Merge pull request #4211 from safwanrahman/search

Upgrade Elasticsearch to version 6.x

Author: ericholscher

.travis.yml

@@ -4,7 +4,7 @@ python:
   - 3.6
 sudo: false
 env:
-  - ES_VERSION=1.3.9 ES_DOWNLOAD_URL=https://download.elastic.co/elasticsearch/elasticsearch/elasticsearch-${ES_VERSION}.tar.gz
+  - ES_VERSION=6.2.4 ES_DOWNLOAD_URL=https://artifacts.elastic.co/downloads/elasticsearch/elasticsearch-${ES_VERSION}.tar.gz
 matrix:
   include:
     - python: 2.7
@@ -42,3 +42,4 @@ notifications:
 branches:
   only:
   - master
+  - search_upgrade

readthedocs/projects/admin.py

@@ -15,7 +15,7 @@
 
 from .forms import FeatureForm
 from .models import (Project, ImportedFile, Feature,
-                     ProjectRelationship, EmailHook, WebHook, Domain)
+                     ProjectRelationship, EmailHook, WebHook, Domain, HTMLFile)
 from .notifications import ResourceUsageNotification
 from .tasks import remove_dir
 
@@ -206,3 +206,4 @@ def project_count(self, feature):
 admin.site.register(Feature, FeatureAdmin)
 admin.site.register(EmailHook)
 admin.site.register(WebHook)
+admin.site.register(HTMLFile)

readthedocs/projects/apps.py

@@ -9,5 +9,6 @@ class ProjectsConfig(AppConfig):
     def ready(self):
         from readthedocs.projects import tasks
         from readthedocs.worker import app
+
         app.tasks.register(tasks.SyncRepositoryTask)
         app.tasks.register(tasks.UpdateDocsTask)

readthedocs/projects/managers.py

@@ -0,0 +1,7 @@
+from django.db import models
+
+
+class HTMLFileManager(models.Manager):
+
+    def get_queryset(self):
+        return super(HTMLFileManager, self).get_queryset().filter(name__endswith='.html')

readthedocs/projects/migrations/0026_add_htmlfile_model.py

@@ -0,0 +1,54 @@
+# -*- coding: utf-8 -*-
+# Generated by Django 1.9.13 on 2018-06-18 16:45
+from __future__ import unicode_literals
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('projects', '0025_show-version-warning-existing-projects'),
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name='HTMLFile',
+            fields=[
+            ],
+            options={
+                'proxy': True,
+            },
+            bases=('projects.importedfile',),
+        ),
+        migrations.AlterField(
+            model_name='project',
+            name='comment_moderation',
+            field=models.BooleanField(default=False, verbose_name='Comment Moderation'),
+        ),
+        migrations.AlterField(
+            model_name='project',
+            name='documentation_type',
+            field=models.CharField(choices=[('auto', 'Automatically Choose'), ('sphinx', 'Sphinx Html'), ('mkdocs', 'Mkdocs (Markdown)'), ('sphinx_htmldir', 'Sphinx HtmlDir'), ('sphinx_singlehtml', 'Sphinx Single Page HTML')], default='sphinx', help_text='Type of documentation you are building. <a href="http://www.sphinx-doc.org/en/stable/builders.html#sphinx.builders.html.DirectoryHTMLBuilder">More info</a>.', max_length=20, verbose_name='Documentation type'),
+        ),
+        migrations.AlterField(
+            model_name='project',
+            name='language',
+            field=models.CharField(choices=[('aa', 'Afar'), ('ab', 'Abkhaz'), ('af', 'Afrikaans'), ('am', 'Amharic'), ('ar', 'Arabic'), ('as', 'Assamese'), ('ay', 'Aymara'), ('az', 'Azerbaijani'), ('ba', 'Bashkir'), ('be', 'Belarusian'), ('bg', 'Bulgarian'), ('bh', 'Bihari'), ('bi', 'Bislama'), ('bn', 'Bengali'), ('bo', 'Tibetan'), ('br', 'Breton'), ('ca', 'Catalan'), ('co', 'Corsican'), ('cs', 'Czech'), ('cy', 'Welsh'), ('da', 'Danish'), ('de', 'German'), ('dz', 'Dzongkha'), ('el', 'Greek'), ('en', 'English'), ('eo', 'Esperanto'), ('es', 'Spanish'), ('et', 'Estonian'), ('eu', 'Basque'), ('fa', 'Iranian'), ('fi', 'Finnish'), ('fj', 'Fijian'), ('fo', 'Faroese'), ('fr', 'French'), ('fy', 'Western Frisian'), ('ga', 'Irish'), ('gd', 'Scottish Gaelic'), ('gl', 'Galician'), ('gn', 'Guarani'), ('gu', 'Gujarati'), ('ha', 'Hausa'), ('hi', 'Hindi'), ('he', 'Hebrew'), ('hr', 'Croatian'), ('hu', 'Hungarian'), ('hy', 'Armenian'), ('ia', 'Interlingua'), ('id', 'Indonesian'), ('ie', 'Interlingue'), ('ik', 'Inupiaq'), ('is', 'Icelandic'), ('it', 'Italian'), ('iu', 'Inuktitut'), ('ja', 'Japanese'), ('jv', 'Javanese'), ('ka', 'Georgian'), ('kk', 'Kazakh'), ('kl', 'Kalaallisut'), ('km', 'Khmer'), ('kn', 'Kannada'), ('ko', 'Korean'), ('ks', 'Kashmiri'), ('ku', 'Kurdish'), ('ky', 'Kyrgyz'), ('la', 'Latin'), ('ln', 'Lingala'), ('lo', 'Lao'), ('lt', 'Lithuanian'), ('lv', 'Latvian'), ('mg', 'Malagasy'), ('mi', 'Maori'), ('mk', 'Macedonian'), ('ml', 'Malayalam'), ('mn', 'Mongolian'), ('mr', 'Marathi'), ('ms', 'Malay'), ('mt', 'Maltese'), ('my', 'Burmese'), ('na', 'Nauru'), ('ne', 'Nepali'), ('nl', 'Dutch'), ('no', 'Norwegian'), ('oc', 'Occitan'), ('om', 'Oromo'), ('or', 'Oriya'), ('pa', 'Panjabi'), ('pl', 'Polish'), ('ps', 'Pashto'), ('pt', 'Portuguese'), ('qu', 'Quechua'), ('rm', 'Romansh'), ('rn', 'Kirundi'), ('ro', 'Romanian'), ('ru', 'Russian'), ('rw', 'Kinyarwanda'), ('sa', 'Sanskrit'), ('sd', 'Sindhi'), ('sg', 'Sango'), ('si', 'Sinhala'), ('sk', 'Slovak'), ('sl', 'Slovenian'), ('sm', 'Samoan'), ('sn', 'Shona'), ('so', 'Somali'), ('sq', 'Albanian'), ('sr', 'Serbian'), ('ss', 'Swati'), ('st', 'Southern Sotho'), ('su', 'Sudanese'), ('sv', 'Swedish'), ('sw', 'Swahili'), ('ta', 'Tamil'), ('te', 'Telugu'), ('tg', 'Tajik'), ('th', 'Thai'), ('ti', 'Tigrinya'), ('tk', 'Turkmen'), ('tl', 'Tagalog'), ('tn', 'Tswana'), ('to', 'Tonga'), ('tr', 'Turkish'), ('ts', 'Tsonga'), ('tt', 'Tatar'), ('tw', 'Twi'), ('ug', 'Uyghur'), ('uk', 'Ukrainian'), ('ur', 'Urdu'), ('uz', 'Uzbek'), ('vi', 'Vietnamese'), ('vo', 'Volapuk'), ('wo', 'Wolof'), ('xh', 'Xhosa'), ('yi', 'Yiddish'), ('yo', 'Yoruba'), ('za', 'Zhuang'), ('zh', 'Chinese'), ('zu', 'Zulu'), ('nb_NO', 'Norwegian Bokmal'), ('pt_BR', 'Brazilian Portuguese'), ('es_MX', 'Mexican Spanish'), ('uk_UA', 'Ukrainian'), ('zh_CN', 'Simplified Chinese'), ('zh_TW', 'Traditional Chinese')], default='en', help_text="The language the project documentation is rendered in. Note: this affects your project's URL.", max_length=20, verbose_name='Language'),
+        ),
+        migrations.AlterField(
+            model_name='project',
+            name='privacy_level',
+            field=models.CharField(choices=[('public', 'Public'), ('protected', 'Protected'), ('private', 'Private')], default='public', help_text='Level of privacy that you want on the repository. Protected means public but not in listings.', max_length=20, verbose_name='Privacy Level'),
+        ),
+        migrations.AlterField(
+            model_name='project',
+            name='python_interpreter',
+            field=models.CharField(choices=[('python', 'CPython 2.x'), ('python3', 'CPython 3.x')], default='python', help_text='The Python interpreter used to create the virtual environment.', max_length=20, verbose_name='Python Interpreter'),
+        ),
+        migrations.AlterField(
+            model_name='project',
+            name='version_privacy_level',
+            field=models.CharField(choices=[('public', 'Public'), ('protected', 'Protected'), ('private', 'Private')], default='public', help_text='Default level of privacy you want on built versions of documentation.', max_length=20, verbose_name='Version Privacy Level'),
+        ),
+    ]

readthedocs/projects/models.py

@@ -7,13 +7,14 @@
 import fnmatch
 import logging
 import os
-from builtins import object  # pylint: disable=redefined-builtin
 
+from builtins import object  # pylint: disable=redefined-builtin
 from django.conf import settings
 from django.contrib.auth.models import User
 from django.core.urlresolvers import NoReverseMatch, reverse
 from django.db import models
 from django.utils.encoding import python_2_unicode_compatible
+from django.utils.functional import cached_property
 from django.utils.translation import ugettext_lazy as _
 from future.backports.urllib.parse import urlparse  # noqa
 from guardian.shortcuts import assign
@@ -24,6 +25,7 @@
 from readthedocs.core.utils import broadcast, slugify
 from readthedocs.projects import constants
 from readthedocs.projects.exceptions import ProjectConfigurationError
+from readthedocs.projects.managers import HTMLFileManager
 from readthedocs.projects.querysets import (
     ChildRelatedProjectQuerySet, FeatureQuerySet, ProjectQuerySet,
     RelatedProjectQuerySet)
@@ -32,6 +34,7 @@
 from readthedocs.projects.version_handling import (
     determine_stable_version, version_windows)
 from readthedocs.restapi.client import api
+from readthedocs.search.parse_json import process_file
 from readthedocs.vcs_support.backends import backend_cls
 from readthedocs.vcs_support.utils import Lock, NonBlockingLock
 
@@ -910,6 +913,40 @@ def __str__(self):
         return '%s: %s' % (self.name, self.project)
 
 
+class HTMLFile(ImportedFile):
+
+    """
+    Imported HTML file Proxy model.
+
+    This tracks only the HTML files for indexing to search.
+    """
+
+    class Meta(object):
+        proxy = True
+
+    objects = HTMLFileManager()
+
+    @cached_property
+    def json_file_path(self):
+        basename = os.path.splitext(self.path)[0]
+        file_path = basename + '.fjson'
+
+        full_json_path = self.project.get_production_media_path(type_='json',
+                                                                version_slug=self.version.slug,
+                                                                include_file=False)
+
+        file_path = os.path.join(full_json_path, file_path)
+        return file_path
+
+    def get_processed_json(self):
+        file_path = self.json_file_path
+        return process_file(file_path)
+
+    @cached_property
+    def processed_json(self):
+        return self.get_processed_json()
+
+
 class Notification(models.Model):
     project = models.ForeignKey(Project,
                                 related_name='%(class)s_notifications')

readthedocs/projects/signals.py

@@ -2,10 +2,11 @@
 
 from __future__ import absolute_import
 import django.dispatch
+from django.db.models.signals import pre_save
 from django.dispatch import receiver
 
 from readthedocs.oauth.utils import attach_webhook
-
+from .models import HTMLFile
 
 before_vcs = django.dispatch.Signal(providing_args=["version"])
 after_vcs = django.dispatch.Signal(providing_args=["version"])

readthedocs/projects/tasks.py

@@ -8,6 +8,7 @@
 from __future__ import absolute_import
 
 import datetime
+import fnmatch
 import hashlib
 import json
 import logging
@@ -29,7 +30,7 @@
 
 from .constants import LOG_TEMPLATE
 from .exceptions import RepositoryError
-from .models import ImportedFile, Project, Domain
+from .models import ImportedFile, Project, Domain, HTMLFile
 from .signals import before_vcs, after_vcs, before_build, after_build, files_changed
 from readthedocs.builds.constants import (LATEST,
                                           BUILD_STATE_CLONING,
@@ -943,18 +944,24 @@ def _manage_imported_files(version, path, commit):
     changed_files = set()
     for root, __, filenames in os.walk(path):
         for filename in filenames:
+            if fnmatch.fnmatch(filename, '*.html'):
+                model_class = HTMLFile
+            else:
+                model_class = ImportedFile
+
             dirpath = os.path.join(root.replace(path, '').lstrip('/'),
                                    filename.lstrip('/'))
             full_path = os.path.join(root, filename)
             md5 = hashlib.md5(open(full_path, 'rb').read()).hexdigest()
             try:
-                obj, __ = ImportedFile.objects.get_or_create(
+                # pylint: disable=unpacking-non-sequence
+                obj, __ = model_class.objects.get_or_create(
                     project=version.project,
                     version=version,
                     path=dirpath,
                     name=filename,
                 )
-            except ImportedFile.MultipleObjectsReturned:
+            except model_class.MultipleObjectsReturned:
                 log.warning('Error creating ImportedFile')
                 continue
             if obj.md5 != md5:
@@ -963,6 +970,12 @@ def _manage_imported_files(version, path, commit):
             if obj.commit != commit:
                 obj.commit = commit
             obj.save()
+
+    # Delete the HTMLFile first from previous versions
+    HTMLFile.objects.filter(project=version.project,
+                            version=version
+                            ).exclude(commit=commit).delete()
+
     # Delete ImportedFiles from previous versions
     ImportedFile.objects.filter(project=version.project,
                                 version=version
@@ -1173,7 +1186,6 @@ def sync_callback(_, version_pk, commit, *args, **kwargs):
     The first argument is the result from previous tasks, which we discard.
     """
     fileify(version_pk, commit=commit)
-    update_search(version_pk, commit=commit)
 
 
 @app.task()

readthedocs/projects/utils.py

@@ -32,18 +32,21 @@ def version_from_slug(slug, version):
     return v
 
 
-def find_file(filename):
+def find_file(basename, pattern, path):
     """
-    Recursively find matching file from the current working path.
+    Recursively find matching file.
 
-    :param file: Filename to match
-    :returns: A list of matching filenames.
+    :param basename: Basename of a file to match
+    :param pattern: Pattern to match
+    :param path: the directory to search for the file
+    :returns: path of matching file
     """
-    matches = []
-    for root, __, filenames in os.walk('.'):
-        for match in fnmatch.filter(filenames, filename):
-            matches.append(os.path.join(root, match))
-    return matches
+    for root, _, files in os.walk(path):
+        for filename in files:
+            file_basename = os.path.splitext(filename)[0]
+
+            if fnmatch.fnmatch(filename, pattern) and file_basename == basename:
+                return os.path.join(root, filename)
 
 
 def run(*commands):

readthedocs/search/conf.py

@@ -0,0 +1 @@
+SEARCH_EXCLUDED_FILE = ['search.html', 'genindex.html', 'py-modindex.html']

readthedocs/search/documents.py

@@ -0,0 +1,106 @@
+from django.conf import settings
+from django_elasticsearch_dsl import DocType, Index, fields
+
+from readthedocs.projects.models import Project, HTMLFile
+from .conf import SEARCH_EXCLUDED_FILE
+
+from readthedocs.search.faceted_search import ProjectSearch, FileSearch
+
+project_conf = settings.ES_INDEXES['project']
+project_index = Index(project_conf['name'])
+project_index.settings(**project_conf['settings'])
+
+page_conf = settings.ES_INDEXES['page']
+page_index = Index(page_conf['name'])
+page_index.settings(**page_conf['settings'])
+
+
+@project_index.doc_type
+class ProjectDocument(DocType):
+
+    class Meta(object):
+        model = Project
+        fields = ('name', 'slug', 'description')
+
+    url = fields.TextField(attr='get_absolute_url')
+    users = fields.NestedField(properties={
+        'username': fields.TextField(),
+        'id': fields.IntegerField(),
+    })
+    language = fields.KeywordField()
+
+    @classmethod
+    def faceted_search(cls, query, language=None, using=None, index=None):
+        kwargs = {
+            'using': using or cls._doc_type.using,
+            'index': index or cls._doc_type.index,
+            'doc_types': [cls],
+            'model': cls._doc_type.model,
+            'query': query
+        }
+
+        if language:
+            kwargs['filters'] = {'language': language}
+
+        return ProjectSearch(**kwargs)
+
+
+@page_index.doc_type
+class PageDocument(DocType):
+
+    class Meta(object):
+        model = HTMLFile
+        fields = ('commit',)
+
+    project = fields.KeywordField(attr='project.slug')
+    version = fields.KeywordField(attr='version.slug')
+
+    title = fields.TextField(attr='processed_json.title')
+    headers = fields.TextField(attr='processed_json.headers')
+    content = fields.TextField(attr='processed_json.content')
+    path = fields.TextField(attr='processed_json.path')
+
+    @classmethod
+    def faceted_search(cls, query, projects_list=None, versions_list=None, using=None, index=None):
+        kwargs = {
+            'using': using or cls._doc_type.using,
+            'index': index or cls._doc_type.index,
+            'doc_types': [cls],
+            'model': cls._doc_type.model,
+            'query': query
+        }
+        filters = {}
+
+        if projects_list:
+            filters['project'] = projects_list
+        if versions_list:
+            filters['version'] = versions_list
+
+        kwargs['filters'] = filters
+
+        return FileSearch(**kwargs)
+
+    def get_queryset(self):
+        """Overwrite default queryset to filter certain files to index"""
+        queryset = super(PageDocument, self).get_queryset()
+
+        # Do not index files that belong to non sphinx project
+        # Also do not index certain files
+        queryset = (queryset.filter(project__documentation_type='sphinx')
+                            .exclude(name__in=SEARCH_EXCLUDED_FILE))
+        return queryset
+
+    def update(self, thing, refresh=None, action='index', **kwargs):
+        """Overwrite in order to index only certain files"""
+        # Object not exist in the provided queryset should not be indexed
+        # TODO: remove this overwrite when the issue has been fixed
+        # See below link for more information
+        # https://github.com/sabricot/django-elasticsearch-dsl/issues/111
+        if isinstance(thing, HTMLFile):
+            # Its a model instance.
+            queryset = self.get_queryset()
+            obj = queryset.filter(pk=thing.pk)
+            if not obj.exists():
+                return None
+
+        return super(PageDocument, self).update(thing=thing, refresh=None, action='index', **kwargs)