From c6a8587c7904afe64c31f74da1f9d4edc7139ed2 Mon Sep 17 00:00:00 2001
From: Dhairya Khanna <31827131+Dhairya3124@users.noreply.github.com>
Date: Thu, 27 Jan 2022 09:32:15 +0530
Subject: [PATCH 01/17] Fix: AttributeError: 'InstallRequirement' object has no
attribute 'requirement' (#606)
---
setup.py | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/setup.py b/setup.py
index d46f92df..a789defe 100644
--- a/setup.py
+++ b/setup.py
@@ -9,7 +9,10 @@
from pip._internal.req import parse_requirements
install_requires = parse_requirements("requirements.txt", session=PipSession())
- dependencies = [str(package.requirement) for package in install_requires]
+ try:
+ dependencies = [str(package.requirement) for package in install_requires]
+ except:
+ dependencies = [str(package.req) for package in install_requires]
except ImportError:
msg = "Your pip version is out of date, please run `pip install --upgrade pip setuptools`"
raise ImportError(msg)
From da2635fd3ba55b2fc654de4c0cb4679a185c9c93 Mon Sep 17 00:00:00 2001
From: Krish Bhardwaj <72294477+Krish-bhardwaj@users.noreply.github.com>
Date: Wed, 26 Jan 2022 15:32:21 +0530
Subject: [PATCH 02/17] Update README.md
---
README.md | 52 ++++++++++++++++++++++++++--------------------------
1 file changed, 26 insertions(+), 26 deletions(-)
diff --git a/README.md b/README.md
index d1460354..a28e8db5 100644
--- a/README.md
+++ b/README.md
@@ -1,43 +1,43 @@
-hydrus [](https://travis-ci.com/HTTP-APIs/hydrus)
+Hydrus [](https://travis-ci.com/HTTP-APIs/hydrus)
===================
hydrus is a set of **Python** based tools for easier and efficient creation of Hypermedia driven REST-APIs. hydrus utilises the power of [Linked Data](https://en.wikipedia.org/wiki/Linked_data) to create a powerful REST APIs to serve data.
hydrus uses the [Hydra(W3C)](http://www.hydra-cg.com/) standard for creation and documentation of it's APIs.
-Start-up the demo
------------------
-* with *Docker* and *docker-compose* installed, run `docker-compose up --build`
-* open the browser at `http://localhost:8080/api/vocab`
+## Start-up the demo
+- With [*Docker*](https://www.docker.com/) and [*docker-compose*](https://docs.docker.com/compose/) installed, run
+``` bash
+docker-compose up --build
+```
+- Open the browser at `http://localhost:8080/api/vocab`.
You should be displaying the example API as served by the server.
-Add your own Hydra documentation file
--------------------------------------
+## Add your own Hydra documentation file
To serve your own Hydra-RDF documentation file:
-* create a `doc.py` file as the ones in `examples/` directory containing your own *ApiDoc*
-* set the `APIDOC_REL_PATH` variable in `docker-compose.yml`. This should the relative path from the project root
-* start-up the demo as above.
+
+1. Create a `doc.py` file as the ones in `examples/` directory containing your own *ApiDoc*.
+2. Set the `APIDOC_REL_PATH` variable in `docker-compose.yml`. This should the relative path from the project root.
+3. Start-up the demo as above.
You should be displaying your API as served by the server.
-Table of contents
--------------
+## Table of contents
* [Features](#features)
* [Requirements](#req)
* [Demo](#demo)
* [Usage](#usage)
-Features
--------------
-hydrus supports the following features:
+## Features
+Hydrus supports the following features:
- A client that can understand Hydra vocabulary and interacts with a Hydra supporting server to basic [CRUD](https://en.wikipedia.org/wiki/Create,_read,_update_and_delete) operations on data.
- A generic server that can serve required data and metadata(in the form of API documentation) to a client over HTTP.
- A middleware that allows users to use the client to interact with the server using Natural Language which is processed machine consumable language. **(under development)**
-Requirements
--------------
+## Requirements
The system is built over the following standards and tools:
+- [Python](https://www.python.org/downloads/) 3.6 and above
- [Flask](http://flask.pocoo.org/) a Python based micro-framework for handling server requests and responses.
- [JSON-LD](http://json-ld.org/spec/latest/json-ld/) as the preferred data format.
- [Hydra](http://www.hydra-cg.com/) as the API standard.
@@ -46,12 +46,14 @@ The system is built over the following standards and tools:
Apart from this, there are also various Python packages that hydrus uses. Using `python setup.py install` installs all the required dependencies.
**NOTE:** You'll need to use `python3` not `python2`. Hydrus does not support python < 3.6
+To check your python version run
+```bash
+python --version
+```
-Demo
--------------
+## Demo
To run a demo for hydrus using the sample API, just do the following:
-
1. Clone hydrus:
```bash
git clone https://github.com/HTTP-APIs/hydrus
@@ -78,7 +80,8 @@ NOTE: there is an alternative way to install dependencies with `poetry`:
pip3 install poetry
poetry install
```
-This is mostly used to check dependencies conflicts among packages and to release to `PyPi`.
+
+This is mostly used to check dependencies conflicts among packages and to release to [`PyPi`](https://pypi.org/).
After installation is successful, to *run the server*:
```bash
@@ -88,14 +91,11 @@ hydrus serve
The demo should be up and running on `http://localhost:8080/serverapi/`.
-Usage
--------------
+## Usage
For more info, head to the [Usage](http://www.hydraecosystem.org/01-Usage.html) section of the [wiki](http://www.hydraecosystem.org/).
-Development
--------------
-
+## Development
From the `hydrus` directory:
* To run formatter: `pip install black && black *.py`
* To test for formatting: `flake8 *.py`
From 6fd7a21d06fbcdd291ec11d9d9bb51116ca7faea Mon Sep 17 00:00:00 2001
From: Krish Bhardwaj <72294477+Krish-bhardwaj@users.noreply.github.com>
Date: Sat, 29 Jan 2022 09:42:54 +0530
Subject: [PATCH 03/17] Update STARTING.md
---
docs/STARTING.md | 28 +++++++++++++++-------------
1 file changed, 15 insertions(+), 13 deletions(-)
diff --git a/docs/STARTING.md b/docs/STARTING.md
index 0fffa123..8928be1d 100644
--- a/docs/STARTING.md
+++ b/docs/STARTING.md
@@ -3,14 +3,16 @@ This document is just a collection of annotations at the moment.
### Objectives
This project in basically aimed to solve these issues:
-* provide an attractive demo API to show HYDRA specs, we decided to leverage space sciences as they provide a subject of great interest;
-* define a multi-layered (and multi-peer?) approach to make the API usable and adaptable to different filtering/querying needs;
-* make useful experiments about the right way to approach aggregation/filtering procedures among data served by multiple HYDRA-enhanced endpoints (starting from well-know industry standards for database management and data warehousing, and some glimpses to GraphQL).
+* Provide an attractive demo API to show hydra specs, we decided to leverage space sciences as they provide a subject of great interest;
+* Define a multi-layered (and multi-peer?) approach to make the API usable and adaptable to different filtering/querying needs;
+* Make useful experiments about the right way to approach aggregation/filtering procedures among data served by multiple hydra-enhanced endpoints (starting from well-know industry standards for database management and data warehousing, and some glimpses to [GraphQL](https://graphql.org/)).
### Starting assumptions
-We suppose the server has to be build with at least two layers of abstraction: a low-level strict-HYDRA-RDF API that can map as directly as possible the data to resources/collections ("Server"), and a middleware API ("Client") to perform some sort of opinionated intermediation between the lower level and the "final" client ("UI") above. The middleware has to implement data querying on the lower endpoints, leveraging HYDRA's metadata framework; to do this it has in some way to implement precise choices about the aggregation mechanism that is going to fetch and "map-reduce" (aggregate) the data from the lower level (this part is really open for discussions, as a starting example you can check Mongo's pipelines and its matching/grouping system); because of the different patterns that makes possible this kind of results, some choices about tools and techniques have to be discussed and options defined.
+#### We suppose the server has to be build with at least **two layers of abstraction**:
+
+A **low-level strict-HYDRA-RDF API** that can map as directly as possible the data to resources/collections **("Server")**, and a middleware API **("Client")** to perform some sort of opinionated intermediation between the lower level and the **"final"** **client ("UI")** above. The middleware has to implement data querying on the lower endpoints, leveraging HYDRA's metadata framework. To do this it has in some way to implement precise choices about the aggregation mechanism that is going to fetch and "map-reduce" (aggregate) the data from the lower level (this part is really open for discussions, as a starting example you can check Mongo's pipelines and its matching/grouping system), because of the different patterns that makes possible this kind of results, some choices about tools and techniques have to be discussed and options defined.
### About the option of having endpoints that are type- (class-) based
As we discussed, the fun idea would be to have somenthing that can "mix" REST and RPC in some way, not technically but at least conceptually. This approach comes from the observation that the statefulness of a REST API can be in some sort (at a level we are trying to understand for the purpose of our implementation) related to the concept of invariability of the output of a pure-function. This would be practically accomplished by dedicating some endpoints to resources obviously but also to "operations" (functions that accepts some kind of paramethers and run some kind of reasoning on a data structure, we temporarly define these endpoints `hydra:Operation`s).
@@ -19,14 +21,14 @@ As we discussed, the fun idea would be to have somenthing that can "mix" REST an
What we mean with "aggregation" and "filtering" with HYDRA? Let's see a simple example.
As HYDRA is meant to let clients to interoperate automatically, we try here to subset the problem posing it on this shape: starting from an initial input from a human/user, how can different layers of HYDRA-featured APIs respond consistantly by querying the provided endpoints?
-* "UI" layer: a user (or a machine from another network) is wishful to know "what is most distant from the Sun, Earth or Mars?"
-* "Client" layer: the client knows that some endpoints are available at a lower level, and we suppose that it knows it has to look for some kind of length value. It looks for the endpoints that can help, we suppose it can understand the fact that it needs the `/api/planet/calculate_average_au` (look for the right operation to perform on the "planet" class, that is basically a documentation problem); so it pass the parameters (Earth and Mars) to it. This layer is commonly referred to as *middleware*; we temporarly define `/api/planet/calculate_average_au` as a `hydra:Operation` (an endpoint that performs come kind of aggregation on other endpoints, something more similar to an RPC call more than a REST call probably), all the `hydra:Operation`s referred to a `hydra:Class` type has to be listed in the class' definition;
-* "Server" layer: the server endpoints are queried and they serve the required data to calculate and respond: "Mars!". This server is the one that works like a traditional HYDRA-enhanced API, with all the descriptive features provided by the spec.
+* **"UI" layer:** a user (or a machine from another network) is wishful to know "what is most distant from the Sun, Earth or Mars?"
+* **"Client" layer:** the client knows that some endpoints are available at a lower level, and we suppose that it knows it has to look for some kind of length value. It looks for the endpoints that can help, we suppose it can understand the fact that it needs the `/api/planet/calculate_average_au` (look for the right operation to perform on the "planet" class, that is basically a documentation problem); so it pass the parameters (Earth and Mars) to it. This layer is commonly referred to as *middleware*; we temporarly define `/api/planet/calculate_average_au` as a `hydra:Operation` (an endpoint that performs come kind of aggregation on other endpoints, something more similar to an RPC call more than a REST call probably), all the `hydra:Operation`s referred to a `hydra:Class` type has to be listed in the class' definition;
+* **"Server" layer:** the server endpoints are queried and they serve the required data to calculate and respond: "Mars!". This server is the one that works like a traditional HYDRA-enhanced API, with all the descriptive features provided by the spec.
In this scenario, allowed operations are defined by the possible interaction of different (or same-kind) classes (this idea is inspired by the Scala programming language's typing system). For example, two instances of a `Planet` (of the Solar System) class can have in common the defined "operation" of `calculate_average_au` (compute the average distance of two planets using the Sun-Earth distance as unit(AU)). This way in the API we could do: `/api/planet/calculate_average_au` and pass to it the related parameters.
This kind and body of the request:
-```
+```json
POST /api/planet/calculate_average_au
{
@type: "hydra:Collection",
@@ -37,7 +39,7 @@ POST /api/planet/calculate_average_au
}
```
This endpoint could respond with:
-```
+```json
{
@type: "vocab:Result", # or the given type in the HYDRA framework
@returns: "umbel:Float",
@@ -46,7 +48,7 @@ This endpoint could respond with:
}
```
This RPC-like endpoint could be meta-described as:
-```
+```json
{
"@context": {
"hydra": "http://www.w3.org/ns/hydra/context.jsonld"
@@ -69,9 +71,9 @@ This RPC-like endpoint could be meta-described as:
### Stack
-* initial version: local Flask ("Server") and in-memory low-footprint actors ("Client") reading from local vocabularies, use ZeroMQ;
-* development version: use a cache (Mongo or Redis) for documents and try to represent a graph;
-* stable version: add a some kind of graph database under the cache layer;
+* **Initial version:** local Flask ("Server") and in-memory low-footprint actors ("Client") reading from local vocabularies, use ZeroMQ;
+* **Development version:** use a cache (Mongo or Redis) for documents and try to represent a graph;
+* **Stable version:** add a some kind of graph database under the cache layer;
* ...
### Implementation
From a06e53de6ae71720b82dad2be729b9e30d90bef9 Mon Sep 17 00:00:00 2001
From: Krish Bhardwaj <72294477+Krish-bhardwaj@users.noreply.github.com>
Date: Sat, 29 Jan 2022 09:44:21 +0530
Subject: [PATCH 04/17] Update TASKS.md
---
docs/TASKS.md | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/docs/TASKS.md b/docs/TASKS.md
index 55efc685..dd0f2dc8 100644
--- a/docs/TASKS.md
+++ b/docs/TASKS.md
@@ -1,7 +1,7 @@
-Establish a protocol for building routes in a way that can fit the HYDRA spec:
+# Establish a protocol for building routes in a way that can fit the HYDRA spec:
-* Every route (see `server.routes.py`) has to be documented with a `hydra:ApiDocumentation` (automatically): it can be used the Werkzeug `Rule.endpoints` property to name each route (within the `add_url_rule` method) and map to an apidoc response;
+* Every route (see `server.routes.py`) has to be documented with a `hydra:ApiDocumentation` (automatically): it can be used the Werkzeug `Rule.endpoints` property to name each route (within the `add_url_rule` method) and map to an apidoc response;.
* Rejoin all the different vocabularies in just one big vocabulary and make the server filter by argument if necessary (example vocabulary are "astronomy" and "solarsystem");
* Routes, outside the basic collections, has to be opinionated considering the concern of filtering and data creation required by the instantiation of the objects based on the provided classes (*or do it in a separate layer above*); in this case we would have a "smart middleware" that relies on a lower server layer.
-* Start with the lower layer, that should be meant to provide resource one-by-one or by collections
\ No newline at end of file
+* Start with the lower layer, that should be meant to provide resource one-by-one or by collections.
From 254ecd72a2ca8010594ea3a9be845a559a0126aa Mon Sep 17 00:00:00 2001
From: Krish Bhardwaj <72294477+Krish-bhardwaj@users.noreply.github.com>
Date: Sat, 29 Jan 2022 09:57:33 +0530
Subject: [PATCH 05/17] Update TUTORIALS.md
---
docs/TUTORIALS.md | 36 ++++++++++++++++++------------------
1 file changed, 18 insertions(+), 18 deletions(-)
diff --git a/docs/TUTORIALS.md b/docs/TUTORIALS.md
index f0451f07..11d8c3c3 100644
--- a/docs/TUTORIALS.md
+++ b/docs/TUTORIALS.md
@@ -6,30 +6,30 @@ https://www.hydraecosystem.org/Starting-Material
### How is an ```ApiDoc``` defined:
- There are two ways of defining an ApiDoc:
- 1. Writing a Hydra-compliant JSON-LD like [here]( https://github.com/HTTP-APIs/hydrus/blob/master/hydrus/samples/doc_writer_sample_output.py)
- 1. Defining the ApiDoc programmatically using the code like in [here](https://github.com/HTTP-APIs/hydrus/blob/master/hydrus/samples/doc_writer_sample.py)
+ 1. Writing a Hydra-compliant JSON-LD like [here]( https://github.com/HTTP-APIs/hydrus/blob/master/hydrus/samples/doc_writer_sample_output.py).
+ 2. Defining the ApiDoc programmatically using the code like in [here](https://github.com/HTTP-APIs/hydrus/blob/master/hydrus/samples/doc_writer_sample.py).
- Any valid ApiDoc can be passed to the server so that it can parse the doc, generate a database and start the server.
- The document containing the context is generated by the doc_maker file in the hydra_python_core repo. hydrus provides the link in its response for the context.
### Points :
#### Description of some files:
-1. hydrus/app.py
- 1. Used for logging. (Line 22)
- 1. Starts the SQLAlchemy engine and creates a session (Line 25-26)
- 1. Creates an Api Documentation using the doc_maker() of hydra_python_core
- 1. Gets the parsed classes from hydrus/data/doc_parse.py
- 1. Creates database tables and its metadata Line 34-36
- 1. Checks if Authentication is True, adds user to the session.Check: https://www.hydraecosystem.org/Auth
- 1. Calls app_factory class' constructor
- 1.Sets authentication,api_name,session etc. which can be later retrieved using the corresponding get functions.
- 1. Runs the API on the Port mentioned.
-1. hydrus/app_factory.py
- 1. Used to add resources to the API from resources.py.
+1. [*hydrus/app.py*](https://github.com/HTTP-APIs/hydrus/blob/develop/hydrus/app.py).
+ 1. Used for logging. [Line 22](https://github.com/HTTP-APIs/hydrus/blob/c6a8587c7904afe64c31f74da1f9d4edc7139ed2/hydrus/app.py#L22).
+ 2. Starts the SQLAlchemy engine and creates a session [Line 30](https://github.com/HTTP-APIs/hydrus/blob/c6a8587c7904afe64c31f74da1f9d4edc7139ed2/hydrus/app.py#L30).
+ 3. Creates an Api Documentation using the [doc_maker() of hydra_python_core](https://hydra-python-core.readthedocs.io/_/downloads/en/develop/pdf/).
+ 4. Gets the parsed classes from [hydrus/data/doc_parse.py](https://github.com/HTTP-APIs/hydrus/blob/develop/hydrus/data/doc_parse.py).
+ 5. Creates database tables and its metadata [Line 34-36](https://github.com/HTTP-APIs/hydrus/blob/c6a8587c7904afe64c31f74da1f9d4edc7139ed2/hydrus/data/doc_parse.py#L34).
+ 6. Checks if Authentication is True, adds user to the session.[Check](https://www.hydraecosystem.org/Auth).
+ 7. Calls app_factory class' constructor.
+ 8. Sets authentication,api_name,session etc. which can be later retrieved using the corresponding get functions.
+ 9. Runs the API on the Port mentioned.
+1. [*hydrus/app_factory.py*](https://github.com/HTTP-APIs/hydrus/blob/develop/hydrus/app_factory.py)
+ 1. Used to add resources to the API from `resources.py`.
1. Configures properties of API.
- 1. Enables CORS
-1. hydrus/resources.py
+ 1. Enables CORS.
+1. [*hydrus/resources.py*](https://github.com/HTTP-APIs/hydrus/blob/develop/hydrus/resources.py)
1. Contains classes which are added as resources to the API.
-1. hydrus/samples/hydra_doc_sample.py
- 1. This file contains the sample ApiDoc that is displayed when the demo for hydrus is run. It contains the JSON-LD document that is displayed.
+1. [*hydrus/samples/hydra_doc_sample.py*](https://github.com/HTTP-APIs/hydrus/blob/develop/hydrus/samples/hydra_doc_sample.py).
+ 1. This file contains the sample ApiDoc that is displayed when the demo for hydrus is run. It contains the [JSON-LD](https://json-ld.org/) document that is displayed.
From dd45a06ce6fe6b6f6d27e0da8f26e84c2326d19c Mon Sep 17 00:00:00 2001
From: Krish Bhardwaj <72294477+Krish-bhardwaj@users.noreply.github.com>
Date: Sat, 29 Jan 2022 09:59:54 +0530
Subject: [PATCH 06/17] Update hydrus_database.md
---
docs/hydrus_database.md | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/docs/hydrus_database.md b/docs/hydrus_database.md
index 1653e900..5f06508e 100644
--- a/docs/hydrus_database.md
+++ b/docs/hydrus_database.md
@@ -21,4 +21,4 @@ The overview of the process of making a database from `apidoc` object is:
### Example ER Diagram
It might be easier to understand what we discussed above if we could see a ER diagram for a Hydra API Doc.
So, let us take the sample [Drone API Doc](https://github.com/HTTP-APIs/hydrus/blob/develop/hydrus/samples/hydra_doc_sample.py) as our example.
-The ER diagram for the database generated for this Hydra doc can be found [here](er_diagram.pdf).
\ No newline at end of file
+The [ER diagram](https://en.wikipedia.org/wiki/Entity%E2%80%93relationship_model) for the database generated for this Hydra doc can be found [here](er_diagram.pdf).
From 6a004fef1c20b45cf347b4624c1df06acface572 Mon Sep 17 00:00:00 2001
From: Krish Bhardwaj <72294477+Krish-bhardwaj@users.noreply.github.com>
Date: Sat, 29 Jan 2022 11:30:15 +0530
Subject: [PATCH 07/17] Update TUTORIALS.md
---
docs/TUTORIALS.md | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/docs/TUTORIALS.md b/docs/TUTORIALS.md
index 11d8c3c3..e6e2f847 100644
--- a/docs/TUTORIALS.md
+++ b/docs/TUTORIALS.md
@@ -1,6 +1,6 @@
-# hydrus Workflow
+# Hydra Workflow
-This document can be used to understand how hydrus works.
+This document can be used to understand how hydra works.
To understand the concepts related to hydrus:
https://www.hydraecosystem.org/Starting-Material
From 70fe61e4dbbc0c8b10e0ec4c0a8c8fec21f80ac2 Mon Sep 17 00:00:00 2001
From: Krish Bhardwaj <72294477+Krish-bhardwaj@users.noreply.github.com>
Date: Sat, 29 Jan 2022 11:30:47 +0530
Subject: [PATCH 08/17] Update hydrus_database.md
---
docs/hydrus_database.md | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/docs/hydrus_database.md b/docs/hydrus_database.md
index 5f06508e..8be8e11f 100644
--- a/docs/hydrus_database.md
+++ b/docs/hydrus_database.md
@@ -1,4 +1,4 @@
-## hydrus
+## Hydrus
`hydrus` is a core part of the HydraEcosystem. It is the server which powers Hydra-based API Docs.
From 6f199de58411ee99a772ce97b92ad069ae35d7a6 Mon Sep 17 00:00:00 2001
From: Krish-bhardwaj
Date: Sat, 29 Jan 2022 23:20:56 +0530
Subject: [PATCH 09/17] test
---
docs/wiki/images/db_schema.svg | 4 ++++
1 file changed, 4 insertions(+)
create mode 100644 docs/wiki/images/db_schema.svg
diff --git a/docs/wiki/images/db_schema.svg b/docs/wiki/images/db_schema.svg
new file mode 100644
index 00000000..cb97d57b
--- /dev/null
+++ b/docs/wiki/images/db_schema.svg
@@ -0,0 +1,4 @@
+
+
+
+
\ No newline at end of file
From d055aade7de72a4de10a802c6ac3c90a029eaafa Mon Sep 17 00:00:00 2001
From: Krish-bhardwaj
Date: Sat, 29 Jan 2022 23:24:51 +0530
Subject: [PATCH 10/17] test over
---
docs/wiki/images/db_schema.svg | 4 ----
1 file changed, 4 deletions(-)
delete mode 100644 docs/wiki/images/db_schema.svg
diff --git a/docs/wiki/images/db_schema.svg b/docs/wiki/images/db_schema.svg
deleted file mode 100644
index cb97d57b..00000000
--- a/docs/wiki/images/db_schema.svg
+++ /dev/null
@@ -1,4 +0,0 @@
-
-
-
-
\ No newline at end of file
From 43bb0454b33a69f916971e9dc096dbf350df5ee7 Mon Sep 17 00:00:00 2001
From: Krish-bhardwaj
Date: Sat, 29 Jan 2022 23:28:19 +0530
Subject: [PATCH 11/17] db_schema update
---
docs/wiki/images/db_schema.png | Bin 28836 -> 95912 bytes
1 file changed, 0 insertions(+), 0 deletions(-)
diff --git a/docs/wiki/images/db_schema.png b/docs/wiki/images/db_schema.png
index ca239748b3383aae30e1832652e0acf5154d06b1..7d3832d30c2742a640669fb9d167a34acc45ab2f 100644
GIT binary patch
literal 95912
zcmcG#2{@GR+dmwsv?2*fwonnqzAu9r+ZbbvElY+OX2#4I#*DEmWtWn)NhwseEG2u(
zP?pG2qNGHKl6}eYUh}Pfzvp@0=lTDS_x-=`ahTgZ_kG>xeJ-DKxvuk?1Up;!?i~kq
zY}&MGx21)d{iaPkGT`ILyA@pFx*1Z<`4?mlhi)ovKRCW=)2<@Axig&{f+zUnHp!})
z{yvjcRq>++(Ph=lWK~to{c+y(Kq>{Cg6m|e4*^fW`TV}8s-miTTm^DmRmTCMC97tl
zp$QJET1pT#E!W@Uy?k+$KZG=tQ~-e^P6Ub%l^Fz1Z5_Y?q5>|fLBScgp{f4+G(=PV
zD7a!05a5q#M;6EVI!+*3eNorBpmn$$9a>eoJV{Dy_iY_
zDi{vYRMJp_fJ>%+fmAXW1k=$}Qc+UZ)Kpi3Ksc=YTQbgI6(ucj4ef>Z3MBl8w{e8V
z(Y<_s3vA=$ZK7g!+~zpRni=9~8_4k1A^#SXfeQ>G04@Eav%l}=JWOW=;C^5A!7&Kl
zfDu($HFKbAAA*-}pci;th44q6fxdXOpBmD_Tvdxi4Gd7hYK72#Q>SAAeN<;
zkF%Dxsg)Jo4jRG;Rri9DtcWhomN+w-CNh9UB%$EyaJUXfSyn1O;E1s&>zLrIt*kZR
z>SnGe7~BL(cF;6;B&w1y7*`tFl;s3fXIW~{wQVgzpbVBtC_0Fxi9%yl%$XRg5TYv-
zOShuBAaGD9iQ?je03^+9wEYQaRVF1=RuvjZg9R~N7=8pz3f2*PN3mR8?Hyfh;r3u8
zU5f>0IWj^KK}4FB4lM-Y96;A)AI`xhJ~XI~mLHxP
zY)6JV2id^@MOUbw21-L62?kICwdst|5O5Zx>5LClN8wqnBpo|#Z51DyHim*z(=t=B
zc2Elq^$M|vU|pD0jbOVFO}dJv9mGLB#K(!mB8QNDELm3S5Q>(%AK5Cv$rY{|%5XNd
zU;#M7RD!|5J~+h23%nN!VE~LnRhiCeI>b;j7aw)-O~(()#9`IInT;0F1!wMLr>bdX
z;b7wo(e_1}0Un`#3>SiZI
zNE}OG11KnGYZ}@GocmJgDrj}06G@F`Zh@nlIhp!U$u=tX6g6*zrav5Mp+*VDF{s*(
z-excs0k|tP5PTeMeW@mnE{+=B5GcjYNyWuG0AcNAO=RJS=Gx8y{$L1E8*fTiwN&wO
zARx6vgAnT8!M4tHGiN(f6bwc;Czv|oNJt3U)(=N=a?Ini>gLL`#9Q$YEVKkFb*d)M^hhLX0V38ABw209!&Ma
z!!)Rt<{GLDf)h*2p5W!L9qjMo=cj=*!#nyp1fYQyp?EBc6Dds8oG9L|WDB?%i3s)O
zC=UEtz>!o-q6OC4kE0$qN`vKX?{CHd582j?NyVzEIAHxOY&EQPthB7uaGFGP2*m`c
z5nyT`6l}son`;HYObM}?xiwfxC8YMN+YEY%rD
zfjVgEQ0O56C>Bc9Il#u*J`m#XWkb_sngoOfS*Zai(p2#d762%CN(g`w%vs09mFe$A
zQz2j&I*ur^C6TBJ5Af1bcLq0u?KsY*%>YN3w>H9t=;&&yW2&a(t>U7iVMbzT(*13;
zt)U?(k~+-aE_l*Mv{`B34GnkWi~)fQhNQW;+G_aw(0tSpcr^_ZPPCzi*qK|AF?gtT5IhJIXacwKwlXs#!3iu|
zOXpyOD~V=D2<8Ca%!TZ44nv_Su70Nebo&6BIn~$8U)>&KO^18=nE2bPdy&8}6bc;d
zAyf+-)s$xD=-}weBA`Mwy
zqih+Dr~sfqvaQNLFn48`dz+g2+5UqJ7y@A)jKBm~*fPirCl`{AS^zUp1LaEv%%I_+
zK2~;tC|4xX5)4MuEYN@te_y1XtG6YA$O6|$bQ3QJqzR2o@rNVG=GqjZg&h(X?5hn6
z)KItaRwHX8eQaPs7S4DrO(#r%D;5<*GIgL>1_e24t5X?x6P7(DKm!ASZsiTg;Z<}T
z{H?sy(E$M{Yb#$bbuC{AQxipSMrxb+`#BKkbU#yPTL_7$;cVvPqls{JRtZ)Ci4(>I
z5BoUTlAuHt3vY^wi+U)Yphd>nd&6B(P<3Z5vjAkMZ=g%4kFyTM3*n1_I9Opqg6#tQ
zZ3F0dlm(sUq@_(qSi>FEfeXNKC=dxT>ZT}+Sr8$FLB(RsnM4AFNb|;fbNt6u)sKz=
zo&|Fyz^xcWKUa>=Qt*Lrh64ktl?#Spja6d;fIuNA
z6Exb=#EJ$Zp)}yuRE9m%nr=>YbtWSNRH-IlgccKucVdyuY?vCB7+W~O!58Q1tcrEO
z5FAbLAl*Z?NMsXxlh6PnhGvTjAZk$^98@#{DHtcTKiv|K#xR`FzFruXKSGB_hSMDJ
z5O0<>3gY4gauhNITncc+1$d*$FiUfqwK|MOqq&Bv`8%liYK3b1(#aG%a}q^cjTorr
zWD0l2*<)ZPS`N-W5I=%Xpg&1n(@fo(VTrU3@>91%kW^d);be6kj6YMuOv4Ln7NBiL
zQnw29av>wMEd4F5RqU}$5Q>?qWQwMxEeJYLjR3kf2F*m$$wUi(D+`plYOqa^nm^Hx
z=7=>TQ~lr$SS|ZdRab&D*2Uc4FBGeeMv=8BP%Q{T9T9}pv8OtE2RLa5qXSG?NL3ce
zgpRR6kqCiK7(ZL4HwzU+q+4652KZW80p93=1Xl-wj*10_>=J}!II>h|R1KQ|B+?cg
z=x-gY%20)d;4QH4QZz8%LJB$IcfS5NOYKspO15Js0qZ%
zk4Ewibsz;0X%ri>w~mF4J=!GzML}7b1~~ZmGOSpZnmSZ?2#)NEu?f&ZY1sIK!U(E~
z}j0Hs7SSkb4Y5Jf(GG9`9T;2{LS+!aNwA}h6OOLI
z0bH;nl;Hy6g#&OFs5VC=O*3y5OR@`rN}xFtEr}3@Biae%WT?HVt(JDb+Uz`&H1qv^mr7hD!%i9Y@cQy^MW;sKx0#Obwj=|OxR0!2I$Qlo0;w@CY
zHNZ=8s>~28e+zBAD+B{E(RR|LlC7z(CRlX{6+)+2K$xax2n^9aSlv6oo2FtFpkk^G
z$6zc$bcA8d!NJ!L4S_f!ZJC;yt`v1D-PSpns6lh3T2Qnt)G>4djjo}sqG}#&8{qG4
z?}V@qA^EbfP60GE2M4B#xt$%u3NYb|B0|hvgP7KK9Fm|k2FJhE@js<9`2J^RqGpaq
zEuY@BNph2=nTbP)$7I&FC2{krjZw1=2SlLpryP65_(Tu8T1~sF(WmwaaE*6uxo2W^
z#PZ0kE3d-@&c8O}*|N|7^A6#;Q_-h{%>{;!4PlVRXV@2JqwwhlR|fH|ZSCzp+J21c
zpX;a_u2h?CAI*yJ{aQ2CzQAL|YrKh@KYA07*og4{udfGVC1EOC|26R66z^ti
z!kQI5ChYpnGQVNW(~~M11;SsNQpDP`jT!V8=VwTx|yobM*2}K1~ib
zTqo$-aHfLyFTVpkHoe;po^+{<^W*!Y32my^#2>oj=3n1;&Cw~zoX}JklpX)CzG1mI(Z
zDbIO-vu>=-VFAH!gY{D|7vZ*U`T`;I>~jEc~RAQg-L)ALWc40hUhH
zy+JYkZG}J%Aoc%B7y#b>?5M}FxZl9!0U}CURC=lZrYFf^=JV2yzw8GK7$@INQvd$P
z&;HL?2k_Ufe`fzjQ|`br-=qP;TI#
zXVW+ss*?;$==}XO7q2?(3B9!Q@-6jUe-T3FjENU`SatU7AKAZ4o~(Z72QL}l@>h8*
zo)E<8>HlEhcmdmK`#<)Tv;ev=sos72k10O$15>`ds
zAF3!-BDyCoN)$Ymy#EZLnwyK4em(kn<*&8%o>cFfzxZDs2BGCswDC=z#E(z+RGhf{
z$E88AaWy#!XqYDZolzK%l>xk@KCwT}<0c3q0pjw?-$)W|%mTMY`cP$J05jE}eM9@8
z;KFhGG51q{vdm2lp3u6FJ==krNr<>HUY=ueX&*ZFZ4bFOmFwg$_y&ON08baiBL(XQ
z_%nVY#$D$zML5Z=y}#W
zHq?(wrMF7>qoX%?mad-NFnk6DbdsY9Embbx-FC+;2&fEnq=|OcLyXUH%bpCK-P(_6
z11{leEV;gNPT!@zcOt@=xAYM|R`wwYWWM*r)urjm%9}NGE^hvRMXpS_lhxTK$wyMI
zp5niIDSG_X3Gze_2MNa6#(^<66#M?dW1ci@#evegfJKvTd-Bf?xbv_u7n+L?@=STd
z3lRsrhyHGvF3ueza6~&`@xteT5dVh3Y(eKr6=5Lh6A|t_F@>KdHz#;(rbOBC#pgs;
zrgn3&d8uwOQa}-0F%@m`9wm#dByrK
zo-sZH%yKN&5(AWj=4LD8huk~J&Hpntdc3JcupF|PJ4HaH&X-^2K#vPz2G%9$(-qtWhc`oqFmkAdap=1aH8@X2aJkWeld2Ae!N1>7|NM*&wz}QJkHm)3>|5k
zI~zT|0#cP&ua`ydc9qY;5j?M>cw9ZjTeIi#OnEN<)E)M(-VyU;JJ2Whl+stAFmc%@{e_$ap35mh
zUjDZ*-L#V);phg`6l)#rEL=SGpjMPS?-zg=e$}ji-#K}c;GxqX`tY9NUh;rb
z1MbsLPi93nY*@HYnGqnmNIDtSzKQH^+`pPsGrptb~P*X9pSA;}twy4zy7L9o~*z{_K0
z?E7qI1&I@XDv|*a+@4%L{xKC!HCoy`VQ@V5vSjS#qQt$dQlwbrpk{(y4hls#gcX8+S3HX@Des(G>JO@x6QApU%ov~M?
zj-^1p-oi=JA|UUpL^tLFWU27%Dwlbg$lFvd;wsG5y|$}_&@&?1A7jh~^d@|hBkxt|
zXpzo+tbUM^-2`91N|jW2_E7$=k=bvo@Yf~3Dmn~;=%c;dSs=!rl8hD+h`pw=7ogw7
zSWWL#Vf25to;_S-?mVp$Mxz@ilitK~
zvkMYq?8SK`KZ(Z_PV$yX*9Hk~WtsgRe|}aa_dO?7P`Yy#fg;j@iy^x5$>6~~+@c)n
zJzUwEr(~ay4m_>i^Ss6uSn<2A@(LfQ{%!$&V(IJfw
zfVQillBvw=LY+eAx{+eQ8#L#R?2OsE{lQ}@zaauRHVlx2J$@_#{POaXXOr>pv(G>4
zu0=-KK5>6a=hc=6syImdbzrG4OK;cvUWHJem~Xtb5lPoVa*Gbz4q00U>^mJT&l4Xt
zB5ty((2;t9Kfp5j@z+!(Y^>g+q9PaDOVkq^ci%3YOEr0N_Q}SFu)LF9$B=1D_8^K$
zy;kQM%FfyrAH`TbvYShoC!VBqSLwbW)%b^R_D2T
z*m$-4HgdK63!K3FSkl-n$V3Q?esif8j1Jt^ZJ@wyT+dxl@B^S>=jx28clEqaw6Gc8>BDRuz%lxZOkKYjDk~^*t9xuRHI`z_l`<&>4Ycfr|$(o>TSy^NvBZn=cqOph`M
zb$6WU^iO1@(xK&MWwu8RdEm0Ry4!IerQVHOO~N%E%<{PiKK)~F2mPn+&R(O=a6u+o
z(CnF3#OqU^Pw4lmR&N#-*qFR-11z42>6d%^xHR6?R2tZQ;&^*19)8R~q{0g;VZFvXxY2pLFG=ntH7@=5)1T1#Q+bPBVTjdo
zvE9IUkwE69;5x^ltM3H>)-O>NNPNL%bA;*88BwQrL1f+G0}}G$Oc3{ye(VE$`<^}E
zITft{M|=K%A)+F(D{mpR?kG(G6qK%@@Vfej*iod@bBLQu&N@t43ieKI
z?%`HIse|8ha_NjiKP;K|UbcH##**(SnS)g}uhwd{hXU=@cV18o|B(@ogRPEtD95`m
z72KZa4^^}5<;;&q>Sxy^5{&pf_$TV3#Az>mnR~BM<_xD>oa(X3
zp19=+YQ6P`&|m)3x3@_MMlDZRV_#`F3QDYes$m<3`73;^`Qe(=d*bGodr4EnpDgOt
z*5;^@^6d@7Fxm_E#l&o*WgGf(&07z&f7S8|c50wfprB->AA-5Nmfp7?ey}#_@z-iS
zc1q=k<3FkvUz|r|f4~dUUe{;uEpca^X>9rm<`=gol}ZQFvoLRm05n
zV90#Ir3CD!h4&O`Mt`JQ>7=Ulxm}K5@{_%s-03-wH{eF%jpyGuy14jP&cjfvHcf?h
zL}=fIK
z@%{bKXLyv-Q2o@x_;I%835rVI6y9T^=?#F1k>JMMQnOm+1rnbq$*`<5?Dl-;uaC2P
zAe0Y}Q8@EwqqVFk;mW894_ad;MQG9Br$t5TiDDze#FXrhkwK-6_=yDBOt+l5er2U2
z@0Zv44(;=-T@V|J+gQ7tMw07m-!@qJ=p8HWTq3#hgUY;OdBytIqHQ$~-gb!{edA{C
zR&dta@i>6{@}sf^J9%qy2sTBl=$PT$(_pq!{X&+Xs?~bC{xUy}J<1R&328iKX`9Fs
zCPBA9ta@i{7`kvFNFJVkW#3_XBk$22lcgIM{c@~+H76`22_prlCuVowHY!+4Px_iU
z+Ahe9T8wE-Y%L2^lm2yJV|}5kb?7lcI3Y;6-teJ;j*Ex(tVrn>#(^Twi?7Z=~V
zT+D^zwkF{`8nEABQQP{&z^D&|<4b6>>-&>8Cy0Njp=1176kROXtZHjU_
z1L5CAhYY4d!;X5Gul<~!vT8K(w;n(tvbOP^1l8)10QoDVa1
zIo!OfDyN<~G6g&gDY%iiJn-{PO=IM?>XZI->n-y1P
z-n-MDZbvcTtKLNzeK|Kz6SpAi%bOJkX2>{B4y?w;elX4s9P(|Hul#s;W>D5C0&|cy
z*>+xes_s{dpVW=;x-YGnS%*JXy}fpOqNmwvy|Q{ab0dq1kG-yL^l0E|XKcp4gT+lcul)3LE_P1+;u&5hp*iWMUl*d>ip@GRN-jch(8$A#XGRon0twRCvW&!aDX
z$fquSP#VItDjkoUd;MkT!EN8<_5JT>yIO-9N9D4P7T=gpxm*9ZZOKVlEI|;R=Xp%2
zVTIH=I4qcW^u?t7*V><%kdo(@f9@EEVY|v^9q&#z7Wanc;@+O#r!#^~L2pcL)Ty?d
zy{x`GiE1)CFCsP7*CA2;{>XEaJg)4}x0s<=v{I@E!fs1krpHxK
z#SQnbZCFB}{lV+PhRTMI#zTiBHkR_2zV$XJq#BJ6p&p&Cer=m5aE&|@N&A+tEsDFMzruAile@=pb>7cqyA@geD-RyO>$8C<$VcHqSoHGV!v)F
zXIFG9C9Pi?zCT*^!_9oPd{$#xIXiow*rMT>&-6oTxKVY$pn;J9d2)lCJJP6}9qM>6
zvaN0G&D-GvN3D}XmPc4aKfQh9txKwLrC}%P%O1{td|_TP(USeWAL+j69?|Sj`rh_L
zvWG0{%#LN<>&fYYO$}Br{dc@|Wh^=f^l&7yMBehnTRnZ$uaHGQ@Aq7p}VAdVKdhiW%|C^_F
zUpHA-oxD_H9{#nwd;c}R^rWDw%k}F?-Hl?7C+Nko*Zujl{a3KKhRD&VL^4BHdWE@c
z)tDiCL&f!HtETefNja(P1C+`)O+!X+Bf=_JtRZ1!lK6r)Y;|th6E@}N!5q1VJ<~e}
zcX!x0M3p|AR8|YEKF>+HLY$NvOdGEX+E9`cXD%Gc2wAy!nvIJ44r$q!rg0Z*enaH^
zeu){yv}a>p@F)|v`H>QM=179T3}zRr$>yL@x>y9p43Q|njGWh>E=3q=tD4tqmOg#4
zkhHGH%1a&ZJwG$#t9mJF9ylfA1XaJxbjPC?h#D2$ZGK6wdIbw=DYkN)$J$x{`LKur$PljyH2H=-Bw&(b7TI1Uj;UU8{!|4qn&y>#%M&zFI{JUu*I
zAZipjoG@nmc&csPi+U9Hh|6Vut`-qmgr4H=z@3yP2+Q3|MkKs
zX~-iyjp=sod;iqaxG2rg-AaAdETvwG9$6Gudq3_>`P;IZ?EA+MZUu=xp$7l1Dcm8_
zHv+JuIDm(#z!}|$uWvx*RNN6B1^NEz5s1D@IU{FHJ!Iq?3q1jj7}ep98^N*XkFj~1
zZz-0fQ4WK$n0kg&V7t5ND(`8`9l1r+`@t#XM>
z#LFsq!W^JvCq(+to~We;BY1in!t+=AkzAzsvF2Wh8J_5xhQ;W2k-fKJ={eV(eOeya
zjD*c(HV9qJu^jf_JFjn
z(#w+Y=Evx{)A}s97dy{|+tq&gPEEvOXEb&{EK((jvTPgq
zendIDg+0&eJe_~%r-E7=r7jXi$EOgrNJY}hHNW&GbeaCRSO11WCN}n)-?HInZ=+us
zk0+BVljR=QUL*v2r-duKt$)A!oFNf0s5^DfxKr?SVM9TBn7#V>wF_}t?jg$vXyjYU
z7rqVnoQvL-md9%pINaCra6_4w?pSchmA8Gv^Pu{eGfh!u)F~-tCGVbo`e3?Ob{MV2
zdU9uVGJkcv>H1?_u&I*p^Io^m3t%3w5%1Aw&&Nvd9nBIwuOB4O!pE-C2a?fz!q&$h
znianT8cTb0f0Q`Z)N$u^y1p6|pL~$@K?mC?w1O=-7vWqu+g^F6e$aigap9k+1mQ%b
z%^Pdz(mu^?=6qx&X35{Z0Hm(=WX@~`#(iov@oZ?l^(B#8Z{@x7LLWS`)x0|1uiU4%
zZ1tfea;>4$sjDZsLc6JD{Q-bl{{17snNDFD;{VcM>-|RFc}dwTSViHfKmm!3o*vKn
z^qVtNM11T%FelCSBglxAdp8)$8;%&f+1pS9YG!tv_ljEWzCKo)NR1e6$k|X@xe8m=kY3I>
zcgNkuv-_^J8nl0Zb~1IZ(QJJlg!<8l{ZQ+pXd_Pa{jf5tIPJz_HBAUin>#pu3C5?9
zN`^CQdxA6D3$V>LJ2m923Wj>sTMF0YY(wV#ULEdr^sM=@6SWn#YQ3YLmTYG51%e5(HH^y`iv`?5YN4$9a5
z$fKt0e~`ZCqsEHQzy)h10r!IX0|K`*uEmj*ja)z$d!N(M)||cSpP?u}Q9Cj*Cub5HGdJ+{Baywdpi!j$isHRwqnYk$`=9O+Z?<@v{)d7%L`B5rJLvfiIf8Y>ipT~d
zje<8fP6XXG;Nb=}GA2o&%W0+?xh-nlW;&xJLcTlry?<2-33z|>cb!1*4p<(U&$e8%^;qKNlY^=X$qCBV=EUt%5&tF-)tjjW@_1p@J
z2MbQB)v$W+$b~cIIg00&5^lLV-`I0}aP$MM1_AdRJ<_TW`%C)Kc`%_o@n=F=Kl)K3
z$zUN{_3Om`vI_kZFuasIW4?C>-vb*x5Cpclg)BRG!kWvE50jK;68!p+d1qQ*SXRQP4Ms!j2Dv@A!*~`GlI9?lj~3w%jeYOdzjbAY+k~3{WBgEx>9{$45G}v>goBx
zXCFzGMYk8#IK9+7iNMV=gM`{oW!G|$Ud`sUlS$#bzkFI_Chv(!pQ05ssQ#zA=b_7Z
ze$=cUtv|WoHe$nbeMM`!nXD<0er^?%HT&2gG>Cl_E5324PFFTc*YlNSldR~AVZquu
z=AdI?bxT8n`=X9v%XPcKH@0CVp{oU%LY~@FtwW^2rF!RLBdNp_ap8>o&h+}nl}cc8
z%)GKB`hG4d>LBm3E+=s_=8od3q*zm7AEh3xEuFoWP0*Zm`&A;h-nV>tDgTACN7(hK
z@s!RgN7ZucVuO#15>x!yeZdN)opB%4TFhPv$_x+s_5QnW9iFv||SX@8<^C;h1Mk!-4t8!mK{rcek(@$Kj
zscw&}6CZy~oIDlE`fnF^FOKHt9n8;qYFGbt@$~^
z_G-%;^I+-UiwRd0PmDY2VgxwrtZD2_|48COdN_8YF)2S0>HpB{V(CTwT5`6*Bcah7
zv^&oDE4h;g?d=YLEixZJ9g|=0t4dz-eF4^d15Ch5{;6Zw*c11Umt-}UjqJ$_bL6>P
zQTTKFkhNw*MLE8#*s;OYYxMjoSPfMgKkK9g0=q9~J@mh>5!+lk&QuvFmws{wcZLrl
zy)8Sua`|F;@xJbc0F2V`=EXG!`i1bB!UK98cQRqg0VWzBbQ7W&_1+P<|1~30uzjD>
z<7_^LZOC-Skj@z{B#ot+FSqK|zG|0Q&J~bx>)KsH0YR|C`TzZ*`+wJA;Fp>F30C}5
z_DZhzefIs|bNTKwSn)r=DM+7pthD?mo9)?p`X1<8TCFPo-`9&<`D$Bq+;1MeX9g!7sUM4>BW
z-H!)-7jcWBBu;bVrX{CKv+}>gCH3hQ0UwV(lRLmt(!f%i|F?an|9??%OrBjc59bG+
zq^+OjlKuaYDCGZ4l=`=4-)0rP5y6CQf$xK?qS#X``-bNOug|3%*k!+r>Q_9$N3UH+j*VB0|aPCe72
zbTt|1yxVDpnu{iHc3z#lUE`gR2zl`X7U4AB%==QGE^~TUDt}{TG{P;(9PDsmUgjU_
zOc0c(j6G{gn=nh}?F-x82l|NoAy;>wihgxR$ZVUqm}NL^*X3_gpVBxal>U-1zVi4I
zzm$(-EZ8IAaHj~exD@o#VZx3#+(tby`DsFgc^T
zO~=JO=)&;&Azg1ClQRxKif0S){{$O?W3zl`hO+lb{n&UPl_FtmlQ?2(WRo*=S8>ynaGDj~8$-Yyh`@MdR@#a=+
zN;#i$^rWDtZT(MO+K^6f{41l1;>#Mdex8hs79@JhntD3_Unw8>5x3+otD4)hwPB#nuJ)
z?j^dLYqmaOEWO)`O&>U?R83EI|JHba_qT@pJ`(oChwQ?YH;)|fXT+PUfT2y1K9T)P
zzoM2Og)EJ1OF0(R+chO1Ru_uCt?CB0S18&Vr;y~N*-)t=Q7QHtO*c;cRB<$0Q7>wJ
z)#D?HAM+>|EsTxble^!&TtyGNIoG-=C-xQI#GFtq5HFlTYM4(MYMPZx(ir@?vY*vd
zt3XH^GkV2BO$M?>hl(ZiFIG#nx;~gtboAai!o3*Ar+m+CCO9K!D_dMckBL
z7yk>hSqc#7qY{Ook>&GCzFIBaL!RgPr+;Kr3R7Jv5v4GlF+7)%~m-W}&
zNj#k=26pt8p3b*OPIHpi`hZ9h_=avNJn{L?a|Y<`R?Lod6tj8zjLV9SQIPw;F!mKB
zD(gC4Zj#K>t;!SS*2S9j8w_3)@sd7v*JY>Nh>fZYwxlnyf^AeYot+#t^Cr2*qt>9j
z>!IW7-FK_6UMMas)ca@hnmZ8T&mHK{&x(jWwIOffSUv@W|(lxMw~ZUe;bQE%HL2MOO95tApnJBATC%
zASS;lQfu`D9G}0Q9Obk+{_KOAG$Z6i)zJKx-u)CZkn7dxQ^rzJX`-HmkdTe4Q~{P*{4ozdLW94)`Oe6^Vw$bTIf*6~+Bao0(Q|cV
zF(>)D+lAmuw)CIJW5~?Bkq|N8wzEA~~
z?ASR(y26`z49YRKaPj8ifYhb%#GZ-%*
zcSy0`QQomCG`Fx&1l6m611p+`-4+ccN7{G=QA6V(WAP5OYHaa&Qkk
zyT$*Zh}7V6+D-2#khzMDOVNm8wD<)@;lY)e!R~{ZtIYQkiKZzVPT*$(250ySNY2&o
zCd7&x<8$Q4g11TShNZaWD8MYf*ks?w9EaVV$8Fy+&9hdW;gP<46^6l@iN=`rHpdH8
z-{~qTPf#Xc&EBFtF*l8qAIx6i%T3z9)b(N}3U1S&sH*i{=7ERyqgYSF$n?Rx*|CE;U<7}{d`B37v6VupYW-8;47{X&MSJ8)y1Sb@!1
zm-b%u3bTZ~hCHuOxzpRP9V^y&sB9lx7TMf#UnNbqR!U{B?2VeJji>M5zH9A&P?2K#
zmMfXa4@-8AuD3|Otix&;leAuL2sxED(q(%csf%_>eksQDZEMF@A+dpGOQB=H`^q#W
zaoe<$rpt!+Uo3Dci-`!IDOrz*v*Yn@}BBnqX*?9=O?kA-4^|B
zbBC)ejL%IZd8ed^+^>N=VC_amxU+kr*hXlqcZt&%&rXE{Wy9MX`_xvlSJwsyBSpne
z9`$^a)}U-ujhv&-MhkQ`N-aFGh5SBL{RD$?)n105DznhjVZA7
z!Se3FyYsTnOG4dz!uf9WtjBXJdTI@$wLf&Qz-!7uxOP)?etQkLLru}8kmdJzXOD@EF=Oq=+eBaDme*A7D
zb~b0-VfyM>TbpLI=Q
zEL#b5q~uNy&Hkb&d#p}aXCw$zU)iF2Z-zKk#^xO`54zzzq+5Qr@v*xP$C;5fU@z4y
z{77zaWBiDGYR4)R>5h#hl01$*YB5ZWFfU2Ivur);K6_vD3JBwRuu+W^b*Y@YZ*qFo
zeU=NIvD$m{U8elwOq(6y>1oh(~>j;5QRmfnaY&AGZIiJ`(fRKn?1YY&X%m
zQXL1H7Rt||qucSK)Ql??Efw8wRY6)pguE7bMt?y&ai#K>CkVNQKk
zfb+;uhHCe`!$8SJp#)hBF5t?NwZ{o_Fn!oA#V%Tl+z>#E`4=y`C#yW{7;6KQY5%jS
z*UEHU1kU^;if8xh#hL^WDH6gk0kCzX?9pVx58TB-f_3f54(nG=RaGIQ4-zd4;z`#w
zJRXXi*IYKDE!-PkN3Gr5Yl??3E_Be%CU
z9YaHMemJ$1O`^-g+za0)t_p{`sewp0eRpLlZ(;f>?9HLORrMDOdC3Afdmz$PDpt?Y
zjosI-mg*7O0X|>ZTC7NtySjCr*uPD`uAaEZJ{@g+H1X16?*gN?7;Lrmy)tBJO#lpQ@S;$^D*pe&(~
z*IPjPM=__M&iggzdI@EOP7o=L&bcRYUJ=d2n_ZoH6V+Pw0x-1?e_!9|6!7?yg~#gO
zxp1aGH+OLjd5RK~TJAJ9Ol^#S4PTiOZZ>JoxK-WGOtqQFDc2PP>}@Bi?$--BuL5en
zUf6bWR%5qbR9=P>nD)FV2+u8S3vDsml(F`Zm8*zMel2vaV9{Blgz)%*b)_wvw>|l_
zxI%l$iOWKzwyO#4OA{h_yUSfvWSi#QrJrCww7+Kyb5__mTa2R^bLB&xS0zzB_E-6(
zT7Af8gEg~)jm|55^Nx?w5|Z(e2;NisYL?t~FInV#p;i(LlmJ3F%H60EwcC)n$M~x(
zyTu3`QNkD2^Ziz0zu2y=6TDK7grL{aj+f0FJh$<6ljO>cW{~C;!n`(rXL^MD|6wQjr%A(_E0a#Q46;}6mCWT(pGzx
zOjRhsm%skK6Bxh>6MKuU!OrX&JHi{&#MvP0c2jWU&XMV!TKUG=jd0u55zn1_io@p>
zJY*e1M_Ptbtc%Ydg}!rv5qRmD#$&Ws^m5P2{C{%#d(%UY^3_~I{WRrFZtmy4g>-#r
z+(L+5>Z;i0C&4`eh_{rd%K|+9-kc0QN4hQwmWY&oB*7+p%KY~krG(i^&RG@^YT222
z2FAZQzs^_!zs|tMitB;>#zHZ#zg#7JjZsK-nZKGSdtDS&8K>`m?$xO{+oy-Ayg(L>
z-537T%fJGOH)mh7G=Heu#6^+MVZQFWyNchm;%|`-{7ZkpQwFvoDz$oE`o-c!(WTEbxsv`Uc6)jaWb4noV;z
z&$>u<2OIpL9#P}-%>C|4U4{|){Wl%c=Bsqw-xp`mF22J}q@4MYKi}oAFlpDch26j@
z7-u!s`{on7_<65!EI4O#wz>{;kE5T7v+v(7Tvi+i)%?hQ=~$*|@J$-+9^I-+T&_(?
zYhTPu50($R;RtuUMiATj(D2QV!bVzFgu=wgYx$zGwc`8NOQsur!B6!ND<|K!-4>9M
zOBHT@cQN@Wm~84kgDM_lFJ7`0Wfk2Mxpk?u?i@e@Pee)-qb^6C=b0R^(QHIis))eC
zT4(&c%_T}{LDH}y#Ms?WeoNBuDUGU|DMh^?BtA4Bg6>aItSMW$BLnMsqb0@{XGrLq
zt4l1aZxRNyOsek?Vk5OPIxfBLoFl)(nZ+yT&Q
zO^jrf5bn;m&bSbkXQ<671))a1xu2aLmeRjZ-2Vmy^Eq+<9KA8O_Hph^$Ly{Cjyt^9
zlZ%>53)bt7oRwpJd*`IydwC^($=Aw%avlSgUUwdGzqHAvNP0e4y&0`lprNCcw+#Ki
zID79vti!f{xTLzIaAiieP&T2A$hcHUvXT`UAtZZ`#Dy!7O-2&3_sU9SX0Hel$sXDB
zcYM`-KllB-&;7p7@BKS{uk$?4W1YuwoS%A7tPm9pGuda0dsLSg!Y2P%FRu1zhQP0n4f
zH_}yH7Y!>%bJo>{uNeEIHI+(2izh1=>GeL$G;;(%#VCIu`ZyLH2(uP@$^4eLaCfHS
zGRs(sElMne{oCBJ907KBbSh>*z*
zv++N`lcdD%l`mdW*=fY3A#TXzgC_PvvQ243c`47}={lO3q}d5?+jsq|^}`$7B;w(Ovt6^;Vq*BI~oW>dr5&j3veuQFh4&
zjt@{ox)m&IY72prI(0(GZs2odA)i**D?JSEh6|70@LY;N+Vr9Ji)S-#G4=^Y{$VMx
zsx$|i_lnsC#=Am^(Uf1xb;P#oRNmS+rc*Gllty9moz{L;q=bZDRk39^Ry&!6q#`6k
znMv4qZC_Ft#N=^k
zzAxX+&t}fB^Sx&^T=e-nUy~V^GrQVqIE&VSQ8=-9wza2GwM1|4_rwunl&YZSPd7Bv
z2zE+5tI2tB`(2rhT%dxTS2%fHm(ban>i0AcdyV_k9}hByJLNn|=4)w5>JYEF8ipdLyhU&-lm1nwYofhNJ6$+Q*WQ8!QZeu`jd>n;R61^f0NL2_
za2U#3bjx>KQLV>j_KWp+w7aB7%XP11P@5WjA+2_{w|x|U$=5Ki^3!*i91seCtS;(1!H)r
zcIkHnZC2bnw`5j4obt4V;m-5=OM1HuuER~Q{03)dR)ZlK$$`QnE_Vy^Wu*o@K}?Fj
zk!;TEqN^NgetIKOU2t}*6DdoTNKtiPC0hTU!#g8+gpq;&X{y|
z4`trGA$Kn=$x3#8c3*t;rhG|4c6P?4svA39b@FQ?P1K>LflUv(
zDW8<0HMdA&h8bT{n`h*C?j{ZS`2Eo-m4u0D_mV1hW*MDX)n!69bp6c&Rd(eFoG~*0u<4$j1#V
z9GtUN{9i>}T;Dd3Flj!!g+b9rm?IY>T1P^Sd$
zbQX$LpH-eyDc3Q~%3oxiwAHigzWP?GI3p-8onk}hg@+B?Yq>}jp(*pX8zQuv^N91k
z2b4U@qr%N<`RDKF-)}I+8ZIjs8qrt)hZGq(9fwYqBlc_AHW
zKs`o~3#vPUt{TY_3~i>iT-U%kZrnG=={}#otDer?kU86EM$2LUb6NG_+cxR<={_o?
z^nc<4>*Ms&zMgAsxHNC)?{?riVz6lQs4cv`38#NqF?50Lt$3`_z
zYtVpXUiZDQG^YLOCzqHpwim&}=p!Cx-;Kq9JWsVhQqv@Y#i?(Nm$z8_&?tFUG{$TuBpNHp)7YidDp5;IkwUc(N4s%)+7@MX%gO7;bYpmcx`RhUu0W-y^mAC^nwg3nw$}|5VPm5
zm$Gj^qoL+%N_{_`slW5msV_=d4>_lNe640S%{>3~dD5Q_VV+IN;jB`eb@ob>
z;HO51&gzxB2T99W9ke+NvFsP2)o}X>~_QS8=H0D$PX)X_4N5omW;KSZ#iqp
zMsxd{%WRY_w^a8MgX%>D!laQr$!#dl>i$U2HGbwp&sEOaLXq@Vo|Yj|z!aLG=r`HM
zvW3@iH@LseYV+ECUr!2*C++a}h%davXSS4E~R
zmW0f{JiQ_8*6wmU&qM&ai_Pg|=Fm20hU+iuER~b@wnsEIo)wy}^$BX)zu?%jIO$b!
zGvkKOy&H)InrI6W<{nv>;rGe&rO$rJHsTxZx(qupeX>B%2dViX0=-Mrwx+>AcThD`;s&G6Ujv@C%NYd8Hsc@{o8)cQthsy
zb{DGGNjX2S-LodgcLh`hZ|frsBqYGgnJA>rftU&x}aenMe}OF&1P3IFzu$PPMEZ*`7|HmaSr_
z2R)IW%PB_n{#<<)sa1HCzvA4dLETe|k0*!#h7eZV<*O~^FVT5xuYn)MsT1;UM2O@L
zKQob5)H%L4XZF%Fh_Pt9kQ7>!suIJh><2FN-@561epUY3kx9lo;>b;EamVH`BX`H0
z)L;%ZdcytbJ2gQvRG64%hYu9PElPWv>k5fVg)davrp5a>fWueBVwG!s%DT)VqO!t>6(O1)gv4fEum?(91
z2NTmAYCLE}l(-qZEmsN
ze@a`$_I=}E4@;bPZ*&>^Fd1EN^Ydyt
zZVZWXTWAecTDj?yioHUzg#sB1n<8-{uk2U_KD~0ky?te#
z{Oqs>N5Gs)L!f1xW>t^B2Y17D6+5ywxiHqXCUmEo>rd5T5O&MSMrvN7T^eNWjDjUnOa!lyF+
zZN_o~b&-XoN!ysp{sMlJadY?I-{|Xt{1jda{cO^U&kYfxuxrCLI$PZAOAYXM>aX9}
zqCd%#GDtU+#e8b;8e@(q_YB0>C_OT^i8ZhSGdHIk%#9(U80WF5){jhD&2_a%-x+ry
zcS*lpWSLDQ)_p)_V^7Y7cz(bL>5RyW@;nisq(<|ESQgnm{&&&-K*0=P9CnIw?xD&`s89`5`~tSO}$
z;t?QpPK~>!B>-(*c3C+o<11QdS+=semoLKJH$ddtE1KHOJanY;kLM!U1#O%c&ocMy
zTv6o3&gvMye~0nAOflRJWt!yW8>h%DNSRrrx2l;_^eBeyO>=k+jVj6NPYr9-j&HYA
zNRc+2f{t>&Pf3!1ldazLb=*Cfnuh)dfj{N;0hQ|)DM|;o%K|hmrX+Np>H%>Qht>q
zbeN<+YbVCH+zTqYT|%B}n;phox3Fi49hT=%v72Wbo+Xz?6>q6D$gPZMH29V~sroUG=?+uaR%vCXI}E-Ob3rr$-}d+hkL67#BP?8^`;!_D+9)tjnv(uNdh
z5NF==NWJ+GKY-;I^nhXk9S}oSU&?tsB~5$RA$Bu$2onqKb`Oi6QlktVEdMBz^`)#^
z8_NvSJ?+r_n3{S@)*o&Tna3SFyXEXc9_Mdia}=+IzZ}XKXR$uH(42Kw4tHnt*vlRF
z7AV53bGzifNjh$$e5)dpXZB{^w0FNcj(sKYy|~kw8lP!>k`Kvv9|B_&vS+dVLD99z
z#g5(S-~bXDx!8`0&stMA9XKCM5V({ltY3DTeE&$`{_3juVikA!lpCFiES#Y~ao1u-
zwdIAJI&4zAE5(+?dPp0`&6^(TjEx%T16MkeYzx>>w#8d_lr7YB{}i$l6z7=*k}Qd@
zKH_161ddB3EX^#x^?qH3(!whx`cNe?J7M~6w)qRJ?WLEtPq_8*)M*IXLfb@Dkb>TD
zbkcZodi2_2I9$!5l_Cvf$O)g=DLjYP4eBaJZLltq^P@6#Fpk_wsm`tr+-Z&(^n=hZ
z9`?~6?A994%w(R%X2VoY2KTE`uMe);a~`!$om5%lf@f&;RfI)lE3L=Na@pmX=Z!h2
z+trLek{sY_hYcu);QktM5pZ}6(J1lkXv_6kI)aq-<8kL0MYV*#XRiNjPe^FVKOwp?
zx8|H-T38UpuH0)Fy&=O__I|TATsV-lz05f86Dw>?Bi`Pz@LUrq^9sdV>etj|GH~rM
zUHI0yG1u#m`WrtgDK(i@$>zfAifskx-9ER0P*sPW@vT;L?IdS=Cz)!DnQ&`Ld}O0(
z0#9bl$xL$s7%-fUMHkkU`ZylWG-k=;?i_z!k>gTI!rY~+95%0dq1m_UQ4aU3@Qfzb
z{c!V;y8!co{;OM9IV6?m`PGUwG|BM&y0VQ)sSlpeD
zZy~^(-zFI%q&D`ugL2;fZAVohjFVW!Q8OIzy8I_aa36kK(^K_bfZ6xc8{tv4tf8;u
zMgKDDP+ktt_H=yaOD9i^e->T1x%Vx#>F1}_n^krv&C_b2P3zKMdK(nuXrr0v{E}qo
z7k!+#`HzT+JN0%S{6{1h<6Ar$o7TK7Pk7NSErx@-*rVmUG=Py&AZAVv_#YG
zs@gEtib*#$cEyS`vxoPG)7Z0{RB<Gn=NL3i?pHLM$q(N#slOAqlWP`o+XFxBLzgNp
zIEy73YB-1eTSHvmMN-lpq#H1!{W$h26(nxoer05bDeh*poiYQIqrt*#$R+yc-Bfn&
zho6JFq}1Ls-`=jhZO4{wAS}8;V&i4o&vWFejyXR5k}gMsPhZ
zOH*l3p)L0bBSy9Ch#W3$fhgCX$HJ97LnOZkA7IO*^a?+)g^hYE?wSwULtjga;IIRB
zppHan^^L@};i_`Sf2PLkkQ!4O$mk~gCb}aZ*;QWUJ@e@aLs_x_5<3_v2IZTyDJih8E2Y2?ncVN6m6Q?
z+E3{zv1d`zjX4fIe0}AEC!jvs&W!mdHj??PYZg0F4`3!3yeBcS-zSIqsjMvn8hhK9
zp3}zFmQsg?p+ESwJZ`MKQbye;52YMQ^~P9+g_vJ~Ukdj`+$?Q0>`0(BGG0{zlLPlt
z03`K+Jjg>nc*TN&r26e>{pmTO%^V8^sKVTgJ{o=WWyf<(^lV+iFV2j1Jk#cm9_yd2~GQdoZ}?A1?WOJ|Jwhpw%jq^V_lQ2sUNL&M?P#&%4T
zxZ@%X?cxVM9(lIChmyS7ra?PgzdKQy`O=HlbK|WfmQMgPCbN{m$(_s8jI`zRT
zVf=fzp|Yg5&6R(h91d}~WBHAFZ99}`NayvFjyC(aUUs)3hXqP2x9&>@UxX`kf6Q3w
zzOuLP__rzB-V)7(pxbnbuVq!^O>8Gk{Ro{Z&r8TwyI4gr_Gvj}@kJWZM!goKu(SAJ
z|5@1}GE10Z#_>x5&zIt)eZ#%DO^t~2lY^2B(d6h2{{rU0@54`a4uzKb_vRt?IHwId
z%fk&UsEM-=fEYw_2FG(P5S?d1EVk=Yx?BMxh*v+Q!!WIED9u93;q8cmd^~$LETkWb
zPG;vme9%CRCvaVF_PfQv%)zuk^cDR^r1-e~&i>4ucLp*;B17KGQ^|xd@%q}<^A<7>
z52sOThqN$jvrUZO{m41BOfY6(3D>Wc76o|!KSTL;#W`}d5@y4GwW!tW-g|O9wMF%X^rN6qt_QOELu%FuD0j+
zef0E6?{!o;-Kv+2x!I!Dfxvv{^`WH(5ed)9l{
zT4sv!Xf>7IOjftN_?a0k7YM@uFOMLDyl>&YussPw=$CH8+t#$ZTcaz47e1V~_AQ+~
zNzAn_ay#Zouzia6ZQ~c29Ey%zTh8*YJYjrNrT`~yYIpV^;KuKbVsyus0PK#Uz045C
z^I(q;g5@ytKTQ_;r%So=QTSfX$SS2xY4lcvmT&TXA2!bl*)nY}#GU+Fa3zq@y*7Sh
zP*#Dds!!$tZ&_cEGFf76^3*m`UQZLN0K^0Un8%>lv8{+pEhdoskU
z1XRYdb>uJ={W6M!gW>xQA2fskWVA;f
zG4V!eo6Cvqc0s+q@28kkW7@6c8EELAyl3n-ENwCRv%){=w}PRlNNrRTig7=>V|*}}
z0AiXLm(M*NG+q2OB;OadZ`sQF0;Q@9EUCpj0t%0JOJ;HVu9Yz9@W=F!y4i|FY_J>EYGSOwTVpNu!Zm~pvj&Jlr1!OHE_unQ
z%N*7nzvfFgElv@iICfFBE(sr8B#~#$oDgy*;wn4d6wIy?HE%7aA}!kO3+@!_7)r=;
z{9_iN`8oqWBCQHlS=XD_+|;?6w^#KedS_U)4L^5utmi&ID
z_fD}MH9Uu}R+_sNH)d}UCScNN`7bieQrM(b2pGy^TF;U
z7xZIg{Q(p;b`IMi_v~Oj7M#mM$a}Zzf#{UDtf6yn3N`AD8N;d!*zgqya;bzb2(52l
zg6=u#(uC;RNmpn2yeVaYy0=PVW0Zy0?cqMh8+?x`*Tr5i)LL^i$Nd|>lxvp)4{Qky_)HAPtJ)_50
z_-sf@5j#k^8V&2QsAXgv;hKQN)i0ebbLzvXgh3^=FsP*EIRr`F%fAnc0)jWRYc
z)9ZkizsaoqiTUf2ev%k<8SSt+6fIW!2Nwb)t@BKugdL#2m9bdD`$%tLl87kR<#1y^
zTCqn?GP0Hg9%cdZ5OO9Lv)2Ht78!A~faP_vT6IP8L*#3|^A-n)&#`Wi#mT)ENTzA$
zf?a1wBw=q1z~-=sdMlH1!V{qfI(}m$2*-tYHnakMZKIZQ4vTADC3)M^e|C5cC)RNC$ncaoHSXvlYP0G<5b=x^GzL^Xd|6VBs!!%I$>7>J_z@m|gESUhedJMU#FrvvVH7qV=~3%^4Gcq2Pg-Z5-^mU(VBGPZ;ra$B@ZP!`<(tzk0D9)#gw%jw=Jt(hVSB_6pK1OUVri6e~ExdvR@uK
z*tSA+_vAlxH>Vp>-4Img;)aPlyOZ$Z1KwaMS7c`Zg%7CGfQEYza+#m?$JHFVH0c^PFpg_Z
z$Fvw8U;wHcN2yVQfFkd>Qais?SSM6m?S)*f9mhNWn3U*^BI?)K2V%3Y#><0IH~5%X
zT=K~ANT}q$&ue`P`=IVai9e_Sr}bszOo3I^BBcTp^O#z#vAtY}_{Sti5sZ60M#+j(
zb;?nI4}A}GAh%Xd|0KR^0^%bO$P&`IN?hV@(Q
z7)c(8`F+3e*yjj7AnKC;0Ce?Vct;I%C@apZ(<3&s(T4PAXVyvPuQNOYly7O^u=rP$
z?;}UUkC+R5erEM&;nKJz3m<(Y69nPBZ_V~R-bVnNK0d{N8t&S+NErIUJu&E1d>8}%
za~-}0_9`ypLej_psL*(T1MMlPf`BWBr0T#d_HD#hb)|68;#IX*BlTUMp#HKGc!DH=
zPi@ZY&x+tvC!c-b)rK&fluLTv>p#xp*DE)=LHPHig)-mx7U?k8I4;CAO!MH^UI2Ob
zJKQ@gi%a4n4~ed-aBo
z+Q$IZn|Dm`52`l;sjp(74H(frHvm#SsejO-U~xxJWru`$P+WIfM-l;?_AK{(4>D;%
zk>ntl@5hk+g4I6Q#MEqT^&j3W@O-RFPOc9iGO-v;;zW2YXz*IA5iDrf+Av!)1#;jEr|i0IJt2M5BOVr4Sxar`NZ=GY9qnEO3-O`zZWJ+R&97LVb#FJNuw^
zHgk6R3cfcpNUzO^3ZA%zF-T@|9DTub8}Rkl$F
zXCMG5Q<@HTUiiMV3)KY!pX{kJBD^5bl&dEc)?k-L2-i%wpn@!dByPq6p!Ar5%kN|l
zcM5YP9Zy9muh)nlWfj}NnF7*z8X)Q$yp2(Re;iSpO2FCgZHes*_^aWz#t#h3eE`fK@>^(r-)f8L~(sP*boC6L^gZ7x{Cu|QFB62#i2$~
zhvo)zuT?L2=cYLU%)F!^@z)hlmn)zyKe%g-z|WijH5ZgjAz=O^c&L}_4{9g`=T2&n<$P$4Tg)I`Wb{;Y$wqOCkWNSfgaLe}tI1p5&YxerO^7pfCg=tGNuu;)sn
zKhUiSFzHq%*E_9Yq1GUe&pMt*LEL8`WSV&@kYfJ&D-i+jBl?%^%LZ%^SoNvlk7Drm
z_4rX9T65avz;uB_0cFE!|igsr6#=asW%9{j87NO9O3{H&ocD!(YZ2z
zM?MS90+f(d)iB+_Lut?D$H-%RPsuoB)pmmzEkxMqOp=)y0W52(px=C^ND8F(;`EG{
z=LLj=ApdFrz_*jID|$MrQPUm=2kr^-!+WYOr?BRx0RG$77z8NFkC`@31oie3Au3?(
zxQ=J|)UIRN_8cB9Vmr@nZ&b8Yy}XHBD3@Go^mQE&Up34qtpt<;qu@|nf1}o66M$?!
zt9|F)Z5q74f1G>n%pv)QW7j=Kg>(p0ug}6ZuL$}cayVUqb=%-an8}kaAuEB3-1HmC
z>nK%ua4A{Ec4nZe42en%7vljZ`#tcsM5X8*(?Cc&mJ%RN9Juue61pM61TFs
zKEJ^8qilHpodd}uQm~2Giko`Sz~DsqKyFCY2p%16NJcIIrF48f#77nJI2Q!3Tc0~L
zAbw_QhodK83lI>7@AW=7s|XA=ix0byfF*J*ak@$ZlyrCbcm({(7DN>s3Fsst1Rr&s
z2U7I0;D`PwfN=w9mh=M|&wBA^;8sx2k1SszQ$iRl9!S^NJ~Zm4Fk(!UK0GD84&-vv
zYt7p)nRpSF63e9Y0RqKE2ox9jvVnq1ils)ESth4w$oTKF3tcL@K6rw{5I1T@o$?AI
z{oaZVB0f@1>=lG8+76#^_;Il0Q-$qyL5^-0)kF
z*$YC{g=@$?CnK@Xg|85Uo)(ujmf`Tb75_QqcVUECIFqh_iByi^v``we^
zHBZ8U(uSEr%J3)#5^>}u&CP$V1ZRzMl&abIK2E^e3I-v@AX%{jv?SmQu0Yi|@2Mrp
z8Qb_j+axWd8v7t1;!$!0V$_eXC`w8&k26Sm1DGIS9fMG6%QO?f`HILR(PA~8;69S2
znaFX`5_=<}L1C#N$?Achb5J+;l4J&JoF&NZe^eRxY&H=>W4DWMjw!w{K3{1TzAG`e
z#pLa3h@&KnOXg!f0_fu)$U^
z@K#4_L3{=hR^+UrD(hx~5B=$g0OTW^3%KADLd8N#p3C1Vh;zX_JD5Rj=-9Hs-fMxa
zDDTvVl6&-cN0D3#;mtw4C880%j;x1(t>Y%dmB1s*w}6fW7?`|BK`D=q=(G}h{yD8@
zROp(-Pj!|gupHY@gsYI&bwH-u^`=e`Wb+G=&6^F%oxQ($284454~Ymx5nbp)3wJ2O
z*Pp@HkDZMKmHT-3^-+BhxM}>AJZjPJ{0aN#h~zok6BFQ@d`H`G><9#jr{pH``LT^C
zy3N~AJMc!RgvNHTpQ_ppSMwQdi%{)ACJp$}$KlTj5zV>ZVO0M8>YvX8^Uv?2eon)T
z-(8+e2eJJEV!Asro~>9Wsa;x7<70?0X|@&UcUiyhxPM%YA8SQbH_-IiVUguXXD4RB
zM-eU-Y5adT6C@EB5irDD7L9*IiwB|pBqfl=Ks0BSG{*n$Hnmx$cSx-iL#tFs1h<*b
z;8aP@A04HX{&=+h!N1ZaWikCj!3xbByP|^$HtDpbj05Lk6-D^)0wsd22#(18EF96+
zSHes)(!W9!U>>|IGA;j)o3C^tl6J>E6^?r@X@Za@i4X<^^&cjgtMQBiNyTrZrAZpu
zKqenO;L|0j(-(I*%@T=2;QnI!Lay@%4jZ<(|Hnsu)}Z1
zX&-}(R^)e%;$E?lBG6i)w8q2yBxMm>BDV`QgJ>a%tj8a?zf2kH@EPkKtA#bX@RJ6T
zNMX-BCqbqcdx5s>wPg8=Xe9Jv5EV&$=4p->*9@aEHWMG*vQILvl>{F7QKqw>EDWdY
zWvCq>OK|aWW2Rr}OF;0N0gi)3%}uMjK3I3c(%4*FoPl+~G*vQ;g^G
z+#lYo2z&*E-X|CTHspFWgvd|v;67v3A<$O?(H8jmnXy|_QcSQu0{;ueu(#f`ysbd-
zlXQOW@-^0}nAg|mNt1nA8iIJ<4yQJ}_RpPpX@UHgK68Ch1^SDwpoT`5Hz@lz85L1|z-
z`s18TFueJUz}*Bt!WPAg#F}79WqW*0^cfL9Bser9`g}>)^Auw+l=l&vWfbscBm%l1
zmS^%%ixvT1KGXOpzjKIgOpr$Z)gvFW^MjOlSJmLKpcp(VBYW0%uL5fRA
z1ZZx#{XjM2uS++?{14`T_A4+GsRl+4LbOpr{G1pcjlg{fDu3K!=lI84{k4b38OYiP
z$C*t~#}`<@q#c@qoXth#Y*OZuhZo?Aj3@w{jmcb>N;~3W#2WmZ4$p?dSWb-pJW>Dj
z#*U4dW&BV0secdOhp|B-kkmA
z{5XG^#`=pdoe^V)nAe?gz0^9CiqYEz{fQ3(+RMSDBSU=>&1!2saIL(;vLvw)e^DxW
zNn4-6w~R@?5_zCIpc)^?c~kq4K
zoucGxCD)vJjy#!7S(zhYo*O{0(MVAj_Zcy(J1VhI#@(qB{Y1`2v69_*mk0kFDay3g
zM)n!i&j8LfsYJ1(1}aYWz~(atY(6OjK4b^4c1Cq`gzbPqvOC6@szlG!*k>`;ZFJs7
zko{ic2V{v)lo*l;+R4utL~-Qz5PGfPV-+XA0%uGf!hXa`a(DN)+d&}JnDS5G44S~_
z704eUvk);$+MWNl$-l3V6HW0L>m|VD49@LjpDruU`@i@ZEm)nYksUG1Xxv!9NkS(W
z2dBhaAn>_7R+e=Aw2f?pGvl53BYE$|MB$mEJ-kQn#+M54FvNfpSlz3gdP}`=k3ZGee~^l+2bEP!Yw(V#(x~Vx_*qNN|E)Y5aSQ@U(>u>wv7g1^%E^uh
zFUOVVw>$u0{?~#3n=D4@KWoIC)4#7de5}dAq;&kF)Cl?4+?hDLZsGduo!2DiDi3sr
z{0o#~St4|un`&?W8|%(nbKr8gl{d1}AGCk4L+7ZPS^agS^PqZ*m7b&^TTGU-`kT+M
z>Xu@SM|T>y5nY(1Mh1Fh;Imx^VkmLo4T;O+hp>2
zcju~!!dU^Hs0)^*TinZo;Tj}6i)%NRzgCtY2;DwW!i$UJ>sG@Y*_&Ni<5ef=(1!#@
zW&4hKO}>ld|4NQyU>M_!`E;{^mVszsCltYXC)H@*_6?q|S`=R-ahcn-&f96z3NVV)
zbXJIbMfw9b?H5QdzFCUU!qhpK9I)qG_{|HEpIoeddv8o|Pg#?nx%p+nrCYw2$J9DU
z3q`Q{y{|50tF;tl6rp41#FVtPUeo89qCuQ*<}kv)<3u$&0HxD
zubUs;{+PB&Ah*bza=suTO4KU;%8bwNI@w5%3*9x{skEG|v=teSlN@?zowosKmxcUr
zs#Ygu0R0sY3_s{$yT;TeG0UFU70~ixKLtjY=Md7T^B!HYAxDz-i@!cS2HX%0FPfP!
zVKfqI;ipV%Gt%3eoIxxh7qYnm{Ckz>jTJ%Bt4aGmuP_U71t{JqU1HkaD#)<4j3#XGH+^_)`I*1x*E(qi-WOU3pD0)W=nT#7TV)YJy1j^%}-QEGn_
z(8Nfox+4rRUn%6d^L$DdI7ZJV#M)w37w@(r%vqRF&r2AxwOavHsqv2A*$bq|nFE@2
zpSDf>3Zz-_gYkow`B7CDlf_wIB*~14TYA=X6UUIt6lUIM%v&P#@F*v!>(`DKrh@|(8N@(NoeY<_k64Q=SrosV2)7PPsGmgBQ}
z;|DxRJ^BB{odjh?x9^rz0a-huN?
zZTsFA<2HI?Hdh`pDdB=O$-4ZjZ32afIR@IrH@u@(ww%VlS?P)2dJ-J?$^E3;_1(<~
zJsArVw_7UvtlevESM1UU9N(_h>Ct;!8YEij-J+D8oRC@@CY%cRGE$+udH@eW6e@SR
z{N)19^n1iPP7r&b8(bGS#>;7}0<*BXP5xNsepn;>BnfinHbVe`#QXEXO{
z<9L6>PSI1W{RawH6~3Jx{Sm)E33F{I^3x4%gm%|^?5*^p6QdI(R29UX^Y2;ohS&2f
zlq9k*N0Tt|6rJbr;)!yrY$zxiw)(iC-
z+B_+W63otg#oJ~XSeoqp?In}pz%RS@VMI*_cDHc3!@rd`Y#*2NrH_B~Hc?Wy^w`{F
zWFwMzt>?PmWUw;w;j~7W_7x!Ta1#0y_I917sDy{GdA0i=vj7qg%X
z!P{c;$o^(`dfT3+PnSvtt<-P>6ECc={M-}jFE$rzTe3+Lt~@qg^)OPeJC~|J%8^rX
z*1%_P^;@e
z-(4TQoUbdPXeD{}qe1pS&(i4XYz0TXboI{L%_X{_KnIJh2<53~7fg)Q6}H9$56-f+
z?v`J9`0<@`h+*a>i>JH(dM|bD=c|7AjgJcHdN{T`$%`Hh>SElgPLVhmoH+^Qh412Z
z-X>Osqt|xRqHMq0TK89O5SA5E2+M&V)kmQ-iw+Eg
z@s$KUHuSHigzAdkPYF->qtl}j*JUg-qFg>Z)Rl1W?x#t!^v(hdk)X@9z07+?;d!5AVTN+*ChY~O`%CHLUc?-JobG4%l3qvqP${2~+Z}uwu$eMh
zb&v+&}=SuVtJe_P7=4|-k$&cYpkT&^kjr@26v38>y10wu{VunB7*ZK
z=-UI7Jk~EWmS7qs%IKBc+ia#bsx>OsxU9$ore0?Zr!ck$mi~%mJCnNDcjXG8<$rAd*x&R5T#{^`0l)LCt0
zgfa9-JM3@d)CIZF3hvh90^39{4iYgPD3r)%{vH&p+u(_1+jndE;=26W*Z?!^hVZ~6og5m$BaQSL5Ba5TZoy|t!a!b7bnC*GfQE=vB*E$tV%uIxGHwVWK`*k2h4
zHQ1gjfD}Ma;#*3LNNNC8xw7KHTx#9iwBMNocLnH~-V^YU!##+_L<^8suQzLWZ*tr|
z0+z#=_w|PRW{QSuD4FqXiCF{dZ@8^8QiTB0DYv~qgDyBOA}t61i<7An#G>|XOOylo
z)L{u}kwe~-#YX!u!x4WBs#!NUFAkMjyKy>N8cuEzM(WtTy|O!S^(|4kDYFRFX8QI+f^Iq)%S{s(LW_@92#A>
z8W)Y+f1*mu@3uIrca{6rXdnpa+fT2>yZF0+7SMJbQE<)Q1h5H~Kbx4Wy6KMoL0HgE
zp||tdC_40U->P)Ib%Q(8Szhgjw=~OIKa2lL3r*VOpI55l-T54yKq*>%Ag->r#%}L$
zx!U%|evV2sy0{rLtdceTISp<^@uHe#S4MtI5C`PA&of#7P+B0#$SF=A@aW2Xe^>@yD{(JV6C+A9r(yWw#FtGht!Ubu>qkx)4a=O20
z#qt}IRlTQ@-5q=cr!`q+I^FedJhfz)t7>nD<&%;@%SH15It``85er7sVLr@W$=d6j?h=q_NxWo-|&{<8Zw`o^z>Y
zGdI04D4Vicoqo1Mt5%r2fWiGs5Pels5T%BOJOj%?|7zqi7X7$8ZQb%rsr{N(`wDO4
z@1Gvm8|Tlh@Hk;RTHEO_8OJ}lF?sPZ7JXESy&>DL@cUIFWi4EyPGI9E-fj%vCGH-Q
z8e#HSrp=;|-6h5PpQP_*FOa-CfX7kSiFXt=eACNc+*|tA=ssI(UiVQhA(M-<-Eni|
zogTTU!mh)7kt`34%G~0uSV8;9K{>aO8OJ)i&hx`>uhOMm0MfZb-@pxHcOe#@Jb%nIgH?8hS
zfI1N{8zM)(oCjZ)hlO*uXZQJ9T{3q1qYef(0bOfo6l5-!;IjP8!ucS*{Q145bmW4Z
z7B2}&gxcjLeYKExx~Njp8sRbkJzF-fM_}Nu5%^3TG~&Luq7pW4^#~EhJCJrjVFjuj
zA8qK8Qy8vgUP9^*4tSfa4)%4%^+J);fxYh$P2Bu!YEa_HrZ`#y89t~zcv|5xZb{T(Pl(S64^ycP-@{)y+z`C?Ms$!G
z(Ltm@L-0NJ8K)64Tv6B++eCek?e&p}-zONxlX!uw$L
zuW2q!g78zz48b{vbq!~{L*=RQQ$u$89o*+c8xda;6@mpP>=nGL*PY;Dv39Htw{pdB}hn2-^T1AM7vNZ!v3
zp6IlFsA{JrYeg=+rK;Ae!UH8gMz0l~6|5i>dn|JF`L(MbsHxz|2YGVIeVL|iv89y-zv+HO%08XF=
z&Vuq3(8f-$VbLkcr9^mZXfWJ|FZ4J*1LMfj%|m}Mr8i`Ff+%TlX1KZ6uMOZXLW;N0hr`iq=d;kio|0iPuNJhyr-_+f;rj&WgeI;@U*&wp$IAZF9k^j
z_|gs69&!Je?9*!X&%kPs0^|&VPvLhp*<NU5-`jO`_mvl?s`w{uuO@+q__YT+W@3)>Pq^@`Tb|3
z)%PKflgo<9QUxR7cysufXZ#bkp+p4|OWG
zk%Ac^>{+#Q{?4LukY%+_VA22OZ^3tt2t_jj;!Fw$4`OP<;Vs@9-+uE1T(oIJPwCoM
zUU@vKQ-2O%nK@qS@D&6qap^L#D#dYFR;+PbcbO282m}r0q>tUPI)hK=I-z8c@bD*A
zmRO9F`~R@^)?rn3UHd2?DhdLd4#|zEbfbVstH2gSLKkkq!o}BQ52**q(izx
zT0%-v1f=VXh0ptb@Ar%Ey3VyuTfDw88j$9*$>uP@1&=0Ta
zBCq_Z_{^azS=B=BK}|*^3cvnTaW-7~%)reW%rOQ*=visugR3+CfF?
zV-Lj2X7ud3j5Gpn9edb*lf?~4=a8+`Mb3$-H7Mc1$b%7>b`9cmgW0CGW{&t&?$jAU
zuDs|w!Izir7OVA(oohK}lFP+>Oy95hGR{U(VLWXtM-`9u$H&y;U4Q*)YDFQKN0Y$l
z_oKPcHXm~~z<;pg^?+mCwxig};vVxo&TcxK!IWamDl4JkE^Q*u;x?_PYx^%PM7w3E9HRYGjt-Rttf#zTtBWL2^}1{#7&Jf~5GV
zW>Wk50J?pB8xXi_*IdcNj72{?&EM;M=4h6+7q5n1-?b{XP&KKKTM{Hkho5(c
zsiI{U^|3Vvi=cF%joS(nP}9NKsd|=lV1*xcZz|F0Q;!OQVYPRBa(H9mYp3TU$zRkc
z*Xm76!!f(~pa*WRHAzT?#k4AZ_wr92lwX!T=InZX&8hBQnr`p#HO*$@@$%5hpxNG?
z3?ht$?U{mKCbVnHXDWp!O3j_vwdh^rzfEUyTaH_FLDF~a4yEYgfPX$GL-;YEsP_D+
z6{_?!nYd!-&V{cR)yH>_|2lsXo5Fxb4uJit&SU6X?ePOwd{Vh2oP*+)M`b6DTl^27
zmo;1u`I7C)b^0EUuKM%kulJ6(Me=&jX9S+E9=ZWc1o|AxI{P)04Yr9@
zoB(UPjF%9CLZ|2dVc%SY{10Z`d$fcq%t1Xwo^93Il>VMJFY`ilMm`02>ze;@2mq>g
z=89^yLB8pNA!&q&stbTqjH;Ma-s;feaSM&flpc&0U|%r2rO8TR**3yD8a0%OcuGxT
zGxP=)+L~irACQpbWt)%L6!EkH&sAlZY%porTkH=&JwaV88DX1ZpQw0Le^ikVTotI#
z$-joHPy(nyD@ylBUFvNd>}uk4qW#*5915b!-v+-ratGt9$Py@-N{pMA@pel#8ZY~o
zWEK^CneM?)I&QAUsE?>F^Zx!(CmhvM!TtUAOEw-9dVP!0?DT@j^K58;q9)9D-|YBc
zah;MkDb&4hhjlEt^?2-(P+!sQEY0
zP~SM{mBMl|DOyoqF;eiOWQZ;6ou?$99Ld!wydEr-3QZS
z8)muG(WGetud^Q8^S?E!7D>?dc{=nK(L-&2uoQ-~SAwoaYofxBrGw48V+k0dUC;{HseuLy7+SN<+;GB2XVu1T%2aZDd{Pv)wNuA{zW7pr>JT9=W}~L
z{R4;MoM~%&wcvPFI2WjWy4`?>eT5VG4R}RWnON6;jV=x98P%KRIJn;_2)GMqk%N3I
z!Qevl{fAJ&&h0M9tGV6wAgMeDb7xwyk@pfz$ki&quqbxDzam14Nb@+(I1K_w&mmiv
zbp|ebAgSHRr9K+h+(*pO~6}M|7@o6c_9j|XvCdBb#DNlv5%TW1#$@x6e9ky;sw0{2e`I1
zrxg)p&3mtA62h|xJD^TqKIQ*)ZmOsMs@zP%dmR9LcjPNHR-Q|u0?(y?S!g&Vj8$fS
z&!_-{US$&UkbuYzCA_bDSuR!E$T(#E*b~sREB#R-c8s?5vH-|&hwyP9YW=tz?BbH(
z-@6pd2waSp^XWU604ky^Es$0qV!?u8KWen9XM@c{@)Nd_5^LbZg9OA%|dU
z{uJA%?$m=vyIazxmCg}LnraPI((F1}+4xs2JuDxdkyh-oPcfX1yol8ob*5Y9sLKMsYg^Scb7{
z3#GV)bq04ZjOvA7Y;~i4W44WiQ~Hggi)XV$YVx^$PL(HA9~>y(^8+RqpU-rm4TA!1
z+FuT~;NqU~tvWacK5A$5{n6X>%w}4hf73XGIrCw&Up@Hs
zG=Iz8+vh1QhEJu~;zxLw1TcoV)$qi2<*fk3iF4M-F`vpABrz2Cv*rDbUD=QTT6)eX
zv11J;ewM}%-CHMDa+2&>+2e%G=ptFP#n)@8fI8P&_RCPFceP%ye^>CrrbibrMu^m9
z0+-R`bxu-rY__$Q#=W2^xN|e%gEW`?#{!27f_saBiM5YKR8sG==**zt7GJEkbE4sZ
z9X|>ko|l{!lW8fsdLpL8u9n~{e59wYYNt?z*_*E)x5eunej~=$VkAX>j^~0oGIg&G
z>?miJf^nP_&Q#68IF6)Zm?Fncn?FQ+`u=vMGb@O9j<`pm<~GS1S*%;*%a&6;m$v;y
z=_?sE{QVCGzKk`P8cjrv%oYl70MvSz`}#>WD>
z_A5Z70g*~^dPZRo0H%^FM^d2*@~(fayR_BY5PIifN^rEKKWel6;4WOCxfTzixpYT@
zx}Ll~*49HGSMRNE#=ua44&YX%Tt=E`he#MF8A>texo7`!d_Up_Y&KTj*n1*3K0(1=
zec!5N?<^m8k2Ptt%lphFrfx^V7mk&oo_rtKc1?2$c%_;Yt-t;WNy_X?ek+o*PT0;N
z%C69aHSr1k=GYZLQ1a*@wL;AEH>YLz_;FO)!_sUT91f@c$I0q=-y~0>EqYCH4$SNB
zP07nL&Fh+)#APpmOn}b4V8R+^%a@;vjW*uk-z_v#QN2A7FW71v^~OfLVKW8_7A{4{
zCUi;hg7j_?!)~2rfkT=94htdjWR%;f&`2uQ{9O@`gr2cT=()itczVMv=6k+fQ^8GA
z1wFlR?I1Dq)Qt?Ah*Cjm&bK9((7RY^9;PHv?4ULPjIRc-GJf>8d!GZSH4oDOJLEjS
z+B*hs$gWKUbOa(SgKi`cq=iv?Q~lV`lMMcSH<3`{Jrxo}8KOJ5tVs$L?oIumq$hb*
z{U>IpsV(QCPU+tHjDRBjqR=7pWDF3xr0*$J5}TvJIp)oX+bJ^SRE8Z(4MZYTByW3G
zSyfOn^%)B`g_tQWO!=IIizfdqTr|FOpv8+&mntF2BtA6s#5?E6KuGH>j7{g|#SdR2
z_EYlF`KkZRUfr@%|F7&7ya=PA=cc5ize3u)k2HQ?xr>Nq$Gp$-e@YB>SwB;kNvv$&
zE68f!mRXeF$6eRW6PUI_pQOunrvLMW|1DhxFN7%HHssINxtVwiN|)0xa0+QC)@i%a
zGYoT5A$-KCQCQ$>>i_qZ0kHy-C>i^7dbRYYKH2@bx9HBwWtXt+-6EL~obj|sLmj=fHCe$Dx4_eQ7~j1
zy73@7)Z}2PmRnYDA1F?CZ`8ZBT=H_K{$~|Q!&r30$;w596Kmt)VJX|vo
zUpP6;W$rK)Jh4UT?vz>3rB-5+&)6#d_D}4o`e3O0{tD&ht&>w#=6`C{I&3UgFA?r$
zwM^0LW5&-FZy7)H)oQP!2mzM^m0j|i8AJ^0NO57;P@8OUZomhrfJ%EI{J%k^v~VK(
zab>g5vvnVZh5$3YYe}zKR!AO|mBryC&MBJ6E((Qw-5$eSboyTL!#;OLT6Z{^D%gS!eDm}StIiuB6JCNS`Yj~+2!O73yXrPki6($zo6v61XgE$?&)lGIfs)-P=^_^
zggg5sf?K^~(@+lZfg!)o8t)Vp=)|Em|kI61j)Y22{M
zUENn!&?|(2FF9d)nP5|^*dn;7|e@Y6%dPMHg!Rp8kO?34(nwIur
z$7-nTrM!4h%$dhF{kMvcC61kx{u7@9c13Z~BrH)$_mlnpm~Mmr%xI#(%cv;out?4i
zBC9(;cr5cGkGXtBHWkkLqihW852$Yi8r>Ba72)A4PE&j~;nG`(!jXL!_ioF;kX>#l
zefN=PU?56F_})y2D_k|9zDuiI;_aWZGOs(<&x|z)bL~D{{PEFZgJZ$Wrs`?W@2%p@
zlJ7#cdzeO&I>E9mnnh{nCqa>H&uw#mjr35)<_|6Mh3E4LtQUK;ocbeBtFxW_PT=Ne
zzSzS3`sU#1U5&co?g?>T6m<}MY4RNl#2xCdzlN!Sh$*#9yHIz}y
zi_W6z85mM{_Uz@%vB;+21#J^)Gjk8_mtJi?(X+w5J~uUZnWOela*Y%F
z)QA%pQe*Jc`g)Xw2PZ3j5}Cc~n&g=o+6s>qIbau4!XQLhHv7$S7m)HnhdCvKX>W9<
z^IGII5e23G)2H@Pn00La4yAcFmGP=^Cf;?ru!c*qm0%$Xe2Vu_-!ea7oS1(gQh&$1j}UiX1r7T
z+=wAXMh=*q3`Jh%nIQ$U6-XGxqTW(}i0_I#{#7#b(qLI<)s>24wV*f9_U?868_XelMojC3U(@fxJ+Y
z7}IcS%qOeu3SSWhb`*zC?3
z0g84O_YE4~X}_Bj@s5cL^h14oJoZUa6lMtw%5BNJQO5d-7q$2}YCfO44=a5X3g$a3
zt2HOK(FQ|?b_)B``gWb#b$G9?xKXdHT=d^Ac3D9Gz(f$2lja2BKyupR)1#p+DDSYW_|Ms%ib9A-q4Wj_*0;%p{fVr
z+n5G^FCR*`Q9aUq4b#x9zA$LqWuK=#@K%ZR%!VF)Ox5lDfV}jw;+Td-_7(oEd~);%
zjq5@t-AT!+bfq&;@10$o_=BisJW#MGM{%;e@(;v9ljd^B$b
zWEFqs;H&JI8XPE9kpXyAJb)8Zwx)W8&RQ{U-JZEo(&dK#W1TpDMp;0b`O|5oSwR7T
zKaNcGXI~JT5fTz|aXsmJv+y-oyOsv#~D+ne8;GZ6zT_*8VTy1TwYsR7_
z=~7m2PVCNgT{o7Jv}E^@K0(KV(k6R%OV?#nwA>py*Wo~vbx5VX_(y`{&z2uECt8pm
z1pSjJM!89=Ts$W+CXn5o@OSc$fQ77gchCa
zQ;cso5+B0{1e;ci7kL0r@T&UI8wJwVaH3Z&dw~@Tgch|5BNl5@4frpJAEuQep-&lQu@3>t8t@(7KQA0kKUYmGn|>toK!Cp5l4M+hl2lg1Apfi
zS>KB*yfuExbwHEAd6?7LdI@K{Lb#Q+PU;EQm*6ZmQOY$;9$wGVra1cf0|AeN9gVxI
z%ia1XX1C8o;`aSdzzC41tu7Wd@H4x~M%;)y-YTas=X>n;I`Tk|&erPP
zBgxLD<&5$|Ez;>gv1hzrRvK@r9l@73hM}mRx!57HIbqE8qHyw36F*2SZ*XZWe|dtk
zJdgy;@N?_lyP6ZF(Nz*v0TDz#>lwV7gDrdE8#mbbla`KT$4v94_s5y4);lufc6H+S
zRc5pIaUJ3k6bW;K2kADmGBbe5MJn-aNsIFopHQD`?Uo4T#G4Z;Z7BGTULYit|B-*+
zQ*B3Z?8e7Nu`bL%^$V&<0f(;0;?b_3q&~+Y26Y1oNyvEE-xnOPRPA!BZhdO{w;obMh%rc~S3`T%%E(mhv^lG<7A
zqP%!5&q#c|expuObh=Ia9#c1fNN_q?RKTpq;-$tUnMiStRhGxY!^X;qtWISf7-{$S
z?eaM`br&;}h%AVxHD7v~-q6Xwb{!F_C=6!lSI?e$sg)B%#pK+WNt)sJLP8L=&#I4sAIKr1bL`ZH*4E}
zT~O(?dt4{?0or>lYTN4Gh0{H}B(J*gwKv}|zv>N>Xu=lj?d>Sh)tWYzg5pHV51|_5
zF6)F1ZEb8SvTuiM%DVK^S6ELIk7?H$-_))5N3Pl5KSei|6C8{0h2=%AP;9tG+UP0-QtLYA
z+GTlo?f#i-9P(n0O%J83d>kVRE)t{5uHB&C$Tx6P#J_MuMsS
z%$ZH1AYxx%-=Nve9iCHvi60OO0kph~nOwbVR{t@VNHE*B9_32B)=^gp17F6k;2gev
zC0*iqKG+EU1zV_2wG|A09XdC2L*nxtyf-q8pGe~i{H%5aL+&xl=rD+9erk~sB)$?R
z0kHeQG6?McGX~zI`g3>Gh*QHPXblT!>}G{b$_so>al{*55ArCy`uf%V&;7QQOX3cu
z-mUX0&Sgja!sA?o!)BB#n`K>97VN*BYM*vhUrA}de(SC4@TSxlcTP&qXZI)EoF=*u
z*0T%`jF(%nJ6@Y~8cRm+CF{Kv>k&QTJgLf7FNUjCkGupLqry}7xy--%lsK(Ty^!*^
z=5I{6Dg~%??PS?60C{8GtuH^va$Y9j#dOoFiR^Jb;m<`v4W=UL^)@0>ZJw96Xgb(J
zD|^KQ5-PMxvU}nJ<39$#s_PEHCL$ek8h;wT^%Dm6#r>kTTgH&IYaH45QF`#6h{g4rh!Sz|55g%kus?&3&0&iXFZnX4fg($BPJuwd&4g1ZqDClNW1q
zgeqUa8xBC}hEzi8^bQ~O&)7cU>fZX}*ujAncIeaLL*u;Vw$}RHOHo*+m?S>WEcaF@
z_(n@zG}S{i!YMB_YXzM05*EW$LQjR@WtZ6N9w6KRu~&8Qy8mZi?
ziJQezJJV9Ko`3LT{}(Hvn9aKfe`>xa%}ZO1-S1~nblciKqF-#s*z^hV6(}trpmrG
zTw5(0ioJ2YkfmtVt{PK(woonjwPL}XWK{PjQY?F-{g&$=2BJ%_T3lYm6-nsudX^2@
ze|7ly$ll!!h1OYYFKd$0`yFK>TXEw^#_Yg>K)Xb{{^f9S1Hj05QF}92U1gI0-q%oo
zPG7fV9rk$Co*Xpxt7?@gZ}XEyS_HGwP^ySyTac22omqsDFc38caH_+=-Cv1DlhUXfOVc%|4W(TgLKwaZfcU3yo8J7bU)(RqbDB#Frn}UMoEM
z#3`=1PdZyYvDc;_#Gk^zg%rB3Sk76aTeP~@DMTDNtjW;dvPA8a*CCBW=Mkg=h{M|N
z*KJ0VfDPl4F>iNdkf&KKX_7w?|7^Ph2!{@ujOcP^`qg%eDimuAi}<)56