diff --git a/.gitignore b/.gitignore
index e75beff..4d0f3f4 100644
--- a/.gitignore
+++ b/.gitignore
@@ -66,3 +66,12 @@ profile
# AppCode
.idea
+
+# dist
+/dist/
+
+# MANIFEST
+MANIFEST
+
+#local
+local.properties
diff --git a/LICENSE b/LICENSE
new file mode 100755
index 0000000..0d8fbcb
--- /dev/null
+++ b/LICENSE
@@ -0,0 +1,28 @@
+软件名称：神策分析 SDK
+版本号：所有版本
+许可协议版本：1.0
+
+1. 商业许可协议（用于商业用途需购买许可）
+任何商业用途必须获得商业许可。
+
+商业许可协议条款：
+
+- 商业用途：任何直接或间接产生收入的用途都需要购买商业许可。
+- 付款条款：在使用本软件用于商业用途之前，您必须支付全额许可费用。具体的付款方式将在双方联系后提供。
+- 商业支持：购买商业许可后，您将获得一年的技术支持和软件更新服务。
+- 禁止再许可：商业用户不得再许可、转售或转让本软件。每份商业许可仅适用于单一实体或公司。
+- 源代码访问：购买商业许可的用户将获得本软件的代码访问权限，并可根据业务需求进行内部修改。但不得公开发布或再分发修改后的版本。
+- 使用范围限制：商业许可仅限于购买者的内部使用，不得与第三方共享或用于为第三方提供服务。任何超出许可范围的使用行为均需额外授权，并可能产生额外费用。
+- 联系信息：如需购买商业许可，请联系 dv@sensorsdata.com。
+- 知识产权声明：本软件的版权归神策网络科技（北京）有限公司所有。购买商业许可仅授予您使用权，所有权仍归属本公司。
+- 终止条款： 如果您未支付相关费用或违反本协议的任何条款，商业许可将自动终止。您必须立即停止所有商业用途，并销毁或删除所有软件副本。
+
+2. 附加授权规则条款
+授权规则条款：
+
+- 功能限制：未经本软件作者的明确书面许可，您不得移除、绕过或规避本软件中的任何功能限制或试用限制。
+- 商标使用：未经授权，您不得在宣传、市场推广或销售产品时使用本软件的名称、商标或品牌标识。任何商标使用必须得到明确的书面许可。
+- 修改条款：本协议的条款可能会不时更新，用户有责任定期检查最新版本。任何重大更改将通过项目主页或电子邮件通知用户。
+
+3. 联系方式
+如需更多信息或申请商业许可，请联系 dv@sensorsdata.com。
diff --git a/LICENSE.txt b/LICENSE.txt
deleted file mode 100644
index 8b48d57..0000000
--- a/LICENSE.txt
+++ /dev/null
@@ -1,13 +0,0 @@
-Copyright 2015 SensorsData Inc.
-
-Licensed under the Apache License, Version 2.0 (the "License");
-you may not use this file except in compliance with the License.
-You may obtain a copy of the License at
-
- http://www.apache.org/licenses/LICENSE-2.0
-
-Unless required by applicable law or agreed to in writing, software
-distributed under the License is distributed on an "AS IS" BASIS,
-WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-See the License for the specific language governing permissions and
-limitations under the License.
diff --git a/README.md b/README.md
index 0a2cfad..06fb0cc 100644
--- a/README.md
+++ b/README.md
@@ -1,32 +1,39 @@
-# Sensors Analytics
+
-This is the official Python SDK for Sensors Analytics.
+[](https://github.com/sensorsdata/sa-sdk-c/blob/master/LICENSE)
+[](https://github.com/sensorsdata/sa-sdk-c/releases)
+[](https://github.com/sensorsdata/sa-sdk-c/releases)
-## Easy Installation
+## 神策简介
-You can get Sensors Analytics SDK using pip.
+[**神策数据**](https://www.sensorsdata.cn/)
+（Sensors Data），隶属于神策网络科技（北京）有限公司，是一家专业的大数据分析服务公司，大数据分析行业开拓者，为客户提供深度用户行为分析平台、以及专业的咨询服务和行业解决方案，致力于帮助客户实现数据驱动。神策数据立足大数据及用户行为分析的技术与实践前沿，业务现已覆盖以互联网、金融、零售快消、高科技、制造等为代表的十多个主要行业、并可支持企业多个职能部门。公司总部在北京，并在上海、深圳、合肥、武汉等地拥有本地化的服务团队，覆盖东区及南区市场；公司拥有专业的服务团队，为客户提供一对一的客户服务。公司在大数据领域积累的核心关键技术，包括在海量数据采集、存储、清洗、分析挖掘、可视化、智能应用、安全与隐私保护等领域。 [**More**](https://www.sensorsdata.cn/about/aboutus.html)
-```
- pip install SensorsAnalyticsSDK
-```
+## SDK 简介
-Once the SDK is successfully installed, use the Sensors Analytics SDK likes:
+SensorsAnalytics SDK 是国内第一家开源商用版用户行为采集 SDK，目前支持代码埋点、全埋点、App 点击图、可视化全埋点等。目前已累计有 1500 多家付费客户，2500+ 的 App 集成使用，作为 App 数据采集利器，致力于帮助客户挖掘更多的商业价值，为其精准运营和业务支撑提供了可靠的数据来源。其采集全面而灵活、性能良好，并一直保持稳定的迭代，经受住了时间和客户的考验。
-```python
- import sensorsanalytics
+## 神策埋点 SDK 官网
+如需了解神策埋点 SDK 的更多商业授权信息，请访问[神策埋点 SDK 官网](https://jssdk.debugbox.sensorsdata.cn/)获取更多详细信息。
- // Gets the url of Sensors Analytics in the home page.
- SA_SERVER_URL = 'YOUR_SERVER_URL'
+## 联系我们
+若您有商业合作或产品集成需求，请通过下面的渠道联系我们获取专业服务与支持。
- // Initialized the Sensors Analytics SDK with Default Consumer
- consumer = sensorsanalytics.DefaultConsumer(SA_SERVER_URL)
- sa = sensorsanalytics.SensorsAnalytics(consumer)
+| 加微信号：skycode008，或扫码添加联系人 
| 扫码关注「神策埋点 SDK」公众号  |
+| ------ | ------ |
- // Track the event 'ServerStart'
- sa.track("ABCDEFG1234567", "ServerStart")
-```
+## 基本要求
+SDK 兼容 Python 2.6+ 和 Python3 3.X，不依赖第三方库。
-## To learn more
+## 集成文档
-See our [full manual](http://www.sensorsdata.cn/manual/python_sdk.html)
+请参考神策官网 [Python SDK 集成文档](https://manual.sensorsdata.cn/sa/latest/tech_sdk_server_python-1573931.html)。
+
+## 新书推荐
+
+| [《数据驱动：从方法到实践》](https://item.jd.com/12322322.html) | [《Android 全埋点解决方案》](https://item.jd.com/12574672.html) | [《iOS 全埋点解决方案》](https://item.jd.com/12867068.html)
+| ------ | ------ | ------ |
+
+## License
+[License 协议](https://github.com/sensorsdata/sa-sdk-python/blob/master/LICENSE)
diff --git a/README.txt b/README.txt
deleted file mode 100644
index 4fecb8c..0000000
--- a/README.txt
+++ /dev/null
@@ -1,14 +0,0 @@
-=====================
-Sensors Analytics SDK
-=====================
-
-Sensors Analytics SDK
-=====================
-
-This is the official Python SDK for Sensors Analytics.
-
-To Learn More
--------------
-
-See our `full manual `_.
-
diff --git a/sensorsanalytics/sdk.py b/sensorsanalytics/sdk.py
index 71d9310..75ab62b 100644
--- a/sensorsanalytics/sdk.py
+++ b/sensorsanalytics/sdk.py
@@ -13,6 +13,7 @@
import threading
import time
import traceback
+from enum import Enum
try:
from urllib.parse import urlparse
@@ -25,10 +26,13 @@
import urllib2
import urllib
-SDK_VERSION = '1.7.5'
+SDK_VERSION = '1.11.0'
+batch_consumer_lock = threading.RLock()
try:
isinstance("", basestring)
+
+
def is_str(s):
return isinstance(s, basestring)
except NameError:
@@ -36,6 +40,8 @@ def is_str(s):
return isinstance(s, str)
try:
isinstance(1, long)
+
+
def is_int(n):
return isinstance(n, int) or isinstance(n, long)
except NameError:
@@ -78,56 +84,60 @@ class SensorsAnalyticsDebugException(Exception):
if os.name == 'nt': # pragma: no cover
import msvcrt
+
def lock(file_):
try:
savepos = file_.tell()
-
+
file_.seek(0)
try:
msvcrt.locking(file_.fileno(), msvcrt.LK_LOCK, 1)
except IOError as e:
- raise SensorsAnalyticsFileLockException(e)
+ raise SensorsAnalyticsFileLockException(e)
finally:
if savepos:
file_.seek(savepos)
except IOError as e:
- raise SensorsAnalyticsFileLockException(e)
+ raise SensorsAnalyticsFileLockException(e)
+
def unlock(file_):
try:
savepos = file_.tell()
if savepos:
file_.seek(0)
-
+
try:
msvcrt.locking(file_.fileno(), msvcrt.LK_UNLCK, 1)
except IOError as e:
- raise SensorsAnalyticsFileLockException(e)
+ raise SensorsAnalyticsFileLockException(e)
finally:
if savepos:
file_.seek(savepos)
except IOError as e:
- raise SensorsAnalyticsFileLockException(e)
+ raise SensorsAnalyticsFileLockException(e)
elif os.name == 'posix': # pragma: no cover
import fcntl
+
def lock(file_):
try:
fcntl.flock(file_.fileno(), fcntl.LOCK_EX)
except IOError as e:
- raise SensorsAnalyticsFileLockException(e)
+ raise SensorsAnalyticsFileLockException(e)
+
def unlock(file_):
fcntl.flock(file_.fileno(), fcntl.LOCK_UN)
else:
- raise SensorsAnalyticsFileLockException("SensorsAnalytics SDK is defined for NT and POSIX system.")
+ raise SensorsAnalyticsFileLockException("SensorsAnalytics SDK is defined for NT and POSIX system.")
class SAFileLock(object):
-
+
def __init__(self, file_handler):
self._file_handler = file_handler
@@ -138,12 +148,15 @@ def __enter__(self):
def __exit__(self, t, v, tb):
unlock(self._file_handler)
+
class SensorsAnalytics(object):
"""
使用一个 SensorsAnalytics 的实例来进行数据发送。
"""
- NAME_PATTERN = re.compile(r"^((?!^distinct_id$|^original_id$|^time$|^properties$|^id$|^first_id$|^second_id$|^users$|^events$|^event$|^user_id$|^date$|^datetime$)[a-zA-Z_$][a-zA-Z\d_$]{0,99})$", re.I)
+ NAME_PATTERN = re.compile(
+ r"^((?!^distinct_id$|^original_id$|^time$|^properties$|^id$|^first_id$|^second_id$|^users$|^events$|^event$|^event_id$|^device_id$|^user_id$|^date$|^datetime$|^user_group|^user_tag)[a-zA-Z_$][a-zA-Z\d_$]{0,99})$",
+ re.I)
class DatetimeSerializer(json.JSONEncoder):
"""
@@ -154,8 +167,8 @@ def default(self, obj):
if isinstance(obj, datetime.datetime):
head_fmt = "%Y-%m-%d %H:%M:%S"
return "{main_part}.{ms_part}".format(
- main_part=obj.strftime(head_fmt),
- ms_part=int(obj.microsecond/1000))
+ main_part=obj.strftime(head_fmt),
+ ms_part=int(obj.microsecond / 1000))
elif isinstance(obj, datetime.date):
fmt = '%Y-%m-%d'
return obj.strftime(fmt)
@@ -179,7 +192,7 @@ def __init__(self, consumer=None, project_name=None, enable_time_free=False):
self._default_project_name = project_name
self._enable_time_free = enable_time_free
self._super_properties = {}
- self.clear_super_properties();
+ self.clear_super_properties()
@staticmethod
def _now():
@@ -189,7 +202,7 @@ def _now():
def _json_dumps(data):
return json.dumps(data, separators=(',', ':'), cls=SensorsAnalytics.DatetimeSerializer)
- def register_super_properties(self, super_properties):
+ def register_super_properties(self, super_properties):
"""
设置每个事件都带有的一些公共属性，当 track 的 properties 和 super properties 有相同的 key 时，将采用 track 的
@@ -213,12 +226,13 @@ def track(self, distinct_id, event_name, properties=None, is_login_id=False):
:param distinct_id: 用户的唯一标识
:param event_name: 事件名称
:param properties: 事件的属性
+ :param is_login_id 是否是登陆 ID
"""
- all_properties = self._super_properties.copy()
+ all_properties = self._super_properties.copy()
if properties:
all_properties.update(properties)
self._track_event('track', event_name, distinct_id, None, all_properties, is_login_id)
-
+
def track_signup(self, distinct_id, original_id, properties=None):
"""
这个接口是一个较为复杂的功能，请在使用前先阅读相关说明:http://www.sensorsdata.cn/manual/track_signup.html，
@@ -233,13 +247,48 @@ def track_signup(self, distinct_id, original_id, properties=None):
raise SensorsAnalyticsIllegalDataException("property [original_id] must not be empty")
if len(str(original_id)) > 255:
raise SensorsAnalyticsIllegalDataException("the max length of property [original_id] is 255")
-
- all_properties = self._super_properties.copy()
+
+ all_properties = self._super_properties.copy()
if properties:
all_properties.update(properties)
-
+
self._track_event('track_signup', '$SignUp', distinct_id, original_id, all_properties, False)
+ @staticmethod
+ def _normalize_properties(data):
+ if "properties" in data and data["properties"] is not None:
+ for key, value in data["properties"].items():
+ SensorsAnalytics._assert_key(key)
+ SensorsAnalytics._assert_value(value, key)
+
+ @staticmethod
+ def _assert_key(key):
+ if not is_str(key):
+ raise SensorsAnalyticsIllegalDataException("property key must be a str. [key=%s]" % str(key))
+ if len(key) > 255:
+ raise SensorsAnalyticsIllegalDataException(
+ "the max length of property key is 256. [key=%s]" % str(key))
+ if not SensorsAnalytics.NAME_PATTERN.match(key):
+ raise SensorsAnalyticsIllegalDataException(
+ "property key must be a valid variable name. [key=%s]" % str(key))
+
+ @staticmethod
+ def _assert_value(value, key=None):
+ if is_str(value) and len(value) > 8191:
+ raise SensorsAnalyticsIllegalDataException(
+ "the max length of property key is 8192. [key=%s]" % str(key))
+
+ if not is_str(value) and not is_int(value) and not isinstance(value, float) \
+ and not isinstance(value, datetime.datetime) and not isinstance(value, datetime.date) \
+ and not isinstance(value, list) and value is not None:
+ raise SensorsAnalyticsIllegalDataException(
+ "property value must be a str/int/float/datetime/date/list. [value=%s]" % type(value))
+ if isinstance(value, list):
+ for lvalue in value:
+ if not is_str(lvalue):
+ raise SensorsAnalyticsIllegalDataException(
+ "[list] property's value must be a str. [value=%s]" % type(lvalue))
+
@staticmethod
def _normalize_data(data):
# 检查 distinct_id
@@ -256,7 +305,7 @@ def _normalize_data(data):
ts = int(data['time'])
ts_num = len(str(ts))
if ts_num < 10 or ts_num > 13:
- raise SensorsAnalyticsIllegalDataException("property [time] must be a timestamp in microseconds")
+ raise SensorsAnalyticsIllegalDataException("property [time] must be a timestamp in microseconds")
if ts_num == 10:
ts *= 1000
@@ -264,47 +313,26 @@ def _normalize_data(data):
# 检查 Event Name
if 'event' in data and not SensorsAnalytics.NAME_PATTERN.match(data['event']):
- raise SensorsAnalyticsIllegalDataException("event name must be a valid variable name. [name=%s]" % data['event'])
+ raise SensorsAnalyticsIllegalDataException(
+ "event name must be a valid variable name. [name=%s]" % data['event'])
# 检查 Event Name
if 'project' in data and not SensorsAnalytics.NAME_PATTERN.match(data['project']):
- raise SensorsAnalyticsIllegalDataException("project name must be a valid variable name. [project=%s]" % data['project'])
+ raise SensorsAnalyticsIllegalDataException(
+ "project name must be a valid variable name. [project=%s]" % data['project'])
# 检查 properties
- if "properties" in data and data["properties"] is not None:
- for key, value in data["properties"].items():
- if not is_str(key):
- raise SensorsAnalyticsIllegalDataException("property key must be a str. [key=%s]" % str(key))
- if len(key) > 255:
- raise SensorsAnalyticsIllegalDataException("the max length of property key is 256. [key=%s]" % str(key))
- if not SensorsAnalytics.NAME_PATTERN.match(key):
- raise SensorsAnalyticsIllegalDataException(
- "property key must be a valid variable name. [key=%s]" % str(key))
-
- if is_str(value) and len(value) > 8191:
- raise SensorsAnalyticsIllegalDataException("the max length of property key is 8192. [key=%s]" % str(key))
-
- if not is_str(value) and not is_int(value) and not isinstance(value, float)\
- and not isinstance(value, datetime.datetime) and not isinstance(value, datetime.date)\
- and not isinstance(value, list) and value is not None:
- raise SensorsAnalyticsIllegalDataException(
- "property value must be a str/int/float/datetime/date/list. [value=%s]" % type(value))
- if isinstance(value, list):
- for lvalue in value:
- if not is_str(lvalue):
- raise SensorsAnalyticsIllegalDataException(
- "[list] property's value must be a str. [value=%s]" % type(lvalue))
-
+ SensorsAnalytics._normalize_properties(data)
return data
def _get_lib_properties(self):
lib_properties = {
- '$lib' : 'python',
- '$lib_version' : SDK_VERSION,
- '$lib_method' : 'code',
+ '$lib': 'python',
+ '$lib_version': SDK_VERSION,
+ '$lib_method': 'code',
}
- if '$app_version' in self._super_properties:
+ if '$app_version' in self._super_properties:
lib_properties['$app_version'] = self._super_properties['$app_version']
try:
@@ -315,21 +343,22 @@ def _get_lib_properties(self):
try:
file_name = trace[-4][0]
line_number = trace[-4][1]
-
+
if trace[-4][2].startswith('<'):
function_name = ''
else:
function_name = trace[-4][2]
-
+
try:
if len(trace) > 4 and trace[-5][3]:
- class_name = trace[-5][3].split('(')[0]
+ class_name = trace[-5][3].split('(')[0]
else:
class_name = ''
except:
- print(trace.format())
+ print(trace.format())
- lib_properties['$lib_detail'] = '%s##%s##%s##%s' % (class_name, function_name, file_name, line_number)
+ lib_properties['$lib_detail'] = '%s##%s##%s##%s' % (
+ class_name, function_name, file_name, line_number)
except:
pass
@@ -349,7 +378,6 @@ def _get_common_properties(self):
return common_properties
-
@staticmethod
def _extract_user_time(properties):
"""
@@ -361,12 +389,35 @@ def _extract_user_time(properties):
return t
return None
+ @staticmethod
+ def _extract_token(properties):
+ """
+ 如果用户传入了 $token 字段，则在 properties 外层加上token，并删除 $token 字段
+ """
+ if properties is not None and '$token' in properties:
+ t = properties['$token']
+ del (properties['$token'])
+ return t
+ return None
+
+ @staticmethod
+ def _extract_project(properties):
+ """
+ 如果用户传入了 $project 字段，则在 properties 外层加上 project，并删除 $project 字段
+ """
+ if properties is not None and '$project' in properties:
+ t = properties['$project']
+ del (properties['$project'])
+ return t
+ return None
+
def profile_set(self, distinct_id, profiles, is_login_id=False):
"""
直接设置一个用户的 Profile，如果已存在则覆盖
:param distinct_id: 用户的唯一标识
:param profiles: 用户属性
+ :param is_login_id: distinct_id 是否是登陆 id
"""
return self._track_event('profile_set', None, distinct_id, None, profiles, is_login_id)
@@ -376,8 +427,9 @@ def profile_set_once(self, distinct_id, profiles, is_login_id=False):
:param distinct_id: 用户的唯一标识
:param profiles: 用户属性
+ :param is_login_id: distinct_id 是否是登陆 id
"""
- return self._track_event('profile_set_once', None, distinct_id, None, profiles, is_login_id)
+ return self._track_event('profile_set_once', None, distinct_id, None, profiles, is_login_id)
def profile_increment(self, distinct_id, profiles, is_login_id=False):
"""
@@ -385,7 +437,14 @@ def profile_increment(self, distinct_id, profiles, is_login_id=False):
:param distinct_id: 用户的唯一标识
:param profiles: 用户属性
+ :param is_login_id: distinct_id 是否是登陆 id
"""
+ if isinstance(profiles, dict):
+ for key, value in profiles.items():
+ if not is_int(value):
+ raise SensorsAnalyticsIllegalDataException("property value must be Number. [key=%s]" % str(key))
+ else:
+ raise SensorsAnalyticsIllegalDataException("profiles must be dict type.")
return self._track_event('profile_increment', None, distinct_id, None, profiles, is_login_id)
def profile_append(self, distinct_id, profiles, is_login_id=False):
@@ -394,7 +453,14 @@ def profile_append(self, distinct_id, profiles, is_login_id=False):
:param distinct_id: 用户的唯一标识
:param profiles: 用户属性
+ :param is_login_id: distinct_id 是否是登陆 id
"""
+ if isinstance(profiles, dict):
+ for key, value in profiles.items():
+ if not isinstance(value,list):
+ raise SensorsAnalyticsIllegalDataException("property value must be list. [key=%s]" % str(key))
+ else:
+ raise SensorsAnalyticsIllegalDataException("profiles must be dict type.")
return self._track_event('profile_append', None, distinct_id, None, profiles, is_login_id)
def profile_unset(self, distinct_id, profile_keys, is_login_id=False):
@@ -403,6 +469,7 @@ def profile_unset(self, distinct_id, profile_keys, is_login_id=False):
:param distinct_id: 用户的唯一标识
:param profile_keys: 用户属性键值列表
+ :param is_login_id: distinct_id 是否是登陆 id
"""
if isinstance(profile_keys, list):
profile_keys = dict((key, True) for key in profile_keys)
@@ -413,11 +480,74 @@ def profile_delete(self, distinct_id, is_login_id=False):
删除整个用户的信息。
:param distinct_id: 用户的唯一标识
+ :param is_login_id: distinct_id 是否是登陆 id
"""
return self._track_event('profile_delete', None, distinct_id, None, {}, is_login_id)
- def _track_event(self, event_type, event_name, distinct_id, original_id, properties, is_login_id):
+ def item_set(self, item_type, item_id, properties=None):
+ """
+ 直接设置一个物品，如果已存在则覆盖。
+
+ :param item_type: 物品类型
+ :param item_id: 物品的唯一标识
+ :param properties: 物品属性
+ """
+ return self._track_item('item_set', item_type, item_id, properties)
+
+ def item_delete(self, item_type, item_id, properties=None):
+ """
+ 删除一个物品。
+
+ :param item_type: 物品类型
+ :param item_id: 物品的唯一标识
+ :param properties: 物品属性
+ """
+ return self._track_item('item_delete', item_type, item_id, properties)
+
+ @staticmethod
+ def _normalize_item_data(data):
+ # 检查 item_type
+ if not SensorsAnalytics.NAME_PATTERN.match(data['item_type']):
+ raise SensorsAnalyticsIllegalDataException(
+ "item_type must be a valid variable name. [key=%s]" % str(data['item_type']))
+
+ # 检查 item_id
+ if data['item_id'] is None or len(str(data['item_id'])) == 0:
+ raise SensorsAnalyticsIllegalDataException("item_id must not be empty")
+ if len(str(data['item_id'])) > 255:
+ raise SensorsAnalyticsIllegalDataException("the max length of item_id is 255")
+ # 检查 properties
+ SensorsAnalytics._normalize_properties(data)
+ return data
+
+ def _track_item(self, event_type, item_type, item_id, properties):
+ if properties is None:
+ properties = {}
+ data = {
+ 'type': event_type,
+ 'time': self._now(),
+ 'lib': self._get_lib_properties(),
+ 'item_type': item_type,
+ 'item_id': item_id,
+ }
+
+ if self._default_project_name is not None:
+ data['project'] = self._default_project_name
+
+ if properties and '$project' in properties and len(str(properties['$project'])) != 0:
+ data['project'] = properties['$project']
+ properties.pop('$project')
+
+ data['properties'] = properties
+
+ data = self._normalize_item_data(data)
+ self._json_dumps(data)
+ self._consumer.send(self._json_dumps(data))
+
+ def _track_event(self, event_type, event_name, distinct_id, original_id, properties, is_login_id, *identities):
event_time = self._extract_user_time(properties) or self._now()
+ event_token = self._extract_token(properties)
+ event_project = self._extract_project(properties)
data = {
'type': event_type,
@@ -426,24 +556,48 @@ def _track_event(self, event_type, event_name, distinct_id, original_id, propert
'properties': properties,
'lib': self._get_lib_properties(),
}
+
+ if identities:
+ identities_data = dict()
+ for identity in identities:
+ identities_data[identity.key] = identity.value
+ data["identities"] = identities_data
+
if self._default_project_name is not None:
data['project'] = self._default_project_name
- if event_type == "track" or event_type == "track_signup":
+ if event_type == EventType.TRACK.value or event_type == EventType.TRACK_SIGNUP.value \
+ or event_type == EventType.TRACK_ID_BIND.value \
+ or event_type == EventType.TRACK_ID_UNBIND.value:
data["event"] = event_name
- if event_type == "track_signup":
+ if event_type == EventType.TRACK_SIGNUP.value:
data["original_id"] = original_id
if self._enable_time_free:
data["time_free"] = True
- if is_login_id:
+ if is_login_id or self._is_identity_has_login_id(*identities):
properties["$is_login_id"] = True
+ if event_token is not None:
+ data["token"] = event_token
+
+ if event_project is not None:
+ data["project"] = event_project
+
data = self._normalize_data(data)
self._consumer.send(self._json_dumps(data))
+ @staticmethod
+ def _is_identity_has_login_id(*identities):
+ if not identities:
+ return False
+ for identity in identities:
+ if identity.key == SensorsAnalyticsIdentity.LOGIN_ID:
+ return True
+ return False
+
def flush(self):
"""
对于不立即发送数据的 Consumer，调用此接口应当立即进行已有数据的发送。
@@ -457,6 +611,159 @@ def close(self):
"""
self._consumer.close()
+ @staticmethod
+ def _check_identity_type(*identities):
+ if not identities:
+ raise SensorsAnalyticsIllegalDataException("Identity (or list) can not be none or empty")
+ key_repeat_map = {}
+ duplicate_keys = set()
+ for identity in identities:
+ if not isinstance(identity, SensorsAnalyticsIdentity):
+ raise SensorsAnalyticsIllegalDataException("Identity type must be SensorsAnalyticsIdentity")
+ SensorsAnalytics._assert_key(identity.key)
+ if not is_str(identity.value):
+ raise SensorsAnalyticsIllegalDataException("identity value must be a str. [key=%s]" % str(identity.key))
+ if not len(identity.value.strip()):
+ raise SensorsAnalyticsIllegalDataException("identity value is empty. [key=%s]" % str(identity.key))
+ if len(identity.value) > 255:
+ raise SensorsAnalyticsIllegalDataException(
+ "the max length of property value is 256. [key=%s]" % str(identity.key))
+ count = key_repeat_map.get(identity.key, 0)
+ count += 1
+ key_repeat_map[identity.key] = count
+ if count > 1:
+ duplicate_keys.add(identity.key)
+ if duplicate_keys:
+ raise SensorsAnalyticsIllegalDataException("Identity has duplicate key. [key=%s]" % str(duplicate_keys))
+
+ @staticmethod
+ def _get_distinct_id(*identities):
+ if not identities:
+ return None
+ distinct_id = "%s+%s" % (identities[0].key, identities[0].value)
+ for identity in identities:
+ if SensorsAnalyticsIdentity.LOGIN_ID == identity.key:
+ distinct_id = identity.value
+ break
+ return distinct_id
+
+ def bind(self, first_identity, second_identity, *other_identities):
+ """
+ 绑定用户标识。至少需要提供两个用户标识信息。identity 的数据类型为 SensorsAnalyticsIdentity
+
+ :param first_identity 待绑定的用户标识，
+ :param second_identity 待绑定的用户标识
+ :param other_identities 其他需要绑定的用户标识
+ """
+ SensorsAnalytics._check_identity_type(first_identity, second_identity, *other_identities)
+ all_properties = self._super_properties.copy()
+ self._track_event(EventType.TRACK_ID_BIND.value, "$BindID",
+ SensorsAnalytics._get_distinct_id(first_identity, second_identity, *other_identities),
+ None, all_properties, None, first_identity, second_identity, *other_identities)
+
+ def unbind(self, identity):
+ """
+ 解绑用户标识
+ :param identity SensorsAnalyticsIdentity
+ """
+ SensorsAnalytics._check_identity_type(identity)
+ # if identity.key == SensorsAnalyticsIdentity.LOGIN_ID:
+ # raise SensorsAnalyticsIllegalDataException("Can not unbind login id.")
+ all_properties = self._super_properties.copy()
+ self._track_event(EventType.TRACK_ID_UNBIND.value, "$UnbindID", SensorsAnalytics._get_distinct_id(identity),
+ None, all_properties, None, identity)
+
+ def track_by_id(self, event_name, properties, *identities):
+ """
+ 使用用户标识 3.0 进行事件埋点
+ :param event_name 事件名
+ :param properties 事件属性，数据类型为 dict
+ :param identities 用户标识
+ """
+ SensorsAnalytics._check_identity_type(*identities)
+ all_properties = self._super_properties.copy()
+ if properties:
+ if not isinstance(properties, dict):
+ raise SensorsAnalyticsIllegalDataException("properties must be a dict type.")
+ all_properties.update(properties)
+ self._track_event(EventType.TRACK.value, event_name,
+ SensorsAnalytics._get_distinct_id(*identities),
+ None, all_properties, None, *identities)
+
+ def profile_set_by_id(self, properties, *identities):
+ """
+ 设置用户的属性。如果要设置的 properties 的 key，之前在这个用户的 profile 中已经存在，则覆盖，否则，新创建
+ :param properties 用户属性
+ :param identities 用户标识，类型是 SensorsAnalyticsIdentity
+ """
+ self._profile_options_by_id(EventType.PROFILE_SET, properties, *identities)
+
+ def profile_set_once_by_id(self, properties, *identities):
+ """
+ 首次设置用户的属性。与 profile_set_by_id 接口不同的是：如果被设置的用户属性已存在，则这条记录会被忽略而不会覆盖已有数据，如果属性不存在则会自动创建
+ :param properties 用户属性
+ :param identities 用户标识，类型是 SensorsAnalyticsIdentity
+ """
+ self._profile_options_by_id(EventType.PROFILE_SET_ONCE, properties, *identities)
+
+ def profile_unset_by_id(self, profile_keys, *identities):
+ """
+ 删除用户某一个属性
+ :param profile_keys: 用户属性键值列表
+ :param identities 用户标识，类型是 SensorsAnalyticsIdentity
+ """
+ if isinstance(profile_keys, list):
+ profile_keys = dict((key, True) for key in profile_keys)
+ else:
+ raise SensorsAnalyticsIllegalDataException("profile_keys must be a list.")
+ self._profile_options_by_id(EventType.PROFILE_UNSET, profile_keys, *identities)
+
+ def profile_append_by_id(self, profiles, *identities):
+ """
+ 追加一个用户的某一个或者多个集合类型的 Profile。
+ :param profiles 用户属性，其 key 必须是 str 类型，value 必须是 str 集合类型
+ :param identities 用户标识，类型是 SensorsAnalyticsIdentity
+ """
+ if isinstance(profiles, dict):
+ for key, value in profiles.items():
+ if not isinstance(value, list):
+ raise SensorsAnalyticsIllegalDataException("property value must be list. [key=%s]" % str(key))
+ else:
+ raise SensorsAnalyticsIllegalDataException("profiles must be dict type.")
+ self._profile_options_by_id(EventType.PROFILE_APPEND, profiles, *identities)
+
+ def profile_delete_by_id(self, *identities):
+ """
+ 删除用户的所有属性
+ :param identities 用户标识，类型是 SensorsAnalyticsIdentity
+ """
+ self._profile_options_by_id(EventType.PROFILE_DELETE, {}, *identities)
+
+ def profile_increment_by_id(self, profiles, *identities):
+ """
+ 为用户的一个或多个数值类型的属性累加一个数值，若该属性不存在，则创建它并设置默认值为 0。属性取值只接受 Number类型。
+ :param profiles 用户属性，类型是 dict，value 必须是 Number 类型
+ :param identities 用户标识，可以是 identity、list、tuple
+ """
+ if isinstance(profiles, dict):
+ for key, value in profiles.items():
+ if not is_int(value):
+ raise SensorsAnalyticsIllegalDataException("property value must be Number. [key=%s]" % str(key))
+ else:
+ raise SensorsAnalyticsIllegalDataException("profiles must be dict type.")
+ self._profile_options_by_id(EventType.PROFILE_INCREMENT, profiles, *identities)
+
+ def _profile_options_by_id(self, event_type, properties, *identities):
+ if not (event_type == EventType.PROFILE_DELETE):
+ if not properties:
+ raise SensorsAnalyticsIllegalDataException("Properties can not be None or Empty")
+ if not isinstance(properties, dict):
+ raise SensorsAnalyticsIllegalDataException("Properties type must be dict")
+ SensorsAnalytics._check_identity_type(*identities)
+ self._track_event(event_type.value, None, SensorsAnalytics._get_distinct_id(*identities), None, properties,
+ None, *identities)
+
+
class DefaultConsumer(object):
"""
默认的 Consumer实现，逐条、同步的发送数据给接收服务器。
@@ -467,7 +774,7 @@ def __init__(self, url_prefix, request_timeout=None):
初始化 Consumer。
:param url_prefix: 服务器的 URL 地址。
- :param request_timeout: 请求的超时时间，单位毫秒。
+ :param request_timeout: 请求的超时时间，单位为秒。
"""
self._url_prefix = url_prefix
self._request_timeout = request_timeout
@@ -531,7 +838,7 @@ def __init__(self, url_prefix, max_size=50, request_timeout=None):
:param url_prefix: 服务器的 URL 地址。
:param max_size: 批量发送的阈值。
- :param request_timeout: 请求服务器的超时时间，单位毫秒。
+ :param request_timeout: 请求服务器的超时时间，单位为秒。
:return:
"""
super(BatchConsumer, self).__init__(url_prefix, request_timeout)
@@ -539,21 +846,23 @@ def __init__(self, url_prefix, max_size=50, request_timeout=None):
self._max_size = min(50, max_size)
def send(self, json_message):
- self._buffers.append(json_message)
- if len(self._buffers) >= self._max_size:
- self.flush()
+ with batch_consumer_lock:
+ self._buffers.append(json_message)
+ if len(self._buffers) >= self._max_size:
+ self.flush()
def flush(self):
"""
用户可以主动调用 flush 接口，以便在需要的时候立即进行数据发送。
"""
- while self._buffers:
- msg_list = self._buffers[:self._max_size]
- self._do_request({
- 'data_list': self._encode_msg_list(msg_list),
- 'gzip': 1
- })
- self._buffers = self._buffers[self._max_size:]
+ with batch_consumer_lock:
+ while self._buffers:
+ msg_list = self._buffers[:self._max_size]
+ self._do_request({
+ 'data_list': self._encode_msg_list(msg_list),
+ 'gzip': 1
+ })
+ self._buffers = self._buffers[self._max_size:]
def close(self):
"""
@@ -610,7 +919,7 @@ def __init__(self, url_prefix, flush_max_time=3, flush_size=20,
:param flush_size: 队列缓存的阈值，超过此值将立即进行发送。
:param max_batch_size: 单个请求发送的最大大小。
:param max_size: 整个缓存队列的最大大小。
- :param request_timeout: 请求的超时时间，单位毫秒。
+ :param request_timeout: 请求的超时时间，单位为秒。
"""
super(AsyncBatchConsumer, self).__init__(url_prefix, request_timeout)
@@ -687,13 +996,16 @@ def __init__(self, url_prefix, write_data=True, request_timeout=None):
初始化Consumer
:param url_prefix: 服务器提供的用于Debug的API的URL地址,特别注意,它与导入数据的API并不是同一个
:param write_data: 发送过去的数据,是真正写入,还是仅仅进行检查
- :param request_timeout:请求的超时时间,单位毫秒
+ :param request_timeout:请求的超时时间,单位为秒
:return:
"""
debug_url = urlparse(url_prefix)
## 将 URI Path 替换成 Debug 模式的 '/debug'
- debug_url = debug_url._replace(path = '/debug')
-
+ url_path = debug_url.path
+ index = url_path.rfind('/')
+ debug_url_path = url_path[0:index] + '/debug'
+ debug_url = debug_url._replace(path=debug_url_path)
+
self._debug_url_prefix = debug_url.geturl()
self._request_timeout = request_timeout
self._debug_write_data = write_data
@@ -719,7 +1031,7 @@ def _do_request(self, data):
encoded_data = urllib.urlencode(data).encode('utf8')
try:
request = urllib2.Request(self._debug_url_prefix, encoded_data)
- if not self._debug_write_data: # 说明只检查,不真正写入数据
+ if not self._debug_write_data: # 说明只检查,不真正写入数据
request.add_header('Dry-Run', 'true')
if self._request_timeout is not None:
response = urllib2.urlopen(request, timeout=self._request_timeout)
@@ -796,6 +1108,7 @@ def flush(self):
def close(self):
self.logger.handlers[0].close()
+
class ConcurrentLoggingConsumer(object):
"""
将数据输出到指定路径，并按天切割，支持多进程并行输出到同一个文件名
@@ -811,7 +1124,7 @@ def close(self):
self._file.close()
def isValid(self, filename):
- return self._filename == filename
+ return self._filename == filename
def write(self, messages):
with SAFileLock(self._file):
@@ -822,7 +1135,7 @@ def write(self, messages):
@classmethod
def construct_filename(cls, prefix):
- return prefix + '.' + datetime.datetime.now().strftime('%Y-%m-%d')
+ return prefix + '.' + datetime.datetime.now().strftime('%Y-%m-%d')
def __init__(self, prefix, bufferSize=8192):
self._prefix = prefix
@@ -846,7 +1159,7 @@ def send(self, msg):
if len(self._buffer) > self._bufferSize:
messages = self._buffer
- filename = ConcurrentLoggingConsumer.construct_filename(self._prefix)
+ filename = ConcurrentLoggingConsumer.construct_filename(self._prefix)
if not self._writer.isValid(filename):
self._writer.close()
self._writer = ConcurrentLoggingConsumer.ConcurrentFileWriter(filename)
@@ -854,7 +1167,7 @@ def send(self, msg):
self._buffer = []
self._mutex.put(1)
-
+
if messages:
self._writer.write(messages)
@@ -866,7 +1179,7 @@ def flush(self):
if len(self._buffer) > 0:
messages = self._buffer
- filename = ConcurrentLoggingConsumer.construct_filename(self._prefix)
+ filename = ConcurrentLoggingConsumer.construct_filename(self._prefix)
if not self._writer.isValid(filename):
self._writer.close()
self._writer = ConcurrentLoggingConsumer.ConcurrentFileWriter(filename)
@@ -874,10 +1187,43 @@ def flush(self):
self._buffer = []
self._mutex.put(1)
-
+
if messages:
self._writer.write(messages)
def close(self):
self.flush()
self._writer.close()
+
+
+# ID-Mapping 3 业务逻辑
+class SensorsAnalyticsIdentity(object):
+ LOGIN_ID = "$identity_login_id"
+ """
+ 用户登录 id
+ """
+ MOBILE = "$identity_mobile"
+ """
+ 手机号
+ """
+ EMAIL = "$identity_email"
+ """
+ 邮箱
+ """
+
+ def __init__(self, key, value):
+ self.key = key
+ self.value = value
+
+
+class EventType(Enum):
+ TRACK = "track"
+ TRACK_SIGNUP = "track_signup"
+ TRACK_ID_BIND = "track_id_bind"
+ TRACK_ID_UNBIND = "track_id_unbind"
+ PROFILE_SET = "profile_set"
+ PROFILE_SET_ONCE = "profile_set_once"
+ PROFILE_UNSET = "profile_unset"
+ PROFILE_APPEND = "profile_append"
+ PROFILE_DELETE = "profile_delete"
+ PROFILE_INCREMENT = "profile_increment"
diff --git a/sensorsanalytics/test_sdk.py b/sensorsanalytics/test_sdk.py
index 3783a04..a5cd4f2 100644
--- a/sensorsanalytics/test_sdk.py
+++ b/sensorsanalytics/test_sdk.py
@@ -6,8 +6,8 @@
from sdk import *
-TEST_URL_PREFIX = 'http://10.10.11.209:8006/sa?token=bbb'
-TEST_DEBUG_URL_PREFIX = 'http://10.10.11.209:8006/debug?token=bbb'
+TEST_URL_PREFIX = 'https://sdk-test.datasink.sensorsdata.cn/sa?project=yuejianzhong&token=95c73ae661f85aa0'
+TEST_DEBUG_URL_PREFIX = 'https://sdk-test.datasink.sensorsdata.cn/sa?project=yuejianzhong&token=95c73ae661f85aa0'
class NormalTest(unittest.TestCase):
@@ -53,11 +53,14 @@ def clear_msg_counter(self):
def testDebug(self):
consumer = DebugConsumer(TEST_DEBUG_URL_PREFIX, False)
sa = SensorsAnalytics(consumer)
- sa.track(1234, 'Test', {'From': 'Baidu'}, is_login_id=True)
+ sa.track(1234, 'Test', {'From1': 'Baidu'}, is_login_id=True)
consumer = DebugConsumer(TEST_DEBUG_URL_PREFIX, True)
sa = SensorsAnalytics(consumer)
- sa.track(1234, 'Test', {'From': 456})
- sa.track(1234, 'Test', {'From': 'Baidu'})
+ sa.track(1234, 'Test', {'From2': 456})
+ sa.track(1234, 'Test', {'From1': 'Baidu'})
+ sa.track(1234, 'Test', {'From1': 'Baidu', '$project': "yuejianzhong"})
+ sa.track(1234, 'Test', {'From1': 'Baidu', '$token': "dhuw393jdcioj39"})
+ sa.track(1234, 'Test', {'From1': 'Baidu', '$token': "dhuw393jdcioj39",'$project': "yuejianzhong"})
def testNormal(self):
consumer = DefaultConsumer(TEST_URL_PREFIX)
@@ -124,7 +127,7 @@ def testDefaultConsumer(self):
sa.track('1234', 'Test', {'From': 'Baidu'})
sa.track_signup('1234', 'abcd', {'Channel': 'Hongbao'})
sa.profile_delete('1234')
- sa.profile_append('1234', {'Gender': 'Male'})
+ sa.profile_append('1234', {'Gender': ['Male']})
sa.profile_increment('1234', {'CardNum': 1})
sa.profile_set('1234', {'City': '北京'})
sa.profile_unset('1234', ['City'])
@@ -138,7 +141,7 @@ def testBatchConsumer(self):
sa.track('1234', 'Test', {'From': 'Baidu'})
sa.track_signup('1234', 'abcd', {'Channel': 'Hongbao'})
sa.profile_delete('1234')
- sa.profile_append('1234', {'Gender': 'Male'})
+ sa.profile_append('1234', {'Gender': ['Male']})
self.assertEqual(self.msg_counter, 0)
sa.profile_increment('1234', {'CardNum': 1})
self.assertEqual(self.msg_counter, 5)
@@ -158,7 +161,7 @@ def testAsyncBatchConsumer(self):
sa.track('1234', 'Test', {'From': 'Baidu'})
sa.track_signup('1234', 'abcd', {'Channel': 'Hongbao'})
sa.profile_delete('1234')
- sa.profile_append('1234', {'Gender': 'Male'})
+ sa.profile_append('1234', {'Gender': ["male", "femal"]})
self.assertEqual(self.msg_counter, 0)
sa.profile_increment('1234', {'CardNum': 1})
time.sleep(0.1)
@@ -177,6 +180,34 @@ def testAsyncBatchConsumer(self):
time.sleep(0.1)
self.assertEqual(self.msg_counter, 9)
+ def testIDM3(self):
+ consumer = DefaultConsumer(TEST_URL_PREFIX)
+ # consumer._do_request = self.mock_request
+ sa = SensorsAnalytics(consumer)
+
+ # sa.bind(SensorsAnalyticsIdentity("s1", "sv1"), SensorsAnalyticsIdentity("s2", "sv2"), SensorsAnalyticsIdentity("s3", "sv3"))
+ # sa.bind(SensorsAnalyticsIdentity("s1", "sv1"), SensorsAnalyticsIdentity("s2", "sv2"), SensorsAnalyticsIdentity(SensorsAnalyticsIdentity.LOGIN_ID, "sv3"))
+ # sa.unbind(SensorsAnalyticsIdentity(SensorsAnalyticsIdentity.EMAIL, "sv1"))
+
+ # sa.track_by_id("hello", None, SensorsAnalyticsIdentity(SensorsAnalyticsIdentity.LOGIN_ID, "sv1"), SensorsAnalyticsIdentity("s2", "sv2"))
+
+ # sa.profile_set_by_id({"p1": "v1"}, SensorsAnalyticsIdentity("s1", "sv1"))
+ # sa.profile_unset_by_id(["k1", "k2"], SensorsAnalyticsIdentity("s1", "sv1"))
+
+ # sa.profile_append("sss", {"k1": "ss"}, False)
+
+ # sa.profile_append_by_id({"k1": ["a1", "a2", "a3"]}, SensorsAnalyticsIdentity("s1", "sv1"),
+ # SensorsAnalyticsIdentity("s2", "sv2"))
+
+ # sa.profile_delete_by_id(SensorsAnalyticsIdentity("s1", "sv1"),
+ # SensorsAnalyticsIdentity("s2", "sv2"))
+
+ # sa.profile_increment_by_id({"age": "123"}, SensorsAnalyticsIdentity("s1", "sv1"))
+
+ sa.flush()
+ time.sleep(2)
+ pass
+
if __name__ == '__main__':
unittest.main()
diff --git a/setup.py b/setup.py
index 436a138..828babb 100644
--- a/setup.py
+++ b/setup.py
@@ -1,13 +1,16 @@
-from distutils.core import setup
+import setuptools
-setup(
- name='SensorsAnalyticsSDK',
- version='1.7.5',
- author='Yuhan ZOU',
- author_email='zouyuhan@sensorsdata.cn',
- url='http://www.sensorsdata.cn',
- license='LICENSE.txt',
- packages=['sensorsanalytics'],
- description='This is the official Python SDK for Sensors Analytics.',
- long_description=open('README.txt').read(),
-)
+# 读取项目的readme介绍
+with open("README.md", "r") as fh:
+ long_description = fh.read()
+setuptools.setup(
+ name="SensorsAnalyticsSDK",
+ version="1.11.0",
+ author="Jianzhong YUE", # 项目作者
+ author_email="yuejianzhong@sensorsdata.cn",
+ description="This is the official Python SDK for Sensors Analytics.",
+ long_description=long_description,
+ long_description_content_type="text/markdown",
+ url="https://github.com/sensorsdata/sa-sdk-python",
+ packages=setuptools.find_packages(),
+)
\ No newline at end of file