You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
567 lines
22 KiB
Python
567 lines
22 KiB
Python
"""
|
|
Contains all the units for the spec.
|
|
|
|
This file loads swagger and JSON schema files and parses out the useful bits
|
|
and returns them as Units for use in Batesian.
|
|
|
|
For the actual conversion of data -> RST (including templates), see the sections
|
|
file instead.
|
|
"""
|
|
from batesian.units import Units
|
|
import inspect
|
|
import json
|
|
import os
|
|
import re
|
|
import subprocess
|
|
import urllib
|
|
import yaml
|
|
|
|
V1_CLIENT_API = "../api/client-server/v1"
|
|
V1_EVENT_EXAMPLES = "../event-schemas/examples/v1"
|
|
V1_EVENT_SCHEMA = "../event-schemas/schema/v1"
|
|
CORE_EVENT_SCHEMA = "../event-schemas/schema/v1/core-event-schema"
|
|
CHANGELOG = "../CHANGELOG.rst"
|
|
TARGETS = "../specification/targets.yaml"
|
|
|
|
ROOM_EVENT = "core-event-schema/room_event.json"
|
|
STATE_EVENT = "core-event-schema/state_event.json"
|
|
|
|
|
|
def get_json_schema_object_fields(obj, enforce_title=False):
|
|
# Algorithm:
|
|
# f.e. property => add field info (if field is object then recurse)
|
|
if obj.get("type") != "object":
|
|
raise Exception(
|
|
"get_json_schema_object_fields: Object %s isn't an object." % obj
|
|
)
|
|
if enforce_title and not obj.get("title"):
|
|
raise Exception(
|
|
"get_json_schema_object_fields: Nested object %s doesn't have a title." % obj
|
|
)
|
|
|
|
required_keys = obj.get("required")
|
|
if not required_keys:
|
|
required_keys = []
|
|
|
|
fields = {
|
|
"title": obj.get("title"),
|
|
"rows": []
|
|
}
|
|
tables = [fields]
|
|
|
|
props = obj.get("properties", obj.get("patternProperties"))
|
|
parents = obj.get("allOf")
|
|
if not props and not parents:
|
|
raise Exception(
|
|
"Object %s has no properties or parents." % obj
|
|
)
|
|
if not props: # parents only
|
|
return [{
|
|
"title": obj["title"],
|
|
"parent": parents[0]["$ref"],
|
|
"no-table": True
|
|
}]
|
|
|
|
for key_name in sorted(props):
|
|
value_type = None
|
|
required = key_name in required_keys
|
|
desc = props[key_name].get("description", "")
|
|
|
|
if props[key_name]["type"] == "object":
|
|
if props[key_name].get("additionalProperties"):
|
|
# not "really" an object, just a KV store
|
|
value_type = (
|
|
"{string: %s}" %
|
|
props[key_name]["additionalProperties"]["type"]
|
|
)
|
|
else:
|
|
nested_object = get_json_schema_object_fields(
|
|
props[key_name],
|
|
enforce_title=True
|
|
)
|
|
value_type = "{%s}" % nested_object[0]["title"]
|
|
|
|
if not nested_object[0].get("no-table"):
|
|
tables += nested_object
|
|
elif props[key_name]["type"] == "array":
|
|
# if the items of the array are objects then recurse
|
|
if props[key_name]["items"]["type"] == "object":
|
|
nested_object = get_json_schema_object_fields(
|
|
props[key_name]["items"],
|
|
enforce_title=True
|
|
)
|
|
value_type = "[%s]" % nested_object[0]["title"]
|
|
tables += nested_object
|
|
else:
|
|
value_type = "[%s]" % props[key_name]["items"]["type"]
|
|
else:
|
|
value_type = props[key_name]["type"]
|
|
if props[key_name].get("enum"):
|
|
if len(props[key_name].get("enum")) > 1:
|
|
value_type = "enum"
|
|
desc += (
|
|
" One of: %s" % json.dumps(props[key_name]["enum"])
|
|
)
|
|
else:
|
|
desc += (
|
|
" Must be '%s'." % props[key_name]["enum"][0]
|
|
)
|
|
if isinstance(value_type, list):
|
|
value_type = " or ".join(value_type)
|
|
|
|
fields["rows"].append({
|
|
"key": key_name,
|
|
"type": value_type,
|
|
"required": required,
|
|
"desc": desc,
|
|
"req_str": "**Required.** " if required else ""
|
|
})
|
|
return tables
|
|
|
|
|
|
class MatrixUnits(Units):
|
|
|
|
def _load_swagger_meta(self, api, group_name):
|
|
endpoints = []
|
|
for path in api["paths"]:
|
|
for method in api["paths"][path]:
|
|
single_api = api["paths"][path][method]
|
|
full_path = api.get("basePath", "") + path
|
|
endpoint = {
|
|
"title": single_api.get("summary", ""),
|
|
"desc": single_api.get("description", single_api.get("summary", "")),
|
|
"method": method.upper(),
|
|
"path": full_path,
|
|
"requires_auth": "security" in single_api,
|
|
"rate_limited": 429 in single_api.get("responses", {}),
|
|
"req_params": [],
|
|
"res_tables": [],
|
|
"example": {
|
|
"req": "",
|
|
"responses": [],
|
|
"good_response": ""
|
|
}
|
|
}
|
|
self.log(" ------- Endpoint: %s %s ------- " % (method, path))
|
|
for param in single_api.get("parameters", []):
|
|
# description
|
|
desc = param.get("description", "")
|
|
if param.get("required"):
|
|
desc = "**Required.** " + desc
|
|
|
|
# assign value expected for this param
|
|
val_type = param.get("type") # integer/string
|
|
refType = Units.prop(param, "schema/$ref/") # Error,Event
|
|
schemaFmt = Units.prop(param, "schema/format") # bytes e.g. uploads
|
|
if not val_type and refType:
|
|
val_type = refType # TODO: Resolve to human-readable.
|
|
if not val_type and schemaFmt:
|
|
val_type = schemaFmt
|
|
if val_type:
|
|
endpoint["req_params"].append({
|
|
"key": param["name"],
|
|
"loc": param["in"],
|
|
"type": val_type,
|
|
"desc": desc
|
|
})
|
|
continue
|
|
# If we're here, either the param has no value or it is an
|
|
# object which we haven't $reffed (so probably just a json
|
|
# object with some keys; we'll add entries f.e one)
|
|
if "schema" not in param:
|
|
raise Exception(
|
|
("API endpoint group=%s path=%s method=%s param=%s"+
|
|
" has no valid parameter value.") % (
|
|
group_name, path, method, param
|
|
)
|
|
)
|
|
if Units.prop(param, "schema/type") != "object":
|
|
raise Exception(
|
|
("API endpoint group=%s path=%s method=%s defines a"+
|
|
" param with a schema which isn't an object. Array?")
|
|
% (group_name, path, method)
|
|
)
|
|
# loop top-level json keys
|
|
json_body = Units.prop(param, "schema/properties")
|
|
for key in json_body:
|
|
endpoint["req_params"].append({
|
|
"key": key,
|
|
"loc": "JSON body",
|
|
"type": json_body[key]["type"],
|
|
"desc": json_body[key]["description"]
|
|
})
|
|
# endfor[param]
|
|
for row in endpoint["req_params"]:
|
|
self.log("Request parameter: %s" % row)
|
|
|
|
# group params by location to ease templating
|
|
endpoint["req_param_by_loc"] = {
|
|
# path: [...], query: [...], body: [...]
|
|
}
|
|
for p in endpoint["req_params"]:
|
|
if p["loc"] not in endpoint["req_param_by_loc"]:
|
|
endpoint["req_param_by_loc"][p["loc"]] = []
|
|
endpoint["req_param_by_loc"][p["loc"]].append(p)
|
|
|
|
good_response = None
|
|
for code, res in single_api.get("responses", {}).items():
|
|
if not good_response and code == 200:
|
|
good_response = res
|
|
description = res.get("description", "")
|
|
example = res.get("examples", {}).get("application/json", "")
|
|
if description and example:
|
|
endpoint["example"]["responses"].append({
|
|
"code": code,
|
|
"description": description,
|
|
"example": example,
|
|
})
|
|
|
|
# form example request if it has one. It "has one" if all params
|
|
# have either "x-example" or a "schema" with an "example".
|
|
params_missing_examples = [
|
|
p for p in single_api.get("parameters", []) if (
|
|
"x-example" not in p and
|
|
not Units.prop(p, "schema/example")
|
|
)
|
|
]
|
|
if len(params_missing_examples) == 0:
|
|
path_template = api.get("basePath", "") + path
|
|
qps = {}
|
|
body = ""
|
|
for param in single_api.get("parameters", []):
|
|
if param["in"] == "path":
|
|
path_template = path_template.replace(
|
|
"{%s}" % param["name"], urllib.quote(
|
|
param["x-example"]
|
|
)
|
|
)
|
|
elif param["in"] == "body":
|
|
body = param["schema"]["example"]
|
|
elif param["in"] == "query":
|
|
qps[param["name"]] = param["x-example"]
|
|
query_string = "" if len(qps) == 0 else "?"+urllib.urlencode(qps)
|
|
endpoint["example"]["req"] = "%s %s%s\n%s" % (
|
|
method.upper(), path_template, query_string, body
|
|
)
|
|
else:
|
|
self.log(
|
|
"The following parameters are missing examples :( \n %s" %
|
|
[ p["name"] for p in params_missing_examples ]
|
|
)
|
|
|
|
# add response params if this API has any.
|
|
if good_response:
|
|
self.log("Found a 200 response for this API")
|
|
res_type = Units.prop(good_response, "schema/type")
|
|
if res_type and res_type not in ["object", "array"]:
|
|
# response is a raw string or something like that
|
|
endpoint["res_tables"].append({
|
|
"title": None,
|
|
"rows": [{
|
|
"key": good_response["schema"].get("name", ""),
|
|
"type": res_type,
|
|
"desc": res.get("description", ""),
|
|
"req_str": ""
|
|
}]
|
|
})
|
|
elif res_type and Units.prop(good_response, "schema/properties"):
|
|
# response is an object:
|
|
schema = good_response["schema"]
|
|
res_tables = get_json_schema_object_fields(schema)
|
|
for table in res_tables:
|
|
if "no-table" not in table:
|
|
endpoint["res_tables"].append(table)
|
|
elif res_type and Units.prop(good_response, "schema/items"):
|
|
# response is an array:
|
|
# FIXME: Doesn't recurse at all.
|
|
schema = good_response["schema"]
|
|
array_type = Units.prop(schema, "items/type")
|
|
if Units.prop(schema, "items/allOf"):
|
|
array_type = (
|
|
Units.prop(schema, "items/title")
|
|
)
|
|
endpoint["res_tables"].append({
|
|
"title": schema.get("title", ""),
|
|
"rows": [{
|
|
"key": "N/A",
|
|
"type": ("[%s]" % array_type),
|
|
"desc": schema.get("description", ""),
|
|
"req_str": ""
|
|
}]
|
|
})
|
|
|
|
for response_table in endpoint["res_tables"]:
|
|
self.log("Response: %s" % response_table["title"])
|
|
for r in response_table["rows"]:
|
|
self.log("Row: %s" % r)
|
|
if len(endpoint["res_tables"]) == 0:
|
|
self.log(
|
|
"This API appears to have no response table. Are you " +
|
|
"sure this API returns no parameters?"
|
|
)
|
|
|
|
endpoints.append(endpoint)
|
|
|
|
aliases = single_api.get("x-alias", None)
|
|
if aliases:
|
|
alias_link = aliases["canonical-link"]
|
|
for alias in aliases["aliases"]:
|
|
endpoints.append({
|
|
"method": method.upper(),
|
|
"path": alias,
|
|
"alias_for_path": full_path,
|
|
"alias_link": alias_link
|
|
})
|
|
|
|
return {
|
|
"base": api.get("basePath"),
|
|
"group": group_name,
|
|
"endpoints": endpoints,
|
|
}
|
|
|
|
def load_swagger_apis(self):
|
|
path = V1_CLIENT_API
|
|
apis = {}
|
|
for filename in os.listdir(path):
|
|
if not filename.endswith(".yaml"):
|
|
continue
|
|
self.log("Reading swagger API: %s" % filename)
|
|
with open(os.path.join(path, filename), "r") as f:
|
|
# strip .yaml
|
|
group_name = filename[:-5]
|
|
api = yaml.load(f.read())
|
|
api["__meta"] = self._load_swagger_meta(api, group_name)
|
|
apis[group_name] = api
|
|
return apis
|
|
|
|
def load_common_event_fields(self):
|
|
path = CORE_EVENT_SCHEMA
|
|
event_types = {}
|
|
|
|
for (root, dirs, files) in os.walk(path):
|
|
for filename in files:
|
|
if not filename.endswith(".json"):
|
|
continue
|
|
|
|
event_type = filename[:-5] # strip the ".json"
|
|
filepath = os.path.join(root, filename)
|
|
with open(filepath) as f:
|
|
try:
|
|
event_info = json.load(f)
|
|
except Exception as e:
|
|
raise ValueError(
|
|
"Error reading file %r" % (filepath,), e
|
|
)
|
|
|
|
if "event" not in event_type:
|
|
continue # filter ImageInfo and co
|
|
|
|
table = {
|
|
"title": event_info["title"],
|
|
"desc": event_info["description"],
|
|
"rows": []
|
|
}
|
|
|
|
for prop in sorted(event_info["properties"]):
|
|
row = {
|
|
"key": prop,
|
|
"type": event_info["properties"][prop]["type"],
|
|
"desc": event_info["properties"][prop].get("description","")
|
|
}
|
|
table["rows"].append(row)
|
|
|
|
event_types[event_type] = table
|
|
return event_types
|
|
|
|
def load_event_examples(self):
|
|
path = V1_EVENT_EXAMPLES
|
|
examples = {}
|
|
for filename in os.listdir(path):
|
|
if not filename.startswith("m."):
|
|
continue
|
|
with open(os.path.join(path, filename), "r") as f:
|
|
examples[filename] = json.loads(f.read())
|
|
if filename == "m.room.message#m.text":
|
|
examples["m.room.message"] = examples[filename]
|
|
return examples
|
|
|
|
def load_event_schemas(self):
|
|
path = V1_EVENT_SCHEMA
|
|
schemata = {}
|
|
|
|
for filename in os.listdir(path):
|
|
if not filename.startswith("m."):
|
|
continue
|
|
self.log("Reading %s" % os.path.join(path, filename))
|
|
with open(os.path.join(path, filename), "r") as f:
|
|
json_schema = json.loads(f.read())
|
|
schema = {
|
|
"typeof": None,
|
|
"typeof_info": "",
|
|
"type": None,
|
|
"title": None,
|
|
"desc": None,
|
|
"msgtype": None,
|
|
"content_fields": [
|
|
# {
|
|
# title: "<title> key"
|
|
# rows: [
|
|
# { key: <key_name>, type: <string>,
|
|
# desc: <desc>, required: <bool> }
|
|
# ]
|
|
# }
|
|
]
|
|
}
|
|
|
|
# add typeof
|
|
base_defs = {
|
|
ROOM_EVENT: "Message Event",
|
|
STATE_EVENT: "State Event"
|
|
}
|
|
if type(json_schema.get("allOf")) == list:
|
|
schema["typeof"] = base_defs.get(
|
|
json_schema["allOf"][0].get("$ref")
|
|
)
|
|
elif json_schema.get("title"):
|
|
schema["typeof"] = json_schema["title"]
|
|
|
|
# add type
|
|
schema["type"] = Units.prop(
|
|
json_schema, "properties/type/enum"
|
|
)[0]
|
|
|
|
# add summary and desc
|
|
schema["title"] = json_schema.get("title")
|
|
schema["desc"] = json_schema.get("description", "")
|
|
|
|
# walk the object for field info
|
|
schema["content_fields"] = get_json_schema_object_fields(
|
|
Units.prop(json_schema, "properties/content")
|
|
)
|
|
|
|
# grab msgtype if it is the right kind of event
|
|
msgtype = Units.prop(
|
|
json_schema, "properties/content/properties/msgtype/enum"
|
|
)
|
|
if msgtype:
|
|
schema["msgtype"] = msgtype[0] # enum prop
|
|
|
|
# link to msgtypes for m.room.message
|
|
if schema["type"] == "m.room.message" and not msgtype:
|
|
schema["desc"] += (
|
|
" For more information on ``msgtypes``, see "+
|
|
"`m.room.message msgtypes`_."
|
|
)
|
|
|
|
# Assign state key info if it has some
|
|
if schema["typeof"] == "State Event":
|
|
skey_desc = Units.prop(
|
|
json_schema, "properties/state_key/description"
|
|
)
|
|
if not skey_desc:
|
|
raise Exception("Missing description for state_key")
|
|
schema["typeof_info"] = "``state_key``: %s" % skey_desc
|
|
|
|
schemata[filename] = schema
|
|
return schemata
|
|
|
|
def load_spec_meta(self):
|
|
path = CHANGELOG
|
|
title_part = None
|
|
version = None
|
|
changelog_lines = []
|
|
with open(path, "r") as f:
|
|
prev_line = None
|
|
for line in f.readlines():
|
|
if line.strip().startswith(".. "):
|
|
continue # comment
|
|
if prev_line is None:
|
|
prev_line = line
|
|
continue
|
|
if not title_part:
|
|
# find the title underline (at least 3 =)
|
|
if re.match("^[=]{3,}$", line.strip()):
|
|
title_part = prev_line
|
|
continue
|
|
prev_line = line
|
|
else: # have title, get body (stop on next title or EOF)
|
|
if re.match("^[=]{3,}$", line.strip()):
|
|
# we added the title in the previous iteration, pop it
|
|
# then bail out.
|
|
changelog_lines.pop()
|
|
break
|
|
changelog_lines.append(line)
|
|
|
|
# parse out version from title
|
|
for word in title_part.split():
|
|
if re.match("^v[0-9\.]+$", word):
|
|
version = word[1:] # strip the 'v'
|
|
|
|
self.log("Version: %s Title part: %s Changelog line count: %s" % (
|
|
version, title_part, len(changelog_lines)
|
|
))
|
|
if not version or len(changelog_lines) == 0:
|
|
raise Exception("Failed to parse CHANGELOG.rst")
|
|
|
|
return {
|
|
"version": version,
|
|
"changelog": "".join(changelog_lines)
|
|
}
|
|
|
|
|
|
def load_spec_targets(self):
|
|
with open(TARGETS, "r") as f:
|
|
return yaml.load(f.read())
|
|
|
|
|
|
def load_git_version(self):
|
|
null = open(os.devnull, 'w')
|
|
cwd = os.path.dirname(os.path.abspath(__file__))
|
|
try:
|
|
git_branch = subprocess.check_output(
|
|
['git', 'rev-parse', '--abbrev-ref', 'HEAD'],
|
|
stderr=null,
|
|
cwd=cwd,
|
|
).strip()
|
|
except subprocess.CalledProcessError:
|
|
git_branch = ""
|
|
try:
|
|
git_tag = subprocess.check_output(
|
|
['git', 'describe', '--exact-match'],
|
|
stderr=null,
|
|
cwd=cwd,
|
|
).strip()
|
|
git_tag = "tag=" + git_tag
|
|
except subprocess.CalledProcessError:
|
|
git_tag = ""
|
|
try:
|
|
git_commit = subprocess.check_output(
|
|
['git', 'rev-parse', '--short', 'HEAD'],
|
|
stderr=null,
|
|
cwd=cwd,
|
|
).strip()
|
|
except subprocess.CalledProcessError:
|
|
git_commit = ""
|
|
try:
|
|
dirty_string = "-this_is_a_dirty_checkout"
|
|
is_dirty = subprocess.check_output(
|
|
['git', 'describe', '--dirty=' + dirty_string, "--all"],
|
|
stderr=null,
|
|
cwd=cwd,
|
|
).strip().endswith(dirty_string)
|
|
git_dirty = "dirty" if is_dirty else ""
|
|
except subprocess.CalledProcessError:
|
|
git_dirty = ""
|
|
|
|
git_version = "Unknown"
|
|
if git_branch or git_tag or git_commit or git_dirty:
|
|
git_version = ",".join(
|
|
s for s in
|
|
(git_branch, git_tag, git_commit, git_dirty,)
|
|
if s
|
|
).encode("ascii")
|
|
return {
|
|
"string": git_version,
|
|
"revision": git_commit
|
|
}
|