Refactor: remove Version.documentation_type

readthedocs/api/v2/views/footer_views.py

@@ -155,6 +155,7 @@ def _get_context(self):
 
         page_slug = self.request.GET.get('page', '')
         path = ''
+        # TODO: review custom logic based con `documentation_type`
         if page_slug and page_slug != 'index':
             if version.documentation_type in {SPHINX_HTMLDIR, MKDOCS}:
                 path = re.sub('/index$', '', page_slug) + '/'

readthedocs/builds/models.py

@@ -352,7 +352,7 @@ def commit_name(self):
 
         # By now we must have handled all special versions.
         if self.slug in NON_REPOSITORY_VERSIONS:
-            raise Exception('All special versions must be handled by now.')
+            raise Exception("All special versions must be handled by now.")  # noqa
 
         if self.type in (BRANCH, TAG):
             # If this version is a branch or a tag, the verbose_name will
@@ -407,7 +407,7 @@ def get_absolute_url(self):
             external=external,
         )
 
-    def delete(self, *args, **kwargs):  # pylint: disable=arguments-differ
+    def delete(self, *args, **kwargs):
         from readthedocs.projects.tasks.utils import clean_project_resources
         log.info('Removing files for version.', version_slug=self.slug)
         clean_project_resources(self.project, self)
@@ -475,6 +475,8 @@ def is_editable(self):
 
     @property
     def is_sphinx_type(self):
+        # TODO: this method should no be required anymore.
+        # I'm not removing it yet because there may be places still using it.
         return self.documentation_type in {SPHINX, SPHINX_HTMLDIR, SPHINX_SINGLEHTML}
 
     @property
@@ -699,7 +701,7 @@ def __init__(self, *args, **kwargs):
 
         super().__init__(*args, **valid_attributes)
 
-    def save(self, *args, **kwargs):  # pylint: disable=arguments-differ
+    def save(self, *args, **kwargs):
         return 0
 
 
@@ -1378,7 +1380,7 @@ def move(self, steps):
         self.save()
         return True
 
-    def delete(self, *args, **kwargs):  # pylint: disable=arguments-differ
+    def delete(self, *args, **kwargs):
         """Override method to update the other priorities after delete."""
         current_priority = self.priority
         project = self.project
@@ -1473,7 +1475,6 @@ def get_edit_url(self):
 
 
 class AutomationRuleMatch(TimeStampedModel):
-
     ACTIONS_PAST_TENSE = {
         VersionAutomationRule.ACTIVATE_VERSION_ACTION: _("Version activated"),
         VersionAutomationRule.HIDE_VERSION_ACTION: _("Version hidden"),

readthedocs/doc_builder/director.py

@@ -441,10 +441,6 @@ def run_build_commands(self):
         if not os.path.exists(html_output_path):
             raise BuildUserError(BuildUserError.BUILD_COMMANDS_WITHOUT_OUTPUT)
 
-        # Update the `Version.documentation_type` to match the doctype defined
-        # by the config file. When using `build.commands` it will be `GENERIC`
-        self.data.version.documentation_type = self.data.config.doctype
-
         # Mark this version to inject the new js client when serving it via El Proxito
         self.data.version.addons = True
 
@@ -696,10 +692,6 @@ def get_build_env_vars(self):
 
         return env
 
-    def is_type_sphinx(self):
-        """Is documentation type Sphinx."""
-        return "sphinx" in self.data.config.doctype
-
     def store_readthedocs_build_yaml(self):
         # load YAML from user
         yaml_path = os.path.join(

readthedocs/embed/views.py

@@ -145,6 +145,11 @@ def do_embed(*, project, version, doc=None, path=None, section=None, url=None):
 
     content = None
     headers = None
+    # TODO: EmbedAPI v1 is already deprecated and shouldn't be used.
+    # So, we may be able to remove the endpoint completely, together with the usage of
+    # `is_sphinx_type` as well :)
+    # It's disabled on commercial already. We should check the stats on Cloudflare/NewRelic
+    # looking for EmbedAPI v1 and decide what to do depending on usage.
     if version.is_sphinx_type:
         file_content = _get_doc_content(
             project=project,

readthedocs/projects/models.py

@@ -51,7 +51,7 @@
     validate_repository_url,
 )
 from readthedocs.projects.version_handling import determine_stable_version
-from readthedocs.search.parsers import GenericParser, MkDocsParser, SphinxParser
+from readthedocs.search.parsers import GenericParser
 from readthedocs.storage import build_media_storage
 from readthedocs.vcs_support.backends import backend_cls
 
@@ -1525,24 +1525,8 @@ class Meta:
     objects = HTMLFileManager()
 
     def get_processed_json(self):
-        if (
-            self.version.documentation_type == constants.GENERIC
-            or self.project.has_feature(Feature.INDEX_FROM_HTML_FILES)
-        ):
-            parser_class = GenericParser
-        elif self.version.is_sphinx_type:
-            parser_class = SphinxParser
-        elif self.version.is_mkdocs_type:
-            parser_class = MkDocsParser
-        else:
-            log.warning(
-                "Invalid documentation type",
-                documentation_type=self.version.documentation_type,
-                version_slug=self.version.slug,
-                project_slug=self.project.slug,
-            )
-            return {}
-        parser = parser_class(self.version)
+        # See https://github.com/readthedocs/readthedocs.org/issues/10272
+        parser = GenericParser(self.version)
         return parser.parse(self.path)
 
     @cached_property
@@ -1934,7 +1918,6 @@ def add_features(sender, **kwargs):
     DISABLE_SERVER_SIDE_SEARCH = 'disable_server_side_search'
     ENABLE_MKDOCS_SERVER_SIDE_SEARCH = 'enable_mkdocs_server_side_search'
     DEFAULT_TO_FUZZY_SEARCH = 'default_to_fuzzy_search'
-    INDEX_FROM_HTML_FILES = 'index_from_html_files'
 
     # Build related features
     DONT_CREATE_INDEX = "dont_create_index"
@@ -2054,14 +2037,6 @@ def add_features(sender, **kwargs):
             DEFAULT_TO_FUZZY_SEARCH,
             _("Search: Default to fuzzy search for simple search queries"),
         ),
-        (
-            INDEX_FROM_HTML_FILES,
-            _(
-                "Search: Index content directly from html files instead or relying in other "
-                "sources"
-            ),
-        ),
-
         (
             DONT_CREATE_INDEX,
             _(

readthedocs/search/api/v2/serializers.py

@@ -145,6 +145,7 @@ def _get_full_path(self, obj):
             docs_url = project_data.version.docs_url
             path = obj.full_path
 
+            # TODO: review this logic that uses doctype
             # Generate an appropriate link for the doctypes that use htmldir,
             # and always end it with / so it goes directly to proxito.
             # For a generic doctype we just strip the index.html part if it exists.

readthedocs/search/documents.py

@@ -85,7 +85,6 @@ class PageDocument(RTDDocTypeMixin, Document):
     # Metadata
     project = fields.KeywordField(attr='project.slug')
     version = fields.KeywordField(attr='version.slug')
-    doctype = fields.KeywordField(attr='version.documentation_type')
     path = fields.KeywordField(attr='processed_json.path')
     full_path = fields.KeywordField(attr='path')
     rank = fields.IntegerField()

readthedocs/telemetry/collectors.py

@@ -36,7 +36,7 @@ def __init__(self, environment):
 
     @staticmethod
     def _safe_json_loads(content, default=None):
-        def lowercase(d):  # pylint: disable=invalid-name
+        def lowercase(d):
             """Convert all dictionary keys to lowercase."""
             return {k.lower(): i for k, i in d.items()}
 
@@ -96,13 +96,9 @@ def collect(self):
         return data
 
     def _get_doctool_name(self):
-        if self.version.is_sphinx_type:
-            return "sphinx"
-
-        if self.version.is_mkdocs_type:
-            return "mkdocs"
-
-        return "generic"
+        # TODO: this data should be communicated from the builder itself homehow.
+        # This is a place where the build contract could be useful.
+        return "unknown"
 
     def _get_doctool(self):
         data = {
@@ -111,6 +107,9 @@ def _get_doctool(self):
             "html_theme": "",
         }
 
+        # TODO: adapt this logic to work with a generic builder. This requires
+        # a build contract that explains how to create the JSON file we are
+        # expecting at this point
         if self._get_doctool_name() != "sphinx":
             return data