asterisk/rest-api-templates/asterisk_processor.py

255 lines
8.9 KiB
Python
Raw Permalink Normal View History

#
# Asterisk -- An open source telephony toolkit.
#
# Copyright (C) 2013, Digium, Inc.
#
# David M. Lee, II <dlee@digium.com>
#
# See http://www.asterisk.org for more information about
# the Asterisk project. Please do not directly contact
# any of the maintainers of this project for assistance;
# the project provides a web site, mailing lists and IRC
# channels for your use.
#
# This program is free software, distributed under the terms of
# the GNU General Public License Version 2. See the LICENSE file
# at the top of the source tree.
#
"""Implementation of SwaggerPostProcessor which adds fields needed to generate
Asterisk RESTful HTTP binding code.
"""
import os
import re
from swagger_model import Stringify, SwaggerError, SwaggerPostProcessor
Update events to use Swagger 1.3 subtyping, and related aftermath This patch started with the simple idea of changing the /events data model to be more sane. The original model would send out events like: { "stasis_start": { "args": [], "channel": { ... } } } The event discriminator was the field name instead of being a value in the object, due to limitations in how Swagger 1.1 could model objects. While technically sufficient in communicating event information, it was really difficult to deal with in terms of client side JSON handling. This patch takes advantage of a proposed extension[1] to Swagger which allows type variance through the use of a discriminator field. This had a domino effect that made this a surprisingly large patch. [1]: https://groups.google.com/d/msg/wordnik-api/EC3rGajE0os/ey_5dBI_jWcJ In changing the models, I also had to change the swagger_model.py processor so it can handle the type discriminator and subtyping. I took that a big step forward, and using that information to generate an ari_model module, which can validate a JSON object against the Swagger model. The REST and WebSocket generators were changed to take advantage of the validators. If compiled with AST_DEVMODE enabled, JSON objects that don't match their corresponding models will not be sent out. For REST API calls, a 500 Internal Server response is sent. For WebSockets, the invalid JSON message is replaced with an error message. Since this took over about half of the job of the existing JSON generators, and the .to_json virtual function on messages took over the other half, I reluctantly removed the generators. The validators turned up all sorts of errors and inconsistencies in our data models, and the code. These were cleaned up, with checks in the code generator avoid some of the consistency problems in the future. * The model for a channel snapshot was trimmed down to match the information sent via AMI. Many of the field being sent were not useful in the general case. * The model for a bridge snapshot was updated to be more consistent with the other ARI models. Another impact of introducing subtyping was that the swagger-codegen documentation generator was insufficient (at least until it catches up with Swagger 1.2). I wanted it to be easier to generate docs for the API anyways, so I ported the wiki pages to use the Asterisk Swagger generator. In the process, I was able to clean up many of the model links, which would occasionally give inconsistent results on the wiki. I also added error responses to the wiki docs, making the wiki documentation more complete. Finally, since Stasis-HTTP will now be named Asterisk REST Interface (ARI), any new functions and files I created carry the ari_ prefix. I changed a few stasis_http references to ari where it was non-intrusive and made sense. (closes issue ASTERISK-21885) Review: https://reviewboard.asterisk.org/r/2639/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@393529 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-07-03 16:32:41 +00:00
try:
from collections import OrderedDict
except ImportError:
from odict import OrderedDict
def simple_name(name):
"""Removes the {markers} from a path segement.
@param name: Swagger path segement, with {pathVar} markers.
"""
if name.startswith('{') and name.endswith('}'):
return name[1:-1]
return name
Update events to use Swagger 1.3 subtyping, and related aftermath This patch started with the simple idea of changing the /events data model to be more sane. The original model would send out events like: { "stasis_start": { "args": [], "channel": { ... } } } The event discriminator was the field name instead of being a value in the object, due to limitations in how Swagger 1.1 could model objects. While technically sufficient in communicating event information, it was really difficult to deal with in terms of client side JSON handling. This patch takes advantage of a proposed extension[1] to Swagger which allows type variance through the use of a discriminator field. This had a domino effect that made this a surprisingly large patch. [1]: https://groups.google.com/d/msg/wordnik-api/EC3rGajE0os/ey_5dBI_jWcJ In changing the models, I also had to change the swagger_model.py processor so it can handle the type discriminator and subtyping. I took that a big step forward, and using that information to generate an ari_model module, which can validate a JSON object against the Swagger model. The REST and WebSocket generators were changed to take advantage of the validators. If compiled with AST_DEVMODE enabled, JSON objects that don't match their corresponding models will not be sent out. For REST API calls, a 500 Internal Server response is sent. For WebSockets, the invalid JSON message is replaced with an error message. Since this took over about half of the job of the existing JSON generators, and the .to_json virtual function on messages took over the other half, I reluctantly removed the generators. The validators turned up all sorts of errors and inconsistencies in our data models, and the code. These were cleaned up, with checks in the code generator avoid some of the consistency problems in the future. * The model for a channel snapshot was trimmed down to match the information sent via AMI. Many of the field being sent were not useful in the general case. * The model for a bridge snapshot was updated to be more consistent with the other ARI models. Another impact of introducing subtyping was that the swagger-codegen documentation generator was insufficient (at least until it catches up with Swagger 1.2). I wanted it to be easier to generate docs for the API anyways, so I ported the wiki pages to use the Asterisk Swagger generator. In the process, I was able to clean up many of the model links, which would occasionally give inconsistent results on the wiki. I also added error responses to the wiki docs, making the wiki documentation more complete. Finally, since Stasis-HTTP will now be named Asterisk REST Interface (ARI), any new functions and files I created carry the ari_ prefix. I changed a few stasis_http references to ari where it was non-intrusive and made sense. (closes issue ASTERISK-21885) Review: https://reviewboard.asterisk.org/r/2639/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@393529 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-07-03 16:32:41 +00:00
def wikify(str):
"""Escapes a string for the wiki.
@param str: String to escape
"""
# Replace all line breaks with line feeds
str = re.sub(r'<br\s*/?>', '\n', str)
Update events to use Swagger 1.3 subtyping, and related aftermath This patch started with the simple idea of changing the /events data model to be more sane. The original model would send out events like: { "stasis_start": { "args": [], "channel": { ... } } } The event discriminator was the field name instead of being a value in the object, due to limitations in how Swagger 1.1 could model objects. While technically sufficient in communicating event information, it was really difficult to deal with in terms of client side JSON handling. This patch takes advantage of a proposed extension[1] to Swagger which allows type variance through the use of a discriminator field. This had a domino effect that made this a surprisingly large patch. [1]: https://groups.google.com/d/msg/wordnik-api/EC3rGajE0os/ey_5dBI_jWcJ In changing the models, I also had to change the swagger_model.py processor so it can handle the type discriminator and subtyping. I took that a big step forward, and using that information to generate an ari_model module, which can validate a JSON object against the Swagger model. The REST and WebSocket generators were changed to take advantage of the validators. If compiled with AST_DEVMODE enabled, JSON objects that don't match their corresponding models will not be sent out. For REST API calls, a 500 Internal Server response is sent. For WebSockets, the invalid JSON message is replaced with an error message. Since this took over about half of the job of the existing JSON generators, and the .to_json virtual function on messages took over the other half, I reluctantly removed the generators. The validators turned up all sorts of errors and inconsistencies in our data models, and the code. These were cleaned up, with checks in the code generator avoid some of the consistency problems in the future. * The model for a channel snapshot was trimmed down to match the information sent via AMI. Many of the field being sent were not useful in the general case. * The model for a bridge snapshot was updated to be more consistent with the other ARI models. Another impact of introducing subtyping was that the swagger-codegen documentation generator was insufficient (at least until it catches up with Swagger 1.2). I wanted it to be easier to generate docs for the API anyways, so I ported the wiki pages to use the Asterisk Swagger generator. In the process, I was able to clean up many of the model links, which would occasionally give inconsistent results on the wiki. I also added error responses to the wiki docs, making the wiki documentation more complete. Finally, since Stasis-HTTP will now be named Asterisk REST Interface (ARI), any new functions and files I created carry the ari_ prefix. I changed a few stasis_http references to ari where it was non-intrusive and made sense. (closes issue ASTERISK-21885) Review: https://reviewboard.asterisk.org/r/2639/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@393529 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-07-03 16:32:41 +00:00
return re.sub(r'([{}\[\]])', r'\\\1', str)
def snakify(name):
"""Helper to take a camelCase or dash-seperated name and make it
snake_case.
"""
r = ''
prior_lower = False
for c in name:
if c.isupper() and prior_lower:
r += "_"
if c == '-':
c = '_'
prior_lower = c.islower()
r += c.lower()
return r
class PathSegment(Stringify):
"""Tree representation of a Swagger API declaration.
"""
def __init__(self, name, parent):
"""Ctor.
@param name: Name of this path segment. May have {pathVar} markers.
@param parent: Parent PathSegment.
"""
#: Segment name, with {pathVar} markers removed
self.name = simple_name(name)
#: True if segment is a {pathVar}, else None.
self.is_wildcard = None
#: Underscore seperated name all ancestor segments
self.full_name = None
#: Dictionary of child PathSegements
self.__children = OrderedDict()
#: List of operations on this segement
self.operations = []
if self.name != name:
self.is_wildcard = True
if not self.name:
assert(not parent)
self.full_name = ''
if not parent or not parent.name:
self.full_name = name
else:
self.full_name = "%s_%s" % (parent.full_name, self.name)
def get_child(self, path):
"""Walks descendants to get path, creating it if necessary.
@param path: List of path names.
@return: PageSegment corresponding to path.
"""
assert simple_name(path[0]) == self.name
if (len(path) == 1):
return self
child = self.__children.get(path[1])
if not child:
child = PathSegment(path[1], self)
self.__children[path[1]] = child
return child.get_child(path[1:])
def children(self):
"""Gets list of children.
"""
return self.__children.values()
def num_children(self):
"""Gets count of children.
"""
return len(self.__children)
Update events to use Swagger 1.3 subtyping, and related aftermath This patch started with the simple idea of changing the /events data model to be more sane. The original model would send out events like: { "stasis_start": { "args": [], "channel": { ... } } } The event discriminator was the field name instead of being a value in the object, due to limitations in how Swagger 1.1 could model objects. While technically sufficient in communicating event information, it was really difficult to deal with in terms of client side JSON handling. This patch takes advantage of a proposed extension[1] to Swagger which allows type variance through the use of a discriminator field. This had a domino effect that made this a surprisingly large patch. [1]: https://groups.google.com/d/msg/wordnik-api/EC3rGajE0os/ey_5dBI_jWcJ In changing the models, I also had to change the swagger_model.py processor so it can handle the type discriminator and subtyping. I took that a big step forward, and using that information to generate an ari_model module, which can validate a JSON object against the Swagger model. The REST and WebSocket generators were changed to take advantage of the validators. If compiled with AST_DEVMODE enabled, JSON objects that don't match their corresponding models will not be sent out. For REST API calls, a 500 Internal Server response is sent. For WebSockets, the invalid JSON message is replaced with an error message. Since this took over about half of the job of the existing JSON generators, and the .to_json virtual function on messages took over the other half, I reluctantly removed the generators. The validators turned up all sorts of errors and inconsistencies in our data models, and the code. These were cleaned up, with checks in the code generator avoid some of the consistency problems in the future. * The model for a channel snapshot was trimmed down to match the information sent via AMI. Many of the field being sent were not useful in the general case. * The model for a bridge snapshot was updated to be more consistent with the other ARI models. Another impact of introducing subtyping was that the swagger-codegen documentation generator was insufficient (at least until it catches up with Swagger 1.2). I wanted it to be easier to generate docs for the API anyways, so I ported the wiki pages to use the Asterisk Swagger generator. In the process, I was able to clean up many of the model links, which would occasionally give inconsistent results on the wiki. I also added error responses to the wiki docs, making the wiki documentation more complete. Finally, since Stasis-HTTP will now be named Asterisk REST Interface (ARI), any new functions and files I created carry the ari_ prefix. I changed a few stasis_http references to ari where it was non-intrusive and made sense. (closes issue ASTERISK-21885) Review: https://reviewboard.asterisk.org/r/2639/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@393529 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-07-03 16:32:41 +00:00
class AsteriskProcessor(SwaggerPostProcessor):
"""A SwaggerPostProcessor which adds fields needed to generate Asterisk
RESTful HTTP binding code.
"""
#: How Swagger types map to C.
type_mapping = {
'string': 'const char *',
'boolean': 'int',
'number': 'int',
'int': 'int',
'long': 'long',
'double': 'double',
'float': 'float',
}
#: String conversion functions for string to C type.
convert_mapping = {
'string': '',
'int': 'atoi',
'long': 'atol',
'double': 'atof',
'boolean': 'ast_true',
}
#: JSON conversion functions
json_convert_mapping = {
'string': 'ast_json_string_get',
'int': 'ast_json_integer_get',
'long': 'ast_json_integer_get',
'double': 'ast_json_real_get',
'boolean': 'ast_json_is_true',
}
Update events to use Swagger 1.3 subtyping, and related aftermath This patch started with the simple idea of changing the /events data model to be more sane. The original model would send out events like: { "stasis_start": { "args": [], "channel": { ... } } } The event discriminator was the field name instead of being a value in the object, due to limitations in how Swagger 1.1 could model objects. While technically sufficient in communicating event information, it was really difficult to deal with in terms of client side JSON handling. This patch takes advantage of a proposed extension[1] to Swagger which allows type variance through the use of a discriminator field. This had a domino effect that made this a surprisingly large patch. [1]: https://groups.google.com/d/msg/wordnik-api/EC3rGajE0os/ey_5dBI_jWcJ In changing the models, I also had to change the swagger_model.py processor so it can handle the type discriminator and subtyping. I took that a big step forward, and using that information to generate an ari_model module, which can validate a JSON object against the Swagger model. The REST and WebSocket generators were changed to take advantage of the validators. If compiled with AST_DEVMODE enabled, JSON objects that don't match their corresponding models will not be sent out. For REST API calls, a 500 Internal Server response is sent. For WebSockets, the invalid JSON message is replaced with an error message. Since this took over about half of the job of the existing JSON generators, and the .to_json virtual function on messages took over the other half, I reluctantly removed the generators. The validators turned up all sorts of errors and inconsistencies in our data models, and the code. These were cleaned up, with checks in the code generator avoid some of the consistency problems in the future. * The model for a channel snapshot was trimmed down to match the information sent via AMI. Many of the field being sent were not useful in the general case. * The model for a bridge snapshot was updated to be more consistent with the other ARI models. Another impact of introducing subtyping was that the swagger-codegen documentation generator was insufficient (at least until it catches up with Swagger 1.2). I wanted it to be easier to generate docs for the API anyways, so I ported the wiki pages to use the Asterisk Swagger generator. In the process, I was able to clean up many of the model links, which would occasionally give inconsistent results on the wiki. I also added error responses to the wiki docs, making the wiki documentation more complete. Finally, since Stasis-HTTP will now be named Asterisk REST Interface (ARI), any new functions and files I created carry the ari_ prefix. I changed a few stasis_http references to ari where it was non-intrusive and made sense. (closes issue ASTERISK-21885) Review: https://reviewboard.asterisk.org/r/2639/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@393529 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-07-03 16:32:41 +00:00
def __init__(self, wiki_prefix):
self.wiki_prefix = wiki_prefix
def process_resource_api(self, resource_api, context):
resource_api.wiki_prefix = self.wiki_prefix
# Derive a resource name from the API declaration's filename
resource_api.name = re.sub('\..*', '',
os.path.basename(resource_api.path))
Update events to use Swagger 1.3 subtyping, and related aftermath This patch started with the simple idea of changing the /events data model to be more sane. The original model would send out events like: { "stasis_start": { "args": [], "channel": { ... } } } The event discriminator was the field name instead of being a value in the object, due to limitations in how Swagger 1.1 could model objects. While technically sufficient in communicating event information, it was really difficult to deal with in terms of client side JSON handling. This patch takes advantage of a proposed extension[1] to Swagger which allows type variance through the use of a discriminator field. This had a domino effect that made this a surprisingly large patch. [1]: https://groups.google.com/d/msg/wordnik-api/EC3rGajE0os/ey_5dBI_jWcJ In changing the models, I also had to change the swagger_model.py processor so it can handle the type discriminator and subtyping. I took that a big step forward, and using that information to generate an ari_model module, which can validate a JSON object against the Swagger model. The REST and WebSocket generators were changed to take advantage of the validators. If compiled with AST_DEVMODE enabled, JSON objects that don't match their corresponding models will not be sent out. For REST API calls, a 500 Internal Server response is sent. For WebSockets, the invalid JSON message is replaced with an error message. Since this took over about half of the job of the existing JSON generators, and the .to_json virtual function on messages took over the other half, I reluctantly removed the generators. The validators turned up all sorts of errors and inconsistencies in our data models, and the code. These were cleaned up, with checks in the code generator avoid some of the consistency problems in the future. * The model for a channel snapshot was trimmed down to match the information sent via AMI. Many of the field being sent were not useful in the general case. * The model for a bridge snapshot was updated to be more consistent with the other ARI models. Another impact of introducing subtyping was that the swagger-codegen documentation generator was insufficient (at least until it catches up with Swagger 1.2). I wanted it to be easier to generate docs for the API anyways, so I ported the wiki pages to use the Asterisk Swagger generator. In the process, I was able to clean up many of the model links, which would occasionally give inconsistent results on the wiki. I also added error responses to the wiki docs, making the wiki documentation more complete. Finally, since Stasis-HTTP will now be named Asterisk REST Interface (ARI), any new functions and files I created carry the ari_ prefix. I changed a few stasis_http references to ari where it was non-intrusive and made sense. (closes issue ASTERISK-21885) Review: https://reviewboard.asterisk.org/r/2639/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@393529 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-07-03 16:32:41 +00:00
# Now in all caps, for include guard
resource_api.name_caps = resource_api.name.upper()
Update events to use Swagger 1.3 subtyping, and related aftermath This patch started with the simple idea of changing the /events data model to be more sane. The original model would send out events like: { "stasis_start": { "args": [], "channel": { ... } } } The event discriminator was the field name instead of being a value in the object, due to limitations in how Swagger 1.1 could model objects. While technically sufficient in communicating event information, it was really difficult to deal with in terms of client side JSON handling. This patch takes advantage of a proposed extension[1] to Swagger which allows type variance through the use of a discriminator field. This had a domino effect that made this a surprisingly large patch. [1]: https://groups.google.com/d/msg/wordnik-api/EC3rGajE0os/ey_5dBI_jWcJ In changing the models, I also had to change the swagger_model.py processor so it can handle the type discriminator and subtyping. I took that a big step forward, and using that information to generate an ari_model module, which can validate a JSON object against the Swagger model. The REST and WebSocket generators were changed to take advantage of the validators. If compiled with AST_DEVMODE enabled, JSON objects that don't match their corresponding models will not be sent out. For REST API calls, a 500 Internal Server response is sent. For WebSockets, the invalid JSON message is replaced with an error message. Since this took over about half of the job of the existing JSON generators, and the .to_json virtual function on messages took over the other half, I reluctantly removed the generators. The validators turned up all sorts of errors and inconsistencies in our data models, and the code. These were cleaned up, with checks in the code generator avoid some of the consistency problems in the future. * The model for a channel snapshot was trimmed down to match the information sent via AMI. Many of the field being sent were not useful in the general case. * The model for a bridge snapshot was updated to be more consistent with the other ARI models. Another impact of introducing subtyping was that the swagger-codegen documentation generator was insufficient (at least until it catches up with Swagger 1.2). I wanted it to be easier to generate docs for the API anyways, so I ported the wiki pages to use the Asterisk Swagger generator. In the process, I was able to clean up many of the model links, which would occasionally give inconsistent results on the wiki. I also added error responses to the wiki docs, making the wiki documentation more complete. Finally, since Stasis-HTTP will now be named Asterisk REST Interface (ARI), any new functions and files I created carry the ari_ prefix. I changed a few stasis_http references to ari where it was non-intrusive and made sense. (closes issue ASTERISK-21885) Review: https://reviewboard.asterisk.org/r/2639/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@393529 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-07-03 16:32:41 +00:00
resource_api.name_title = resource_api.name.capitalize()
resource_api.c_name = snakify(resource_api.name)
# Construct the PathSegement tree for the API.
if resource_api.api_declaration:
resource_api.root_path = PathSegment('', None)
for api in resource_api.api_declaration.apis:
segment = resource_api.root_path.get_child(api.path.split('/'))
for operation in api.operations:
segment.operations.append(operation)
api.full_name = segment.full_name
# Since every API path should start with /[resource], root should
# have exactly one child.
if resource_api.root_path.num_children() != 1:
raise SwaggerError(
"Should not mix resources in one API declaration", context)
# root_path isn't needed any more
resource_api.root_path = list(resource_api.root_path.children())[0]
if resource_api.name != resource_api.root_path.name:
raise SwaggerError(
"API declaration name should match", context)
resource_api.root_full_name = resource_api.root_path.full_name
Update events to use Swagger 1.3 subtyping, and related aftermath This patch started with the simple idea of changing the /events data model to be more sane. The original model would send out events like: { "stasis_start": { "args": [], "channel": { ... } } } The event discriminator was the field name instead of being a value in the object, due to limitations in how Swagger 1.1 could model objects. While technically sufficient in communicating event information, it was really difficult to deal with in terms of client side JSON handling. This patch takes advantage of a proposed extension[1] to Swagger which allows type variance through the use of a discriminator field. This had a domino effect that made this a surprisingly large patch. [1]: https://groups.google.com/d/msg/wordnik-api/EC3rGajE0os/ey_5dBI_jWcJ In changing the models, I also had to change the swagger_model.py processor so it can handle the type discriminator and subtyping. I took that a big step forward, and using that information to generate an ari_model module, which can validate a JSON object against the Swagger model. The REST and WebSocket generators were changed to take advantage of the validators. If compiled with AST_DEVMODE enabled, JSON objects that don't match their corresponding models will not be sent out. For REST API calls, a 500 Internal Server response is sent. For WebSockets, the invalid JSON message is replaced with an error message. Since this took over about half of the job of the existing JSON generators, and the .to_json virtual function on messages took over the other half, I reluctantly removed the generators. The validators turned up all sorts of errors and inconsistencies in our data models, and the code. These were cleaned up, with checks in the code generator avoid some of the consistency problems in the future. * The model for a channel snapshot was trimmed down to match the information sent via AMI. Many of the field being sent were not useful in the general case. * The model for a bridge snapshot was updated to be more consistent with the other ARI models. Another impact of introducing subtyping was that the swagger-codegen documentation generator was insufficient (at least until it catches up with Swagger 1.2). I wanted it to be easier to generate docs for the API anyways, so I ported the wiki pages to use the Asterisk Swagger generator. In the process, I was able to clean up many of the model links, which would occasionally give inconsistent results on the wiki. I also added error responses to the wiki docs, making the wiki documentation more complete. Finally, since Stasis-HTTP will now be named Asterisk REST Interface (ARI), any new functions and files I created carry the ari_ prefix. I changed a few stasis_http references to ari where it was non-intrusive and made sense. (closes issue ASTERISK-21885) Review: https://reviewboard.asterisk.org/r/2639/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@393529 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-07-03 16:32:41 +00:00
def process_api(self, api, context):
api.wiki_path = wikify(api.path)
def process_operation(self, operation, context):
# Nicknames are camelCase, Asterisk coding is snake case
operation.c_nickname = snakify(operation.nickname)
operation.c_http_method = 'AST_HTTP_' + operation.http_method
if not operation.summary.endswith("."):
raise SwaggerError("Summary should end with .", context)
operation.wiki_summary = wikify(operation.summary or "")
operation.wiki_notes = wikify(operation.notes or "")
for error_response in operation.error_responses:
error_response.wiki_reason = wikify(error_response.reason or "")
operation.parse_body = (operation.body_parameter or operation.has_query_parameters) and True
def process_parameter(self, parameter, context):
if parameter.param_type == 'body':
parameter.is_body_parameter = True;
parameter.c_data_type = 'struct ast_json *'
else:
parameter.is_body_parameter = False;
if not parameter.data_type in self.type_mapping:
raise SwaggerError(
"Invalid parameter type %s" % parameter.data_type, context)
# Type conversions
parameter.c_data_type = self.type_mapping[parameter.data_type]
parameter.c_convert = self.convert_mapping[parameter.data_type]
parameter.json_convert = self.json_convert_mapping[parameter.data_type]
# Parameter names are camelcase, Asterisk convention is snake case
parameter.c_name = snakify(parameter.name)
# You shouldn't put a space between 'char *' and the variable
if parameter.c_data_type.endswith('*'):
parameter.c_space = ''
else:
parameter.c_space = ' '
parameter.wiki_description = wikify(parameter.description)
if parameter.allowable_values:
parameter.wiki_allowable_values = parameter.allowable_values.to_wiki()
else:
parameter.wiki_allowable_values = None
def process_model(self, model, context):
Update events to use Swagger 1.3 subtyping, and related aftermath This patch started with the simple idea of changing the /events data model to be more sane. The original model would send out events like: { "stasis_start": { "args": [], "channel": { ... } } } The event discriminator was the field name instead of being a value in the object, due to limitations in how Swagger 1.1 could model objects. While technically sufficient in communicating event information, it was really difficult to deal with in terms of client side JSON handling. This patch takes advantage of a proposed extension[1] to Swagger which allows type variance through the use of a discriminator field. This had a domino effect that made this a surprisingly large patch. [1]: https://groups.google.com/d/msg/wordnik-api/EC3rGajE0os/ey_5dBI_jWcJ In changing the models, I also had to change the swagger_model.py processor so it can handle the type discriminator and subtyping. I took that a big step forward, and using that information to generate an ari_model module, which can validate a JSON object against the Swagger model. The REST and WebSocket generators were changed to take advantage of the validators. If compiled with AST_DEVMODE enabled, JSON objects that don't match their corresponding models will not be sent out. For REST API calls, a 500 Internal Server response is sent. For WebSockets, the invalid JSON message is replaced with an error message. Since this took over about half of the job of the existing JSON generators, and the .to_json virtual function on messages took over the other half, I reluctantly removed the generators. The validators turned up all sorts of errors and inconsistencies in our data models, and the code. These were cleaned up, with checks in the code generator avoid some of the consistency problems in the future. * The model for a channel snapshot was trimmed down to match the information sent via AMI. Many of the field being sent were not useful in the general case. * The model for a bridge snapshot was updated to be more consistent with the other ARI models. Another impact of introducing subtyping was that the swagger-codegen documentation generator was insufficient (at least until it catches up with Swagger 1.2). I wanted it to be easier to generate docs for the API anyways, so I ported the wiki pages to use the Asterisk Swagger generator. In the process, I was able to clean up many of the model links, which would occasionally give inconsistent results on the wiki. I also added error responses to the wiki docs, making the wiki documentation more complete. Finally, since Stasis-HTTP will now be named Asterisk REST Interface (ARI), any new functions and files I created carry the ari_ prefix. I changed a few stasis_http references to ari where it was non-intrusive and made sense. (closes issue ASTERISK-21885) Review: https://reviewboard.asterisk.org/r/2639/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@393529 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-07-03 16:32:41 +00:00
model.description_dox = model.description.replace('\n', '\n * ')
model.description_dox = re.sub(' *\n', '\n', model.description_dox)
model.wiki_description = wikify(model.description)
model.c_id = snakify(model.id)
return model
Update events to use Swagger 1.3 subtyping, and related aftermath This patch started with the simple idea of changing the /events data model to be more sane. The original model would send out events like: { "stasis_start": { "args": [], "channel": { ... } } } The event discriminator was the field name instead of being a value in the object, due to limitations in how Swagger 1.1 could model objects. While technically sufficient in communicating event information, it was really difficult to deal with in terms of client side JSON handling. This patch takes advantage of a proposed extension[1] to Swagger which allows type variance through the use of a discriminator field. This had a domino effect that made this a surprisingly large patch. [1]: https://groups.google.com/d/msg/wordnik-api/EC3rGajE0os/ey_5dBI_jWcJ In changing the models, I also had to change the swagger_model.py processor so it can handle the type discriminator and subtyping. I took that a big step forward, and using that information to generate an ari_model module, which can validate a JSON object against the Swagger model. The REST and WebSocket generators were changed to take advantage of the validators. If compiled with AST_DEVMODE enabled, JSON objects that don't match their corresponding models will not be sent out. For REST API calls, a 500 Internal Server response is sent. For WebSockets, the invalid JSON message is replaced with an error message. Since this took over about half of the job of the existing JSON generators, and the .to_json virtual function on messages took over the other half, I reluctantly removed the generators. The validators turned up all sorts of errors and inconsistencies in our data models, and the code. These were cleaned up, with checks in the code generator avoid some of the consistency problems in the future. * The model for a channel snapshot was trimmed down to match the information sent via AMI. Many of the field being sent were not useful in the general case. * The model for a bridge snapshot was updated to be more consistent with the other ARI models. Another impact of introducing subtyping was that the swagger-codegen documentation generator was insufficient (at least until it catches up with Swagger 1.2). I wanted it to be easier to generate docs for the API anyways, so I ported the wiki pages to use the Asterisk Swagger generator. In the process, I was able to clean up many of the model links, which would occasionally give inconsistent results on the wiki. I also added error responses to the wiki docs, making the wiki documentation more complete. Finally, since Stasis-HTTP will now be named Asterisk REST Interface (ARI), any new functions and files I created carry the ari_ prefix. I changed a few stasis_http references to ari where it was non-intrusive and made sense. (closes issue ASTERISK-21885) Review: https://reviewboard.asterisk.org/r/2639/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@393529 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-07-03 16:32:41 +00:00
def process_property(self, prop, context):
if "-" in prop.name:
raise SwaggerError("Property names cannot have dashes", context)
if prop.name != prop.name.lower():
raise SwaggerError("Property name should be all lowercase",
context)
prop.wiki_description = wikify(prop.description)
Update events to use Swagger 1.3 subtyping, and related aftermath This patch started with the simple idea of changing the /events data model to be more sane. The original model would send out events like: { "stasis_start": { "args": [], "channel": { ... } } } The event discriminator was the field name instead of being a value in the object, due to limitations in how Swagger 1.1 could model objects. While technically sufficient in communicating event information, it was really difficult to deal with in terms of client side JSON handling. This patch takes advantage of a proposed extension[1] to Swagger which allows type variance through the use of a discriminator field. This had a domino effect that made this a surprisingly large patch. [1]: https://groups.google.com/d/msg/wordnik-api/EC3rGajE0os/ey_5dBI_jWcJ In changing the models, I also had to change the swagger_model.py processor so it can handle the type discriminator and subtyping. I took that a big step forward, and using that information to generate an ari_model module, which can validate a JSON object against the Swagger model. The REST and WebSocket generators were changed to take advantage of the validators. If compiled with AST_DEVMODE enabled, JSON objects that don't match their corresponding models will not be sent out. For REST API calls, a 500 Internal Server response is sent. For WebSockets, the invalid JSON message is replaced with an error message. Since this took over about half of the job of the existing JSON generators, and the .to_json virtual function on messages took over the other half, I reluctantly removed the generators. The validators turned up all sorts of errors and inconsistencies in our data models, and the code. These were cleaned up, with checks in the code generator avoid some of the consistency problems in the future. * The model for a channel snapshot was trimmed down to match the information sent via AMI. Many of the field being sent were not useful in the general case. * The model for a bridge snapshot was updated to be more consistent with the other ARI models. Another impact of introducing subtyping was that the swagger-codegen documentation generator was insufficient (at least until it catches up with Swagger 1.2). I wanted it to be easier to generate docs for the API anyways, so I ported the wiki pages to use the Asterisk Swagger generator. In the process, I was able to clean up many of the model links, which would occasionally give inconsistent results on the wiki. I also added error responses to the wiki docs, making the wiki documentation more complete. Finally, since Stasis-HTTP will now be named Asterisk REST Interface (ARI), any new functions and files I created carry the ari_ prefix. I changed a few stasis_http references to ari where it was non-intrusive and made sense. (closes issue ASTERISK-21885) Review: https://reviewboard.asterisk.org/r/2639/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@393529 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-07-03 16:32:41 +00:00
def process_type(self, swagger_type, context):
swagger_type.c_name = snakify(swagger_type.name)
swagger_type.c_singular_name = snakify(swagger_type.singular_name)
swagger_type.wiki_name = wikify(swagger_type.name)