: if you have a `.yamllint` file in your working directory, it will be
+automatically loaded as configuration by yamllint.

bin/yamllint

@@ -15,68 +15,10 @@
 # You should have received a copy of the GNU General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
-from __future__ import print_function
-import os.path
 import sys
 
-import argparse
-
-from yamllint import APP_DESCRIPTION, APP_NAME, APP_VERSION
-from yamllint import config
-from yamllint.errors import YamlLintConfigError
-from yamllint import lint
-from yamllint import output
+from yamllint import cli
 
 
 if __name__ == '__main__':
-    parser = argparse.ArgumentParser(prog=APP_NAME,
-                                     description=APP_DESCRIPTION)
-    parser.add_argument('files', metavar='FILES', nargs='+',
-                        help='files to check')
-    parser.add_argument('-c', '--config', dest='config_file', action='store',
-                        help='path to a custom configuration')
-    parser.add_argument('-f', '--format',
-                        choices=('parsable', 'standard'), default='standard',
-                        help='format for parsing output')
-    parser.add_argument('-v', '--version', action='version',
-                        version='%s %s' % (APP_NAME, APP_VERSION))
-
-    # TODO: read from stdin when no filename?
-
-    args = parser.parse_args()
-
-    try:
-        if args.config_file is not None:
-            conf = config.parse_config_from_file(args.config_file)
-        elif os.path.isfile('.yamllint'):
-            conf = config.parse_config_from_file('.yamllint')
-        else:
-            conf = config.parse_config('extends: default')
-    except YamlLintConfigError as e:
-        print(e, file=sys.stderr)
-        sys.exit(-1)
-
-    return_code = 0
-
-    for file in args.files:
-        if args.format != 'parsable':
-            print('\033[4m%s\033[0m' % file)
-
-        try:
-            with open(file) as f:
-                for problem in lint(f, conf):
-                    if args.format == 'parsable':
-                        print(output.parsable_format(problem, file))
-                    else:
-                        print(output.standard_format(problem, file))
-
-                    if return_code == 0 and problem.level == 'error':
-                        return_code = 1
-        except EnvironmentError as e:
-            print(e)
-            return_code = -1
-
-        if args.format != 'parsable':
-            print('')
-
-    sys.exit(return_code)
+    cli.run(sys.argv[1:])

setup.py

@@ -44,7 +44,7 @@ setup(
     packages=find_packages(),
     scripts=['bin/yamllint'],
     package_data={'yamllint': ['conf/*.yml']},
-    install_requires=[
-        'pyyaml>=3'
-    ],
+    install_requires=['pyyaml'],
+    tests_require=['nose'],
+    test_suite='nose.collector',
 )

tests/rules/common.py

@@ -33,12 +33,19 @@ class RuleTestCase(unittest.TestCase):
                 'rules': conf}
         return parse_config(yaml.safe_dump(conf))
 
-    def check(self, source, conf, line=None, column=None, **kwargs):
+    def check(self, source, conf, **kwargs):
         expected_problems = []
         for key in kwargs:
             assert key.startswith('problem')
+            if len(kwargs[key]) > 2:
+                if kwargs[key][2] == 'syntax':
+                    rule_id = None
+                else:
+                    rule_id = kwargs[key][2]
+            else:
+                rule_id = self.rule_id
             expected_problems.append(
-                LintProblem(kwargs[key][0], kwargs[key][1], rule=self.rule_id))
+                LintProblem(kwargs[key][0], kwargs[key][1], rule=rule_id))
         expected_problems.sort()
 
         real_problems = list(lint(source, self.build_fake_config(conf)))

tests/rules/test_braces.py

@@ -0,0 +1,108 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+from tests.rules.common import RuleTestCase
+
+
+class ColonTestCase(RuleTestCase):
+    rule_id = 'braces'
+
+    def test_disabled(self):
+        conf = 'braces: disable'
+        self.check('---\n'
+                   'dict1: {}\n'
+                   'dict2: { }\n'
+                   'dict3: {   a: 1, b}\n'
+                   'dict4: {a: 1, b, c: 3 }\n'
+                   'dict5: {a: 1, b, c: 3 }\n'
+                   'dict6: {  a: 1, b, c: 3 }\n'
+                   'dict7: {   a: 1, b, c: 3 }\n', conf)
+
+    def test_min_spaces(self):
+        conf = 'braces: {max-spaces-inside: -1, min-spaces-inside: 0}'
+        self.check('---\n'
+                   'dict: {}\n', conf)
+
+        conf = 'braces: {max-spaces-inside: -1, min-spaces-inside: 1}'
+        self.check('---\n'
+                   'dict: {}\n', conf, problem=(2, 8))
+        self.check('---\n'
+                   'dict: { }\n', conf)
+        self.check('---\n'
+                   'dict: {a: 1, b}\n', conf,
+                   problem1=(2, 8), problem2=(2, 15))
+        self.check('---\n'
+                   'dict: { a: 1, b }\n', conf)
+        self.check('---\n'
+                   'dict: {\n'
+                   '  a: 1,\n'
+                   '  b\n'
+                   '}\n', conf)
+
+        conf = 'braces: {max-spaces-inside: -1, min-spaces-inside: 3}'
+        self.check('---\n'
+                   'dict: { a: 1, b }\n', conf,
+                   problem1=(2, 9), problem2=(2, 17))
+        self.check('---\n'
+                   'dict: {   a: 1, b   }\n', conf)
+
+    def test_max_spaces(self):
+        conf = 'braces: {max-spaces-inside: 0, min-spaces-inside: -1}'
+        self.check('---\n'
+                   'dict: {}\n', conf)
+        self.check('---\n'
+                   'dict: { }\n', conf, problem=(2, 8))
+        self.check('---\n'
+                   'dict: {a: 1, b}\n', conf)
+        self.check('---\n'
+                   'dict: { a: 1, b }\n', conf,
+                   problem1=(2, 8), problem2=(2, 16))
+        self.check('---\n'
+                   'dict: {   a: 1, b   }\n', conf,
+                   problem1=(2, 10), problem2=(2, 20))
+        self.check('---\n'
+                   'dict: {\n'
+                   '  a: 1,\n'
+                   '  b\n'
+                   '}\n', conf)
+
+        conf = 'braces: {max-spaces-inside: 3, min-spaces-inside: -1}'
+        self.check('---\n'
+                   'dict: {   a: 1, b   }\n', conf)
+        self.check('---\n'
+                   'dict: {    a: 1, b     }\n', conf,
+                   problem1=(2, 11), problem2=(2, 23))
+
+    def test_min_and_max_spaces(self):
+        conf = 'braces: {max-spaces-inside: 0, min-spaces-inside: 0}'
+        self.check('---\n'
+                   'dict: {}\n', conf)
+        self.check('---\n'
+                   'dict: { }\n', conf, problem=(2, 8))
+        self.check('---\n'
+                   'dict: {   a: 1, b}\n', conf, problem=(2, 10))
+
+        conf = 'braces: {max-spaces-inside: 1, min-spaces-inside: 1}'
+        self.check('---\n'
+                   'dict: {a: 1, b, c: 3 }\n', conf, problem=(2, 8))
+
+        conf = 'braces: {max-spaces-inside: 2, min-spaces-inside: 0}'
+        self.check('---\n'
+                   'dict: {a: 1, b, c: 3 }\n', conf)
+        self.check('---\n'
+                   'dict: {  a: 1, b, c: 3 }\n', conf)
+        self.check('---\n'
+                   'dict: {   a: 1, b, c: 3 }\n', conf, problem=(2, 10))

tests/rules/test_brackets.py

@@ -0,0 +1,107 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+from tests.rules.common import RuleTestCase
+
+
+class ColonTestCase(RuleTestCase):
+    rule_id = 'brackets'
+
+    def test_disabled(self):
+        conf = 'brackets: disable'
+        self.check('---\n'
+                   'array1: []\n'
+                   'array2: [ ]\n'
+                   'array3: [   a, b]\n'
+                   'array4: [a, b, c ]\n'
+                   'array5: [a, b, c ]\n'
+                   'array6: [  a, b, c ]\n'
+                   'array7: [   a, b, c ]\n', conf)
+
+    def test_min_spaces(self):
+        conf = 'brackets: {max-spaces-inside: -1, min-spaces-inside: 0}'
+        self.check('---\n'
+                   'array: []\n', conf)
+
+        conf = 'brackets: {max-spaces-inside: -1, min-spaces-inside: 1}'
+        self.check('---\n'
+                   'array: []\n', conf, problem=(2, 9))
+        self.check('---\n'
+                   'array: [ ]\n', conf)
+        self.check('---\n'
+                   'array: [a, b]\n', conf, problem1=(2, 9), problem2=(2, 13))
+        self.check('---\n'
+                   'array: [ a, b ]\n', conf)
+        self.check('---\n'
+                   'array: [\n'
+                   '  a,\n'
+                   '  b\n'
+                   ']\n', conf)
+
+        conf = 'brackets: {max-spaces-inside: -1, min-spaces-inside: 3}'
+        self.check('---\n'
+                   'array: [ a, b ]\n', conf,
+                   problem1=(2, 10), problem2=(2, 15))
+        self.check('---\n'
+                   'array: [   a, b   ]\n', conf)
+
+    def test_max_spaces(self):
+        conf = 'brackets: {max-spaces-inside: 0, min-spaces-inside: -1}'
+        self.check('---\n'
+                   'array: []\n', conf)
+        self.check('---\n'
+                   'array: [ ]\n', conf, problem=(2, 9))
+        self.check('---\n'
+                   'array: [a, b]\n', conf)
+        self.check('---\n'
+                   'array: [ a, b ]\n', conf,
+                   problem1=(2, 9), problem2=(2, 14))
+        self.check('---\n'
+                   'array: [   a, b   ]\n', conf,
+                   problem1=(2, 11), problem2=(2, 18))
+        self.check('---\n'
+                   'array: [\n'
+                   '  a,\n'
+                   '  b\n'
+                   ']\n', conf)
+
+        conf = 'brackets: {max-spaces-inside: 3, min-spaces-inside: -1}'
+        self.check('---\n'
+                   'array: [   a, b   ]\n', conf)
+        self.check('---\n'
+                   'array: [    a, b     ]\n', conf,
+                   problem1=(2, 12), problem2=(2, 21))
+
+    def test_min_and_max_spaces(self):
+        conf = 'brackets: {max-spaces-inside: 0, min-spaces-inside: 0}'
+        self.check('---\n'
+                   'array: []\n', conf)
+        self.check('---\n'
+                   'array: [ ]\n', conf, problem=(2, 9))
+        self.check('---\n'
+                   'array: [   a, b]\n', conf, problem=(2, 11))
+
+        conf = 'brackets: {max-spaces-inside: 1, min-spaces-inside: 1}'
+        self.check('---\n'
+                   'array: [a, b, c ]\n', conf, problem=(2, 9))
+
+        conf = 'brackets: {max-spaces-inside: 2, min-spaces-inside: 0}'
+        self.check('---\n'
+                   'array: [a, b, c ]\n', conf)
+        self.check('---\n'
+                   'array: [  a, b, c ]\n', conf)
+        self.check('---\n'
+                   'array: [   a, b, c ]\n', conf, problem=(2, 11))

tests/rules/test_colons.py

@@ -32,16 +32,16 @@ class ColonTestCase(RuleTestCase):
                    '    val\n'
                    '  property   : value\n'
                    '  prop2      : val2\n'
-                   '  propriété  : [ valeur ]\n'
+                   '  propriété  : [valeur]\n'
                    '  o:\n'
                    '    k1: [v1, v2]\n'
                    '  p:\n'
                    '    - k3: >\n'
                    '        val\n'
-                   '    - o: { k1: v1 }\n'
+                   '    - o: {k1: v1}\n'
                    '    - p: kdjf\n'
                    '    - q: val0\n'
-                   '      q2:\n'
+                   '    - q2:\n'
                    '        - val1\n'
                    '...\n', conf)
         self.check('---\n'
@@ -54,7 +54,7 @@ class ColonTestCase(RuleTestCase):
                    '    val\n'
                    '  property:     value\n'
                    '  prop2:        val2\n'
-                   '  propriété:    [ valeur ]\n'
+                   '  propriété:    [valeur]\n'
                    '  o:\n'
                    '    k1:  [v1, v2]\n', conf)
         self.check('---\n'
@@ -64,9 +64,9 @@ class ColonTestCase(RuleTestCase):
                    '        val\n'
                    '    - k3:  >\n'
                    '        val\n'
-                   '    - o: { k1: v1 }\n'
-                   '    - o:  { k1: v1 }\n'
-                   '      q2:\n'
+                   '    - o: {k1: v1}\n'
+                   '    - o:  {k1: v1}\n'
+                   '    - q2:\n'
                    '        - val1\n'
                    '...\n', conf)
         self.check('---\n'
@@ -94,7 +94,7 @@ class ColonTestCase(RuleTestCase):
                    '...\n', conf, problem=(2, 4))
         self.check('---\n'
                    '- lib :\n'
-                   '  - var\n'
+                   '    - var\n'
                    '...\n', conf, problem=(2, 6))
         self.check('---\n'
                    'a: {b: {c : d, e : f}}\n', conf,
@@ -118,7 +118,7 @@ class ColonTestCase(RuleTestCase):
                    '...\n', conf, problem=(3, 8))
 
     def test_before_with_explicit_block_mappings(self):
-        conf = 'colons: {max-spaces-before: 0, max-spaces-after: -1}'
+        conf = 'colons: {max-spaces-before: 0, max-spaces-after: 1}'
         self.check('---\n'
                    'object:\n'
                    '  ? key\n'
@@ -129,6 +129,30 @@ class ColonTestCase(RuleTestCase):
                    '  ? key\n'
                    '  : value\n'
                    '...\n', conf, problem=(2, 7))
+        self.check('---\n'
+                   '? >\n'
+                   '    multi-line\n'
+                   '    key\n'
+                   ': >\n'
+                   '    multi-line\n'
+                   '    value\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   '- ? >\n'
+                   '      multi-line\n'
+                   '      key\n'
+                   '  : >\n'
+                   '      multi-line\n'
+                   '      value\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   '- ? >\n'
+                   '      multi-line\n'
+                   '      key\n'
+                   '  :  >\n'
+                   '       multi-line\n'
+                   '       value\n'
+                   '...\n', conf, problem=(5, 5))
 
     def test_after_enabled(self):
         conf = 'colons: {max-spaces-before: -1, max-spaces-after: 1}'
@@ -152,6 +176,21 @@ class ColonTestCase(RuleTestCase):
                    'a: {b: {c:  d, e :  f}}\n', conf,
                    problem1=(2, 12), problem2=(2, 20))
 
+    def test_after_enabled_question_mark(self):
+        conf = 'colons: {max-spaces-before: -1, max-spaces-after: 1}'
+        self.check('---\n'
+                   '? key\n'
+                   ': value\n', conf)
+        self.check('---\n'
+                   '?  key\n'
+                   ': value\n', conf, problem=(2, 3))
+        self.check('---\n'
+                   '?  key\n'
+                   ':  value\n', conf, problem1=(2, 3), problem2=(3, 3))
+        self.check('---\n'
+                   '- ?  key\n'
+                   '  :  value\n', conf, problem1=(2, 5), problem2=(3, 5))
+
     def test_after_max(self):
         conf = 'colons: {max-spaces-before: -1, max-spaces-after: 3}'
         self.check('---\n'

tests/rules/test_commas.py

@@ -0,0 +1,187 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+from tests.rules.common import RuleTestCase
+
+
+class CommaTestCase(RuleTestCase):
+    rule_id = 'commas'
+
+    def test_disabled(self):
+        conf = 'commas: disable'
+        self.check('---\n'
+                   'dict: {a: b ,   c: "1 2 3",    d: e , f: [g,      h]}\n'
+                   'array: [\n'
+                   '  elem  ,\n'
+                   '  key: val ,\n'
+                   ']\n'
+                   'map: {\n'
+                   '  key1: val1 ,\n'
+                   '  key2: val2,\n'
+                   '}\n'
+                   '...\n', conf)
+
+    def test_before_enabled(self):
+        conf = 'commas: {max-spaces-before: 0, max-spaces-after: -1}'
+        self.check('---\n'
+                   'array: [1, 2,  3, 4]\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   'array: [1, 2 ,  3, 4]\n'
+                   '...\n', conf, problem=(2, 13))
+        self.check('---\n'
+                   'array: [1 , 2,  3      , 4]\n'
+                   '...\n', conf, problem1=(2, 10), problem2=(2, 23))
+        self.check('---\n'
+                   'dict: {a: b, c: "1 2 3", d: e,  f: [g, h]}\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   'dict: {a: b, c: "1 2 3" , d: e,  f: [g, h]}\n'
+                   '...\n', conf, problem=(2, 24))
+        self.check('---\n'
+                   'dict: {a: b , c: "1 2 3", d: e,  f: [g    , h]}\n'
+                   '...\n', conf, problem1=(2, 12), problem2=(2, 42))
+        self.check('---\n'
+                   'array: [\n'
+                   '  elem,\n'
+                   '  key: val,\n'
+                   ']\n', conf)
+        self.check('---\n'
+                   'array: [\n'
+                   '  elem ,\n'
+                   '  key: val,\n'
+                   ']\n', conf, problem=(3, 7))
+        self.check('---\n'
+                   'map: {\n'
+                   '  key1: val1,\n'
+                   '  key2: val2,\n'
+                   '}\n', conf)
+        self.check('---\n'
+                   'map: {\n'
+                   '  key1: val1,\n'
+                   '  key2: val2 ,\n'
+                   '}\n', conf, problem=(4, 13))
+
+    def test_before_max(self):
+        conf = 'commas: {max-spaces-before: 3, max-spaces-after: -1}'
+        self.check('---\n'
+                   'array: [1 , 2, 3   , 4]\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   'array: [1 , 2, 3    , 4]\n'
+                   '...\n', conf, problem=(2, 20))
+        self.check('---\n'
+                   'array: [\n'
+                   '  elem1   ,\n'
+                   '  elem2    ,\n'
+                   '  key: val,\n'
+                   ']\n', conf, problem=(4, 11))
+
+    def test_after_enabled(self):
+        conf = 'commas: {max-spaces-before: -1, max-spaces-after: 1}'
+        self.check('---\n'
+                   'array: [1, 2, 3, 4]\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   'array: [1, 2,  3, 4]\n'
+                   '...\n', conf, problem=(2, 15))
+        self.check('---\n'
+                   'array: [1,  2, 3,     4]\n'
+                   '...\n', conf, problem1=(2, 12), problem2=(2, 22))
+        self.check('---\n'
+                   'dict: {a: b , c: "1 2 3", d: e, f: [g, h]}\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   'dict: {a: b , c: "1 2 3",  d: e, f: [g, h]}\n'
+                   '...\n', conf, problem=(2, 27))
+        self.check('---\n'
+                   'dict: {a: b ,  c: "1 2 3", d: e, f: [g,     h]}\n'
+                   '...\n', conf, problem1=(2, 15), problem2=(2, 44))
+        self.check('---\n'
+                   'array: [\n'
+                   '  elem,\n'
+                   '  key: val,\n'
+                   ']\n', conf)
+        self.check('---\n'
+                   'array: [\n'
+                   '  elem,  key: val,\n'
+                   ']\n', conf, problem=(3, 9))
+        self.check('---\n'
+                   'map: {\n'
+                   '  key1: val1,   key2: [val2,  val3]\n'
+                   '}\n', conf, problem1=(3, 16), problem2=(3, 30))
+
+    def test_after_max(self):
+        conf = 'commas: {max-spaces-before: -1, max-spaces-after: 3}'
+        self.check('---\n'
+                   'array: [1,  2, 3,   4]\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   'array: [1,  2, 3,    4]\n'
+                   '...\n', conf, problem=(2, 21))
+        self.check('---\n'
+                   'dict: {a: b ,   c: "1 2 3",    d: e, f: [g,      h]}\n'
+                   '...\n', conf, problem1=(2, 31), problem2=(2, 49))
+
+    def test_both_before_and_after(self):
+        conf = 'commas: {max-spaces-before: 0, max-spaces-after: 1}'
+        self.check('---\n'
+                   'dict: {a: b ,   c: "1 2 3",    d: e , f: [g,      h]}\n'
+                   'array: [\n'
+                   '  elem  ,\n'
+                   '  key: val ,\n'
+                   ']\n'
+                   'map: {\n'
+                   '  key1: val1 ,\n'
+                   '  key2: val2,\n'
+                   '}\n'
+                   '...\n', conf,
+                   problem1=(2, 12), problem2=(2, 16), problem3=(2, 31),
+                   problem4=(2, 36), problem5=(2, 50), problem6=(4, 8),
+                   problem7=(5, 11), problem8=(8, 13))
+
+    def test_comma_on_new_line(self):
+        conf = 'commas: {max-spaces-before: 0, max-spaces-after: 1}'
+        self.check('---\n'
+                   'flow-seq: [1, 2, 3\n'
+                   '           , 4, 5, 6]\n'
+                   '...\n', conf, problem=(3, 11))
+        self.check('---\n'
+                   'flow-map: {a: 1, b: 2\n'
+                   '           , c: 3}\n'
+                   '...\n', conf, problem=(3, 11))
+        conf = ('commas: {max-spaces-before: 0, max-spaces-after: 1}\n'
+                'indentation: disable\n')
+        self.check('---\n'
+                   'flow-seq: [1, 2, 3\n'
+                   '         , 4, 5, 6]\n'
+                   '...\n', conf, problem=(3, 9))
+        self.check('---\n'
+                   'flow-map: {a: 1, b: 2\n'
+                   '         , c: 3}\n'
+                   '...\n', conf, problem=(3, 9))
+        self.check('---\n'
+                   '[\n'
+                   '1,\n'
+                   '2\n'
+                   ', 3\n'
+                   ']\n', conf, problem=(5, 1))
+        self.check('---\n'
+                   '{\n'
+                   'a: 1,\n'
+                   'b: 2\n'
+                   ', c: 3\n'
+                   '}\n', conf, problem=(5, 1))

tests/rules/test_comments.py

@@ -0,0 +1,149 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+from tests.rules.common import RuleTestCase
+
+
+class CommentsTestCase(RuleTestCase):
+    rule_id = 'comments'
+
+    def test_disabled(self):
+        conf = ('comments: disable\n'
+                'comments-indentation: disable\n')
+        self.check('---\n'
+                   '#comment\n'
+                   '\n'
+                   'test: #    description\n'
+                   '  - foo  # bar\n'
+                   '  - hello #world\n'
+                   '\n'
+                   '# comment 2\n'
+                   '#comment 3\n'
+                   '  #comment 3 bis\n'
+                   '  #  comment 3 ter\n'
+                   '\n'
+                   'string: "Une longue phrase." # this is French\n', conf)
+
+    def test_starting_space(self):
+        conf = ('comments:\n'
+                '  require-starting-space: yes\n'
+                '  min-spaces-from-content: -1\n'
+                'comments-indentation: disable\n')
+        self.check('---\n'
+                   '# comment\n'
+                   '\n'
+                   'test:  #     description\n'
+                   '  - foo  #   bar\n'
+                   '  - hello  # world\n'
+                   '\n'
+                   '# comment 2\n'
+                   '# comment 3\n'
+                   '  #  comment 3 bis\n'
+                   '  #  comment 3 ter\n', conf)
+        self.check('---\n'
+                   '#comment\n'
+                   '\n'
+                   'test:  #    description\n'
+                   '  - foo  #  bar\n'
+                   '  - hello  #world\n'
+                   '\n'
+                   '# comment 2\n'
+                   '#comment 3\n'
+                   '  #comment 3 bis\n'
+                   '  #  comment 3 ter\n', conf,
+                   problem1=(2, 2), problem2=(6, 13),
+                   problem4=(9, 2), problem5=(10, 4))
+
+    def test_spaces_from_content(self):
+        conf = ('comments:\n'
+                '  require-starting-space: no\n'
+                '  min-spaces-from-content: 2\n')
+        self.check('---\n'
+                   '# comment\n'
+                   '\n'
+                   'test:  #    description\n'
+                   '  - foo  #  bar\n'
+                   '  - hello  #world\n'
+                   '\n'
+                   'string: "Une longue phrase."  # this is French\n', conf)
+        self.check('---\n'
+                   '# comment\n'
+                   '\n'
+                   'test: #    description\n'
+                   '  - foo  # bar\n'
+                   '  - hello #world\n'
+                   '\n'
+                   'string: "Une longue phrase." # this is French\n', conf,
+                   problem1=(4, 7), problem2=(6, 11), problem3=(8, 30))
+
+    def test_both(self):
+        conf = ('comments:\n'
+                '  require-starting-space: yes\n'
+                '  min-spaces-from-content: 2\n'
+                'comments-indentation: disable\n')
+        self.check('---\n'
+                   '#comment\n'
+                   '\n'
+                   'test: #    description\n'
+                   '  - foo  # bar\n'
+                   '  - hello #world\n'
+                   '\n'
+                   '# comment 2\n'
+                   '#comment 3\n'
+                   '  #comment 3 bis\n'
+                   '  #  comment 3 ter\n'
+                   '\n'
+                   'string: "Une longue phrase." # this is French\n', conf,
+                   problem1=(2, 2),
+                   problem2=(4, 7),
+                   problem3=(6, 11), problem4=(6, 12),
+                   problem5=(9, 2),
+                   problem6=(10, 4),
+                   problem7=(13, 30))
+
+    def test_empty_comment(self):
+        conf = ('comments:\n'
+                '  require-starting-space: yes\n'
+                '  min-spaces-from-content: 2\n')
+        self.check('---\n'
+                   '# This is paragraph 1.\n'
+                   '#\n'
+                   '# This is paragraph 2.\n', conf)
+        self.check('---\n'
+                   'inline: comment  #\n'
+                   'foo: bar\n', conf)
+
+    def test_first_line(self):
+        conf = ('comments:\n'
+                '  require-starting-space: yes\n'
+                '  min-spaces-from-content: 2\n')
+        self.check('# comment\n', conf)
+
+    def test_multi_line_scalar(self):
+        conf = ('comments:\n'
+                '  require-starting-space: yes\n'
+                '  min-spaces-from-content: 2\n'
+                'trailing-spaces: disable\n')
+        self.check('---\n'
+                   'string: >\n'
+                   '  this is plain text\n'
+                   '\n'
+                   '# comment\n', conf)
+        self.check('---\n'
+                   '- string: >\n'
+                   '    this is plain text\n'
+                   '  \n'
+                   '  # comment\n', conf)

tests/rules/test_comments_indentation.py

@@ -0,0 +1,145 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+from tests.rules.common import RuleTestCase
+
+
+class CommentsIndentationTestCase(RuleTestCase):
+    rule_id = 'comments-indentation'
+
+    def test_disable(self):
+        conf = 'comments-indentation: disable'
+        self.check('---\n'
+                   ' # line 1\n'
+                   '# line 2\n'
+                   '  # line 3\n'
+                   '  # line 4\n'
+                   '\n'
+                   'obj:\n'
+                   ' # these\n'
+                   '   # are\n'
+                   '  # [good]\n'
+                   '# bad\n'
+                   '      # comments\n'
+                   '  a: b\n'
+                   '\n'
+                   'obj1:\n'
+                   '  a: 1\n'
+                   '  # comments\n'
+                   '\n'
+                   'obj2:\n'
+                   '  b: 2\n'
+                   '\n'
+                   '# empty\n'
+                   '#\n'
+                   '# comment\n'
+                   '...\n', conf)
+
+    def test_enabled(self):
+        conf = 'comments-indentation: {}'
+        self.check('---\n'
+                   '# line 1\n'
+                   '# line 2\n', conf)
+        self.check('---\n'
+                   ' # line 1\n'
+                   '# line 2\n', conf, problem=(2, 2))
+        self.check('---\n'
+                   '  # line 1\n'
+                   '  # line 2\n', conf, problem1=(2, 3), problem2=(3, 3))
+        self.check('---\n'
+                   'obj:\n'
+                   '  # normal\n'
+                   '  a: b\n', conf)
+        self.check('---\n'
+                   'obj:\n'
+                   ' # bad\n'
+                   '  a: b\n', conf, problem=(3, 2))
+        self.check('---\n'
+                   'obj:\n'
+                   '# bad\n'
+                   '  a: b\n', conf, problem=(3, 1))
+        self.check('---\n'
+                   'obj:\n'
+                   '   # bad\n'
+                   '  a: b\n', conf, problem=(3, 4))
+        self.check('---\n'
+                   'obj:\n'
+                   ' # these\n'
+                   '   # are\n'
+                   '  # [good]\n'
+                   '# bad\n'
+                   '      # comments\n'
+                   '  a: b\n', conf,
+                   problem1=(3, 2), problem2=(4, 4),
+                   problem3=(6, 1), problem4=(7, 7))
+        self.check('---\n'
+                   'obj1:\n'
+                   '  a: 1\n'
+                   '  # the following line is disabled\n'
+                   '  # b: 2\n', conf)
+        self.check('---\n'
+                   'obj1:\n'
+                   '  a: 1\n'
+                   '  # b: 2\n'
+                   '\n'
+                   'obj2:\n'
+                   '  b: 2\n', conf)
+        self.check('---\n'
+                   'obj1:\n'
+                   '  a: 1\n'
+                   '  # b: 2\n'
+                   '# this object is useless\n'
+                   'obj2: no\n', conf)
+        self.check('---\n'
+                   'obj1:\n'
+                   '  a: 1\n'
+                   '# this object is useless\n'
+                   '  # b: 2\n'
+                   'obj2: no\n', conf, problem=(5, 3))
+        self.check('---\n'
+                   'obj1:\n'
+                   '  a: 1\n'
+                   '  # comments\n'
+                   '  b: 2\n', conf)
+        self.check('---\n'
+                   'my list for today:\n'
+                   '  - todo 1\n'
+                   '  - todo 2\n'
+                   '  # commented for now\n'
+                   '  # - todo 3\n'
+                   '...\n', conf)
+
+    def test_first_line(self):
+        conf = 'comments-indentation: {}'
+        self.check('# comment\n', conf)
+        self.check('  # comment\n', conf, problem=(1, 3))
+
+    def test_no_newline_at_end(self):
+        conf = ('comments-indentation: {}\n'
+                'new-line-at-end-of-file: disable\n')
+        self.check('# comment', conf)
+        self.check('  # comment', conf, problem=(1, 3))
+
+    def test_empty_comment(self):
+        conf = 'comments-indentation: {}'
+        self.check('---\n'
+                   '# hey\n'
+                   '# normal\n'
+                   '#\n', conf)
+        self.check('---\n'
+                   '# hey\n'
+                   '# normal\n'
+                   ' #\n', conf, problem=(4, 2))

tests/rules/test_common.py

@@ -0,0 +1,96 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import unittest
+
+import yaml
+
+from yamllint.rules.common import (Comment, get_line_indent,
+                                   get_comments_between_tokens)
+
+
+class CommonTestCase(unittest.TestCase):
+    def test_get_line_indent(self):
+        tokens = list(yaml.scan('a: 1\n'
+                                'b:\n'
+                                '  - c: [2, 3, {d: 4}]\n'))
+
+        self.assertEqual(tokens[3].value, 'a')
+        self.assertEqual(tokens[5].value, '1')
+        self.assertEqual(tokens[7].value, 'b')
+        self.assertEqual(tokens[13].value, 'c')
+        self.assertEqual(tokens[16].value, '2')
+        self.assertEqual(tokens[18].value, '3')
+        self.assertEqual(tokens[22].value, 'd')
+        self.assertEqual(tokens[24].value, '4')
+
+        for i in (3, 5):
+            self.assertEqual(get_line_indent(tokens[i]), 0)
+        for i in (7,):
+            self.assertEqual(get_line_indent(tokens[i]), 0)
+        for i in (13, 16, 18, 22, 24):
+            self.assertEqual(get_line_indent(tokens[i]), 2)
+
+    def check_comments(self, buffer, *expected):
+        yaml_loader = yaml.BaseLoader(buffer)
+
+        comments = []
+
+        next = yaml_loader.peek_token()
+        while next is not None:
+            curr = yaml_loader.get_token()
+            next = yaml_loader.peek_token()
+            for comment in get_comments_between_tokens(curr, next):
+                comments.append(comment)
+
+        self.assertEqual(comments, list(expected))
+
+    def test_get_comments_between_tokens(self):
+        self.check_comments('# comment\n',
+                            Comment(1, 1, '# comment', 0))
+        self.check_comments('---\n'
+                            '# comment\n'
+                            '...\n',
+                            Comment(2, 1, '# comment', 0))
+        self.check_comments('---\n'
+                            '# no newline char',
+                            Comment(2, 1, '# no newline char', 0))
+        self.check_comments('# just comment',
+                            Comment(1, 1, '# just comment', 0))
+        self.check_comments('\n'
+                            '   # indented comment\n',
+                            Comment(2, 4, '# indented comment', 0))
+        self.check_comments('\n'
+                            '# trailing spaces    \n',
+                            Comment(2, 1, '# trailing spaces    ', 0))
+        self.check_comments('# comment one\n'
+                            '\n'
+                            'key: val  # key=val\n'
+                            '\n'
+                            '# this is\n'
+                            '# a block     \n'
+                            '# comment\n'
+                            '\n'
+                            'other:\n'
+                            '  - foo  # equals\n'
+                            '         # bar\n',
+                            Comment(1, 1, '# comment one', 0),
+                            Comment(3, 11, '# key=val', 0),
+                            Comment(5, 1, '# this is', 0),
+                            Comment(6, 1, '# a block     ', 0),
+                            Comment(7, 1, '# comment', 0),
+                            Comment(10, 10, '# equals', 0),
+                            Comment(11, 10, '# bar', 0))

tests/rules/test_document_end.py

@@ -64,12 +64,6 @@ class DocumentEndTestCase(RuleTestCase):
                    '---\n'
                    'third: document\n'
                    '...\n', conf)
-        self.check('first: document\n'
-                   '...\n'
-                   'second: document\n'
-                   '...\n'
-                   'third: document\n'
-                   '...\n', conf)
         self.check('---\n'
                    'first: document\n'
                    '...\n'

tests/rules/test_document_start.py

@@ -82,9 +82,23 @@ class DocumentStartTestCase(RuleTestCase):
                    '...\n'
                    'second: document\n'
                    '---\n'
-                   'third: document\n', conf, problem=(4, 1))
+                   'third: document\n', conf, problem=(4, 1, 'syntax'))
 
     def test_directives(self):
-        # TODO
-        # %YAML 1.2
-        pass
+        conf = 'document-start: {present: yes}'
+        self.check('%YAML 1.2\n'
+                   '---\n'
+                   'doc: ument\n'
+                   '...\n', conf)
+        self.check('%YAML 1.2\n'
+                   '%TAG ! tag:clarkevans.com,2002:\n'
+                   '---\n'
+                   'doc: ument\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   'doc: 1\n'
+                   '...\n'
+                   '%YAML 1.2\n'
+                   '---\n'
+                   'doc: 2\n'
+                   '...\n', conf)

tests/rules/test_hyphens.py

@@ -36,12 +36,12 @@ class HyphenTestCase(RuleTestCase):
                    '- elem2\n', conf)
         self.check('---\n'
                    'object:\n'
-                   '- elem1\n'
-                   '-  elem2\n', conf)
+                   '  - elem1\n'
+                   '  -  elem2\n', conf)
         self.check('---\n'
                    'object:\n'
-                   '-  elem1\n'
-                   '-  elem2\n', conf)
+                   '  -  elem1\n'
+                   '  -  elem2\n', conf)
         self.check('---\n'
                    'object:\n'
                    '  subobject:\n'
@@ -69,12 +69,12 @@ class HyphenTestCase(RuleTestCase):
                    '- elem2\n', conf, problem=(2, 3))
         self.check('---\n'
                    'object:\n'
-                   '- elem1\n'
-                   '-  elem2\n', conf, problem=(4, 3))
+                   '  - elem1\n'
+                   '  -  elem2\n', conf, problem=(4, 5))
         self.check('---\n'
                    'object:\n'
-                   '-  elem1\n'
-                   '-  elem2\n', conf, problem1=(3, 3), problem2=(4, 3))
+                   '  -  elem1\n'
+                   '  -  elem2\n', conf, problem1=(3, 5), problem2=(4, 5))
         self.check('---\n'
                    'object:\n'
                    '  subobject:\n'

tests/rules/test_indentation.py

@@ -48,37 +48,205 @@ class IndentationTestCase(RuleTestCase):
                    '...\n', conf)
 
     def test_one_space(self):
-        conf = 'indentation: {spaces: 1}'
+        conf = 'indentation: {spaces: 1, indent-sequences: no}'
+        self.check('---\n'
+                   'object:\n'
+                   ' k1:\n'
+                   ' - a\n'
+                   ' - b\n'
+                   ' k2: v2\n'
+                   ' k3:\n'
+                   ' - name: Unix\n'
+                   '   date: 1969\n'
+                   ' - name: Linux\n'
+                   '   date: 1991\n'
+                   '...\n', conf)
+        conf = 'indentation: {spaces: 1, indent-sequences: yes}'
         self.check('---\n'
                    'object:\n'
                    ' k1:\n'
                    '  - a\n'
                    '  - b\n'
                    ' k2: v2\n'
+                   ' k3:\n'
+                   '  - name: Unix\n'
+                   '    date: 1969\n'
+                   '  - name: Linux\n'
+                   '    date: 1991\n'
                    '...\n', conf)
 
     def test_two_spaces(self):
-        conf = 'indentation: {spaces: 2}'
+        conf = 'indentation: {spaces: 2, indent-sequences: no}'
+        self.check('---\n'
+                   'object:\n'
+                   '  k1:\n'
+                   '  - a\n'
+                   '  - b\n'
+                   '  k2: v2\n'
+                   '  k3:\n'
+                   '  - name: Unix\n'
+                   '    date: 1969\n'
+                   '  - name: Linux\n'
+                   '    date: 1991\n'
+                   '...\n', conf)
+        conf = 'indentation: {spaces: 2, indent-sequences: yes}'
         self.check('---\n'
                    'object:\n'
                    '  k1:\n'
                    '    - a\n'
                    '    - b\n'
                    '  k2: v2\n'
+                   '  k3:\n'
+                   '    - name: Unix\n'
+                   '      date: 1969\n'
+                   '    - name: Linux\n'
+                   '      date: 1991\n'
                    '...\n', conf)
 
     def test_three_spaces(self):
-        conf = 'indentation: {spaces: 3}'
+        conf = 'indentation: {spaces: 3, indent-sequences: no}'
+        self.check('---\n'
+                   'object:\n'
+                   '   k1:\n'
+                   '   - a\n'
+                   '   - b\n'
+                   '   k2: v2\n'
+                   '   k3:\n'
+                   '   - name: Unix\n'
+                   '     date: 1969\n'
+                   '   - name: Linux\n'
+                   '     date: 1991\n'
+                   '...\n', conf)
+        conf = 'indentation: {spaces: 3, indent-sequences: yes}'
         self.check('---\n'
                    'object:\n'
                    '   k1:\n'
                    '      - a\n'
                    '      - b\n'
                    '   k2: v2\n'
+                   '   k3:\n'
+                   '      - name: Unix\n'
+                   '        date: 1969\n'
+                   '      - name: Linux\n'
+                   '        date: 1991\n'
                    '...\n', conf)
 
-    def test_under_indented(self):
+    def test_indent_sequences_whatever(self):
+        conf = 'indentation: {spaces: 4, indent-sequences: whatever}'
+        self.check('---\n'
+                   'list one:\n'
+                   '- 1\n'
+                   '- 2\n'
+                   '- 3\n'
+                   'list two:\n'
+                   '    - a\n'
+                   '    - b\n'
+                   '    - c\n', conf)
+        self.check('---\n'
+                   'list one:\n'
+                   '  - 1\n'
+                   '  - 2\n'
+                   '  - 3\n'
+                   'list two:\n'
+                   '    - a\n'
+                   '    - b\n'
+                   '    - c\n', conf, problem=(3, 3))
+        self.check('---\n'
+                   'list one:\n'
+                   '- 1\n'
+                   '- 2\n'
+                   '- 3\n'
+                   'list two:\n'
+                   '  - a\n'
+                   '  - b\n'
+                   '  - c\n', conf, problem=(7, 3))
+        self.check('---\n'
+                   'list:\n'
+                   '    - 1\n'
+                   '    - 2\n'
+                   '    - 3\n'
+                   '- a\n'
+                   '- b\n'
+                   '- c\n', conf, problem=(6, 1, 'syntax'))
+
+    def test_flow_mappings(self):
         conf = 'indentation: {spaces: 2}'
+        self.check('---\n'
+                   'a: {x: 1,\n'
+                   '    y,\n'
+                   '    z: 1}\n', conf)
+        self.check('---\n'
+                   'a: {x: 1,\n'
+                   '   y,\n'
+                   '    z: 1}\n', conf, problem=(3, 4))
+        self.check('---\n'
+                   'a: {x: 1,\n'
+                   '     y,\n'
+                   '    z: 1}\n', conf, problem=(3, 6))
+        self.check('---\n'
+                   'a: {x: 1,\n'
+                   '  y, z: 1\n'
+                   '}\n', conf, problem=(3, 3))
+        self.check('---\n'
+                   'a: {\n'
+                   '  x: 1,\n'
+                   '  y, z: 1\n'
+                   '}\n', conf)
+        self.check('---\n'
+                   'a: {\n'
+                   '  x: 1,\n'
+                   '  y, z: 1}\n', conf)
+        self.check('---\n'
+                   'a: {\n'
+                   '   x: 1,\n'
+                   '  y, z: 1\n'
+                   '}\n', conf, problem=(3, 4))
+        self.check('---\n'
+                   'a: {\n'
+                   '  x: 1,\n'
+                   '  y, z: 1\n'
+                   '  }\n', conf, problem=(5, 3))
+
+    def test_flow_sequences(self):
+        conf = 'indentation: {spaces: 2}'
+        self.check('---\n'
+                   'a: [x,\n'
+                   '    y,\n'
+                   '    z]\n', conf)
+        self.check('---\n'
+                   'a: [x,\n'
+                   '   y,\n'
+                   '    z]\n', conf, problem=(3, 4))
+        self.check('---\n'
+                   'a: [x,\n'
+                   '     y,\n'
+                   '    z]\n', conf, problem=(3, 6))
+        self.check('---\n'
+                   'a: [x,\n'
+                   '  y, z\n'
+                   ']\n', conf, problem=(3, 3))
+        self.check('---\n'
+                   'a: [\n'
+                   '  x,\n'
+                   '  y, z\n'
+                   ']\n', conf)
+        self.check('---\n'
+                   'a: [\n'
+                   '  x,\n'
+                   '  y, z]\n', conf)
+        self.check('---\n'
+                   'a: [\n'
+                   '   x,\n'
+                   '  y, z\n'
+                   ']\n', conf, problem=(3, 4))
+        self.check('---\n'
+                   'a: [\n'
+                   '  x,\n'
+                   '  y, z\n'
+                   '  ]\n', conf, problem=(5, 3))
+
+    def test_under_indented(self):
+        conf = 'indentation: {spaces: 2, indent-sequences: yes}'
         self.check('---\n'
                    'object:\n'
                    ' val: 1\n'
@@ -88,7 +256,13 @@ class IndentationTestCase(RuleTestCase):
                    '  k1:\n'
                    '   - a\n'
                    '...\n', conf, problem=(4, 4))
-        conf = 'indentation: {spaces: 4}'
+        self.check('---\n'
+                   'object:\n'
+                   '  k3:\n'
+                   '    - name: Unix\n'
+                   '     date: 1969\n'
+                   '...\n', conf, problem=(5, 6, 'syntax'))
+        conf = 'indentation: {spaces: 4, indent-sequences: yes}'
         self.check('---\n'
                    'object:\n'
                    '   val: 1\n'
@@ -98,9 +272,15 @@ class IndentationTestCase(RuleTestCase):
                    '- el2:\n'
                    '   - subel\n'
                    '...\n', conf, problem=(4, 4))
+        self.check('---\n'
+                   'object:\n'
+                   '    k3:\n'
+                   '        - name: Linux\n'
+                   '         date: 1991\n'
+                   '...\n', conf, problem=(5, 10, 'syntax'))
 
     def test_over_indented(self):
-        conf = 'indentation: {spaces: 2}'
+        conf = 'indentation: {spaces: 2, indent-sequences: yes}'
         self.check('---\n'
                    'object:\n'
                    '   val: 1\n'
@@ -110,7 +290,13 @@ class IndentationTestCase(RuleTestCase):
                    '  k1:\n'
                    '     - a\n'
                    '...\n', conf, problem=(4, 6))
-        conf = 'indentation: {spaces: 4}'
+        self.check('---\n'
+                   'object:\n'
+                   '  k3:\n'
+                   '    - name: Unix\n'
+                   '       date: 1969\n'
+                   '...\n', conf, problem=(5, 12, 'syntax'))
+        conf = 'indentation: {spaces: 4, indent-sequences: yes}'
         self.check('---\n'
                    'object:\n'
                    '     val: 1\n'
@@ -132,42 +318,99 @@ class IndentationTestCase(RuleTestCase):
         self.check('---\n'
                    '  - el1\n'
                    '  - el2:\n'
-                   '      - subel\n'
-                   '...\n', conf, problem1=(2, 3), problem2=(4, 7))
+                   '        - subel\n'
+                   '...\n', conf,
+                   problem=(2, 3))
+        self.check('---\n'
+                   'object:\n'
+                   '    k3:\n'
+                   '        - name: Linux\n'
+                   '           date: 1991\n'
+                   '...\n', conf, problem=(5, 16, 'syntax'))
+        conf = 'indentation: {spaces: 4, indent-sequences: whatever}'
+        self.check('---\n'
+                   '  - el1\n'
+                   '  - el2:\n'
+                   '    - subel\n'
+                   '...\n', conf,
+                   problem=(2, 3))
 
     def test_multi_lines(self):
+        conf = 'indentation: {spaces: 2, indent-sequences: yes}'
         self.check('---\n'
                    'long_string: >\n'
                    '  bla bla blah\n'
                    '  blah bla bla\n'
-                   '...\n', None)
+                   '...\n', conf)
         self.check('---\n'
                    '- long_string: >\n'
                    '    bla bla blah\n'
                    '    blah bla bla\n'
-                   '...\n', None)
+                   '...\n', conf)
         self.check('---\n'
                    'obj:\n'
                    '  - long_string: >\n'
                    '      bla bla blah\n'
                    '      blah bla bla\n'
-                   '...\n', None)
+                   '...\n', conf)
+
+    def test_empty_value(self):
+        conf = 'indentation: {spaces: 2}'
+        self.check('---\n'
+                   'key1:\n'
+                   'key2: not empty\n'
+                   'key3:\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   '-\n'
+                   '- item 2\n'
+                   '-\n'
+                   '...\n', conf)
 
     def test_nested_collections(self):
+        conf = 'indentation: {spaces: 2}'
         self.check('---\n'
                    '- o:\n'
-                   '    k1: v1\n'
-                   '...\n', None)
+                   '  k1: v1\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   '- o:\n'
+                   ' k1: v1\n'
+                   '...\n', conf, problem=(3, 2, 'syntax'))
         self.check('---\n'
                    '- o:\n'
                    '   k1: v1\n'
-                   '...\n', None, problem=(3, 4))
+                   '...\n', conf, problem=(3, 4))
+        conf = 'indentation: {spaces: 4}'
+        self.check('---\n'
+                   '- o:\n'
+                   '      k1: v1\n'
+                   '...\n', conf)
         self.check('---\n'
                    '- o:\n'
                    '     k1: v1\n'
-                   '...\n', None, problem=(3, 6))
+                   '...\n', conf, problem=(3, 6))
+        self.check('---\n'
+                   '- o:\n'
+                   '       k1: v1\n'
+                   '...\n', conf, problem=(3, 8))
+        self.check('---\n'
+                   '- - - - item\n'
+                   '    - elem 1\n'
+                   '    - elem 2\n'
+                   '    - - - - - very nested: a\n'
+                   '              key: value\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   ' - - - - item\n'
+                   '     - elem 1\n'
+                   '     - elem 2\n'
+                   '     - - - - - very nested: a\n'
+                   '               key: value\n'
+                   '...\n', conf, problem=(2, 2))
 
     def test_return(self):
+        conf = 'indentation: {spaces: 2}'
         self.check('---\n'
                    'a:\n'
                    '  b:\n'
@@ -176,16 +419,303 @@ class IndentationTestCase(RuleTestCase):
                    '    e:\n'
                    '      f:\n'
                    'g:\n'
-                   '...\n', None)
-        # self.check('---\n'
-        #            'a:\n'
-        #            '  b:\n'
-        #            '    c:\n'
-        #            '   d:\n'
-        #            '...\n', None, problem=(5, 5))
-        # self.check('---\n'
-        #            'a:\n'
-        #            '  b:\n'
-        #            '    c:\n'
-        #            ' d:\n'
-        #            '...\n', None, problem=(5, 2))
+                   '...\n', conf)
+        self.check('---\n'
+                   'a:\n'
+                   '  b:\n'
+                   '    c:\n'
+                   '   d:\n'
+                   '...\n', conf, problem=(5, 4, 'syntax'))
+        self.check('---\n'
+                   'a:\n'
+                   '  b:\n'
+                   '    c:\n'
+                   ' d:\n'
+                   '...\n', conf, problem=(5, 2, 'syntax'))
+
+    def test_first_line(self):
+        conf = ('indentation: {spaces: 2}\n'
+                'document-start: disable\n')
+        self.check('  a: 1\n', conf, problem=(1, 3))
+
+    def test_broken_inline_flows(self):
+        conf = 'indentation: {spaces: 2}'
+        self.check('---\n'
+                   'obj: {\n'
+                   '  a: 1,\n'
+                   '   b: 2,\n'
+                   ' c: 3\n'
+                   '}\n', conf, problem1=(4, 4), problem2=(5, 2))
+        self.check('---\n'
+                   'list: [\n'
+                   '  1,\n'
+                   '   2,\n'
+                   ' 3\n'
+                   ']\n', conf, problem1=(4, 4), problem2=(5, 2))
+
+    def test_explicit_block_mappings(self):
+        conf = 'indentation: {spaces: 4}'
+        self.check('---\n'
+                   'object:\n'
+                   '    ? key\n'
+                   '    :\n'
+                   '        value\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   'object:\n'
+                   '    ? key\n'
+                   '    :\n'
+                   '       value\n'
+                   '...\n', conf, problem=(5, 8))
+        self.check('---\n'
+                   'object:\n'
+                   '    ?\n'
+                   '        key\n'
+                   '    :\n'
+                   '        value\n'
+                   '...\n', conf)
+        self.check('---\n'
+                   'object:\n'
+                   '    ?\n'
+                   '       key\n'
+                   '    :\n'
+                   '         value\n'
+                   '...\n', conf, problem1=(4, 8), problem2=(6, 10))
+        self.check('---\n'
+                   'object:\n'
+                   '    ?\n'
+                   '         key\n'
+                   '    :\n'
+                   '       value\n'
+                   '...\n', conf, problem1=(4, 10), problem2=(6, 8))
+
+
+class ScalarIndentationTestCase(RuleTestCase):
+    rule_id = 'indentation'
+
+    def test_basics_plain(self):
+        conf = ('indentation: {spaces: 2}\n'
+                'document-start: disable\n')
+        self.check('multi\n'
+                   'line\n', conf)
+        self.check('multi\n'
+                   ' line\n', conf, problem=(2, 2))
+        self.check('- multi\n'
+                   '  line\n', conf)
+        self.check('- multi\n'
+                   '   line\n', conf, problem=(2, 4))
+        self.check('a key: multi\n'
+                   '       line\n', conf)
+        self.check('a key: multi\n'
+                   '        line\n', conf, problem=(2, 9))
+        self.check('a key:\n'
+                   '  multi\n'
+                   '  line\n', conf)
+        self.check('a key:\n'
+                   '  multi\n'
+                   '   line\n', conf, problem=(3, 4))
+
+    def test_basics_quoted(self):
+        conf = ('indentation: {spaces: 2}\n'
+                'document-start: disable\n')
+        self.check('"multi\n'
+                   ' line"\n', conf)
+        self.check('"multi\n'
+                   'line"\n', conf, problem=(2, 1))
+        self.check('"multi\n'
+                   '  line"\n', conf, problem=(2, 3))
+        self.check('- "multi\n'
+                   '   line"\n', conf)
+        self.check('- "multi\n'
+                   '  line"\n', conf, problem=(2, 3))
+        self.check('- "multi\n'
+                   '    line"\n', conf, problem=(2, 5))
+        self.check('a key: "multi\n'
+                   '        line"\n', conf)
+        self.check('a key: "multi\n'
+                   '       line"\n', conf, problem=(2, 8))
+        self.check('a key: "multi\n'
+                   '         line"\n', conf, problem=(2, 10))
+        self.check('a key:\n'
+                   '  "multi\n'
+                   '   line"\n', conf)
+        self.check('a key:\n'
+                   '  "multi\n'
+                   '  line"\n', conf, problem=(3, 3))
+        self.check('a key:\n'
+                   '  "multi\n'
+                   '    line"\n', conf, problem=(3, 5))
+
+    def test_basics_folded_style(self):
+        conf = ('indentation: {spaces: 2}\n'
+                'document-start: disable\n')
+        self.check('>\n'
+                   '  multi\n'
+                   '  line\n', conf)
+        self.check('- >\n'
+                   '    multi\n'
+                   '    line\n', conf)
+        self.check('- key: >\n'
+                   '    multi\n'
+                   '    line\n', conf)
+        self.check('- key:\n'
+                   '    >\n'
+                   '      multi\n'
+                   '      line\n', conf)
+        self.check('- ? >\n'
+                   '      multi-line\n'
+                   '      key\n'
+                   '  : >\n'
+                   '      multi-line\n'
+                   '      value\n', conf)
+        self.check('- ?\n'
+                   '    >\n'
+                   '      multi-line\n'
+                   '      key\n'
+                   '  :\n'
+                   '    >\n'
+                   '      multi-line\n'
+                   '      value\n', conf)
+
+    def test_basics_literal_style(self):
+        conf = ('indentation: {spaces: 2}\n'
+                'document-start: disable\n')
+        self.check('|\n'
+                   '  multi\n'
+                   '  line\n', conf)
+        self.check('- |\n'
+                   '    multi\n'
+                   '    line\n', conf)
+        self.check('- key: |\n'
+                   '    multi\n'
+                   '    line\n', conf)
+        self.check('- key:\n'
+                   '    |\n'
+                   '      multi\n'
+                   '      line\n', conf)
+        self.check('- ? |\n'
+                   '      multi-line\n'
+                   '      key\n'
+                   '  : |\n'
+                   '      multi-line\n'
+                   '      value\n', conf)
+        self.check('- ?\n'
+                   '    |\n'
+                   '      multi-line\n'
+                   '      key\n'
+                   '  :\n'
+                   '    |\n'
+                   '      multi-line\n'
+                   '      value\n', conf)
+
+    # The following "paragraph" examples are inspired from
+    # http://stackoverflow.com/questions/3790454/in-yaml-how-do-i-break-a-string-over-multiple-lines
+
+    def test_paragraph_plain(self):
+        conf = ('indentation: {spaces: 2}\n'
+                'document-start: disable\n')
+        self.check('- long text: very "long"\n'
+                   '             \'string\' with\n'
+                   '\n'
+                   '             paragraph gap, \\n and\n'
+                   '             spaces.\n', conf)
+        self.check('- long text: very "long"\n'
+                   '    \'string\' with\n'
+                   '\n'
+                   '    paragraph gap, \\n and\n'
+                   '    spaces.\n', conf,
+                   problem1=(2, 5), problem2=(4, 5), problem3=(5, 5))
+        self.check('- long text:\n'
+                   '    very "long"\n'
+                   '    \'string\' with\n'
+                   '\n'
+                   '    paragraph gap, \\n and\n'
+                   '    spaces.\n', conf)
+
+    def test_paragraph_double_quoted(self):
+        conf = ('indentation: {spaces: 2}\n'
+                'document-start: disable\n')
+        self.check('- long text: "very \\"long\\"\n'
+                   '              \'string\' with\n'
+                   '\n'
+                   '              paragraph gap, \\n and\n'
+                   '              spaces."\n', conf)
+        self.check('- long text: "very \\"long\\"\n'
+                   '    \'string\' with\n'
+                   '\n'
+                   '    paragraph gap, \\n and\n'
+                   '    spaces."\n', conf,
+                   problem1=(2, 5), problem2=(4, 5), problem3=(5, 5))
+        self.check('- long text: "very \\"long\\"\n'
+                   '\'string\' with\n'
+                   '\n'
+                   'paragraph gap, \\n and\n'
+                   'spaces."\n', conf,
+                   problem1=(2, 1), problem2=(4, 1), problem3=(5, 1))
+        self.check('- long text:\n'
+                   '    "very \\"long\\"\n'
+                   '     \'string\' with\n'
+                   '\n'
+                   '     paragraph gap, \\n and\n'
+                   '     spaces."\n', conf)
+
+    def test_paragraph_single_quoted(self):
+        conf = ('indentation: {spaces: 2}\n'
+                'document-start: disable\n')
+        self.check('- long text: \'very "long"\n'
+                   '              \'\'string\'\' with\n'
+                   '\n'
+                   '              paragraph gap, \\n and\n'
+                   '              spaces.\'\n', conf)
+        self.check('- long text: \'very "long"\n'
+                   '    \'\'string\'\' with\n'
+                   '\n'
+                   '    paragraph gap, \\n and\n'
+                   '    spaces.\'\n', conf,
+                   problem1=(2, 5), problem2=(4, 5), problem3=(5, 5))
+        self.check('- long text: \'very "long"\n'
+                   '\'\'string\'\' with\n'
+                   '\n'
+                   'paragraph gap, \\n and\n'
+                   'spaces.\'\n', conf,
+                   problem1=(2, 1), problem2=(4, 1), problem3=(5, 1))
+        self.check('- long text:\n'
+                   '    \'very "long"\n'
+                   '     \'\'string\'\' with\n'
+                   '\n'
+                   '     paragraph gap, \\n and\n'
+                   '     spaces.\'\n', conf)
+
+    def test_paragraph_folded(self):
+        conf = ('indentation: {spaces: 2}\n'
+                'document-start: disable\n')
+        self.check('- long text: >\n'
+                   '    very "long"\n'
+                   '    \'string\' with\n'
+                   '\n'
+                   '    paragraph gap, \\n and\n'
+                   '    spaces.\n', conf)
+        self.check('- long text: >\n'
+                   '    very "long"\n'
+                   '     \'string\' with\n'
+                   '\n'
+                   '      paragraph gap, \\n and\n'
+                   '       spaces.\n', conf,
+                   problem1=(3, 6), problem2=(5, 7), problem3=(6, 8))
+
+    def test_paragraph_literal(self):
+        conf = ('indentation: {spaces: 2}\n'
+                'document-start: disable\n')
+        self.check('- long text: |\n'
+                   '    very "long"\n'
+                   '    \'string\' with\n'
+                   '\n'
+                   '    paragraph gap, \\n and\n'
+                   '    spaces.\n', conf)
+        self.check('- long text: |\n'
+                   '    very "long"\n'
+                   '     \'string\' with\n'
+                   '\n'
+                   '      paragraph gap, \\n and\n'
+                   '       spaces.\n', conf,
+                   problem1=(3, 6), problem2=(5, 7), problem3=(6, 8))

tests/rules/test_syntax_error.py

@@ -0,0 +1,69 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+from tests.rules.common import RuleTestCase
+
+
+class YamlLintTestCase(RuleTestCase):
+    rule_id = None  # syntax error
+
+    def test_lint(self):
+        self.check('---\n'
+                   'this is not: valid: YAML\n', None, problem=(2, 19))
+        self.check('---\n'
+                   'this is: valid YAML\n'
+                   '\n'
+                   'this is an error: [\n'
+                   '\n'
+                   '...\n', None, problem=(6, 1))
+
+    def test_directives(self):
+        self.check('%YAML 1.2\n'
+                   '%TAG ! tag:clarkevans.com,2002:\n'
+                   'doc: ument\n'
+                   '...\n', None, problem=(3, 1))
+
+    def test_explicit_mapping(self):
+        self.check('---\n'
+                   '? key\n'
+                   ': - value 1\n'
+                   '  - value 2\n'
+                   '...\n', None)
+        self.check('---\n'
+                   '?\n'
+                   '  key\n'
+                   ': {a: 1}\n'
+                   '...\n', None)
+        self.check('---\n'
+                   '?\n'
+                   '  key\n'
+                   ':\n'
+                   '  val\n'
+                   '...\n', None)
+
+    def test_mapping_between_sequences(self):
+        # This is valid YAML. See http://www.yaml.org/spec/1.2/spec.html,
+        # example 2.11
+        self.check('---\n'
+                   '? - Detroit Tigers\n'
+                   '  - Chicago cubs\n'
+                   ':\n'
+                   '  - 2001-07-23\n'
+                   '\n'
+                   '? [New York Yankees,\n'
+                   '   Atlanta Braves]\n'
+                   ': [2001-07-02, 2001-08-12,\n'
+                   '   2001-08-14]\n', None)

tests/rules/test_trailing_spaces.py

@@ -33,11 +33,11 @@ class TrailingSpacesTestCase(RuleTestCase):
         self.check('', conf)
         self.check('\n', conf)
         self.check('    \n', conf, problem=(1, 1))
-        self.check('\t\t\t\n', conf, problem=(1, 1))
+        self.check('\t\t\t\n', conf, problem=(1, 1, 'syntax'))
         self.check('---\n'
                    'some: text \n', conf, problem=(2, 11))
         self.check('---\n'
-                   'some: text\t\n', conf, problem=(2, 11))
+                   'some: text\t\n', conf, problem=(2, 11, 'syntax'))
 
     def test_with_dos_new_lines(self):
         conf = ('trailing-spaces: {}\n'

tests/test_config.py

@@ -0,0 +1,69 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import os
+import unittest
+
+from yamllint import config
+
+
+class ConfigTestCase(unittest.TestCase):
+    def setUp(self):
+        self.base = config.parse_config_from_file(os.path.join(
+            os.path.dirname(os.path.dirname(os.path.realpath(__file__))),
+            'yamllint', 'conf', 'default.yml'))
+
+    def test_extend_config_disable_rule(self):
+        new = config.parse_config('extends: default\n'
+                                  'rules:\n'
+                                  '  trailing-spaces: disable\n')
+
+        base = self.base.copy()
+        del base['trailing-spaces']
+
+        self.assertEqual(sorted(new.keys()), sorted(base.keys()))
+        for rule in new:
+            self.assertEqual(new[rule], base[rule])
+
+    def test_extend_config_override_whole_rule(self):
+        new = config.parse_config('extends: default\n'
+                                  'rules:\n'
+                                  '  empty-lines:\n'
+                                  '    max: 42\n'
+                                  '    max-start: 43\n'
+                                  '    max-end: 44\n')
+
+        base = self.base.copy()
+        base['empty-lines']['max'] = 42
+        base['empty-lines']['max-start'] = 43
+        base['empty-lines']['max-end'] = 44
+
+        self.assertEqual(sorted(new.keys()), sorted(base.keys()))
+        for rule in new:
+            self.assertEqual(new[rule], base[rule])
+
+    def test_extend_config_override_rule_partly(self):
+        new = config.parse_config('extends: default\n'
+                                  'rules:\n'
+                                  '  empty-lines:\n'
+                                  '    max-start: 42\n')
+
+        base = self.base.copy()
+        base['empty-lines']['max-start'] = 42
+
+        self.assertEqual(sorted(new.keys()), sorted(base.keys()))
+        for rule in new:
+            self.assertEqual(new[rule], base[rule])

tests/test_parser.py

@@ -65,8 +65,8 @@ class ParserTestCase(unittest.TestCase):
         e = list(token_generator(''))
         self.assertEqual(len(e), 2)
         self.assertEqual(e[0].prev, None)
-        self.assertIsInstance(e[0].curr, yaml.Token)
-        self.assertIsInstance(e[0].next, yaml.Token)
+        self.assertTrue(isinstance(e[0].curr, yaml.Token))
+        self.assertTrue(isinstance(e[0].next, yaml.Token))
         self.assertEqual(e[1].prev, e[0].curr)
         self.assertEqual(e[1].curr, e[0].next)
         self.assertEqual(e[1].next, None)
@@ -74,20 +74,20 @@ class ParserTestCase(unittest.TestCase):
         e = list(token_generator('---\n'
                                  'k: v\n'))
         self.assertEqual(len(e), 9)
-        self.assertIsInstance(e[3].curr, yaml.KeyToken)
-        self.assertIsInstance(e[5].curr, yaml.ValueToken)
+        self.assertTrue(isinstance(e[3].curr, yaml.KeyToken))
+        self.assertTrue(isinstance(e[5].curr, yaml.ValueToken))
 
     def test_token_or_line_generator(self):
         e = list(token_or_line_generator('---\n'
                                          'k: v\n'))
         self.assertEqual(len(e), 12)
-        self.assertIsInstance(e[0], Token)
-        self.assertIsInstance(e[0].curr, yaml.StreamStartToken)
-        self.assertIsInstance(e[1], Token)
-        self.assertIsInstance(e[1].curr, yaml.DocumentStartToken)
-        self.assertIsInstance(e[2], Line)
-        self.assertIsInstance(e[3].curr, yaml.BlockMappingStartToken)
-        self.assertIsInstance(e[4].curr, yaml.KeyToken)
-        self.assertIsInstance(e[6].curr, yaml.ValueToken)
-        self.assertIsInstance(e[8], Line)
-        self.assertIsInstance(e[11], Line)
+        self.assertTrue(isinstance(e[0], Token))
+        self.assertTrue(isinstance(e[0].curr, yaml.StreamStartToken))
+        self.assertTrue(isinstance(e[1], Token))
+        self.assertTrue(isinstance(e[1].curr, yaml.DocumentStartToken))
+        self.assertTrue(isinstance(e[2], Line))
+        self.assertTrue(isinstance(e[3].curr, yaml.BlockMappingStartToken))
+        self.assertTrue(isinstance(e[4].curr, yaml.KeyToken))
+        self.assertTrue(isinstance(e[6].curr, yaml.ValueToken))
+        self.assertTrue(isinstance(e[8], Line))
+        self.assertTrue(isinstance(e[11], Line))

yamllint/__init__.py

@@ -14,13 +14,16 @@
 # You should have received a copy of the GNU General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
+import yaml
+
 from yamllint import config
+from yamllint.errors import LintProblem
 from yamllint import parser
 
 
 APP_NAME = 'yamllint'
-APP_VERSION = '0.1.0'
-APP_DESCRIPTION = 'Lint YAML files.'
+APP_VERSION = '0.4.0'
+APP_DESCRIPTION = 'A linter for YAML files.'
 
 __author__ = 'Adrien Vergé'
 __copyright__ = 'Copyright 2016, Adrien Vergé'
@@ -28,28 +31,24 @@ __license__ = 'GPLv3'
 __version__ = APP_VERSION
 
 
-def _lint(buffer, conf):
+def get_costemic_problems(buffer, conf):
     rules = config.get_enabled_rules(conf)
 
     # Split token rules from line rules
     token_rules = [r for r in rules if r.TYPE == 'token']
     line_rules = [r for r in rules if r.TYPE == 'line']
 
-    # If the document contains a syntax error, save it and yield it at the
-    # right line
-    syntax_error = parser.get_syntax_error(buffer)
+    context = {}
+    for rule in token_rules:
+        context[rule.ID] = {}
 
     for elem in parser.token_or_line_generator(buffer):
-        if syntax_error and syntax_error.line <= elem.line_no:
-            syntax_error.level = 'error'
-            yield syntax_error
-            syntax_error = None
-
         if isinstance(elem, parser.Token):
             for rule in token_rules:
                 rule_conf = conf[rule.ID]
                 for problem in rule.check(rule_conf,
-                                          elem.curr, elem.prev, elem.next):
+                                          elem.curr, elem.prev, elem.next,
+                                          context[rule.ID]):
                     problem.rule = rule.ID
                     problem.level = rule_conf['level']
                     yield problem
@@ -62,6 +61,44 @@ def _lint(buffer, conf):
                     yield problem
 
 
+def get_syntax_error(buffer):
+    try:
+        list(yaml.parse(buffer, Loader=yaml.BaseLoader))
+    except yaml.error.MarkedYAMLError as e:
+        problem = LintProblem(e.problem_mark.line + 1,
+                              e.problem_mark.column + 1,
+                              'syntax error: ' + e.problem)
+        problem.level = 'error'
+        return problem
+
+
+def _lint(buffer, conf):
+    # If the document contains a syntax error, save it and yield it at the
+    # right line
+    syntax_error = get_syntax_error(buffer)
+
+    for problem in get_costemic_problems(buffer, conf):
+        # Insert the syntax error (if any) at the right place...
+        if (syntax_error and syntax_error.line <= problem.line and
+                syntax_error.column <= problem.column):
+            yield syntax_error
+
+            # If there is already a yamllint error at the same place, discard
+            # it as it is probably redundant (and maybe it's just a 'warning',
+            # in which case the script won't even exit with a failure status).
+            if (syntax_error.line == problem.line and
+                    syntax_error.column == problem.column):
+                syntax_error = None
+                continue
+
+            syntax_error = None
+
+        yield problem
+
+    if syntax_error:
+        yield syntax_error
+
+
 def lint(input, conf):
     """Lints a YAML source.
 

yamllint/cli.py

@@ -0,0 +1,119 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+from __future__ import print_function
+import os.path
+import sys
+
+import argparse
+
+from yamllint import APP_DESCRIPTION, APP_NAME, APP_VERSION
+from yamllint import config
+from yamllint.errors import YamlLintConfigError
+from yamllint import lint
+
+
+def find_files_recursively(items):
+    for item in items:
+        if os.path.isdir(item):
+            for root, dirnames, filenames in os.walk(item):
+                for filename in [f for f in filenames
+                                 if f.endswith(('.yml', '.yaml'))]:
+                    yield os.path.join(root, filename)
+        else:
+            yield item
+
+
+class Format(object):
+    @staticmethod
+    def parsable(problem, filename):
+        return ('%(file)s:%(line)s:%(column)s: [%(level)s] %(message)s' %
+                {'file': filename,
+                 'line': problem.line,
+                 'column': problem.column,
+                 'level': problem.level,
+                 'message': problem.message})
+
+    @staticmethod
+    def standard(problem, filename):
+        line = '  \033[2m%d:%d\033[0m' % (problem.line, problem.column)
+        line += max(20 - len(line), 0) * ' '
+        if problem.level == 'warning':
+            line += '\033[33m%s\033[0m' % problem.level
+        else:
+            line += '\033[31m%s\033[0m' % problem.level
+        line += max(38 - len(line), 0) * ' '
+        line += problem.desc
+        if problem.rule:
+            line += '  \033[2m(%s)\033[0m' % problem.rule
+        return line
+
+
+def run(argv):
+    parser = argparse.ArgumentParser(prog=APP_NAME,
+                                     description=APP_DESCRIPTION)
+    parser.add_argument('files', metavar='FILE_OR_DIR', nargs='+',
+                        help='files to check')
+    parser.add_argument('-c', '--config', dest='config_file', action='store',
+                        help='path to a custom configuration')
+    parser.add_argument('-f', '--format',
+                        choices=('parsable', 'standard'), default='standard',
+                        help='format for parsing output')
+    parser.add_argument('-v', '--version', action='version',
+                        version='%s %s' % (APP_NAME, APP_VERSION))
+
+    # TODO: read from stdin when no filename?
+
+    args = parser.parse_args(argv)
+
+    try:
+        if args.config_file is not None:
+            conf = config.parse_config_from_file(args.config_file)
+        elif os.path.isfile('.yamllint'):
+            conf = config.parse_config_from_file('.yamllint')
+        else:
+            conf = config.parse_config('extends: default')
+    except YamlLintConfigError as e:
+        print(e, file=sys.stderr)
+        sys.exit(-1)
+
+    return_code = 0
+
+    for file in find_files_recursively(args.files):
+        try:
+            first = True
+            with open(file) as f:
+                for problem in lint(f, conf):
+                    if args.format == 'parsable':
+                        print(Format.parsable(problem, file))
+                    else:
+                        if first:
+                            print('\033[4m%s\033[0m' % file)
+                            first = False
+
+                        print(Format.standard(problem, file))
+
+                    if return_code == 0 and problem.level == 'error':
+                        return_code = 1
+
+            if not first and args.format != 'parsable':
+                print('')
+        except EnvironmentError as e:
+            print(e)
+            return_code = -1
+
+    sys.exit(return_code)

yamllint/conf/default.yml

@@ -1,17 +1,24 @@
 ---
 
 rules:
-  #block-sequence-indentation:
-  #  present: yes
+  braces:
+    min-spaces-inside: 0
+    max-spaces-inside: 0
+  brackets:
+    min-spaces-inside: 0
+    max-spaces-inside: 0
   colons:
     max-spaces-before: 0
     max-spaces-after: 1
-  #comma:
-  #  max-spaces-before: 0
-  #  max-spaces-after: 1
-  #comment:
-  #  min-spaces-after: 1
-  #  min-spaces-before: 2
+  commas:
+    max-spaces-before: 0
+    max-spaces-after: 1
+  comments:
+    level: warning
+    require-starting-space: yes
+    min-spaces-from-content: 2
+  comments-indentation:
+    level: warning
   document-end: disable
   document-start:
     level: warning
@@ -24,16 +31,10 @@ rules:
     max-spaces-after: 1
   indentation:
     spaces: 2
-    #comments: yes
+    indent-sequences: yes
   line-length:
     max: 80
   new-line-at-end-of-file: {level: error}
   new-lines:
     type: unix
-  #spaces-in-brackets:  [ 1, 2 ]
-  #  min: 1
-  #  max: 1
-  #spaces-in-braces:  { df: d }
-  #  min: 1
-  #  max: 1
   trailing-spaces: {}

yamllint/config.py

@@ -46,7 +46,11 @@ def extend_config(content):
         if 'extends' in conf:
             base = parse_config_from_file(get_extended_conf(conf['extends']))
 
-            base.update(conf['rules'])
+            for rule in conf['rules']:
+                if type(conf['rules'][rule]) == dict and rule in base:
+                    base[rule].update(conf['rules'][rule])
+                else:
+                    base[rule] = conf['rules'][rule]
             conf['rules'] = base
 
         return conf
@@ -83,10 +87,16 @@ def parse_config(content):
                     raise YamlLintConfigError(
                         'invalid config: unknown option "%s" for rule "%s"' %
                         (optkey, id))
-                if type(conf['rules'][id][optkey]) != options[optkey]:
-                    raise YamlLintConfigError(
-                        'invalid config: option "%s" of "%s" should be %s' %
-                        (optkey, id, options[optkey].__name__))
+                if type(options[optkey]) == tuple:
+                    if conf['rules'][id][optkey] not in options[optkey]:
+                        raise YamlLintConfigError(
+                            ('invalid config: option "%s" of "%s" should be '
+                             'in %s') % (optkey, id, options[optkey]))
+                else:
+                    if type(conf['rules'][id][optkey]) != options[optkey]:
+                        raise YamlLintConfigError(
+                            ('invalid config: option "%s" of "%s" should be '
+                             '%s' % (optkey, id, options[optkey].__name__)))
                 rules[id][optkey] = conf['rules'][id][optkey]
         else:
             raise YamlLintConfigError(('invalid config: rule "%s": should be '

yamllint/output.py

@@ -1,38 +0,0 @@
-# -*- coding: utf-8 -*-
-# Copyright (C) 2016 Adrien Vergé
-#
-# This program is free software: you can redistribute it and/or modify
-# it under the terms of the GNU General Public License as published by
-# the Free Software Foundation, either version 3 of the License, or
-# (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-# GNU General Public License for more details.
-#
-# You should have received a copy of the GNU General Public License
-# along with this program.  If not, see <http://www.gnu.org/licenses/>.
-
-
-def parsable_format(problem, filename):
-    return ('%(file)s:%(line)s:%(column)s: [%(level)s] %(message)s' %
-            {'file': filename,
-             'line': problem.line,
-             'column': problem.column,
-             'level': problem.level,
-             'message': problem.message})
-
-
-def standard_format(problem, filename):
-    line = '  \033[2m%d:%d\033[0m' % (problem.line, problem.column)
-    line += max(20 - len(line), 0) * ' '
-    if problem.level == 'warning':
-        line += '\033[33m%s\033[0m' % problem.level
-    else:
-        line += '\033[31m%s\033[0m' % problem.level
-    line += max(38 - len(line), 0) * ' '
-    line += problem.desc
-    if problem.rule:
-        line += '  \033[2m(%s)\033[0m' % problem.rule
-    return line

yamllint/parser.py

@@ -16,8 +16,6 @@
 
 import yaml
 
-from yamllint.errors import LintProblem
-
 
 class Line(object):
     def __init__(self, line_no, buffer, start, end):
@@ -85,11 +83,3 @@ def token_or_line_generator(buffer):
         else:
             yield token
             token = next(token_gen, None)
-
-
-def get_syntax_error(buffer):
-    try:
-        yaml.safe_load_all(buffer)
-    except yaml.error.MarkedYAMLError as e:
-        return LintProblem(e.problem_mark.line + 1, e.problem_mark.column + 1,
-                           'syntax error: ' + e.problem)

yamllint/rules/__init__.py

@@ -15,7 +15,12 @@
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
 from yamllint.rules import (
+    braces,
+    brackets,
     colons,
+    commas,
+    comments,
+    comments_indentation,
     document_end,
     document_start,
     empty_lines,
@@ -28,7 +33,12 @@ from yamllint.rules import (
 )
 
 _RULES = {
+    braces.ID: braces,
+    brackets.ID: brackets,
     colons.ID: colons,
+    commas.ID: commas,
+    comments.ID: comments,
+    comments_indentation.ID: comments_indentation,
     document_end.ID: document_end,
     document_start.ID: document_start,
     empty_lines.ID: empty_lines,

yamllint/rules/braces.py

@@ -0,0 +1,47 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import yaml
+
+from yamllint.rules.common import spaces_after, spaces_before
+
+
+ID = 'braces'
+TYPE = 'token'
+CONF = {'min-spaces-inside': int,
+        'max-spaces-inside': int}
+
+
+def check(conf, token, prev, next, context):
+    if isinstance(token, yaml.FlowMappingStartToken):
+        problem = spaces_after(token, prev, next,
+                               min=conf['min-spaces-inside'],
+                               max=conf['max-spaces-inside'],
+                               min_desc='too few spaces inside braces',
+                               max_desc='too many spaces inside braces')
+        if problem is not None:
+            yield problem
+
+    elif (isinstance(token, yaml.FlowMappingEndToken) and
+            (prev is None or
+             not isinstance(prev, yaml.FlowMappingStartToken))):
+        problem = spaces_before(token, prev, next,
+                                min=conf['min-spaces-inside'],
+                                max=conf['max-spaces-inside'],
+                                min_desc='too few spaces inside braces',
+                                max_desc='too many spaces inside braces')
+        if problem is not None:
+            yield problem

yamllint/rules/brackets.py

@@ -0,0 +1,47 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import yaml
+
+from yamllint.rules.common import spaces_after, spaces_before
+
+
+ID = 'brackets'
+TYPE = 'token'
+CONF = {'min-spaces-inside': int,
+        'max-spaces-inside': int}
+
+
+def check(conf, token, prev, next, context):
+    if isinstance(token, yaml.FlowSequenceStartToken):
+        problem = spaces_after(token, prev, next,
+                               min=conf['min-spaces-inside'],
+                               max=conf['max-spaces-inside'],
+                               min_desc='too few spaces inside brackets',
+                               max_desc='too many spaces inside brackets')
+        if problem is not None:
+            yield problem
+
+    elif (isinstance(token, yaml.FlowSequenceEndToken) and
+            (prev is None or
+             not isinstance(prev, yaml.FlowSequenceStartToken))):
+        problem = spaces_before(token, prev, next,
+                                min=conf['min-spaces-inside'],
+                                max=conf['max-spaces-inside'],
+                                min_desc='too few spaces inside brackets',
+                                max_desc='too many spaces inside brackets')
+        if problem is not None:
+            yield problem

yamllint/rules/colons.py

@@ -16,7 +16,7 @@
 
 import yaml
 
-from yamllint.errors import LintProblem
+from yamllint.rules.common import spaces_after, spaces_before, is_explicit_key
 
 
 ID = 'colons'
@@ -25,22 +25,23 @@ CONF = {'max-spaces-before': int,
         'max-spaces-after': int}
 
 
-def check(conf, token, prev, next):
+def check(conf, token, prev, next, context):
     if isinstance(token, yaml.ValueToken):
-        if (prev is not None and
-                prev.end_mark.line == token.start_mark.line and
-                conf['max-spaces-before'] != - 1 and
-                (prev.end_mark.pointer + conf['max-spaces-before'] <
-                 token.start_mark.pointer)):
-            yield LintProblem(token.start_mark.line + 1,
-                              token.start_mark.column,
-                              'too many spaces before colon')
+        problem = spaces_before(token, prev, next,
+                                max=conf['max-spaces-before'],
+                                max_desc='too many spaces before colon')
+        if problem is not None:
+            yield problem
 
-        if (next is not None and
-                token.end_mark.line == next.start_mark.line and
-                conf['max-spaces-after'] != - 1 and
-                (next.start_mark.pointer - token.end_mark.pointer >
-                 conf['max-spaces-after'])):
-            yield LintProblem(token.start_mark.line + 1,
-                              next.start_mark.column,
-                              'too many spaces after colon')
+        problem = spaces_after(token, prev, next,
+                               max=conf['max-spaces-after'],
+                               max_desc='too many spaces after colon')
+        if problem is not None:
+            yield problem
+
+    if isinstance(token, yaml.KeyToken) and is_explicit_key(token):
+        problem = spaces_after(token, prev, next,
+                               max=conf['max-spaces-after'],
+                               max_desc='too many spaces after question mark')
+        if problem is not None:
+            yield problem

yamllint/rules/commas.py

@@ -0,0 +1,46 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import yaml
+
+from yamllint.errors import LintProblem
+from yamllint.rules.common import spaces_after, spaces_before
+
+
+ID = 'commas'
+TYPE = 'token'
+CONF = {'max-spaces-before': int,
+        'max-spaces-after': int}
+
+
+def check(conf, token, prev, next, context):
+    if isinstance(token, yaml.FlowEntryToken):
+        if prev is not None and prev.end_mark.line < token.start_mark.line:
+            yield LintProblem(token.start_mark.line + 1,
+                              max(1, token.start_mark.column),
+                              'too many spaces before comma')
+        else:
+            problem = spaces_before(token, prev, next,
+                                    max=conf['max-spaces-before'],
+                                    max_desc='too many spaces before comma')
+            if problem is not None:
+                yield problem
+
+        problem = spaces_after(token, prev, next,
+                               max=conf['max-spaces-after'],
+                               max_desc='too many spaces after comma')
+        if problem is not None:
+            yield problem

yamllint/rules/comments.py

@@ -0,0 +1,46 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import yaml
+
+from yamllint.errors import LintProblem
+from yamllint.rules.common import get_comments_between_tokens
+
+
+ID = 'comments'
+TYPE = 'token'
+CONF = {'require-starting-space': bool,
+        'min-spaces-from-content': int}
+
+
+def check(conf, token, prev, next, context):
+    for comment in get_comments_between_tokens(token, next):
+        if (conf['min-spaces-from-content'] != -1 and
+                not isinstance(token, yaml.StreamStartToken) and
+                comment.line == token.end_mark.line + 1):
+            # Sometimes token end marks are on the next line
+            if token.end_mark.buffer[token.end_mark.pointer - 1] != '\n':
+                if (comment.pointer - token.end_mark.pointer <
+                        conf['min-spaces-from-content']):
+                    yield LintProblem(comment.line, comment.column,
+                                      'too few spaces before comment')
+
+        if (conf['require-starting-space'] and
+                comment.pointer + 1 < len(comment.buffer) and
+                comment.buffer[comment.pointer + 1] != ' ' and
+                comment.buffer[comment.pointer + 1] != '\n'):
+            yield LintProblem(comment.line, comment.column + 1,
+                              'missing starting space in comment')

yamllint/rules/comments_indentation.py

@@ -0,0 +1,64 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import yaml
+
+from yamllint.errors import LintProblem
+from yamllint.rules.common import get_line_indent, get_comments_between_tokens
+
+
+ID = 'comments-indentation'
+TYPE = 'token'
+
+
+# Case A:
+#
+#     prev: line:
+#       # commented line
+#       current: line
+#
+# Case B:
+#
+#       prev: line
+#       # commented line 1
+#     # commented line 2
+#     current: line
+
+def check(conf, token, prev, next, context):
+    if prev is None:
+        return
+
+    curr_line_indent = token.start_mark.column
+    if isinstance(token, yaml.StreamEndToken):
+        curr_line_indent = 0
+
+    skip_first_line = True
+    if isinstance(prev, yaml.StreamStartToken):
+        skip_first_line = False
+        prev_line_indent = 0
+    else:
+        prev_line_indent = get_line_indent(prev)
+
+    if prev_line_indent <= curr_line_indent:
+        prev_line_indent = -1  # disable it
+
+    for comment in get_comments_between_tokens(
+            prev, token, skip_first_line=skip_first_line):
+        if comment.column - 1 == curr_line_indent:
+            prev_line_indent = -1  # disable it
+        elif comment.column - 1 != prev_line_indent:
+            yield LintProblem(comment.line, comment.column,
+                              'comment not indented like content')

yamllint/rules/common.py

@@ -0,0 +1,118 @@
+# -*- coding: utf-8 -*-
+# Copyright (C) 2016 Adrien Vergé
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import yaml
+
+from yamllint.errors import LintProblem
+
+
+def spaces_after(token, prev, next, min=-1, max=-1,
+                 min_desc=None, max_desc=None):
+    if next is not None and token.end_mark.line == next.start_mark.line:
+        spaces = next.start_mark.pointer - token.end_mark.pointer
+        if max != - 1 and spaces > max:
+            return LintProblem(token.start_mark.line + 1,
+                               next.start_mark.column, max_desc)
+        elif min != - 1 and spaces < min:
+            return LintProblem(token.start_mark.line + 1,
+                               next.start_mark.column + 1, min_desc)
+
+
+def spaces_before(token, prev, next, min=-1, max=-1,
+                  min_desc=None, max_desc=None):
+    if (prev is not None and prev.end_mark.line == token.start_mark.line and
+            # Discard tokens (only scalars?) that end at the start of next line
+            (prev.end_mark.pointer == 0 or
+             prev.end_mark.buffer[prev.end_mark.pointer - 1] != '\n')):
+        spaces = token.start_mark.pointer - prev.end_mark.pointer
+        if max != - 1 and spaces > max:
+            return LintProblem(token.start_mark.line + 1,
+                               token.start_mark.column, max_desc)
+        elif min != - 1 and spaces < min:
+            return LintProblem(token.start_mark.line + 1,
+                               token.start_mark.column + 1, min_desc)
+
+
+class Comment(object):
+    def __init__(self, line, column, buffer, pointer):
+        self.line = line
+        self.column = column
+        self.buffer = buffer
+        self.pointer = pointer
+
+    def __repr__(self):
+        end = self.buffer.find('\n', self.pointer)
+        if end == -1:
+            end = self.buffer.find('\0', self.pointer)
+        if end != -1:
+            return self.buffer[self.pointer:end]
+        return self.buffer[self.pointer:]
+
+    def __eq__(self, other):
+        return (self.line == other.line and
+                self.column == other.column and
+                str(self) == str(other))
+
+
+def get_line_indent(token):
+    """Finds the indent of the line the token starts in."""
+    start = token.start_mark.buffer.rfind('\n', 0,
+                                          token.start_mark.pointer) + 1
+    content = start
+    while token.start_mark.buffer[content] == ' ':
+        content += 1
+    return content - start
+
+
+def get_comments_between_tokens(token1, token2, skip_first_line=False):
+    if token2 is None:
+        buf = token1.end_mark.buffer[token1.end_mark.pointer:]
+    elif (token1.end_mark.line == token2.start_mark.line and
+          not isinstance(token1, yaml.StreamStartToken) and
+          not isinstance(token2, yaml.StreamEndToken)):
+        return
+    else:
+        buf = token1.end_mark.buffer[token1.end_mark.pointer:
+                                     token2.start_mark.pointer]
+
+    line_no = token1.end_mark.line + 1
+    column_no = token1.end_mark.column + 1
+    pointer = token1.end_mark.pointer
+
+    for line in buf.split('\n'):
+        if skip_first_line:
+            skip_first_line = False
+        else:
+            pos = line.find('#')
+            if pos != -1:
+                yield Comment(line_no, column_no + pos,
+                              token1.end_mark.buffer, pointer + pos)
+
+        pointer += len(line) + 1
+        line_no += 1
+        column_no = 1
+
+
+def is_explicit_key(token):
+    # explicit key:
+    #   ? key
+    #   : v
+    # or
+    #   ?
+    #     key
+    #   : v
+    return (token.start_mark.pointer < token.end_mark.pointer and
+            token.start_mark.buffer[token.start_mark.pointer] == '?')

yamllint/rules/document_end.py

@@ -24,7 +24,7 @@ TYPE = 'token'
 CONF = {'present': bool}
 
 
-def check(conf, token, prev, next):
+def check(conf, token, prev, next, context):
     if conf['present']:
         if (isinstance(token, yaml.StreamEndToken) and
                 not (isinstance(prev, yaml.DocumentEndToken) or

yamllint/rules/document_start.py

@@ -24,15 +24,14 @@ TYPE = 'token'
 CONF = {'present': bool}
 
 
-# TODO: Don't fail if document contains directives such as
-# %YAML 1.2
-
-def check(conf, token, prev, next):
+def check(conf, token, prev, next, context):
     if conf['present']:
-        if ((isinstance(prev, yaml.StreamStartToken) or
-             isinstance(prev, yaml.DocumentEndToken)) and
-            not (isinstance(token, yaml.DocumentStartToken) or
-                 isinstance(token, yaml.StreamEndToken))):
+        if (isinstance(prev, (yaml.StreamStartToken,
+                              yaml.DocumentEndToken,
+                              yaml.DirectiveToken)) and
+            not isinstance(token, (yaml.DocumentStartToken,
+                                   yaml.DirectiveToken,
+                                   yaml.StreamEndToken))):
             yield LintProblem(token.start_mark.line + 1, 1,
                               'missing document start "---"')
 

yamllint/rules/hyphens.py

@@ -16,7 +16,7 @@
 
 import yaml
 
-from yamllint.errors import LintProblem
+from yamllint.rules.common import spaces_after
 
 
 ID = 'hyphens'
@@ -24,11 +24,10 @@ TYPE = 'token'
 CONF = {'max-spaces-after': int}
 
 
-def check(conf, token, prev, next):
+def check(conf, token, prev, next, context):
     if isinstance(token, yaml.BlockEntryToken):
-        if token.end_mark.line == next.start_mark.line:
-            if (next.start_mark.pointer - token.end_mark.pointer >
-                    conf['max-spaces-after']):
-                yield LintProblem(token.start_mark.line + 1,
-                                  next.start_mark.column,
-                                  'too many spaces after hyphen')
+        problem = spaces_after(token, prev, next,
+                               max=conf['max-spaces-after'],
+                               max_desc='too many spaces after hyphen')
+        if problem is not None:
+            yield problem

yamllint/rules/indentation.py

@@ -17,47 +17,247 @@
 import yaml
 
 from yamllint.errors import LintProblem
+from yamllint.rules.common import is_explicit_key
 
 
 ID = 'indentation'
 TYPE = 'token'
-CONF = {'spaces': int}
+CONF = {'spaces': int,
+        'indent-sequences': (True, False, 'whatever')}
+
+ROOT, MAP, B_SEQ, F_SEQ, KEY, VAL = range(6)
 
 
-def check(conf, token, prev, next):
-    if isinstance(token, yaml.StreamEndToken):
+class Parent(object):
+    def __init__(self, type, indent):
+        self.type = type
+        self.indent = indent
+        self.explicit_key = False
+
+
+def check_scalar_indentation(conf, token, context):
+    if token.start_mark.line == token.end_mark.line:
         return
 
-    if (prev is None or isinstance(prev, yaml.StreamStartToken) or
-            isinstance(prev, yaml.DirectiveToken) or
-            isinstance(prev, yaml.DocumentStartToken)):
-        if token.start_mark.column != 0:
-            yield LintProblem(
-                token.end_mark.line + 1, token.start_mark.column + 1,
-                'found indentation of %d instead of %d' %
-                (token.start_mark.column, 0))
-        return
+    if token.plain:
+        expected_indent = token.start_mark.column
+    elif token.style in ('"', "'"):
+        expected_indent = token.start_mark.column + 1
+    elif token.style in ('>', '|'):
+        if context['stack'][-1].type == B_SEQ:
+            # - >
+            #     multi
+            #     line
+            expected_indent = token.start_mark.column + conf['spaces']
+        elif context['stack'][-1].type == KEY:
+            assert context['stack'][-1].explicit_key
+            # - ? >
+            #       multi-line
+            #       key
+            #   : >
+            #       multi-line
+            #       value
+            expected_indent = token.start_mark.column + conf['spaces']
+        elif context['stack'][-1].type == VAL:
+            if token.start_mark.line + 1 > context['cur_line']:
+                # - key:
+                #     >
+                #       multi
+                #       line
+                expected_indent = context['stack'][-1].indent + conf['spaces']
+            elif context['stack'][-2].explicit_key:
+                # - ? key
+                #   : >
+                #       multi-line
+                #       value
+                expected_indent = token.start_mark.column + conf['spaces']
+            else:
+                # - key: >
+                #     multi
+                #     line
+                expected_indent = context['stack'][-2].indent + conf['spaces']
+        else:
+            expected_indent = context['stack'][-1].indent + conf['spaces']
 
-    if token.start_mark.line > prev.end_mark.line:
-        buffer = prev.end_mark.buffer
+    line_no = token.start_mark.line + 1
 
-        start = buffer.rfind('\n', 0, prev.end_mark.pointer) + 1
-        prev_indent = 0
+    line_start = token.start_mark.pointer
+    while True:
+        line_start = token.start_mark.buffer.find(
+            '\n', line_start, token.end_mark.pointer - 1) + 1
+        if line_start == 0:
+            break
+        line_no += 1
 
-        # YAML recognizes two white space characters: space and tab.
-        # http://yaml.org/spec/1.2/spec.html#id2775170
-        while buffer[start + prev_indent] in ' \t':
-            prev_indent += 1
+        indent = 0
+        while token.start_mark.buffer[line_start + indent] == ' ':
+            indent += 1
+        if token.start_mark.buffer[line_start + indent] == '\n':
+            continue
 
-        # Discard any leading '- '
-        if (buffer[start + prev_indent:start + prev_indent + 2] == '- '):
-            prev_indent += 2
-            while buffer[start + prev_indent] in ' \t':
-                prev_indent += 1
+        if indent != expected_indent:
+            yield LintProblem(line_no, indent + 1,
+                              'wrong indentation: expected %d but found %d' %
+                              (expected_indent, indent))
 
-        if (token.start_mark.column > prev_indent and
-                token.start_mark.column != prev_indent + conf['spaces']):
-            yield LintProblem(
-                token.end_mark.line + 1, token.start_mark.column + 1,
-                'found indentation of %d instead of %d' %
-                (token.start_mark.column, prev_indent + conf['spaces']))
+
+def check(conf, token, prev, next, context):
+    if 'stack' not in context:
+        context['stack'] = [Parent(ROOT, 0)]
+        context['cur_line'] = -1
+
+    # Step 1: Lint
+
+    needs_lint = (
+        not isinstance(token, (yaml.StreamStartToken, yaml.StreamEndToken)) and
+        not isinstance(token, yaml.BlockEndToken) and
+        not (isinstance(token, yaml.ScalarToken) and token.value == '') and
+        token.start_mark.line + 1 > context['cur_line'])
+
+    if needs_lint:
+        found_indentation = token.start_mark.column
+        expected = context['stack'][-1].indent
+
+        if isinstance(token, (yaml.FlowMappingEndToken,
+                              yaml.FlowSequenceEndToken)):
+            expected = 0
+        elif (context['stack'][-1].type == KEY and
+                context['stack'][-1].explicit_key and
+                not isinstance(token, yaml.ValueToken)):
+            expected += conf['spaces']
+
+        if found_indentation != expected:
+            yield LintProblem(token.start_mark.line + 1, found_indentation + 1,
+                              'wrong indentation: expected %d but found %d' %
+                              (expected, found_indentation))
+
+    if isinstance(token, yaml.ScalarToken):
+        for problem in check_scalar_indentation(conf, token, context):
+            yield problem
+
+    # Step 2.a:
+
+    if needs_lint:
+        context['cur_line_indent'] = found_indentation
+        context['cur_line'] = token.end_mark.line + 1
+
+    # Step 2.b: Update state
+
+    if isinstance(token, yaml.BlockMappingStartToken):
+        assert isinstance(next, yaml.KeyToken)
+        if next.start_mark.line == token.start_mark.line:
+            #   - a: 1
+            #     b: 2
+            # or
+            #   - ? a
+            #     : 1
+            indent = token.start_mark.column
+        else:
+            #   - ?
+            #       a
+            #     : 1
+            indent = token.start_mark.column + conf['spaces']
+
+        context['stack'].append(Parent(MAP, indent))
+
+    elif isinstance(token, yaml.FlowMappingStartToken):
+        if next.start_mark.line == token.start_mark.line:
+            #   - {a: 1, b: 2}
+            indent = next.start_mark.column
+        else:
+            #   - {
+            #   a: 1, b: 2
+            # }
+            indent = context['cur_line_indent'] + conf['spaces']
+
+        context['stack'].append(Parent(MAP, indent))
+
+    elif isinstance(token, yaml.BlockSequenceStartToken):
+        #   - - a
+        #     - b
+        assert next.start_mark.line == token.start_mark.line
+        assert isinstance(next, yaml.BlockEntryToken)
+
+        indent = token.start_mark.column
+
+        context['stack'].append(Parent(B_SEQ, indent))
+
+    elif isinstance(token, yaml.FlowSequenceStartToken):
+        if next.start_mark.line == token.start_mark.line:
+            #   - [a, b]
+            indent = next.start_mark.column
+        else:
+            #   - [
+            #   a, b
+            # ]
+            indent = context['cur_line_indent'] + conf['spaces']
+
+        context['stack'].append(Parent(F_SEQ, indent))
+
+    elif isinstance(token, (yaml.BlockEndToken,
+                            yaml.FlowMappingEndToken,
+                            yaml.FlowSequenceEndToken)):
+        assert context['stack'][-1].type in (MAP, B_SEQ, F_SEQ)
+        context['stack'].pop()
+
+    elif isinstance(token, yaml.KeyToken):
+        indent = context['stack'][-1].indent
+
+        context['stack'].append(Parent(KEY, indent))
+
+        context['stack'][-1].explicit_key = is_explicit_key(token)
+
+    if context['stack'][-1].type == VAL:
+        context['stack'].pop()
+        assert context['stack'][-1].type == KEY
+        context['stack'].pop()
+
+    elif isinstance(token, yaml.ValueToken):
+        assert context['stack'][-1].type == KEY
+
+        # Discard empty values
+        if isinstance(next, (yaml.BlockEndToken,
+                             yaml.FlowMappingEndToken,
+                             yaml.FlowSequenceEndToken,
+                             yaml.KeyToken)):
+            context['stack'].pop()
+        else:
+            if context['stack'][-1].explicit_key:
+                #   ? k
+                #   : value
+                # or
+                #   ? k
+                #   :
+                #     value
+                indent = context['stack'][-1].indent + conf['spaces']
+            elif next.start_mark.line == prev.start_mark.line:
+                #   k: value
+                indent = next.start_mark.column
+            elif isinstance(next, (yaml.BlockSequenceStartToken,
+                                   yaml.BlockEntryToken)):
+                # NOTE: We add BlockEntryToken in the test above because
+                # sometimes BlockSequenceStartToken are not issued. Try
+                # yaml.scan()ning this:
+                #     '- lib:\n'
+                #     '  - var\n'
+                if conf['indent-sequences'] is False:
+                    indent = context['stack'][-1].indent
+                elif conf['indent-sequences'] is True:
+                    indent = context['stack'][-1].indent + conf['spaces']
+                else:  # 'whatever'
+                    if next.start_mark.column == context['stack'][-1].indent:
+                        #   key:
+                        #   - e1
+                        #   - e2
+                        indent = context['stack'][-1].indent
+                    else:
+                        #   key:
+                        #     - e1
+                        #     - e2
+                        indent = context['stack'][-1].indent + conf['spaces']
+            else:
+                #   k:
+                #     value
+                indent = context['stack'][-1].indent + conf['spaces']
+
+            context['stack'].append(Parent(VAL, indent))