From mboxrd@z Thu Jan 1 00:00:00 1970 Path: news.gmane.io!.POSTED.blaine.gmane.org!not-for-mail From: JD Smith Newsgroups: gmane.emacs.devel Subject: Re: Tree-sitter navigation time grows as sqrt(line-number) Date: Fri, 18 Aug 2023 09:39:02 -0400 Message-ID: References: Mime-Version: 1.0 (Mac OS X Mail 16.0 \(3731.700.6\)) Content-Type: multipart/alternative; boundary="Apple-Mail=_E8416D76-63E7-4065-BA11-A39F397F46CE" Injection-Info: ciao.gmane.io; posting-host="blaine.gmane.org:116.202.254.214"; logging-data="28568"; mail-complaints-to="usenet@ciao.gmane.io" Cc: emacs-devel@gnu.org To: Yuan Fu Original-X-From: emacs-devel-bounces+ged-emacs-devel=m.gmane-mx.org@gnu.org Fri Aug 18 15:40:23 2023 Return-path: Envelope-to: ged-emacs-devel@m.gmane-mx.org Original-Received: from lists.gnu.org ([209.51.188.17]) by ciao.gmane.io with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.92) (envelope-from ) id 1qWziA-0007Cb-RR for ged-emacs-devel@m.gmane-mx.org; Fri, 18 Aug 2023 15:40:23 +0200 Original-Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qWzhi-0007IN-6h; Fri, 18 Aug 2023 09:39:54 -0400 Original-Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qWzh9-0007FK-HO for emacs-devel@gnu.org; Fri, 18 Aug 2023 09:39:19 -0400 Original-Received: from mail-il1-x132.google.com ([2607:f8b0:4864:20::132]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1qWzh5-0003m1-US for emacs-devel@gnu.org; Fri, 18 Aug 2023 09:39:19 -0400 Original-Received: by mail-il1-x132.google.com with SMTP id e9e14a558f8ab-34bad74fb3dso3130925ab.1 for ; Fri, 18 Aug 2023 06:39:15 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20221208; t=1692365955; x=1692970755; h=references:to:cc:in-reply-to:date:subject:mime-version:message-id :from:from:to:cc:subject:date:message-id:reply-to; bh=PJBGI4RXdtiYgfrXjxz6pjRE++8AZK+Rm8cALeULXxQ=; b=M1y2VmuyqWmOKwGWRj2LvMoHIXaZ2Ui9/tz/um8q/uAvVgFjGZsa6rfJBM8tnkLSeS C/evjeUv2QOcJmD3VTQlAuU7b1Kp6Ixr/sbBIPqdowxGjHOxSD8W953HxamXekaBQpNQ zYfl60yulLJscr9/IiuYvIw8xXGhDN1chgaSt3UFLffXP9NhdVeingcgGF94ayfe8qw3 /8HIhiWxVBsgJ8wuZpdFqbry08Z5l2U7josET7Vq64a47yJC8epprNW9DC5LHFxg7X1z Zlf4qYB1iwQxw0z93RV4IGcz0qUd1iqixNy0Udq+K5snqiqO5NC4dGYTll5LL4bo4/p7 I/jg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20221208; t=1692365955; x=1692970755; h=references:to:cc:in-reply-to:date:subject:mime-version:message-id :from:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=PJBGI4RXdtiYgfrXjxz6pjRE++8AZK+Rm8cALeULXxQ=; b=CC3FDSp5QeUX/xcVOwV5fYAveREOvMhXSXRdzgPtORAEBybCb6oiLjTPX1kd/yjYU9 A85xEyJBae73T3k031n37eVi7U2/NrumIBpfkH2waDMboG3m5dnG+azNBeiqt8MEpQnv q0mq8vxDUGi/9ztg+0DTlYhGX6Xwmj98xKGBdstwNK+uhv1GDp7W17XSSfmCa0mB+j78 KWYWSYkB7py4ontTcZVeeA6K9Yk5IatQAfSFOCNJjm5+awjjS3UWhpmYsFqCwRMrlE5Q noDk1lvCaeJpLS0mlL4Z3Qwts4/oOGsQpQFW0zTFkSOrnemQueNVUyTxPIkBqI62ATtn WTJw== X-Gm-Message-State: AOJu0YxRmGWjtZ4ppk0QBCFhq214oczedDsVWrMSAuE1KqiZDx4M4c+6 EkoqOdYzYNtYj/O/RBLzZ/U= X-Google-Smtp-Source: AGHT+IFk/1aUo8PKPqKO6RekcmoMiHhELb08U1MiUa/QjApidgPTk+KV3tS3QNsRklIpY6oE7WjENg== X-Received: by 2002:a05:6e02:1294:b0:348:8396:9f88 with SMTP id y20-20020a056e02129400b0034883969f88mr2645606ilq.9.1692365954548; Fri, 18 Aug 2023 06:39:14 -0700 (PDT) Original-Received: from smtpclient.apple (cm-24-53-184-115.buckeyecom.net. [24.53.184.115]) by smtp.gmail.com with ESMTPSA id k11-20020a926f0b000000b00348d2e9701csm526172ilc.63.2023.08.18.06.39.13 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Fri, 18 Aug 2023 06:39:13 -0700 (PDT) In-Reply-To: X-Mailer: Apple Mail (2.3731.700.6) Received-SPF: pass client-ip=2607:f8b0:4864:20::132; envelope-from=jdtsmith@gmail.com; helo=mail-il1-x132.google.com X-Spam_score_int: -20 X-Spam_score: -2.1 X-Spam_bar: -- X-Spam_report: (-2.1 / 5.0 requ) BAYES_00=-1.9, DC_PNG_UNO_LARGO=0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: emacs-devel@gnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: "Emacs development discussions." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: emacs-devel-bounces+ged-emacs-devel=m.gmane-mx.org@gnu.org Original-Sender: emacs-devel-bounces+ged-emacs-devel=m.gmane-mx.org@gnu.org Xref: news.gmane.io gmane.emacs.devel:308894 Archived-At: --Apple-Mail=_E8416D76-63E7-4065-BA11-A39F397F46CE Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=utf-8 > On Aug 18, 2023, at 1:21 AM, Yuan Fu wrote: >=20 > =EF=BB=BF >>> to go to the parent node from a child node, what tree-sitter = actually does is go down from the root node until it hits the parent = node. This process is linear to the height of the tree. >>=20 >> Do you mean linear in the width of the tree? I=E2=80=99ve = color-coded the plot by tree depth (i.e. how many ancestors a given node = has back to root). Or maybe you are thinking of the tree lying on its = side (like vundo ;)? >=20 > Going down from the root node is proportion to the height of the tree, = no? >=20 Sure, but as you can see from the plot, depth (=E2=80=9Cheight=E2=80=9D) = is subdominant to the starting line number. To me that says the = width/breadth of the tree must also matter a lot here.=20 >>=20 >>> Also, getting the node at point isn=E2=80=99t free either. To get = the node at point, we actually iterates from the first child node of the = root node until reaching one that contains the point, then iterate from = the first child node of that node until reaching one that contains the = point, etc, until we reach a leaf node. So log(N) time complexity is = expected. >>=20 >> I tested node-at-point on the same file, and it is quite fast, with = worst case performance only 30=C2=B5s (vs. 3ms for the full parent = navigation), and growing very slowly, between sqrt(log(N)) and log(N). = Check the gist for a new figure. >>=20 >> Unless I am misunderstanding, for the common case of finding parent = of the node at point, it seems the algorithm you describe could be = tweaked to work well. I.e. instead of "stop when reaching a leaf node = containing point", just "stop when you reach a node containing point who = has the original node as a child=E2=80=9D? This should give = (hypothetical) =E2=80=9Cparent-of-node-at-point=E2=80=9D the same great = speed as node-at-point. =20 >=20 > I should=E2=80=99ve been more clear. For finding the parent of a node = x, we stop at the parent of x. We only go to the leaf node when finding = the node at point, which always returns a leaf node. Thanks for the clarification. But does parent of x make use of the = positional information node=E2=80=99s store, or only look for =E2=80=9Csom= e node who has a matching child x=E2=80=9D? I understood the latter, = which would make sense why finding a node=E2=80=99s parent is more = expensive than finding mode at point. But this post seems to indicate = it does use the bounds to accelerate all searches: = https://github.com/tree-sitter/tree-sitter/discussions/2250#discussioncomm= ent-5848411=EF=BF=BC time complexity of TSNode::* APIs =C2=B7 tree-sitter tree-sitter =C2=B7 = Discussion #2250 github.com So I added a plot for the ratio in time taken to find a node=E2=80=99s = (single) direct parent, and find the node at point. Parent is more = expensive, and the ratio does grow to about 10 at high line number. But = better than the 100x growth in time-to-root. So I=E2=80=99d conclude = some of the increase in the navigate to root time with line is due to = the natural growing tree depth with line number, and some is just being = at the =E2=80=9Cend=E2=80=9D of a long list of children at many levels. = That said, I can=E2=80=99t understand why finding one parent is not just = as cheap as finding a leaf node at point. Maybe I=E2=80=99ll take it up = on treesitter to confirm. >>=20 >> Then =E2=80=9Cparent-of-node-at-point-until=E2=80=9D could do = something quite clever: accumulate parent nodes all the way from root to = the child=E2=80=99s direct parent into a list (same low cost, modulo the = node storage). Then run the predicate on that list (in reverse), = returning the first match. Could of course return that full list too = (=E2=80=9Cancestors-of-node-at-point=E2=80=9D), for other uses. These = should all be quite fast compared to a full breadth and depth searching = of every nook and cranny in the syntax tree that node-parent seems to do = (having no positional information to wield). >=20 > Sure, there hasn=E2=80=99t been a clever version because so far no one = have complained they can=E2=80=99t find the parent of a node fast = enough. Also I doubt the amount of gain we can get with a more clever = algorithm. You can try implement one and benchmark it. I=E2=80=99m = curious to know the result :-) Yeah it seems this is baked into the design of TS. >>=20 >>> I=E2=80=99m not too worried tho, because IIRC the absolute time is = very short. The 100x variability doesn=E2=80=99t mean much if the 100x = is still very fast. >>=20 >>=20 >> This is on a brand new fast machine, and 3ms is pretty slow for = things that need to run dozens of times per keystroke (e.g. font-lock). >=20 > Font-lock doesn=E2=80=99t need to find the parent of a node. So that = hasn=E2=80=99t been a problem. In use-cases where finding the parent of = a node is useful, 3ms hasn=E2=80=99t been a problem AFAIK. (Well, I = guess indentation could benefit from a faster parent-finding function.) And any algorithm that responds to containing syntactic units. For = example, I=E2=80=99m working on my indent-bars package to allow it to = find tree-sitter =E2=80=9Ccontaining nodes of certain types=E2=80=9D to = alter the bar depth, inside lists, function parameters/argument, etc. >>=20 >>> These are fundamental limits of tree-sitter, unless it changes its = data structure.=20 >>=20 >> That is an interesting limitation for sure. It seems that perhaps = each node's start..end information can be really helpful here for = winnowing the tree, when you are mostly concerned about nodes relevant = to and covering particular positions in the buffer. >=20 >> BTW, https://tree-sitter.github.io/tree-sitter/using-parsers says: >> =E2=80=A2=20 >> A TSNode represents a single node in the syntax tree. It tracks its = start and end positions in the source code, as well as its relation to = other nodes like its parent, siblings and children. >=20 > I=E2=80=99m not entirely sure what exactly do you have in mind. A = node=E2=80=99s start and end position doesn=E2=80=99t really help us = finding it. Suppose you have an array of numbers, and you know one of = the numbers is 3. How do you find the index of the number 3 in the = array? This was to some degree my confusion over whether node-at-point and = node-parent are really that different in terms of how they search the = tree. But the start/end very much do help as you described in the = node-at-point algorithm, and as mentioned in the post above: it does pretty effective stepping down over a tree to a node by using = checks of predecessor ranges I.e. you can skip searching entire branches that do not contain node=E2=80= =99s start..end range. > While the documentation say it tracks its relation to other nodes, I = think it=E2=80=99s more pedagogical than factual. Behind the scenes, = tree-sitter=E2=80=99s own node API does the same thing as I described: = it goes from the root node and traverses down. I see, kind of an interesting choice to call that =E2=80=9Ctracking=E2=80=9D= .=20 Thanks again for the insights.= --Apple-Mail=_E8416D76-63E7-4065-BA11-A39F397F46CE Content-Type: multipart/related; type="text/html"; boundary="Apple-Mail=_B2A5A624-E9D6-4204-9376-C3C6DD979582" --Apple-Mail=_B2A5A624-E9D6-4204-9376-C3C6DD979582 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=utf-8

On Aug 18, = 2023, at 1:21 AM, Yuan Fu <casouri@gmail.com> = wrote:

=EF=BB=BF
to go to the parent node = from a child node, what tree-sitter actually does is go down from the = root node until it hits the parent node. This process is linear to the = height of the tree.

Do you mean linear in the width of the tree? =  I=E2=80=99ve color-coded the plot by tree depth (i.e. how many = ancestors a given node has back to root).  Or maybe you are = thinking of the tree lying on its side (like vundo = ;)?

Going down from the = root node is proportion to the height of the tree, = no?


Sure, = but as you can see from the plot, depth (=E2=80=9Cheight=E2=80=9D) is = subdominant to the starting line number.  To me that says the = width/breadth of the tree must also matter a lot = here. 


Also, getting the node at = point isn=E2=80=99t free either. To get the node at point, we actually = iterates from the first child node of the root node until reaching one = that contains the point, then iterate from the first child node of that = node until reaching one that contains the point, etc, until we reach a = leaf node. So log(N) time complexity is = expected.

I tested node-at-point on the same file, and it is = quite fast, with worst case performance only 30=C2=B5s (vs. 3ms for the = full parent navigation), and growing very slowly, between sqrt(log(N)) = and log(N).  Check the gist for a new = figure.

Unless I am misunderstanding, for the common case of = finding parent of the node at point, it seems the algorithm you describe = could be tweaked to work well.  I.e. instead of "stop when reaching = a leaf node containing point", just "stop when you reach a node = containing point who has the original node as a child=E2=80=9D? =   This should give (hypothetical) = =E2=80=9Cparent-of-node-at-point=E2=80=9D the same great speed as = node-at-point.  

I = should=E2=80=99ve been more clear. For finding the parent of a node x, = we stop at the parent of x. We only go to the leaf node when finding the = node at point, which always returns a leaf = node.

Thanks for the = clarification. But does parent of x make use of the positional = information node=E2=80=99s store, or only look for =E2=80=9Csome node = who has a matching child x=E2=80=9D?  I understood the latter, = which would make sense why finding a node=E2=80=99s parent is more = expensive than finding mode at point.   But this post seems to = indicate it does use the bounds to accelerate all = searches:

So I added a plot for = the ratio in time taken to find a node=E2=80=99s (single) direct parent, = and find the node at point.  Parent is more expensive, and the = ratio does grow to about 10 at high line number.  But better than = the 100x growth in time-to-root.  So I=E2=80=99d conclude some of = the increase in the navigate to root time with line is due to the = natural growing tree depth with line number, and some is just being at = the =E2=80=9Cend=E2=80=9D of a long list of children at many levels. =  That said, I can=E2=80=99t understand why finding one parent is = not just as cheap as finding a leaf node at point. Maybe I=E2=80=99ll = take it up on treesitter to confirm.


Then =E2=80=9Cparent-of-node-at-point-until=E2=80=9D = could do something quite clever: accumulate parent nodes all the way = from root to the child=E2=80=99s direct parent into a list (same low = cost, modulo the node storage).  Then run the predicate on that = list (in reverse), returning the first match.  Could of course = return that full list too (=E2=80=9Cancestors-of-node-at-point=E2=80=9D), = for other uses.  These should all be quite fast compared to a full = breadth and depth searching of every nook and cranny in the syntax tree = that node-parent seems to do (having no positional information to = wield).

Sure, there = hasn=E2=80=99t been a clever version because so far no one have = complained they can=E2=80=99t find the parent of a node fast enough. = Also I doubt the amount of gain we can get with a more clever algorithm. = You can try implement one and benchmark it. I=E2=80=99m curious to know = the result :-)

Yeah it = seems this is baked into the design of = TS.


I=E2=80=99m not too = worried tho, because IIRC the absolute time is very short. The 100x = variability doesn=E2=80=99t mean much if the 100x is still very = fast.


This is on a brand new fast machine, and 3ms is = pretty slow for things that need to run dozens of times per keystroke = (e.g. = font-lock).

Font-lock = doesn=E2=80=99t need to find the parent of a node. So that hasn=E2=80=99t = been a problem. In use-cases where finding the parent of a node is = useful, 3ms hasn=E2=80=99t been a problem AFAIK. (Well, I guess = indentation could benefit from a faster parent-finding = function.)

And any = algorithm that responds to containing syntactic units. For example, = I=E2=80=99m working on my indent-bars package to allow it to find = tree-sitter =E2=80=9Ccontaining nodes of certain types=E2=80=9D to alter = the bar depth, inside lists, function parameters/argument, = etc.


These are fundamental = limits of tree-sitter, unless it changes its data structure. =

That is an interesting limitation for sure.  It = seems that perhaps each node's start..end information can be really = helpful here for winnowing the tree, when you are mostly concerned about = nodes relevant to and covering particular positions in the = buffer.

BTW, = https://tree-sitter.github.io/tree-sitter/using-parsers = says:
=    =E2=80=A2
A TSNode represents a single node in the syntax = tree. It tracks its start and end positions in the source code, as well = as its relation to other nodes like its parent, siblings and = children.

I=E2=80=99m not = entirely sure what exactly do you have in mind. A node=E2=80=99s start = and end position doesn=E2=80=99t really help us finding it. Suppose you = have an array of numbers, and you know one of the numbers is 3. How do = you find the index of the number 3 in the = array?

This was to some = degree my confusion over whether node-at-point and node-parent are = really that different in terms of how they search the tree.  But = the start/end very much do help as you described in the node-at-point = algorithm, and as mentioned in the post = above:

it does pretty effective stepping = down over a tree to a node by using checks of predecessor = ranges

I.e. you can skip searching entire = branches that do not contain node=E2=80=99s start..end = range.

While the documentation say it tracks its relation to = other nodes, I think it=E2=80=99s more pedagogical than factual. Behind = the scenes, tree-sitter=E2=80=99s own node API does the same thing as I = described: it goes from the root node and traverses = down.

I see, kind of an = interesting choice to call that = =E2=80=9Ctracking=E2=80=9D. 

Thanks = again for the insights.
= --Apple-Mail=_B2A5A624-E9D6-4204-9376-C3C6DD979582 Content-Transfer-Encoding: base64 Content-Disposition: inline; filename=2250.png Content-Type: image/png; x-unix-mode=0666; name="2250.png" Content-Id: <416B9FD5-46D2-48C5-AA64-76AEDC9E8D15> iVBORw0KGgoAAAANSUhEUgAAAoAAAAFACAYAAAAszc0KAAAAAXNSR0IArs4c6QAAAERlWElmTU0A KgAAAAgAAYdpAAQAAAABAAAAGgAAAAAAA6ABAAMAAAABAAEAAKACAAQAAAABAAACgKADAAQAAAAB AAABQAAAAABvE1SKAAABzGlUWHRYTUw6Y29tLmFkb2JlLnhtcAAAAAAAPHg6eG1wbWV0YSB4bWxu czp4PSJhZG9iZTpuczptZXRhLyIgeDp4bXB0az0iWE1QIENvcmUgNi4wLjAiPgogICA8cmRmOlJE RiB4bWxuczpyZGY9Imh0dHA6Ly93d3cudzMub3JnLzE5OTkvMDIvMjItcmRmLXN5bnRheC1ucyMi PgogICAgICA8cmRmOkRlc2NyaXB0aW9uIHJkZjphYm91dD0iIgogICAgICAgICAgICB4bWxuczpl eGlmPSJodHRwOi8vbnMuYWRvYmUuY29tL2V4aWYvMS4wLyI+CiAgICAgICAgIDxleGlmOkNvbG9y U3BhY2U+MTwvZXhpZjpDb2xvclNwYWNlPgogICAgICAgICA8ZXhpZjpQaXhlbFhEaW1lbnNpb24+ MTAyNDwvZXhpZjpQaXhlbFhEaW1lbnNpb24+CiAgICAgICAgIDxleGlmOlBpeGVsWURpbWVuc2lv bj41MTI8L2V4aWY6UGl4ZWxZRGltZW5zaW9uPgogICAgICA8L3JkZjpEZXNjcmlwdGlvbj4KICAg PC9yZGY6UkRGPgo8L3g6eG1wbWV0YT4KMyImCQAAQABJREFUeAHsXQVglEfafiGKJEBwCW4BilNo cWiBlhrU3eWu/1WuV9erXd2vpVcvpUq9OKW4u7sEJxAgIW7/88xmNl82uxshhMg77bKbT0aemW/m +V6bSllIokkRUAQUAUVAEVAEFAFFoMIgULnCtFQbqggoAoqAIqAIKAKKgCJgEFACqANBEVAEFAFF QBFQBBSBCoaAEsAK1uHaXEVAEVAEFAFFQBFQBJQA6hhQBBQBRUARUAQUAUWggiGgBLCCdbg2VxFQ BBQBRUARUAQUASWAOgYUAUVAEVAEFAFFQBGoYAgoAaxgHa7NVQQUAUVAEVAEFAFFQAmgjgFFQBFQ BBQBRUARUAQqGAJKACtYh2tzFQFFQBFQBBQBRUARUAKoY0ARUAQUAUVAEVAEFIEKhoASwArW4dpc RUARUAQUAUVAEVAElADqGFAEFAFFQBFQBBQBRaCCIaAEsIJ1uDZXEVAEFAFFQBFQBBQBJYA6BhQB RUARUAQUAUVAEahgCCgBrGAdrs1VBBQBRUARUAQUAUVACaCOAUVAEVAEFAFFQBFQBCoYAkoAK1iH a3MVAUVAEVAEFAFFQBFQAqhjQBFQBBQBRUARUAQUgQqGgBLACtbh2lxFQBFQBBQBRUARUASUAOoY UAQUAUVAEVAEFAFFoIIhoASwgnW4NlcRUAQUAUVAEVAEFAElgDoGFAFFQBFQBBQBRUARqGAIKAGs YB2uzVUEFAFFQBFQBBQBRUAJoI4BRUARUAQUAUVAEVAEKhgCSgArWIdrcxUBRUARUAQUAUVAEVAC qGNAEVAEFAFFQBFQBBSBCoaAEsAK1uHaXEVAEVAEFAFFQBFQBJQA6hhQBBQBRUARUAQUAUWggiGg BLCCdbg2VxFQBBQBRUARUAQUASWAOgYUAUVAEVAEFAFFQBGoYAgoAaxgHa7NVQQUAUVAEVAEFAFF QAmgjgFFQBFQBBQBRUARUAQqGAJKACtYh2tzFQFFQBFQBBQBRUARUAKoY0ARUAQUAUVAEVAEFIEK hoASwArW4dpcRUARUAQUAUVAEVAElADqGFAEFAFFQBFQBBQBRaCCIaAEsIJ1uDZXEVAEFAFFQBFQ BBQBJYA6BhQBRUARUAQUAUVAEahgCCgBrGAdrs1VBBQBRUARUAQUAUVACaCOAUVAEVAEFAFFQBFQ BCoYAkoAK1iHa3MVAUVAEVAEFAFFQBEIVAgUAUVAEVAEFAFFQBE41QhkZWUJPzY5/65UqZLww+T8 ba/V7+JHQAlg8WOqOSoCioAioAgoAhUaAUvu0tPTJTMz03zsMQLDY0yWEFryV7mySzHJv/nbfgIC Asxvc5P+UywIVAL4OXS8WLLUTBQBRUARUAQUAUWgoiFgiR5JX0ZGhvkUF8UgISQJtJ/AwEC3xLCi 4Vxc7VUCWFxIaj6KgCKgCCgCikAFQ8BJ+tLS0tySvVMFAwmlJYMkgfyQFFoJ4qkqtzzmW+acQDIz syQpKdktNi6NnbJv/0H5fdI0SUhMclcvPT1DkpJz15tvSckex9w3FPHHoZgjMnfBYknH2xcfxpTU 1CLmVDK3LVyyXOYtXJqrsFTUm3W3iQ98SnKKEC9NioAioAgoAqcfARK/lJQUSUhIMB+uZVateypr Z4keJYws05bPupRE+aeybSWdd5kjgNNmzJIvxn1f0jgVqry9+/fJL79PBgFMdN/35dc/yNTps9xv KRyo73/0hcyau9B9zH3xSfyYNW+BbNu+UzLS0uXVtz+Q1Ws3nERup/7WeQuXyJx5C90FHTh4SN54 90PZf+CQ+9ja9Rvl1XfHSOzRY+5j+kMRUAQUAUWg5BFwEj8n6bPErCRrZMt0ksFUCD2KS+1ckm05 HWWVOSeQ/SAIR48fL1bSVNzAd+oQJc8++aBE1KzpznpndLRUq1bF/TcNW7eCqLVq0cx97GR/JCQk yuYt2+Wayy+RzKxM2b5jFy1sTzbbU3r/TddekethTUlJlZ07o41Y3xZ8JPaY7N6zV4Ig6tekCCgC ioAicHoQSEuj1C2x1EraSE6TkpKEJDA0tArWkYDTA1QZKTXgGaSyUNfU1DTZsGmLrFizTuLiTkhY 9WpG718d35QWHYk9KjVrhMt2kAeqCqtVrWqadSjmsGzask1iDh+RkJAQDIqQXM1NS0+TLVt3yM7o 3ZIM8lGrZjjOu1zRc13o5Y9okJKt23bIEUimwsOquwkKB+Dhw7ESEVFTUpHnuo2bZOny1VI524B1 374Dsnf/flm+co1UCQ2FZ1Ml3B8mwcFBKCVLdu7aI9tA3o7HxUmN8DDTThZPtegWlBeKdpDXUdJX vVo1N1lavW4DsDgovXt1lxWr1snK1eukRo0w87DWqlnDkObNW7eb6/mgkCASP9pQZODvbch7B/BL RP1r1nBd72x2LIgYsaSUjvdUrZpDaJ3Xef5OhvqW5e7avQcPZprURF1sOnwk1nUMfbdv/wFZhf7d sHmb1K5dS44eOy77oU7ftGWrRO/Zh74Bnrg/olYN4w1GlfomXLsb51h/YmXTseNxprw6EbWELw1H jmB8OMq11+m3IqAIKAKKgH8EUlOPS1JCLMhfPJZHrlMFWyP953rqzlIAkpJ8XDLS46QS1teAgNzr /qkruWzlXGZEKrSnmzDlT5DArUZi9NV3P8voC0dIg/r1ZO7CxbJ9+y5p3661/DV7vlwx+kKpU7u2 TIW6ePLUGeiRSrCJSzfk79Ybrpb2bVubXjoC8vHx2G9APA5KZXgY0fasR5cz5JorRuHaUJ89SbIx /pc/ZCbKCgkJBuHMlHp1a8utN14tjRs2MCTpf5+Ok5eefcyQ0Z9/mywHQUSPx8fL7t17Tf0zweDi T5yQpStXy04Qo7tvvxkEMFi+/OYHqG3XG7KYhreYhsjv9huvQf51cH2CfDb2OxnQr7fs2BEtMaj/ PX+71bSLIu8Fi5dK757d5OChGPnpt4kgtClGxUypYORdt0qVKqHy6dhvpUfXzoZsHQPBuvfu2/HG lCJffv09yPMuQ6xIsjp1aC/XX3WpIYgEgrZ6bHMG2sqHq3JAJeA/UvqffaZPnHiCxPujz8fJnr0H JDgkSNJS01HHrnI1pJRBQUHy4y8TDbF96P6/yxqoeieiv0hA/4ANJfGgiJ+SzUT0/8+/T5Kodm2k VctmEr17n3zxzfdy7FicKT8DpH9g/7PlEoyJQBBtEmXWd9iQgUKzgY5R7aR5s0i/ddWTioAioAgo AjkIZGGuP350qRw7sgBzf5IEBTeU8NoXYF52CStyrixNv0BOUe8TcfMlNWmzBAaFSUSd/hJW4wxQ gTJn9XZKgS0zBLAmJFn3/f12Q5AOHjwsD9xzpyEQRIfSKErZ6oIk/fP/7pDGjRoI7cYmTJ4mV192 iXQ5o6MkgQxNnjZDxn7zozzyr/+TqiB4Y7/7yUjtHv3nP1zSQ0jExnz6FaRmNeTSi8/3Cfy+fftB KubIpRedBwLU2xC5d8Z8Kn/OnCM3XH25IS186yDJI3F75IH/kxdefVvatm4pV4y60BynlPLJ516R AX17y8jh5xgSNxEEd+eu3fIPkLUmjRtKLKSaX4z7Qb7+/me5D0TNxkdi2ReMGCrXdusCqRYllgLS d1hOxJ8w5JbSvofvv1uefuE1Q2b79OxuiCpJNCV/8xctlVGoe5dOHYyk7/2PPpcTIFkP3Xe3kbzt 2btfPvx0rPw6YYpce+Vo2bErWr4d/6tceN65kC52g2t/psyG3d4PP/8u7dCmevXq+MRqJmwc90Li +fA//w5SHmH65fOvvjNSynZtWrmwyn4ohw7qL00aNZT3xnwm/7jzFmnatLE5P2vuAkP+HkKbGjWo ZwyPP/5inLRq1QLXDQexDZF16zcZctu8WRPpCVwCAiobSeXK1WvlpuuvlOaRSv58dpKeUAQUAUXA A4H0tDiJjZkpcceWQ2iRgbOZ4FXU3pRusyLbjKxMOqWcgBYuUQ7u+0WSEqMlot5g8IXq9pIK/11m CCAlQZS2UWrExZ3SLHcC0apcOUDOHzYEEsG6RsI2H84FjSA9a92yhcRB8lYJNnfdu3SWWXMWQmW4 X2qEVYO93Da56borcW8lOQa7wrqQ4rWGdIlqyFEXjJBFy1ZAqrYcJNFlR5AGxwpKvJgvH4KqUDOz Pg0b1DfE0+n1a41T+V0FaucAlBEcFGiIHuudmZlhyE0w7qdamhIuErOe3TobVSbVn8y7e7cz5Lc/ poAMHsObTICkgTi2bNFczhk8gNm4E6WGkU0aQz3qsjusWjUU+YuEQopm1d4U2qegDc2aNJJB/c4y 99JreOPmrTIahJflsdwa4eEgkm0gidwIopgM9fUqgzelgicghWSbOneMkql/zoId4w6j6v4eZPAQ SCjd8SmN5Pfloy8w0rjAgECo5KsYHHqDjFI6WxuqWSbWkR8mSu4oeTX1Rv9SPc4UHOSSBFbDMUoF V0C1TYeQG886ExLWNJgEpEozSPfqg4iuwjkSQMSUB8ZZMnRgP2nfxiXxNZnpP4qAIqAIKAJ+EUhJ PiCH9v0myUnRmI8ZYoXzOm/Jnqz93l1aTrKulU3dSV6PH10IE6IjUq/hhdBG1S4tlTyt9SgzBNCi 5M27h1KtOrAZq1vH1alUYR6C6pGSq5feeNdI3DhweR3JI8XDJBBUqX717Y94I3CRFhIbqh+bRjYx ASzDqlc3EqeAbAJINSOJZ6OG9Y0kb8asuQhhsliqV60GaVRzGQKy4S85605ywmSPnYArPZ1bpoBU zZm/2KhZK0EylpaWKlVANCkxDAp2dVcbkFRnouPECki6zh8+1H3YM397ggJwF4F1HTl67JjxVv7x lwnye/BUUx+Wy1A7JNN0rad0kVLPl9/8L84zejseLMwGLJf1J26U7pH0UUqZhbYF4JvEb+jAvsY+ 8+33P5Fw2OjRdvNMqKl9Ob9YPOw3a2l/U6LKdAB2juy79//3uSkbV5jv+PgEadHMhQ3vqVatqjSG JFWTIqAIKAKKQMEQSEzYAfL3s6SlxEJwUuYogo9Gcpu5IEk8sVX27/4GJPACqVKtuY9rK87h8tK7 kBIFoYNdHUcSQrUwJUFUYdJFnInHeQ2J3doNm+BAUVX+dtsN0hB2hCQXJDK8hvaAgZDWderQznxc ueb8S3IxbOhA6QdpYAwkaHRC+HXSVKO+vfuOm3BhdkVybjG/mL9nsscCQJb4ufTi4cYOkXaGvJqO I0x0MqFjA7Ng25xp246dkgm1bKvmuYkhr7H58zfpE1XTTm9akl9KVmkb2QJSNIMDrqsMokwSRykn MWnTuhVU0zcb9S/zotSU/1FiR+JHWztviaT7luuvgq3ecdhBxsh6OPJ8CDX7nTdfJ927wibDR3LW 215ijwVBClgbDjYP3fs3kOMqbhLKfqMUk4l9RImivcfmod+KgCKgCCgC3hE4cXytHDowAfZ+1Jrl Xme831G2jlZGm1JTDsqBPT9I3UYXSvWw9mWrAcVc2zJoEVnJeHz6wyEI5K1dm5bGs5cEgGpRfuid ++mX38IhYT/UoI0lBERiN5wJ6B3K87Sd+/OvubA3m+wmOt7KWb9xs7z53v+M80Szpk0METwD6lGS Siul8rwPfMSob53HsS22cTzhMdo4No9sbELDhIVXd9UHddq6bTscNH4wwS5JxLylpStWS+dOUbm8 ctlu5m/Jr7f7eIzklxJMhqRx44By5y1YKt/99LvBoUP7NsbTmhI/i6VkVYI95nijAvaVN0nYt+N/ M7aC9IimA8dF5w8zamlKZ70lkrgMSBBJHN0JpJd2h1aqSbtDqtvp3cs+Y51IkH/+Y5KxTTT35eXa 7uz0hyKgCCgCikBuBOLj1smh/b+B/J3Ai3P5I3+2tVRnp8M7+BDsAhNObLaHK+R3metlOgHMRrDj X+Cg0L1zJ6hrGxuSQw9eZxoEj9Aly1bJq2+9D4LWG+rKZJk1b5HUh1MG1ZUMYTJyxDny7Y+/yv5D h4zTBUOVLMU91145yqiFnfk5f9OxgyFN3vngYzm7Ty/YxZ0wxIPetSQwtO/zDEZJO8EFi5ZJnYgI 6dm9i6lDI5Avei0zfA1t4y4aOVxee2eMvAtyQ+kj7fNmzpkvfWHrVhUkjWFhqN52kjqqsneCTA0d lFv9THLLUC6Tpv1lSCnLJCfyvJ8SvvOGDYajyS8m/5bNm5pQMLR9HAWPWpJpqmwXLF5m6kbbQZJL 7t7BsDQNoQ73lXhdgwZ1QWDHS0JSojSDan39hs1yDO1omS2tpE0jbSttov1hEMLhjP91ggxEv3WH TSRJKq/54Zc/5OzePYz9IZ1v6F3MsD+1EBZm1Zr1xqaTjkJMlIiyrSShmhQBRUARUAR8I5AA1WjM fkj+MpIwv7u0Tr6vLvtn2EYS3Zj9v0vlxldAk1QxnQTLTBxAO+Rof0d7OHr5NobHKMOuMFAw7da6 QaVI0sFEW70e+JuSolVwkGAcuDN7dJOrLr3YLSmjHRoJJGPJbd66zcQOpMq4V4+utjiv34wxSIcI qmTXQep3DHZ0dDa4+ILhRgXJsCWMVUjSxZh9TFSvHkXIEpJMegPTCaJFs6bCOHi7EBqmQ1QbaQqp ZAdIyeg1Sykjt3EbOeJc49xC20VuJ3cAMQ9b437iwMQAyaQ4JJDORFUo82O4FIagodNGMFS9lH62 btFUmjXNGfAtQPpatWqOuII74BCyzbThSuDUFypuosm8unXpZKRwJtYgwuZ0RPuvu2q01K7lcuZw lu383RzlRDZpaNq9GfEWq6JfrrrsYuOZzb5iDEd6MjNMCxOJOft0Pby6ExKTpVNUW9gi1jd4MQ4k MSA5pmc3MeQ4YPiaBvXqyg3XXi6tW7Uw+XAXljh4RfcAgbR9YE7oP4qAIqAIKAJuBJKSdsvBPT+C EHGDBX8yITr3hUtotQ64rjSTRK5asFFP2oo2HUZdvWvOeJyENylxp4RWbWrCxbhBqSA/KkFCoiKS MtzZ7D5LestwM7TqioAioAgoAiWMQGrqYdkf/Y2xi/NP/mhXnYEoDI2lRt3LsOYEo6allTrA/Al1 jYudBBK4IV+yymtDQhtKw8iroH2KKOEeOL3FeafGp7dOWnohEFDyVwiw9FJFQBFQBBQBg0BmRjJU oJMkFSFf8iN/5RkySjNTkvbK4YNTQBxzzJHKc5tt25QAWiT0WxFQBBQBRUARqAAIUOp1JOYvhEXZ DG9fV+SECtBsn00kBifi1mPHk4U+rymPJ5QAlsde1TYpAoqAIqAIKAI+EEiM3ybHYxdD8qcUwEJE LGIPz5KE+IrjGezP4tPiot+KgCKgCCgCioAiUA4QoNX/iRO7EbA/tYjSP5JGfkqvDWDR6oYIHhmJ 2DFkuVQLa1sOejr/JigBzB8jvUIRUAQUAUVAESgXCKSYUGKwdcuOmFGYRiFCK8KcIVSM0FauNBNA 197FhWmb61q0DFFGUlLhGBJcmj2dC98yb3coAfSGih5TBBQBRUARUATKGQIMo8UYtUUhfyZsSnqs HD/8M1BxhVsrzfBkZnLnrMKTOLYsFZseBAYEm12uSnMbT7ZuSgBPFkG9XxFQBBQBRUARKAMIpKam nERwfIZXwUYEaQfLQEtRRWPfWASiilsYXo1EuUqVKmWjrUWspRLAIgKntykCioAioAgoAmUFAao2 +cGO9ydRZbCjIkjVTqLA03Yrd7oKxo5a3Ou+vKaTGQnlFRNtlyKgCCgCioAiUK4QyNmetAhSsXKF RMEaw/3ojbq8YJeXyauUAJbJbtNKKwKKgCKgCCgCBUOA+8dT+qep4AhwkwVKAUkEy2tSAlhee1bb pQgoAoqAIqAIAIHyTmROVSfTFpDYldekBLC89qy2SxFQBBQBRaDCI0ASY2z/ihD2pcKDBwCIHTEs j0kJYHnsVW2TIqAIKAKKgCIABKj+5UdT0RAgASyv+CkBLNqY0LsUAUVAEVAEFIFSjwBVmOVVglUS 4JdnNbASwJIYQVqGIqAIKAKKgCJQwgio+vfkAaczCCWA5ZFEKwE8+fGhOSgCioAioAgoAqUOAXqw lmcv1pICvLziqASwpEaQlqMIKAKKgCKgCJQgAhr6pXjApvSvPBJpJYDFMz40F0VAEVAEFAFFoFQh UF6dF0oaZKtKL+lyT3V5ZX4rOHZMCjZuzsR3aEiIVK7sO8p5Cvb2y8Bm2CEh5Xt7l1M9aE5H/mlp 6XI87nhO0fDKDwioLOHh4eV6q56cBuuv0oAA55r4+Hhsh5VdG47DwACpEV7D79xTGurOXQ2Ox8WL OxoI6h4YFIi6h+OYbVBpqKnWoTgQ8Cu1yoJNW2YqxnEhgxxjf91KlYpCG7JgQ4dA1IUNp4Jx6Sqv 8OMzC20UfgqRWEdTT497+HxQAkhMy9OzUpSe9IDm9P6ZmJQs02bOlYSEJBlxzgCpXaum1wqx4xYu WSlbd+ySoQP6SvOmjbxe5+tg/IkE2X8wRo5j8icZCQoMwsRZXRrWryth1av5us3r8cTEZDl4+LDE Hj1utprhXoPMo37dOhJRs0bO4uLl7uSUFFmxer1kYDDi0fByRd5DJMdVQI7P6NDWkN+8V+Q+wvYd jDkshw7HSnJKsiFYtWrUkIYN6klYtaq5Ly6mv+JPnJBly1dKi+bNpFnTyDy5bty8We5/8HEcdz2A 6SDyjRo2kFdffEYaNSpcX+bJXA8oAgVEYN6ChfLcf14zLx9cCDgOmzdvKm+89LzU8jH3FDDrU37Z Ujxfjz75LPY3DTJl8Tnv2KG9vPaf57Dpfajf8lPhSZoK8lu9kHOd30z15ClFwJfdGolRcGgLCauF 5R+ErqCJewinpR6UlKSNBb0l+zq8aYA0VgvrjeKq4FhBSWclEK40SUlYA8FNXKHqyjJCqrSSoJBm hSgPl2KtDAyKwFde4mgdQZQAZndrafgisUsAoYpPSMxXR0/yRCKXnl7wyN7H40/IShCuHdF7JTE5 yUz4WZkuEhKIN/9qVatIqxbNpEfnDvlOosmYQNes3ySbtm5HPRJNgMlMR14h2Hi6edPG0qNLR6lZ I9wrvCSPy1GfNLOpN+lQ/ikjI1Nq4S2/bevm+RLAvfsOyqIVqyUmJlZS0yBZJdHEJMG2Vgf56xTV BkSyHQhw8bw77Nm7T6bPmCmTpk6XHTuj5fWXn/dKAFNT02T/gYNorCWAIOGQXnAB1qQIlBQCScnJ ZhxS+mwJIElRWbAPSsH8x2cohwCmSYP69YxUwxd+x+PiZO78hfLHxCnSqWMHufvOW31dqsdLGQJW YpW3WlkSFNwEJLBpwRaQ7AwqVQqQpBNri0QAeW+V6t0hLa+FMgs6Z1PqliSpSdtRA4f2J2+D8hwh LwgKjpRqNc4GFyxoeTYbSvry3sM8+SlPqXhW8dOMCNW+lSvn/ybDCZvXFZTBU+I3a95iiTkSazqe 9waC+FTOFgfzjYAEcfnqdXIE1wzp30fCwqp7RSMO182ct0h27d5n8iKpouQvKKgy1NeZeMPJlISk JFm3aascPHREBvfvbaSLnpmlo0ySMSYuQpUru357Xuf8m9LC0Coh+bZ7Pcqet3i5JEGqyhRo6hdk 6ktj4mPH42X+4hVyJPa4DDirZ75k0lkHz98s4533P5Tpf82SA1iUSOQohQjw0Y/sM2JvCSDzI34F 7UvP8vVvRaAoCNhxaAkgxfUch2Uh2bq7niM8SVjMfNV9BzQlf0yeIn/NnCM7dkUb1XHzZiAMmsoM Av7JClXAeUmO/8aBGElh78nJ0aVehdoZ613BkksCyDm/aIlthLDHC5krWn7gklhLfT0zRc3zdN5X LgjgqQDwBCSKM0n+Dh8xpDEC6p0O7VpLvbq1DQGkPWE0yNymbTthg5giu/bsk9kLl8mwwX3zSMdo ezNr/mLZCSliJRCcmuFh0qZVM4ls1BCDCVKsjHTZt/+AbNy6E/ZFJ+Tw0aOm7JHDBkm4h8olw2xL 44ru3ql9G2nftmW+zwcfH5K5qlUofveeduzaI3MWLDNSv6CgIGnRrIm0bdlcqlQNRQykTIlG+zZs 2iaUgGzYsg3ENUD6n9XLJ2HzXkrO0bj4OPntj0lQgx+FRCIYH5LZ/El8Tg76SxFQBE4VAl9+/a18 +sU4CcNcxZcyaif0+TxVaJ+afP0TwFNTZnnPtbxhqgTQx4hduW6Dm/w1bdII0r3eUIHmtvVrHtlY muEzbdY8SU5OgZp4j2wGievYvnWuXLdCtWnIH6R+tWqGy4gh/aVOBEThjhTZqIG0bdVCJkydKcdg qB0D+7vV6zZJv97dHVdhU28GpDT/CRwgwozdYK4LivAHy1uwZIWkQOVL6WbvHl2ka6f2uSb8Jg3r S3PgMGXGXDkBSeV6kMEG9epI+zatilAi5CYoh6ooK8XL78Him5dLMpmjAqZKn/aNmhQBRSB/BKix SMKzaz1D0/C806nFW6LWgM5yfHHUpAgoAi4E8lunyhpOZZ4AOr1+/akDec6er+THU5gdmEQyt3M3 rsebL5wnevfonIf82Y5uFtlIOrZrI0tXrjEqlZ2790oUpHL2bZnEJXrPfkNUgqC67XZGVB7yZ/Oq BQeQziBecxYsRdlZsg+qUU7QnIhtMpM3SA/dPwLyaYe9J7/vLZBiHjl6zEgSmzRpaOz8bP2d9zYC CezcKQrONCsgtcwwJJD2j8VlD+gsy/N3m1Yt5cP3XidzNG0nrlUg0awHxxlNioAikD8CXTqfIZ+M eSfX3EQveuf8YnOxc6X9W7/LHgIkK/xoXxZf33HdKU+pzBFADui4+AQMbDonVJLExCTo5V0Dncfp nctznon30ZGA99Bj+HjcCXMd5UfVqlaVYDgU2MRQCQnIlyQnAhK7OhER9pTXb5LAFWvWG/uAo8fi TDmhoSHmWjqoHIW3Lx1HaOPWrEkTr3nYg5QEhoLw0bv5BBxFWA/nBM0wNqwzVcnFYYtA277ovQcM QaVEriPU3MFQAftK7eBIshEqYBLGmCNHjSdzfajFC5toh+Q5MZHUWvtGz/xqwClm8MD+noeL/DfV 9hwToaH+vR9ZAMdNBtT0fBnwRowLWglKiWnvGYy2B/rBuKD5+bqOkxTbR3ypXj+ZOnOspUDtT0kr x4W1H/NVdmGPs558fvPzQmW+3NOU4zUY/eDLVrSw5fN6SsUq82Uv+5ktSh6FuceMA/QRnZhoblFS qU7tCBkyaECBiqPkj89HToJRfQnWNadc/aUIlA4Ecj8PpaNOJ1uLHNZzsjmV0P20vftt8gyhjR6l f+wUy8qn/jXXqDCd05azWryOpGne4mVG5cnrKFE7b+gAY/Nmr+Wx0JBQ5B8g4XDqoMG3v0TyyHyp NuEixW+bSBx4rho8aGnPFxqaI82z1zi/ucC6FlmGmEhHXrmNbqkCJgOkHKw4FuOjx+Pk2LFjBrdw qLip6vWX6AnM0DckgLRt3HfgENTQBSOACYmJxu7vSGysUeeS5DrJCXH7+bcJsmLlamN36K8e7Euq wC++4DypgRA1Nk2cPE22bKONYpDp2zq1a5trqoLkM5TFgoWLZdHipULvY5KaenXrSrcuZ8igAf1y OfBQ3bx0xQpZumwlbD33mLbWRDntIe0d0O9sr57Ktg72m/23FqYEq9eul+07IGVFu3msGnCm92VU u7bSvVsXhNfxj7nNz983vTXXoJy1+ETv2yvH8SJCCXZ4jTC8dDQ2HpydOkQBK+/e5c68SVAYdmfl qjWyDfXm+KCTDr1dI/EC06F9W+l8RkeDnfM+5+/pM2bJug0b3P0QHhYmoy6+AM9TmMFg0ZJlws8u OBjwRasuJLldkCexre144SJBXL1mnSxYvATX7kZYohTT721atpSBCOfUqmULZ7Hu3/RcXbp8BUhr sHmGamMcXHT+CHcYE/bp/IVLZMPGTXL4yBFDKNknnTpGSZ/eZxarZJnzycZNW2T5ylUYB7tMeRwH VTEOGqPv27RuZcZBY9gEO9PW7Ttk8tTp5tnkE8/xypfD80cMMyGQnNfydxxeXH/89XdJQKQDPlc0 FcEEKRei3U0RWmnHTjh2wJvXvjhybmlQv75ccuFI9FOg/DVrLsbQOhMbcO36jeblwZbBuWbZipXy wUefmpdZ3tuiWTPZs2+feVb58sbE4+cMGihRUe1cBzz+Zb8sX7HKPXfx+oYNGuAZPd/UweNy/fMk ECiPhOUk4CiWW8sbpmWOABryg9mG0o08EiQcw0GB9tRvMkrE7Os88+CNdPS4+PyhZvL0Jw2zhXBx JCHhJGg8c1mP7BRWvboMh81fJiY6TryBmJj9JTpcuEKbZBmCS5s8Z3I5gVCsn1daxvs4QNkmX5I0 Z178zViElFKSTNPRJTSfeGC8py4kCWxHGso7BCcZSrUoQckvJYK0j/36O1m/YZOx/6MK1xJA1pkY /vzrH4a45ZcXF9CmkZEysP/ZuQjgr39Mkj8mTYHDS6jJj+TwrD5nSuOGQfLWe2NQ/rdGrW4lSCSd 33z/o5wzZIA8/vC/DDE7gZiEL7zyhkwAmUyG9It9QDLF+pF9t2jRXJ55/GHpc2ZPn9XcCWLz4adf yAwQIZIzLt4BeKFgvzEffkhSm2ERvfHaK+WyURe5F2afmfo4sRCE9r8ffmwIICXixNKFK1+O+DBQ wlYF6vuOcuetN0nfs3v7yElAVDbLmI8/k3kLFhuHJErTTV7IM4v1RjvoEMB4jXfeeqOcf95wM+49 M2RYn29/+FGqValqXogoET+zZ3epDtJLEvHx519JEl4IKuOZ4AjnuP/2+5/kzF7d5YWnn5DIyMZG MvfKm+/KL79NxBgFqXH3g2ucj/tuvDz12IMYA309i5eZc+bJfz/4yBA+Ys1x0LNbV2kLm9U/Z86W V15/W3ZF78HzwvZB2oX/eB37KAok/5F/3Ye69MiTb2EP7NmzV/732RcybfpfcvTYcdMfnAdc48BV Jk0omjWLlGuvukyuuHSUGRcsh8/j7Lnzhf1LzNmTfMGMjT0mDz1wT56qjP/5N3nh5dfMWGX+yXhB I6G+5qorzLXbtu+U1956zzx7PMC8unftKiNBKEkApyEc05fjvpUq6CtKuyk9tgse60zitnDREpMX zWQYEuav2XNlCzQCdr7hywNfnrwRQBLhL776RiZOmW6eT2bEa2+75UYZjZcDTcWLgLe1rXhLqHi5 lTdMyxwBpDp09AXn8sXWJC54U7MDQZ/HQNBwrrCTlnN48ticBUtky/ZdMnDg2dKiaRP3dXyrdiZO yLUKICmx9xyMOeImbZQYcjK1iYSQAaMLmg5DSpQClSMTF21PFSWlJWw6ByKlG9Ybefe+/Vhg4sy9 IVDl1qldy0h+mkClbCdnk6nHP1R3MxHPmlB3czHOL/G6AGCUjgXzBKQNVJFyF5b8EutMSRxD5VCi 4K2f2GZfyXm9kaAgBqMn8STRYP6sD8vjdQwzM3v2PPn8y3GGZHl6QzPfqVigq4eFy9OPPgiC8jmI 6ATTj57Xsm6UWj374ivywTuve5UErl23Xp569j+yHtKlIPQR+5BlcAFkYtuNhAbHdkIq88LLrxuJ 5N/vuLVAqlCTSfY/v0Bi+spb78rR2KMoKwj4eseP5GbxsuWQ6O2QRx+8Xy4AcfNMS5atkCf+/YJp n6veLgyduPMe/r1123Z5Em3cGb3bkEpPaTQJRBgkXKY/0Q+UFu8/eEgoWfrw489N33AsOBPzpXT2 9Xfek5eee1o+gRcqSSGfIW/9sP/AAfTDqzLm3QZGiubMi+QlLKwa8KhmymL5MYdjEGLpkDzxzAuG lHMu8ZzQWYf1IMEPP/lveef1l+SMjh2c2Rbq9+at2+Txp58zxJx9k3ccuDzfWeYu4PjSa2/L3r37 5d7/+5sx+4ioVUse/ue98o9/PgyNR4IZuxzPEydPlYsg+aY02qboPXvkmx/Gm+eLfcGxVrNmTbnv 7rvgeOaSkHMeoPSZph5MJIBUvVsMiAdfWPkMsU78OBPbQFyZOH5bIAB2ePh58u77H5l7eJxlU6pK yS3HgDPRppmS15qoD+dY5k8p/vChg1wvGc6L9bciUMoQsM9JKavWSVUnf7HNSWVf/DezE7gYMAAz P1zwrISCkxklP/ac85uqy0CEXGHirhjO66xKpCi15Rvshs3bcKtrsqSjhOdiWNB8OSEy1Aonb/6u CzLLejoTCSBOGpUVJ9Q/psyUyTPmyKq1G42zCXfw2LP/oPl74nTXOdog+koMjA1IkbJMoGdf1zmP VzeYQ8qEGyk9JAEsSGKbEmFvxcUsEdIf/u2ZGGaG5xM8PsS5oMnmy2+Ol6+/Gy9fQvLInQ+4gFJK 4awzr+Fi9RdiEr7+9nvy48+/Gky4QHIho9rR5sk68FqqdJmv8zjPUc37LHaK2AASwQWf5IX5VMZ3 86ZNpWULBuMOMYSIhbgW40ogO1/Jd+N/YhYFTtP+nCkvvvqmxCE+I/Pkc0CJH6WWbKOpO75ZR54j KT4Olf+LL78Bdd6qXOVsAaF7/JnnJRpSMVtvXsB8cvKCLSCIJPEiESCWH378mXz3g/d6295l+Qx3 RAnVJ59/ZfA0/QBcScxsYr4se+6CRaYfKEHkuOS1bAvHhkui6bqDddiLFx9KlZ3HbX4o1iSWz7yn 4nl4D1LBo1Bp8xmjSYATI15s6gCcDoKsvvHO+1B/+352XLl7//fQoRh5/qXXQP42ZOMZYMYcnxnu dMNxQKJl209yxT76Ytx3ZhzYcdWje1dI8C4397pwDMAuPTHAcSza7Hqh4PEvxn5jyCPnMv7NQPGj LhopPWBiUNDEuthnj5gTC2fiMftc8jpKbc8ZPBBmBuFmXPBazn1bt+0QSsA9E6XLe9BfJH9MnMua Q5LcsWN7z0v172JAgP3n2YfFkG2FzsJwjXKEQI6oqow2yk6UrL7zt2dzeM7YxORzned9/v6m9G3O wqVyGNIXLB2QPtaUqDYt/d3i8xzrt2zlWuOQwXm3SmgV6dShTZ4H2DqB8HruCEIvYRLfyMYNDYHj pHz0+HFTJy6K2xGChtvXjRjc39TPWQHmQXJjk6ck1B73/KakwDwIqCgXUUNKPS/y8jcX9+HnDJEu UEXyvjlQb7F85sW68Htgv7OlXr16ZoG2WfB4LAJtz1u4yBwqzKTGa0mUgiD1OH/EudIVnpBc6ObM Wyir1qwxZfIalkGC8dU3PwBzgT1aJ+nf9ywjTdy8dbtMmjLNEEjrdU4njjmwZ7od9apTx2UDSVLx xVffwuZvnVsiSvJH1eetN11v1HwBUCXvP3hQxv/0m1FVc8E2eWKbzG9AAM8ZMkiaNG5km+7ze8/e vfI2AmlTXW2lMiRnVBUOP2ewdIhqb1TzG6BupzqU44RklAs0VZFU8/73rVeNYwdJ0LvIi9IZK8ll fzA/qot7de8Gghks27bvkKl//mUcr1z1rmzI2ZhPvpAusKOkjaGvxDbOmj3XOC9xDJCYcHzOX0hb vVVmnLMf+EnDC8VX34w30ugO7dvLgP5nQYpVE2N5F6Rf2Wp59BcTpe3zoZbct3+/X9xY398nTjbj jXafw4a6cCapWQ/7vJkzZ5sQS9Y0gJguXroM90yR669xqVB9tc3zOHH7GARtEWyNrXMLx1zvXj3l 5huuEarPWZ8DGAffjf8ZY2u6o/0in439GmOvD65rbrK+6vLRsM+bY9TzVgrHAOozEKSZ42XmnLnG ttY6FpH8de3cSe6Aip7jOr/EvmYi2UxLhZMNnpVVa9aanXms9oBtatOqFbaOizLjitvCsR2tW7WE KUQPqLhnmjHCdsXApnI57Hjbtc2RUDJ/2t9m8gU2mwDyeaFTSjgk75oUgbKAAOen8pTKPAE8XZ1B 1TN3zdi0baepAifNPj27FHpfYN5MArV0xVqzo4j1bu7epYM08uIcwAULbMkszpy4W7dshtAyHaRu nQgs5i61CqVlm7bukKWr1hmyc/jwUZm7aJlxdrHqH1Np/EPCaBPtnwqS7O4jfBSMtBILeUESVbP3 3n2nufRQzGE4e6wyEkG7SHHxuOn6a+Ss3r3yZEfHkPmLFvsl+Xluyj7ANt8GAnb7LTcaqQsPU4V2 34OPGvWcJVB8uAMCKkkUSMerLz5r7NB4LdsYUasGAuN+jQXVhRElOZTyUAJlCSDVeL/89keOhAN9 RSeFp2Av2MKxi0IjGPtzMWXwbzoquBb1IOPksGjJUhCZi1is3zRh0jTZvn2HuZcXciywHQ/+8x65 9JILDcHgcdb9o8++hI3gJ4assY2URq5Ytdqo6kh0V65eC5u/RYYM8h4mjskLzhsmTzzyL6MW5DGW 0Rt2cc9DZc3xz34joaRalQTAHwHk/ZQCXnf15fJ/d91mVJU8dgmkVA88/IQsgXra2Q/Mu3XrFvLK C09LK5AMJpZP8kYbQpt43WHYoe6GCjQ/4swXtiiQkkcfekB6ds+RjBGj78b/Im//d4yxO2SeTJkZ WbAXnCUXX3iecV6xZeb3vQ9jYvLUPw2R4rW0zaWzxcMP3JtLbcu9rBneiNJGqt85DojnfoyLJSDF lgDSe/eu226Wx6BOJpEnmSah/PTLr80Y/QyqcjrFkaSTqJHE/+22W2AGUjDnLNue0RddIPww/fuF l43jSmCgS03PF5l+IKUP3Hu3vdz9PWzIYPnzr9nuv2mWQWcPEle7YNJUZBmeYT7jTKxnNWhyeuPl SNOpQcBif2pyr5i5ljdMXTNdxezLIreaQZqpdt24ZbuRUjBkQv8+PaQlDLkLmzhx/zl7gSxbtdZM iry/K2IFdunoXZpCwmYHYXsskMOx80jjhvUM+eO9PEe1eDfsTczt2igh4Vs8d/LYuiOal+RKXFQL m+AOgHZnvwkV4X6Wh2oWjszhhqIURcJMScSN11/tJn8sn563A/r1zVUHYkHbt2tgjE8nBJu4aNFL sR4CX3PhYiLOZgcX2KHZtHjpcnh4xroXOUo5zx06KBf5s9fS9mnYuUMMObDHiCk9XvNLVF+TOLK+ diywnf3P7pPHmYR1v+GaK6UPpE9169SRBmg3XyxqwSmC9mZM8yFZ5R7Tlvgwr6aRTeQff7vDTf54 Hcu64PwRxn7Q5ajEoxDooIzFy5aZXV1cR/L+yzxbNI80RNxp+0dyM2hgP9yQPZ7wi+3imL3qstFu 8sccWf6F5w/HeG8AYosXoezEPtkHswd/idfQJpEkzEn+eA8xuubKS40jglOazTpsgfTXtQe1v9xz n1u8dAVCPx1zjwN6uvaGlMxps2fvCAsLk/OGnWNUtvYY27kGEjg71nicTkr0orb7mPO5pkr1ocee kXWwq7QvdiSGF190vvTDWDiZRGcfVMORXE5ajgPun927dUZQ+HrmBYMHKWlm3fiCYxPtIUlsLQEk 6aYq3JcXt71Pv4uOgH2evefALVEZiqtwH1h/4gGlGVIRPlg3ipSKUpbZ/o0ahcK1j9cL9i32lfxj 6uuu0nu8zEsAKYmxyS6G9m/nN8/ZS/1d57zH8zcXpk1bd8qipSvlOFRvzIf2LH16dTXbxHlen9/f e/YdMHvrHoA0zLQDz0f3rh2lN6QTnES9pXZtWpjdP1g27Q0pNfCV2oEg7oWd4NoNWwxh24oQFFFt W7lx4H12QPPRdC44vvLkcQr8rDrdSMRyusDfbbnOEUtvqbDHveXhPMY2tW/X1qsEh2E3aDvHMokn v6lq7NalszML85vkqR4+lPq5MQMXpGTDprXrN9if5pvq2GWQ5LwMj1NPbJkHHTLsws0bWIf9CKtj 65MrM8cfcVDp7wF5s4spT/EeEilbN8flRtr2/DOPm/Agbp4F+CPgKET7Poapsapt3sfFmZI+XxK1 YVDh/vjL76ZN5hnAGNwOD1OSMDoueEusX2tIPSM8dsDhtY0aNoSqNMSQYdsPNRCguHvXHCmdzbM2 CCNJ7G5414K3mcS8uYWiv0T7z4492iPcSlefl1F1/v2Pv7hsNtE/rAtt3mKNiYfP2/KcWINxwF1T 7TJCqR5tRl9+DeMAXsfOxP5ieBaOFZtc44D7Y1Md6zpOL/RbbrhW5kGyxut5PBMvg1TLM3/ewxcD vuzcesN1Rspr8zvV3/VB/vr07ik//fIH6uKaU7iF5vqNG8WGtmGIoniMWz5vTHweKOnn86bp1CDA MeE90S75oGSkkqD7usbLncgvPTVGAoIbmPXEyxU+DmHMV4JDmfuJ8HGZ18O4KwjB/k017RPl9cLc B0EA09OOSVL8ChzP/czlvtDzLzy3gXBSCmme5z4+q74x9cynbPxd5gggJ3tKzfjNxAXM/maAZ39h W9LT8NaCxBAGcY4Fg3Y61jDZXODlH8asm794hWyGgTMnL47HiJo1pH/vHpCW5G+z5cySb+kr4LRB xw3aXzEvOqz06naGIZLeFnF7f/26dQq1/Vu7Vi2wPd0Os7jGQirBWHx0iGHiYKbayPzGP9b72Bzw 8w8XGga2Ju5GwmhXYj/3nK5TbCO33/OWGO7CeDNDUsfE9hAPxvvzTDxOO0I71lzn4dmLRdqmQzDO d9JaEjTakc1bsBCXeE60lHIFufFnHpyv09KBLerhb6JJwPgnCSQpYOL1JATNEBbHV6pfry7jG+U5 TaeQGLyAmMKz82K9ad/lK0U2aQzvzXAj8QtApVlXPpP5ESVfiz37gQHcGVbIosRxZb1XnfXg821J u/O4UyLoPO783QxqeGvT5jxuf9P2lPEIGbrFTvZ81klcCpNiOA5wn020K2Q8SIZR8RwHfJHi3GOJ kb2HhDX3WBMTouif99wNVfCzJnwK60jyx8Rr6VRy5203uUmXzetUf3O80L6WsR9NUG3UiyrjhYuW wklkEJ6RDEiZF5uQP6wLMaWHskvye6prV3Hz53PJvuELnTNx3khN2i4JcbPwvBWcAmC2wxzYSGrW oWqf64ZztnOW4P13pUoQVni8AHm/0h6lXTg802sNw4HClMWX+QyJPzpV4o5OQBsLThyzstIltGqU BIe2Qh45zzBrZOcEW7vy8F3w3i8lrSVJ+XXSn3gzT4RBuWu5sOooqmV5yNdQofqUE+Yc2MNl34qH I9PszduiWROfLaRn7ZwFy0wYC17EXTg6QJJGwsawL4VJ9LqdNX+x2RvY3tcEDhx9z+wGCVPhbHbs /f6+WT/udMKAz1xgiZslgLyvOs4RLyzjxh7PX172XFIy9hPFJM5FpyokN/6kkPae0/ttaUXBakEi 5j35PGGkNakwoPe8wrlI580ztxNOGl4MKBG04zrv9a4jVAOal5DswtANIBAM0eFbGuwrLzoM2Bca ew0XDW9hV+x5kiiaGcTC49mdUBd6H/tLvnH1fpc/Euz9Dv9HrYOLr6s4jkNgAuAkXlSF0u6uwAl9 QcmdJ9FjWyxZ85aX0xmLDhbBeOHg2PFMQwb1h9NFL5ky7U8QvlD3aZZJSfeAvme7j5Xkj949eyBk TpQsgEMO680xRJMISmYZB3PTlq0mcgHrxPHGeIx0yNJ06hCw/eBJAE2JdqKy3wWpBsY2LIgN+SsK AfS9MvsrnC+YhZ3X2CiS3uznpzBtZFW8XM85gWO6uOckFnc6U5kjgJidMYHgXQSTCIYi/3QnTixZ ZpVxHHSfxfDLvjgLNjmZRnqS7UzhuMbz525skzZjzkLjSctxwUm3Z9dO0rlDOwyIvBO05/3Ov2Ph fTkD9n77D8ZgkGFBwIDi1mtnYq/h/BYn5pMOskoCxweJDzeJnVN15yzL/mYZHLhMfKMhbs5UgwQW uJB02JiAzvPefnMbPYM17qsOuyrGHazoiQ4O3OvZOfKI0Vl9epnwL85x6gsrShO7Q/1PMu4v0duT JgM2Tw55SpUpdSlsIpnjpOasNxeMZITr8ZV4nkGcc02GyMBKk33dd1qPA6P8iBwxTE5JztUu4lyo dqEc1/OWgyjnHaqz27Vt7e4zf1hQmtkVZgiUTnsmxnJcCQceqxq251nmNoTyoRPR0MED7eES+ya5 HQ6b1gXYXYWJ8xOdgzZs2gQJ8xGzm4wdLxw/g7DzTul/cSwx+E5ZQewH34nzDD8FTfZajm37Kei9 J3NdzrNU8FzsPUVtY+6SOHbtOpr7TNn+K+8MU8rbw4lv5LkDDQFhVTmpU6LGPXMH9+8D9V2Ya2x6 acci2GPt2LUHzhG9YHdUz1zHYeIrUHP07r0yHeSPHo9MdALoB2NuOl0UNsUi9tjUv+Zh5wyX1CQM 5K1vn+7SpkWzXAuOv3yPI7zL5BlzoapNgUE7dxjpl68Eko4KnHD5GHAyoGrNmWrBu5WSI6qcGM6G 357XOK/nb+4BbKUcEQh9Ux4fDM825/c3SVg9qA/tFMnrKV0bce45xis3v/sLc57mAmFwIqFkhW+5 nJxIXqKhuuziQ6rCbdmOoN8ovWbiy0BLqHlp28jPjl3R5jjz4ksCwwf5Sjt27pZjiD1oF3QSHEoE vdn3+cqjxI+jjgyWzGfB13jl9oDc2caeZ7v4zNBRozDJjAMOiOxEDQVfBP52+y32UJG+aVf3+lvv mTpam0G+ZLCO/NBMhR7SUdiqj3aVJZ2460gknId2I0QR1dq0j6VEkGYnNHWhmpv416ldR/qe5Xs3 mpKud3kuj8Scz6kVfpTntp7KthFDOy+cynJKOu/cbKCkSy9CeZR4Md6eTYzdZiRx6CCG6qiN7ZN8 JU6afBCoFuV2Zv6SsflbugoqDGxBBUkf97sdNqhvoXYIsfmTCMxfvFIO4U2YEzVJ6sC+vSUSu3QU JtHhgvZ3rFNqSpqxf8xPBU2VM8kxZgEJhp0V9yR2JtrH0eCens3M99Dhw8bL0nmN8zfLP3josJFk cHJpVD+vXZnz+sL8dkkVc0soC3P/6b6WYVC4l7FNjHm2afMW+2eeb9qhbt6yxWCJ7jFjkxLmjlHt zTjJc0P2Adrf0UGDO5JYtSInKMaK47ZeHGPOFA+HpedfegMerdgjGX1m3t3xHDx4/z+M1zB3u6Dk yCZOdPx7LwhRYy8xCal+JOG0KmdKHtu1aY09ogs3nm15JfFNadM62OExvp035xKCwhiDKZhPrHSN cwXNJbi7UGESY+VV+ulX9y0c1xs2bDZOG5xLPBODKnOcULPhHgcgS4zl6F50UJfPx46TNWiDJX/M hx7VtLtj/7M/aGv40Wdj5YmHH8i517PAIv2ds5ONr9vpDNK/31kmMDfHGevO/YX5smhfKvm7Q1Rb 8/LhKx89XnwI2JcDEm9NRUeAOPIZK28p72xUxlpIGz4r7OVE6ytxMrdvQf6us/evXg/VhZEGVAZZ rFVk8sf8tu/cY2z+KH2pEhIqQyCpLCz5Yz5cjCIiapgFnqTSqJJ5wk/auj3apRpE++vWjZCqIBjO VAV/u0icyx5tM7ycLU7O6+xvbjlHKSafhZrY4q6odotUc1onBubNh4sOPeuBe1lN3DuWMQHtZEsi MR3x0Rhnz1v66pvv5fa/3yd33/eg+dx+931mRxBv1zqP0RGiJ4L2OickEsHZcxfItwgsTJJuE4na Z4hfyO3pKC2nNIaerfQgbY0YhUwM9swXA9vvNBuI3r1H3hnzkXE2sXmxXSS4E6dOA/HMMazm8V6I 58aQLqU1cQKnicNL8MheuXpNrmoyPuBnIFcTJk/NpZak5K5N69Ymhl+uG/L5g7Ht6A1txwHJzxKE 7eHHWxqH4OO5xgHGBEmc01xjyoy/5AeQSqs2JenuENVO/v3kI8ahwoav4Zj79feJZr9jb2UV9BjJ G6YMd+JY2wiSyp18/KXzzh2KF8owo6FhHpQsU7Jq42cyH6qo7YuLv7z03MkjQLyJtX22Tz7HipcD sSOGzvm2vKBQ5iSAJQE8PRo3Yns3djgn77N7dfPqkViQuqRBBbt5Oz2HM4wXXNfOUV4DPBckL9al BfYw3r3HFb9t1bqNkLrUherRu/PIRnj/btiKWIVoBz/tfexS0rplc7NHMqWpm7FXckvs8dksMq9n MyWJy7n2FfQAAEAASURBVFetc6l/sTi0gvqanodFSYFoS/WwapCKwh4yO1Fq8PlX2NKKuzo0aWyI yuWjL4E6q/RKlmzd+U3P2VEXjpQPP/nckHQugLSDeurfL8o9d98hnSBpYx9yaz7u5zr26+8NliQK XGvTsagPHtA/jwTPWYb9zb18f5swWbiXMBd99i/J3qtvvosdN+aZnSBIIEg+l4F48LyVDJJo9+l9 BnZ1aG+y6961MxwLesKLc7a7P1l3SsR27dptdjKhinfDxs0mYDTLsXlRosPQLNzdI5f+21a0FH0H BwdCCrhe/nH/wyaoMYNzE4tV2LllKQIxc6K37bK/hw7GThUgNIVJHLsjhg2VLzCWuRcy86TX9ouv vCHc75nBt4NQF24zx51Gvv52vHlJc5lVwJ4TRH1Q/37uiAaUyr+H/XZp/0sCyLrRLvGOW2/CdX3x grkT5z+WrMqu+rN/3vvgY7NrB2MmFiVxT2COGZtY7spVa+RfjzwJCWpnoWahb58zhS89ztQRjiC9 e/bEjjEzjL00bShtIqFu0rgxQgxp8GeLSUl8l1fyUhLYsQw+B+X1hUUJoJdRRHUoVcD0AKSEi2pl quu4oHLy5cctdvRyf/XqVd1qmuPHGGaDErPKEo5JtWXzJiYvSiGZv/+86GVbxb2dFItibL+t23aZ +H7HUadJf86RHl06SmSTBsaRhPlxJ5CNW3bI6vWbzcJC1VLnjm2lWZO8pI55NmpQD4tSlCxEfEMG L/5r7kI5s3tns70cFxpO3EeOHoUEYw32IT2CpmdJk0b1TZ68vyiJC0wH2Cpt2bLN/XDRESUO6kpu j8WHjkbwDCNRVgggF/obrrsK6tNlshKqRjr2kNRygX7wsaeFKrLQ0GATLuUIAwXjepekhX2WZIzo zx0yqEBwUgV879/vlCf//YKRynCCYvlUO8+eOw87MSww+bDvqRq0xIY2nrTbvOvWm9zSJBLIe5AX t7xjfD2qGA3+6IPVaAeDU3Md55glEbB5GUk6Ttxxy01yRqcOBar36bqIJHsgHA/27dsHIrvFSDKJ Px9lxuZjm227WEeSqLPP6mNU6oWtM7G77ebrjcqdNnBU63McMGbfI088I/XrcxyEmjA6R7IDh9tx QFu5ISCd5wwZaIolxtzpYwvCT9mtGlm3q6+4TAZC3cp03VVXyMoVq83WhOxLjgUGXn4HO5s8//Tj 7n42FxfwH27xxjpzPmF7+OHvmbPnyqw58wwZrYV50ZMAEsdBA/piZ5BZ7nttkZRaMiB20ya+Iy7Y a/W7+BDg2OLHSqSLL+eKkROfJ+JXHpMSQC+9Go83bapU2Ok0dp8wfZZRq/ENnaSNi6pXBojDnCgH nt3LbNHGrEkkKTmjnSIn7j9nLoATBzw2kT8nd8YLY27eEifcXt06SVfH4soJtv9ZPWQ6vInpjEHJ Ap1gTLgXqvGQXxzezk2A4uxJm6SzT48uuRY4Z3m4DGW0N3mt37jNvN3PnLfYqHLocEBSeBRElvVm nerCJqo/dhmh+rioiQ/V+cOHmS2kSFj5N5MhRSBOxJnSsvxCongrn31gU84veyTvd0GusXfld23t iAh5/JEH5Onn/oNAuJvNIsq2sa+pCiN5Zhvt4mrGFM5RkvMQtnGjpK2giV6Xx+AY9Oa7Hwjj+ZHo se2eMeWYH8tnP1Ja98g/7zX7vjrLadumtTz75KOGUJIEkkiQEPHbM3EMsN7BCJly87VXyLXYOcWJ ub3eiZXztz2f97tgV/G+gl/pKoUvMZR0jxx+rjwB0kzJH8eXZ2LbiFPDBg3k/v+7S0hy/CfvNeGW dY899E959KnnjPqdxJnjgM+6axy4xrpLoucKHcMXTJLOxx9+wC11nDNvgVH9Uu1PjDmHtGzRXG69 6Tr87bLg4cvU3+64RRiAms895y2OBdpqDh86GIRyoP8meDlLO8nWrRBDFC9oIdlSfpZvxwPz97Uo 9kCw7cjIxqad9rkmrjQb4L7Y1hHJS7F66BQgwH7jOFMCWDRwiZ23+a1ouZWuu/LOgKWrfvnXxixG iPqNBQk//SY+ACbuWT4XkpBwQ3oJyjKk6BgWVzxDSN4ne1soJzkumqyLTQw6zUmbiw2lLzTIzxlM /vPjop0CZw/PVBdSyRFDB8gCBKaOBqmgzRclSkdij+FSgIDKkmRUQ4DrtpAYMmxNfmFmSEgG9Olp 4gKugeQwARKpGEgnYhAjmMk1iQRCKtgAKvHuuRxxXFcU/l96Dd560w3YY3ccCMxxU4YNa8PFkOEw iAETvyklY/tYF2LMuHMkis5ErGnnZgPxcjGn5MFb4nhgnlS9Mk/+zTAgvgYSpTPEJdiMNVe5zr62 ZdCp4s1XXpSPPv3SSEJoe0YTAC587PE02N9R6kTMuSXdBSNHyE3XXiU1vASgtnn6+r7yslHCvYU/ wF6/G7D9FscuxyrbwxoSB/6mswAX5jtuu1HOOrOX1+yoBn7rtf/I/6DCZhBfjlXiTqLBPDm+CTel wq1agIRAynXR+SO8LujEnVixwbynIP1AEwSqDE1f4zfL85ZS4QXPvEPQX7yGL2Te+sHzXj4n5w0/ B89Lirz53hio510Bmw0hQVH0jKbUuRNU4w8/cB9U9nm3Y7Rjhi90rnGIMWPq6lmawH6wlbz16ovy Mez5psGGjypf3u8cByyTJKkegnRfAAce7oVtvanptfwSdg9hvEUXUcT8AnJ3DaR/docNWyrDxlw+ +mL575hPQMxcxJCYv/HOB9IWu4NQYmzrboNm87lg/EFvODeGg9r99/zdqK1JWO04MAMYmHOc+cK8 GbbE5AtF9O49tnrmfo51a3bgPqE/SgQBjjEzJ/h4pkqkEmWwEK7n9iWmDFY/3yqXeQJIYtWudXNM wql+JVJceKkC5RssjZT9pbp1Isx+vCRRhUlcrujc4NzBgCFmuLdvYfNiuSQ39RF6xluqVSNchg3u Z1TBu3bvNR68JLdcqIODQwxBa4WJ2Okx7S0f5zEuMr0hKWzetDFsAqNNmBGSVj4ElAQ2BX4tcK44 H4g7b70RxKSLTJ0+w+y56iJ5JBkhxrjdbi3G0BqUNBmM0UY6MURAOkPphzMN6N/XLKCUUDDRSagb bJa8pWZNm2IxvdRIe4kbiQR3d7FSDuc9lHZQ4sb9XM0Cm81NuLh6S9zn9NmnHoWX7cVQzc2XnTCG jzNEEOpGtI27YkTBG5K2Xr62XPOWr7dj3AO46xmdZOGSJbJo8TJDbBJBgulxWrVKVUMWzsJWXdwG jbEj/SV6IL/6IqRW6zcYVV909G6Jp9MIiCDtPevWqSs9gCfDeFCa6Cv1RdgTSqtpd8fEfqCjiLfU uFEjufKy0ZKanupepKojzBHt5zwTydOQgQOMLZlxRDHELct4zHpe6/m3JToXw06TnroTp0zHFm07 jMS7MkguVePdOnfGHs2DfW5p1xxj5rprrjD1dI2ZTJiJ1DVb2XmWx7/Zt08//hDaN0pmzJwNc4Bd 2eMgwz0OGB+Q47Y5xowzRUfvhaS2C+w1XXZ2GemZZjeQURePdF7m/n3DNVeZXTdI3I23MbAhyeMe vKwHP9ddcyWeX9e8xj6JbOL7eWasPu4JPXnaDGOqEX8iHmSR80uQGQtR7dq6y3b+4HaJJH+cN2wi +ezS5QwTE9Me0++SQ4DzF+d3viBrKhgCnC+ImXMcF+zOsnNVJTQyeykrO5XWmuZFgN1IqRmJri/V TN67/B/hwODETalMSTwElIoZySvKJcksrnb4b2XJnaUEykp8TmXbOBY40VNyZ4nwybSSeRmVfCDV fjmL+snkearvfQXOMB9BklkNgcqZ6AF90/VXQ7J3b56ieY7SkeLAKk/mXg6U1DjwUvRJHaJzCucE jgFfL4FHoYl4b8zH8u0PP5rFkwVyPJIsv/XqS8Y+8KQqoTcXGQHO5YmJCeiPynIibr4kHJ+Nfim4 DIjbqwUHN5YadWnyUfit4Ipc8ULfSHvVDImLnSQpSRtQ14AC5+DaCq69hNcehbU0Q6pDwFASa1+B K1jMFxa894u5YM2ueBHgAsbwHcWZSCaLO09/9aMUxpv9mr97ytK5kiIYHAvFiaM3iWhZwj2/uhZq p4/8MivA+ZIaBwWoSqEu4fZ0nomOLZ98/pXZuo5qYcYz3AEpp5Mg8gWiR/duJnSR5/36d8khwJdO 7j+emnYyMQGxKhiTEL4IllbZEc1fWDeuYEVNLnvX8kz+iIwSwKKOD71PEVAEFIEKjgDtoyfD2YTx AWnmYl5EIb23iSYkNNOgqUf16i5prD2n3yWPAM2D0tN92zn7qxHpVGZWsqQm7QS3Yh+XXgII42eY rTjt7f21zOMcmkVb9PL+4stW5zypHhjon4qAIqAIKAKKgD8ESPhoJ0XHKKe0hDajdELhLkMP3Hu3 9O/rClnjLy89d+oRoPqeJLBICZK/zPRjEn90UpFuL+mbYBSFIgtvskLpoWdYqJKue0mVpwSwpJDW chQBRaDEEEhLTzNSKbsDBffP9uUNXmKVKocFGaIHFS/VvMTaqN5g8xcWVt2EfLnlxuvkbASM1lR6 EAgJcamCYShXSC1ptmo1q6w4khRFBUzHjyogycVrTlV6ej93TZQA5sZD/1IEFIFygADD07Rv3dpt i0aiwrh2mooXgUh4Ft8KkheD3Uqg/8V2ldXhbdxQ6CHMreoqghqteBEtmdyqh7WQ+GPc/jEFBRZW SlYUYlUy7TqZUlwOejWkVu2K88KiXsAnM2L0XkVAEVAEFAEQCZc9GFXCmsoCAlly5NAMiT08E64S JIAVvd8ou0bYtYYXS3it3NsbloXeLGodC0v9i1qO3qcIKAKKgCJQThEg8VPyV5Y6F/Fq6/SHY04H kPecjQvKUguKs65ZmXBWCj9Dwmp2Lc5sS31eSgBLfRdpBRUBRUARUAQUgeJFoHLlYKnTYDi2+mtU oUkgyV9I1SZSu945hYoZWLy9cXpyUwJ4enDXUhUBRUARUAQUgdOKQFBwhNRvfClsNWuDBHrfMvO0 VvAUF56VmSbBVRpIg8ajJCi41ikurfRlrwSw9PWJ1kgRUAQUAUVAESgRBEJCG0i9RpeAANWpUJJA Et6QKo1A/i5DIPMGJYJ1aStECWBp6xGtjyKgCCgCioAiUIIIVKnWQuo1vAhe8zUqBAmk3SMJbz04 fZAAV9SkXsAVtee13YqAIqAIKAKKgAOBxISdErPvV8R1jIE9XPmMEkfyR7vH+pB6UgJYkZMSwHx6 PyMjU9Iz0s02R879LfO5TU8rAoqAIqAIKAJlDoGU5ANyYM8PkorvSpWDylz9/VWYat/Q0EjYPY6W 4NB6/i6tEOeUAHrp5oSERJk1Z64sWLJMjhw+grehNBDAStKsWTO56/abJKKWd2PR+BMn5Dj2xjzZ xJhaVatWkdoREV6z4vkjsbFy+Eis+WZ9g7D/Zs0a4VK3Tl2pVaumVKtW1eu9nge5h+cR5OMZwqES 9vVsWL9eru2dPO/l39xdgfXIyMi9wTjrUh17gJamtH7DRvnux1+kMmOV4ZOJOjds0EBuuO4qqVql Smmqar51YZ8fPXYUWxaFSO3aEfn2U74ZOi6Ij8c4jvMYxwjzFhgUgPFVR7ipfH4pHVuDHToUg71D XfHh8rue24hxS7Fw7CAREpL/VlXHjh8X1tOOWz4TDDpct07tYsUiv3rreUWgPCKQlor15cBkORG/ Ac8YLcXKeJxA7A3MQM9htbpJnXrnYi4LL4/dVug2lU8Zb6FhyLlh3fqN8sa778vSZSskOSXFTRYy sKDVjlgvV10x2icB/GPiFPnvmI/x1nRyDwsJ59AhA+T5px7PtZiRZM1dsFCmTP9L1q3bIDExhyUN 9eIuByyRC3PVqlVBVCNlQL+z5fzh50oDkDh/af6CxfLYU8/hrSgYYTCzE35wYX3kwftMHvawt++d u6LlwceeQV1ipDL2mWTKzMySe/52h1x5+Shvt5y2Y9G798hX33zv3rSee5V26hCFeo4ucwRwweIl 8tSzL0qfXj3kpeefKdYdF94b85H89sdETJI5b/9Z6NPgkGB569UXpXOnjvn24b79B+T2v98rJ04k uMeFv5sCMXbDwsKwi0Qj6dL5DBk2dJC0aN7M5y1jx30rn3/1rXlR4kWpKanSAX35zuv/kbBS9uLh sxF6QhEopQhY7+CQ2IVy9PAczOnJWBPyf/Erjc3JysrAOlpVIur0k5q1++B3cGms5mmpkxJAB+y7 QBCe+Pfzsn7jZiOFqBIa6j6bgQUqNDTEECP3QY8fyckpEnv0WIEWPI9bc/3JxSwBC6ebkOHs3n37 5IP/fSaTpk6TxMRkDGLEb4fUxCUBIf3D1SA0CZDoHTx0SJYsXS4//zZBbrnhWrkQ22IFBnp/eEk2 WecqVUPwhpRTjXTk9cXYr7GfZw+pAwmTr0RJz3FIY5hHQHYZmVCbJ6ck+7rltB0nQa4GgkzsiFtQ WmB2n562KhWqYI4vvsVWgbSyRni4xMXFG0mm3W6LUrGaNWoUKk/Piw8cPCR8KTgGSXYApMrOlIoX olmz5xWIAPJl5dix4xIHKZ0dF8688vzG4DsUc0S2bNsuM1HGd+N/kqtAzK+/5kpxPof2vsTEJIy5 o5KCPWiZUoBNHKWWjjFsr9VvRUARKDwClQNCJKLuQLz41ZXDB6dIehq1AmXJb5STQRbs/epJnfoj pGr1VoUHoZzfkXuGL+eNza95EyZOFkoAKUWzidI1qjk5lCgxslse2fPObxIyEi1+n0zK9Mhj+/ad 8vATz8jqtetAWEJBTnPeYGijyDcciukDIIEzJYPoMG3fsVOefu4/sh/SmLvuuMUlzfSoGIkQ60wJ jJMA8u81kDJOmfanXHvV5R535fzJ+0msmIdVDWbi2MlikFNC8f5i/2VlkTBjaihDZIFE+9Mvv5JN m7fKA/feLY0aNgARDJU6UHmyTeN//k0mTp4q/37yUWka2aTIoK1ctUai9+wxEkXXy4UjK6hYZ86Z JzffeC32fK3mOJH3p7dxkfcq30co3X7znfeNlPv+f/zdLemzd7ietUAzbnksA2TVjj97jX4rAorA ySNQPbwDSFRDSU6Ol7TUdKyFmDgxx5fehPrh/8pYE4NgthJapaaqfH10lhLAbGBouzd3/kIQmRxI SP5IBgcP7C/NmkZKenoaFr4ccuiJqSWL3sgPFyfP41y402k758FESDgpQaGtGm29XnjldZCx9Uby Y8vked5fq2ZNszjSlu8oJC7My7aBdoG8bszHn0nr1q2MWs3eX5Bv2gGO+/YH6df3LGl2EqSiIGXp Nf4RoJ3lrDnzZeGiJbJx0xb5Owg9bT2DAoNAlD5AH3+Ofq8M04XlRSaAtNebNuMvY/PqfMmwNauM McyXCpLEfmf3sYcL9c3xyOckd6JE1mXCYEmnfV7GQmVPCeB9//hbnucndx76lyKgCJwqBBgkmZ/0 jCxI25Nd6xYKK200kO/0rFNAALU7ofg+VYiUj3xz2E75aE+RW0F1UvTuvW7yxIxoy/Z/d91mJGB2 YfJXQH3Y2/Xs0c28eeS6DqvbruhoOQo1qc2H5I2OGrRzCsQi7iSBJICtW7Y0186cPVcWQZ3rNIzn ItqyRXO57eYbpGNUOyOtSUlNkU2btsq4736QVavXutvBhTQFqjvaTJ3dp1e+khtnvUkgt+/YJZ9/ +bU8+ei/dAF2glPCv+vVrSOv/edZmQb7zz8g6XvlzXcgmU6XX2GrxynvwpHDZfRFF0j3bp2LXLOD UP+uWbsBk2aOBJvjlOyMkypfSDiWZsycUyQCyLwimzSWevXqCm0KbSIhjE9IkD179kLKkOweu3xW gmGH+P1Pv8qIYedIB4x1TYqAInD6EAgMgManaqjRiqXC/IJrUWlKXO+4VlIIgulDUz4IKAHMBigL ixClaJagcVGqUSNMBg3o5z6WD5ZwmDjHfLxd9+iTz8qvEyaZBY3n09MzQP6ay5h3Xofxu3dv2RTY Ak6bMVMycG1giOtVhvWindfzTz8uZ3TqkKuoVi1aSK+e3eTeBx6RpStWuctim9at3yCboT7s3q1L rnvy+yM4OEj+gGqcbevVs3t+l+v5U4QApcckT7fcdB2cHdrLo089ixcKSHyh6xh18YXyyAP3GCeK kyl+5eo1cuDgATcB4wtQKMwN0hAGyRI2TrDzFy6S/QcOGA/qgpZH8kc1Nm36rr7i0jy30X5xxcrV 8s77H8rmLduhunFNTWx3XHy8TJg8RQlgHtT0gCJQ8giYFzOYg9Brn880Xwq5LvFj18+SrBXLJOFj fVzET5lfQfEvFwSQA3Dnrt0mnEnciXgjTAuHJyDtoyKbNMpl0+cEhmEqYo4cMZKt6OjdRqXqPM/B RJUXw7twAQuEWJketiEY/IVNRpLicROP+XuDojRkF+rlfKj4kDVq1FBatWzhkZvrT4bpuPSSi3Df XkgGs7048TzQSD5PaA+vOeQ+yAWYjiWffjnOEE6K1YsrHT12TPbvOyAxCGmTCFU3vYip0qbTSfNm TQslcUyABIkeyQfQp3RWYAiahg3qS6tWkKQWQ4XZD4dgl8Y60xSAjhgMB9Sgfn2pD4mWr0RHjb37 92dLeCsZ+xmGnKFJgbNf9+EaOk24Qi7w8kxpCDs/4sGUmJgI1etamQybzOl4KagLiSAnX57/a9Yc kKatctHIETBXGABP2oa+quPzOMfhn3/NMlLFkOyXjbS0VBl98UjZvHUbJIPrjY0dxwNtSufMWyBX XFp4L2+OX2+JLzU0taBJxEOPPmXCx1h82H/rNm7CQpOay/7VWz7+jhEveoIfgbT/OLCmB30opAW1 2I/owwYYL5oUAUWgYAjw+bSki/MHny9++Jtrm31+C5Zbwa9i3pyH+OEabT8Fz0GvtAiUaQLIwcYF ccKkqbJmzTo5gUXSLjD09Kxetbq0b99WLr7wPBlx7lAzWG3D+U1Pw48/HwdbgWCj7nUOWg4uehX+ 65EnTVgXLkx1akXIh++/Jc2xeJdEYluo5nOKss0CDOnLxi1bpHsX7+o+qsu6dO6Uq4qU5lBFXZBE KQ9j5FklHVXB8xYsMlhfcuHIgmTh9xrGbxsPtd5U2JtRxcxwO5TAcqGnnVktkIHu3bvKqItGQm19 pt+JhH02e+58Gfv1d3CSWW9UiCyck094eJgMASG6+YZrQSiL1mfMc/xPv8iCxUsR7/CIwYVYsh84 +UUibMngQQNAhi7xGnKH4Vqeee4lSc+ko04leNKlSY9uXeUDSH45cdn04SdfyK+/TzShVjB7wtg6 TZ569EEZdcmF5pIEjO1X33pXlq1YKeyDW2+8Tu5/6HEZOnigtG/XVp789wvy9PMvSzUQ3yaNC99H u/fsk2XLVxqSxwI59khUz8NYIhFdDukcxwXbkAzVD9XAF5wHzzrEqyzO1OWMjtK8eVMQ2m3u55Vl Ho09arx+SeoLmzhGSJJ/+WOSmScobWT77PPOF7smeFHsc2YvhHkaJS0Q71OTIqAIFAwBPp+WhPGZ smSQ3/ww8Tg/TLy+IMl5vb2H5di5l9/2eEHy02vyIpCzAuU9V6qPHISk5+33PoDn4zSEgkjDAHQt TnZAcJE+Hh9niAsN4+cvXCz33n1XrkU6FQQyKYmk0bU4c0A5E8erISfZgzqparJ7EDuvO1W/SYZo z2CJGMthHRly5d/PvyJ33nYjFq2eJiQIF2ebaFvYulVL+2ehvqma7ggVYwQlS7PnmgebmPL4l+O+ k759ehvpU6EydVy8CZKql19/WxaBUPEBZ6gRWpjZfuOEcQjBtydMmmKI3bVXXWZsHRm+JU/C/azT f8d8ZKS0jFtngjzjQvYdJWok+WvhQPPGKy+4YuU5wcyTYc4B2mF++8NP8tGnX0oM6mM8rLPHB+vK /GkDQ+kY20Rbzcceul96du+Wkwl+8cUhKTnJ4Mf7SOgpsfZM9DBPgqSV7ScuLN84CGVfSMku1aeM 73jX7TdDBXvQSHTZxuHnDMYYCDNewINhslCUtGzlSoQPinFhhAxIkBogSHYU7O7oNF0FIZD4TLEN fCFYhReuXbt3SxTIZ3EmSlYpvSW+NrFMPuPEqLCJOH702ZfyCV70EhMT3OPZ5mOwhkR3K8LPUIo6 c/YceeRf90MaWTQcbb76rQhURAT4rDrJIDGwRNC+dPGbyRI8+81jvN9+8zfXO364vtk1zl5jLtR/ TgqBMkkA6RH58OPPyEJIV0iQrMciB5J94+BgsQOHx3/+9Q+org7KGy8/LxERtQxoJAvOgeUNSeZh U4Djtz12Kr+rIsxHx6i2sn7DhlzFMETLtu3bYQf2nFFdcRFuh09LSE4iIyOlCVTE4eG0KyzYm5Yz c6oeufjeevP1Jh7ioZgYgxFtsjZs2iy/I9j1LTde67ylwL83IL7iQ489LVtQd6rebGL/kGAygDbb RphJ6KkCf/9/n4KEp8q/7r3bPQHY+2ZDDfnumI8kGcTJ6STD8yT1nGgYI4/qwxdfeUOGnTsEhLNy LnJh8/L8/uTzr+S9Dz82hzm+WEfGnOOY4QREckYHBUoBmTYCm4cQEPuVF5/JRQJ5LccYMeVvKz00 Nzn+sWPVdW2WqbvnREeJqD1GVaZRoUMazHbyRaAXHJB4f2ET2/bnjFnuvHk/petR7dpgZ44waduq lTRu1Eh2QMVO8scy6NC0YtXqYieAjO/n2uEjpxWsH80ZWHZhE+1XP8AY4rNgxwhJITEj5pwv2Ie2 H3cB1+dfek2aNmkM84EWhS1Or1cEFIFsBOxcZQmhBYbPs03O3/aYvc9+2+P6XfwIFH5GLf46FCpH Bhn+3yefyYJFi42btx0klKwEg6Q0a9HcLNZ79u4z4SxIXHgNJ/+FkDr979Mv5MF/3mN2g+BinIGF wL6ROMmerZQ9x29emzN07RWn7puL0sjzhsnkqX9KEqRGTgLKh4oPz17Y0O2G9+TEqdNNGyMQGqQO tpBr07ql9IA0aiAkRvS6LGhi6Beqvtu1bSPXXX2FvAa1I3EhhgGVA2TsN9/BA7S3tG3T2mRJd/uC 7HzCPF9+/a085I9SNMaUa9WysVDNuXefy16O7WO5tLcch1AgkU0ayjVX5MQjJAF5b8zHsB1McEut WCESFyaqCmtgO7rDkN4xwDD73iXJC8Ci73oDNRd6+WcBQq18AptHJhJS9j3rc8F5w6V7186IKRUo O3dGy9RpM2BzeMicI9Hk7hevvvkuHHveMCFabNbsJzvROX/b8/bbnrPf9rj9tmOdf7OPb77hGpD+ 5obEEKuikD/mtRWEfO16ev+6yCPLZ3vo0c7EF6auMCnYtn0Hmbk5xj6fMXO28TwuTrvQNYh1SY95 WxcWxvrQ3tJfQHJTKY9/OEbGfvOD6T9L8NiXnTp2gGp7KHYMCTMB1idMmSb7MO5YJl9Mdu/dK19/ P16eeORfuUixR/b6pyKgCBQBAec85vxdhKz0lpNEoMwRwNXr1skvv000ZMcOHr7RU215z9/vyCYm lWTL1m3y7vv/M/HzOPnzWpLBnyAJPHfoYNhhdUF4l8vk3CGDjBpyx86d8sQzL4Lg0RaNkhp624bL U489bMgESQPvb9yw8DZIJ9NHvbATxyUXXQAbt2/N4utcGNkmSsoQ9UgoT2MctyNHjhpnhbXY93YC 1OMkarfBc5R2gRYvf/WhzDApKdnYoF02+iKQz+myGipUEjGWRSnqZ2O/lheeecJF0CAdc9XBX64i k6fPkKXLV+VyoCFZGwL7udsRzqYxHBe4KwmJ2ocff2aIoCWBVP2NHfe9cXBoCKcLphmzZssGSPZI VGxifgyCfMdtN0kvkF/WmZ7djO/4yedjIdHclEvyaO9zflPqyPbFg7CS2ABStDPAxN27CYSL9mI2 DYHTwkOPP+0mlpQUrkXw7KWw0+O4YqL3LMcOJU108OA3x5ZnosSSUsUAXEvCY65j4T5Sp45RIDJR Ps4W7vDixdjzGlJ14s3E+tUG6TsLNnE2nQXV/0+/TrB/utXAxLR714J7lnt7yWKmx+Ess2LlKiPx 5cucrQvPEYZOUVG5+prH80t0CqIDjn1m+AJXr25deebxhyUKtsE2tWrV0vRjJsYZOsk8JwsXLTVO J9z9R5MioAgoAuURgZzVrIy0bjpUVfRmpa0QExd9esU+//TjbqkUjzNuGp0e7r7vQdkLaSAXFC4E 3LZs2p9/mXAo9ODkh4kRwzH3Y8HGH/zGokPpVnuowbg/6elKVDnee/edIKkB8tMvv8OujVuuweUd HyehI1WwhBAtNdUlkaBq8rGnn5dY3HftlTkSNN/tgZ0aMKWdGj1N6dxAosW8mD9JztTpM2F3NsSE yKHEhKSIePlK3MJsEsgo+yow0LWgkuz17tVDnnv6MbenK+8fffEFhng/+OhTph4kDFS1MkYjHVEu g4czyRGJIvOziztJC/eSfeaJR7B9Xe5wNddceZnxKn7kyWdM6BR7j7f6Msjy0uUrzAsGz/Plokf3 LlB7X2cIr/MeSsguBz5vvzfGSAo5gFi3ZctyCOBZUM1+MuZd14DCebMhefXq7nrb/O64+UYZdeEF rjKAZSYGYgt4Qp/qRLL/JxwkOOjteGIbGHOvoeNlh/Emm0Y2FkrWLTGnqpbPY0EIoGtsBsrX340H eZ/jDivDZ419R29pEjba3DrJH+sSFlZNzj/v3EJDwe3sqDGw7eIgheAS2OeYdTDTQf3PlpeefcrY GPJ544sUd1nxRVYLXRG9QRFQBBSBUohAmSKA3BWD0hUTODkbTJIAho+wKkknxq0RKuWikefBScBl y8VzlAYuXb4SO2wk5AqKzFh7nomcJh0x0E53olPHw1Bb9zurt0ycMk1Wrlwr+w7sN04DVNlyoeLC xW/3YodK8zfbSwna2+99KG1gy+VJjjzbhluwIGe4HRUoOZwFL1uSZqrRmScdZz77YhyIUVdDxF1S Md8McPfePTCy32kkqCyPZJLS1KuuGJ2L/Nm6nNW7l1Ffz52/wCH1ycIuGEsNAWRYHnoPO4kc+4kx G3uiTt4SvYmHn3OOkaQ67/O8dgnIX0pyqruubC+dEugFS1LmTNYekPnZ1rMPtu3YYaSZtFuj+tTa nDrv9fzNgOD8lHTiDjMk+E5SxPac2aOHA3uBhLaRCQNEkmYJGk0SqC4neaPHdX6JWNKO0KiSPS42 WxniJYcqd5tIDEkA+RISBZOEwiaGgaJ0GpwS45bEL8A4urwAe1CGSmL4nlo1a5hwURcihI4mRUAR UAQqEgJligDGI8ZfdPQeM6mzk0gkSAY7RrX32We9QAhCQoJAalwSLC4C3HGAUsT89jP1melpOtEX BJAfeoBu27ZDtoJorN+wyai7aevG8BbGFhLEwym9oLQwPj7e7AZCRwEuxP4SpSaUxDCRfN4JlSoN /rldHvM1JBpeoz8gPMqN114Nm7icRdtbvoy3eORorDswNfuCpKizRyBrey+lu53P6ChzQABtYp13 7tplpEX0lmWeth1GOoldkPvDNtHZbnuv/aYUy5A1jBt7rz1nv3fsZBl8GXA9GiSqC+BBzuDHbpZn LyaOYH7EwyYSGaozSVxwxh4utd+M50evcm63xmTUv4jDOGBA31x15hjq1aO7CblkpcEkgjtAxJcj 6PigAnrNGoLnIHm5CnH8QfwoiabU+u47b4OdaW6pneNSnz9bglBzbpi/YLErvA6u5PhYsnSF+QSR oIMA1q1XTxpCW9AGTh+DBvQ36mGOE02KgCKgCJRnBMoUATxxIgG7YuRI5LgQ0U6LITB8pWpwMAgJ CTWBdO2iT1sgSi0aN2zo67ZSfZwODvz069vHqNISIJHbB2cQSmcYJ27K9D8lNvZYLjJEkrJh8xZj H+gvcDEbTlcXF4FxwcBFlNuMMZwGbe4MjiA+P/z0u4kFFwp8/aV49BuJhU1Ug7LP/DkPUDLj6fVJ tTQdRegxTHs525/Ml9K28PBwW4TXb5JZfwSRN1HKTHWoM1lJlPOY/c06OOtBSSTLKIhjjM3jdH3T PpKSTScmxLZ7ly5wLskrjaTdLB2KYhGTz7ab8Qqn/TWzwASQWPK59UzmCI6bcQJMOUavuny03HDN lX7HiWc+zr/Z3ySPW/GyRGcdmhKwrTZAOuvBiAIM8L1mzVqosyvLl19/LyNHnCv3/d9dcCKq4cxO fysCioAiUK4QKFMEkMj7XDx8dUvutTznKi+LUM7J0vHLSEGgvsVq664QfwUFBWMhcx0j0aAks22b VuYzDDHhGBz4wcf+v73zAIyqyv7/SSG9h04CofcO0pEqXUQRRVRc7G3V1V3X9b+rv113dV3XdVfs XRErAiJFepHeey8JoaRBQnr/n++dvGRSCAOGwAzfq8Nk5r13y+e+mfedc+457y/mQm1vyYDlLCEx sdI7VxQ3VOYPrKP7efFSE3EMMQnrDyxys378qZT7sMxh5mVJ7+23VvyutUeF81ykG8y28hqilBCz 6rn459IV4xVc30Y0OHDO5OTkacqUesbSePFtV+8R+w8cMu5qy6WL1nGqQWR/+MkXpUQ7BB/WbUJE 2c8Nzj8kkEYkemREg0oHgOOwZhSWN62k1L7IeRmornYk1u7Ysb1J3N6kcVSpfS7lRWeN2v73K3+T t97/yNwjG2lm8LnC58KW29FDP08l1j6siZz+7QzzA+MFDQBzJMDpUvrFY0iABEjgShNwKgGItVhe egGBe9KyQCDJLpLhnq8gVxqEj72VBl/+wUFX/697BDq8/d5HagHRZNB6vdRkIuKmz489cn+lC++x zg93m8B6QXsBCNcughoupSCg5j7NDfjXf7xmBAB4Yu3ftKI0G/btlK0fueTst8Odl5ySbKxt1q3O yh6DlC1wZ1s5HrEdFkM/Xz+BJRgXZnshgnWOmOvKCoIWYGGyt3iV3R/nmH2BWBh781hNQVP+/rX2 +1l/o08IRilrvbS2X03PK1et1jtsqPtXAx6sAnG/avUac1u4MoZQhImYfe3nEn/jLiJbt++sVACC C1hO1kjqEXpXnrIF5xOENiLv/SpK+l32gIt4jbWq/3vtFbNUAkE+MZrm5fjx40a0no5LMOuBMQ70 AaLQ283LJJhH3kVHAlwuoivclQRIgASuGgJOJQCNhUCtDDt22tyb+MLGRQWRriIV3/5q4+atRvRY 6UKwf329tZUtUfJVMw8VdgTCdcOmzWadow7VFCRFvk6jZy90YQKbsjY25K+DW+xSy4hhQ+XHnxaY KFnwxMUS6w5RKhNVuItFuOZxg0DDhRbBE1hPuFtT1VQUYY0AHdx71hoz6oeAQHJetAkrEtKUnC1q G2PFreQQ3DNq+A2lxD6OtQoSUWP+7dfsWdus56iGuAdxCTkEB+E+ylVhjbLauBqe0zSQBom07a1/ Vr8wtxBjFRV70Y3tYI+1ejhP4Tq1F4dlj8exOA+uBMuAAH+BNRAPFHy2EEwEF/CiJcvV9fuNSX2E uce5nKZLAfC9cqHPWdkx8jUJkAAJOAsBpxKAuO8oojyx6ByL81GwqBw3sb9Ro/jatimdFw052X7S O1fYixNYirp37aRCyP+qnyNEhdbTdYpn9eb11oUVKWBmzJoj3bt1Nla+igYBYbVNgzbsL+5IbYGL L+7ocKkFt2ND8mHcEQTr8cDVnu356o3Q9CHNmjbWQIqNZhxGNOg8fPn1d3qB7SAQiPYFt1XD2jT7 /mN7L43kRUHEKUTEfr1frBU1Ci6Lly6X0SoArQTGZueif1aotWv+osWVij/sCq6wNJo7k6gOhCDY ruvDsO6tomhe3GYOycWR7w+K1fqB8cyTj5l6cCcVWJ0s4YRnJKju2L5dKXb7DhzUHIuni99DxDGS cdfX27FdjgK+R44eU8YexdWjb3DzwtZcWcGc24tob3XprtW5jdYArSaNG1V2qOFT6Q5VuPErdeUu XKLR6ypoUSDmB2hwyyRNcI5AIzyQF7Bp48ayWnNFIhm29UNRJ0wDp9KrsDesigRIgASuLgJOJQAh HIYMul6+/X6m3l81y1wsIRJOxsXL//u/v5sF3630oonLF+7r+eY7H8jp0/HFYhEX5xB1/Q7VdXKo 62ovkZrUuEf3LvLj3AXiq0IXBak3EAX8zB//IndPut1YAxFQAbcqLGfbd+ySz6Z9pekuEktd3HP1 jhv99K4gFd5T9yJA9OvTS24YMkiTAs8pvrBe6HBY7EaqdWij5seDyAB7CAisHXtecxTed89dEhFR 39xqbe3ajfLBp5+rUCjJ8Qe3dWREhLkDiWGgLK7r3s1c3K22wQVLAf7813/IA1Pulo4dOhjXOSxd SFXy8edfmmTH57NsWfW0atnCMF22fJU5Hv1Eypl3PvhEHnvoPiPerH1jjsfK62++rYmm1xe5fDVF ji5JwHpJS0is37RF/vj8C0WCUtfRqVu7p87p5x+9Wyz2UB/6992MWYapEWI65pc1Nx0CIS5HwX2e kfTa4oE2sbwCyaXNjw19XVEx7nvljIAj6zOE/U9qwmVESl9IAFZU5+V6D3efgbCzxCoCh07Hx5kf EvZWyITEBLVknysVaYyxQaizkAAJkICrEnAqAYhJQHqQCeNv0tvBfWYsLPii9lIRePDQEXn6j39W a1K4ERlJarFBjkDLUgirBqwbk++caKIcnWFCseD+gSlIwbJLrUOnii9ksNokJCWZ27T5+weYiFpY ZdLS04xrFdfuknGLsdbBmnS7g+vYKmMDwf3gvZNlw8ZNckqFt2WBq+wYbBs5bKjM11varVaxZN1d ARdmJHdG5HK43toMt4VLRISp7m9ZPGG5xNh+M3mSRj6XWMOGDR0o32samj1qjQQnFPQNwQh/+dsr 5lZsvj6+6uZLVSapJio3Ui2RiYlnSgU3mAPt/oFYfXDKZGNBhQUIrOF2/ub7H3TpwS5jXYQwiIk5 ri7n7aY9P7UkoeB8C9cUKkhWbVlGMY4SiyLWmOktC4ssUnbNmrlFKhZsgxjDcY6yta/Hkb+RQme9 ri+1GOMYsO/Tu6f8++W/Go6V1XNYLYdTHnhMz7f04nGq+VMF+XIZP+5GM97Kjq+ubcgP+tV3Pxg3 L1jifDsWfVwe/u3T0q9Pb00CX9ssS8Ba29NxccXrNrFONDg4UL9r2lRXV9kOCZAACVQ7gYtPrlXt XSzdIFa2Ifdc3969jLAxaSN0F8tdGKcXN6R1wEXUeg9557B2rnfP7nLPXRPtLlql674aX+Gevs// 4SnjqsIdNTAuFOuClqUWp1N68Tpx8qS6rNLMmC3xB/EEV219dSM//+zTmvamRED9mrHidmt33jHB rLuz+nOh+rCw/5knHlO3ZjPTJ+s4XJThbj2pVk2IdozLEia4hRosl2jrZl2Qb18QPPLQA1M0f52v EV7WNsw5jsd9YHFfYQSMIK9fO10e8OSjD5t8cNY5Yx1T9rlTh/Ya8HK3svQwdeNHBgQdXISffjFd /vfWe+aWgrEaTGCxzlPLHsby0L2/0YTJbctWeVW93rJtu0kbZHHGXMCFPlhvywe3KMZR2QMpYjrr 7d9s7mLb0MABa+b2HTx41Yy1WdMmJo0Mxgfrv/XDAucF7kjyxtR3zHyi35Zgx7kBMXzjqJHSqkXJ 7eKumkGxIyRAAiRQRQScTgBi3Mjw/9KLf5IbR48wX9wQRrC+WMmeccGG+IGwsEUMa4iIrg372wvP 6w3lwytEh4sE6ij9yC8WXBUedBFvQsyUrjvP4fVQsGQglQXukIF+Ykxwi+KihmLdBQR/4z1swz5w i8Jl+5oei9uuVVQqHnfeBcd90+hRuo6trXEjlh0XXlcksnD/1VdeekF66D1msQ/EKdxy6IN1V43i /qt70tfX29yC7dEH7jWCpGz/IVju18hkRApjvDgW7eKBcwBR02ijkQZ2YE0eokHDQkOMcLH6bDEs VbcqhXvunChPP/GoOV/gKgVTiASIJiR7htBEGxALaKNOnVry3O+fknv1vsuwGFqlIr4VtYnclFaf rOeKGFr1Xuoz2l68bKUgdyT+RlsYQ21dh4n1mI4UWCn79LKtxwQX1IG+4jaFS5etMlVUNG5rP0fa uNA+aM/iZD2X5Yr5ulPX+903+U7j6s4omkfMG8Q95hHpZ7AfIs4xz+763s1jxxh3vyUKL9QXbicB EiABZyTgdC5gCzLcgX9/8f/JQM3cj/x0CExI1eTOuCDpN7u5QAcFBggsaCNuGKLr/gYVuwqtOuyf YfGASwgXLhyPlCmhKhaqKp1HSEiQSd6MdlDytJ+IZHU0YXAXTcI79T+vmkCHX3Rd06EjR82atmy9 eBfohRwXNYgTLxVD4SqQWzVvJtf36yP9VQDCqnO+AncsoqIhomBbRDStZQU63zF4H27Q+/X+tXHx r5m2wcwqYOfvV3GQTeuWLWTq6/80awgRmHFMXalIvoyLOMYAPuDSvm0bGT1ymBG9eL+igvcfUHd0 82ZN5buZs+XAgYOmLogpCDQEbUD43jFhvDkP0Eb7Nm0kPS3D5KLDHOAHAcRn2QKWd2kSYtz9Avdg hps6Ti3LEEtIZO3h7iHeyg4JuXF3lZvGjJKmeuvBssVXk2TX0ftNY47ACG2GqbsbLlP7EqpJrFFX DRVX2oA5j6s6HQraw/rRk2oBi9BgINxfGgXip4/eRSXqIu493Eut6W3btLTlmlTWKEjSjs8hxDjm EamDEH0LkYUxgb/fr4hCN40U/ROo614RIOOrgWEomBfMZdnPE8Tqk48/rPf+7iTzf14k+w4cMJZm 7A8RCcGHvuJcaaHnEda3DtS7mlif06Lm+EQCJEACLkfATQWPzafoxEPDEJA3DhGXKSY1iJsg9xys MlhbZrm6KhsiLk62PHK4MNuQ4OIQonekcOT4yurGtnOpqWphQP7Coj21CbjNcLeB8wmcyupEGhX0 F9GpuKMDrBUI8MAdNEJUuGL8jhRYr3DrsuJ+6UFws2Pclgv9fPWAO9ove49c1BCgF/rKhCfqhHhN MndiSCh2X4dppHKtWuFqqQs9X7MVvo9UJHG6JvHk6dPmXr5YwwXXd01dk2dfcEu8zKJ5wJmPOQhx YA5w5xgk0cYawtz8XJOPEHXj/KostQ74Yu5tgk8bNG3WKJr3kp7BfY/AJts8qPVSO4cfMJXdLaXk aMf/gtsWqXvs5xsccO5UNo6yLaB/sPjlqpXdvi5Y1ZDLD+c0zk/sZyt64uvfEIQXOi/KtlXRa6w/ RFJnq200gx9rOG8r+zxh7GeSzuqtCc8ayy0EIuYf51xocEhZXV5R03yPBEiABFyCgEsIQJeYCQ6C BEiABEiABEiABKqJQMlipWpqkM2QAAmQAAmQAAmQAAlcWQIUgFeWP1snARIgARIgARIggWonQAFY 7cjZIAmQAAmQAAmQAAlcWQIUgFeWP1snARIgARIgARIggWonQAFY7cjZIAmQAAmQAAmQAAlcWQIU gFeWP1snARIgARIgARIggWonQAFY7cjZIAmQAAmQAAmQAAlcWQIUgFeWP1snARIgARIgARIggWon QAFY7cjZIAmQAAmQAAmQAAlcWQIUgFeWP1snARIgARIgARIggWonQAFY7cjZIAmQAAmQAAmQAAlc WQIUgFeWP1snARIgARIgARIggWonQAFY7cjZIAmQAAmQAAmQAAlcWQIUgFeWP1snARIgARIgARIg gWonQAFY7cjZIAmQAAmQAAmQAAlcWQIUgFeWP1snARIgARIgARIggWonQAFY7cjZIAmQAAmQAAmQ AAlcWQIUgFeWP1snARIgARIgARIggWonQAFY7cjZIAmQAAmQAAmQAAlcWQIUgFeWP1snARIgARIg ARIggWonQAFY7cjZIAmQAAmQAAmQAAlcWQIUgFeWP1snARIgARIgARIggWonQAFY7cjZIAmQAAmQ AAmQAAlcWQIUgFeWP1snARIgARIgARIggWonQAFY7cjZIAmQAAmQAAmQAAlcWQIUgFeWP1snARIg ARIgARIggWon4FntLVZxg4WFhXIg8ajEJp+UjNxMCfQOlMZhEdIoNKKKW2J1JEACJEACJEACJOAa BJxWAGbn5cjKI+vl620/yoGEw5KQliSZedni7+0ndQNqSfu6rWRy9/HSpUF715gpjsKpCJw5myxp aRmV9rlQCsXd3V1qhoeKr49PpftyIwmQAAmQAAlUJQE3taAVVmWF1VFXQvoZeWnxf2XunsVG9Hm6 e4i7m7u46X+4qOYXFOgjX8L8QuQ33W+Tx/rcI16eNaqja2yjLAHr9HJzK7vFZV/v2X9Ivp85T+IS ki44Rg8Pd+nYro3cPn60+Pv5XnB/7kACJEACJEACVUHA6SyAaTnp8sd5/5B5e5eKj6e3ediDgAh0 14tqDQ9PSc1Ok3+vfE8ycjLlj4MfFQhFluojkJ96UPJjPpXC/EzxqDNMPPVxLZRtO/ZKfOIZGTqw j3h5eZ13yNDEh47EyPZde2XwgN7i06CupGdkqhD0EwhDFhIgARIgARK4XAScSgDCqvfKkqmyYN9y 8atxYWuJhwo+N73KfrhhujQIqavWwAmVcszNzTUX4MAAf70Au4ZYhIE3NS1dXYzeUqNG9VpBC1N3 ScGJ96Qg55y4Kc7CWkPE7RJEeGpamuTl5YuPt7f4+pZ3lebl50tqapp4q9jyO48VDRzSlEMNrxqm nkpPhF+5MS8vV7y9vWTUsIEXZL5s5TrZs0+Fso5h8fI18vOSldK+TQsZN3qYhIQEVdiTArVwQygG +PuZ87vCnVz4TcwlPtcsJEACJEACl07AqQTgqqMb5Jvtc8Tb4/xWFaCAUIS1D75tuIbzCvPk/XXT pF+T66RZeBR2qbCsWL1epn/7gzx8713SvWvnCvdxtjfz8vLkf+98JDeOHCod1NVYnaUw54w2p655 nS/zd2GuvnZcWCcmnZHZ8xbK/gOHJDcnT8Wdj3Tu2E5GDR9SvGbu1Ol4+WbGbDkSHSP+vr4y8dZx Os7W5YYZn5goL782VdqquLp/8qRy26vyDUuc5OTkXlAAQrxa+zdvGiXxCYmyZsMWadQwQgb171Vh tzZs3qZicZX87tEHzit4KzzQyd9MT8+QRctWSo/uXaRendpOPhp2nwRIgASuLAGn8TMVFBbIrF0/ a6RvVvEFsyJ02fk5EhFcT94Z/4pM7narIFjE091TjmuUMNzG5yvZ2Tmye89+c8HetHWHoD1XKefU Opaj1s3qLm4FyeJew1Pc1VXvLhoQoa5gR8vZ5BSZ+t4ncloF3j133ibPPf2YjBkxVDZs2ipffTfL WMxQFwRBckqq/OHJR4xo/27WT5KVnV2umR0790pmZqYcPHRUzpw5W277FX9DrVpNoiJl+JD++qPF zVgry/bpwKEj8v2sufLRZ18J+FxrJTMrW9as26Q/Bqr/XL7WWHO8JEACrk/AaQRgfFqi7I0/KB7w JRaV3Pxcyc3Ps14asVcnoKa8NubPMrR5P31tJwT0orr1xC6zT/EBdn8ci4lV60uC/GbSBDlyNEZF QnLx1vT0dNl/8LARFnjerAIxOeVc8Xa4J/eplWqjWmaOHIs2lkdcoPEe3MoosGbt2rNPIDRRIMoO aF25aqFDSUhMMvVu37Wn1MU/Rds5qtYtCDgIAFwEUdBfWILs2zAb9J8zycmyedsO2afBCNmwQnna XOHW9mp5VvOrW36SeHi6i7uXj9r9knUtYLrDTc9buETdnBny24emSJuWzaV27ZrS87qu8tiDUwyn 3fsOmLpgFWrcKEIi6teTVi2aGmFYWFA6rglW0LVqVRs6sL+EhYbI9t37HO5Hde+IcwnFsgrat79D 1wpu3b7TuIbh7kbAU2UF/HbqMZu27JDTcQnldo0+Hqvn7HY9hw6qi73kcxQTe8KcjziHN23dLrEn TppjUd+2Hbtl7/6S/eGOPnD4iDlnUR/OO0RAo8Tp5wk/po7FHDevrX/Ddzv9AAA5e0lEQVTg7oYQ x/l76PDRYjGP7THHT0jKuVRJUpGOz9lB3Q6Xb4a6vHfs2q2fn2x93iPHokvXadXNZxIgARIgAccI OI0L+ExGisSdSzDWEQwNF7+m6s7NLciV6LMnxE2vhcj998bYF6RjvTbyt0VvyJdbZhZH/yI45GRK nGSqBdHbs7wLec2GjVK7Vk3jYly4dKVeNLfL8KEDDcXjJ07Jm+9+rK6nzuZCiouTt5e3PPLAZAkP DZVPpn0th49GS3BgoCTpxa9/n57GDYljfv/EwxLVKFLmLVwqcxcskeeeeUw6tW8r6zdukbXrN8nz f3jSXGRnzJ6rbk1fIxg9PT3lvnsmSYN6dWT/ocMya84CadG8qbmwTr5jgsxRtygurEGBAbq+L00a N4yUyXdOMG5RCNQPP50u3rrmD2sZ4SrLzskp5ubYafHr90Lgh2SfNjNlastVd3CuCgPfC+dnhDiG xa5Pj24SqGO0Lw0jGxj3KMRBh7atpW+v69QqNkfWbdgsS1eukRsGXV9uneBRFQtnzp4xAtJT10Gu 37RF+vXuLp4aKHQ5iyOBHLD2OVpuGTtS8MAPh+nfzTTC6HzHxuo5+9EXX5m1k566nhU/WMaNGW7O TQiwH36cZ87BMD1/cQ7Vr19X7r1ronEpz52v0fX6QyNAz5+Tp06bY2+5caQcVyF4JDpaTp2Mk65d Osldt9+iQlXkky++kWaNo8yPGog+nL+jbhgk61XgpZw7J/HxiXLnbTdLL51P/JD55vvZsnvvfq0/ wNTdrUsHufWm0Wbd7ZwFi83xebofBCjaHzywn/Tt2V2tv1pfaqoR82gDnysWEiABEiCBSyNwea+A l9anCo/KVddulj4sy4gaBeSvw5+WeoF15IlZL0hCepL896b/k24RHeTFha/Lxxu+Nq5fCD8U/IsI 4jxdH1i2IDhgz94DxkLkrYEGHdq2MtaJgf37mMX8uNjgAlozLEwmjr/JWNX++fpUWbV6nfTp1UMt EnvlyUfvl2ZNomSvLuhfvmqNDLq+j7E2wSKI9VxwO9YMD5Fjx2I07UdbtRTGSEu1WGWoW3LWTz/L 2FHDpGf3riaFzQcfT5Of5i+SB6fcaURK7MlTcn2/XnLHrTcZSyQsiU88fK800Is2gh9efeNt2bJt p1kb9d3MudJYXYn3Tb5DAy7czUV+yYrVGhldYjktO/7L8boQ4i8nRsEXtVuYJgXp+8U9qN0FmwOT NLW61q1Tp8J9w0JCVByoRVFPgjatmkv9enXl3/97V8aOHi4DlBPeh3gICQ4yefY2q5ivU7uWRDSo J7BYLVq6QqKjY6WpztflKDhHIbKOn4iT0NAgCQsJNv1AW3G6xs+y8nmqZfasnlem2E5T29/n+dc6 9wt0fDj/bWd1+Z1zVPB/rSIrJDBIHrz3TrOsYZmK41VrNkg3FW4H9UfF6nUb5ZH7JpsfFskpKfL6 m+9rAMpyFYkjjBA7dvy4/P7xh8059q261T/8fLo5p24fP9ZYAd/+8DO5YXB/Y3lFe9m5OfKQrp3F B+2NqR/IzJ8WyO8ee1B/VIWrQP9JFixebgT4+o2b1aJ9XF36j0uwzg/E4etvva+BL62lbesW4qHn 7CHt3+8ef0jnv5YsWfGLzJ67UAZf31fu1XP69anv6Y+jO9TqS/FXfub5DgmQAAk4TsBpBKC/t78E +waqFTDLXGQKCvPlq62z5dVRz8vbt/xd0jXVS/NajeWvavn7ZOM3eiEp7fbE9bKmf5ha/8pHwu4/ eMQIAywuR0EAyCJdZH/seKy0bNZECvRiHuDvr+93MhdTRNNGRtRX60ScXtyDjMD7WS9wp9q3loYR DfSie5e54Lds0UzdvgekXZtWalHJkv59ewrcZMkpyerCipF77rpNEMQQF5+gFqwtKuJ2mePgMvNQ 0QkRgYt9uArPHt26mJQiO3fv1eNT5Md5P2ufC80FMyEpSeuL1YtiQ0nUYIcJN48uTj/SvUtH+VEv oPYuPjPIy/xPYcYxccs+KYUQ3PkatZmXJpK6W6TeLRdsGYE77jp/5+szuHjouk4IvS+/mSkItpgy eaKsWLVOjqqwDgoKlLc++EyeUPcxBP1WdVvepNYvFMwbhPOaDZsumwBsp4Em23bukdfe/EBq1wyT Jx/5jc5hqCz/Zb1a3hboUoKSHzIKyIgZbEdxxGpodqzknwRdbgA37gP6A8KnKMH04AF95bpunTUC 2kt/3GxX4dzSiD9UExIcrBbRHgKRCAGI9a8tmjaRyMj6phUEp2zQ/kFsI3E1zrNAtd6lqjWusLCu eQ+5DK0I7Sj9ARJ2LtQIOFTQvGljdV3r8gsVirt0nS2WQ3z1PdZxFhgLIgJfovWchwCEQO/auaPU q2sL8mivnx2cvxnqfjYpdWzK1/SL/5AACZAACVw6AacRgDX9w6VBUF3jxvXQFWUQeD/uXqT5/mrI P0Y8q+5fP5Mc+sP10404KOtaK8SFNjRSfGuUTyOyYfNWSU/PNFY3WFnydF1higYWwNUGAYhSQwWZ fRoVY+XRixUusLB8wLpy5Ei0rFm7SSIi6sntt4yVluq2nffzYmPtg2WjW+dOZgE/1jUhHUpEvXpm fRQsQaG6Ns2kntELXKcO7Yy1ChdbiBzcJQJpXFAgHvxVjAYHBZmLJdQwrCNwhyL9CASj/V0lIKRQ f3WV2JQctfwlS72E+eKmUcCFWNOGC71aIN2Sf5Ez8bskyb2lzoWnWjcrNnvBdY27Yxw6elT69r6u VNfB43RcvHTs0NaI3kNHjspv1RpaMzxMBUWhRnHPVP71JVT5wH0MV/mp+HgjQCC63VRcnlHr4Jmz KTJ2ZKoRi6UaqIIXsCD/ZtKtgqjyffsPq9tT1yCu32LEX5Rag1s2b2Is2XDN4jxq3qyxuvMD5az2 6bRaxGzGvYrZONI9CORCPTcDNCraKmjHW1PgoGAdHQScfYFQttarov1S6XT0DSxPwLpDFCy/wLmq TZiCz5pP0fmJN2B1R4oa+wLumDuIQGwL1PGij3qojFR3cWsVl6ZoW/52x6LfKDhW/y/+2/YX/yUB EiABErhUAk4jAEN8AqVPVDfZELPNjBUXGdzd4/sd8yQp/ayE+AbLnD2LjDAsL/4K1fLnLQOa9TJu YXtYsJZgQXovDTAIUVedsS7pxa2nWgO36IL7YYMHmIu0/THW37gIwqW36pd1Mn7caPM21gv+41// VXfy9cYljAvX2vWbdb1SQ6lft44EBgXIUnXJ1qtby1zo0Gagf4BxGUc2sFlc9mrwxgld+2QV2zXQ diGEOxkWk9tuvlGthDZht2rNehWGuVJH1/vhQrx7335j6cLxx9USFK93pIBgvpzl2JlsWXwoQ5bG uEvzvF/k+dCPVCSoGMxT6x8CdTw0EOTcclm86kv5KvdZ6V63QG5o4i2dIvx0Tmxjs/oHS1L/Pj00 vcscnZduRkhb235Zu0EwZ3CXw2oEVWAJ3tHDB6sFNElmq/vxGV17CQGNtYFNlD3c4rkqxCAoemud CxYvk8Pqnu+sYruqC9pt27q5CrpkjSw/aH4crFm32VjOptw93qwVRZtY04Y5a6SWttiTp+XDz74x rmsvtTBj/eelllC16MG9ulOtbVFRDU01BzWAaN6ipXLPpNuN5W+FLlOAq92vSCRu1wCLxkX7QoFa Ysu+D0X6y/6t8/5d0b74zMJCDuE3acK44mOR0iYvV38oFBV8Zioq+BzgLj/4McVCAiRAAiTw6wg4 jQDEMG/rOEZmq9Xv2Jnj5k4fuKD4aEDHisPrjFUCghDvlS05Gi18fZMeckOL68tuki1bd6p1yk1v xXWjcRdaO8Ct+peXXjNRthAkWLxuf2FC/jZY3LzUqrJRIyXjVXjAkrNn70Hj7sNtvWDJgCUFQQdP //ZB4yKrq2vR5sxbJFPuvt00BVGIoJD3P5omPXt0NdYZCDqsIYRYgfsZwsUqEKbrVFD+5+0PTF6/ U6dPq3Vrt3E7Y80bXHk/zl1koibhMoM7GS49WAYvV8lWkffK6kxZd8pN/H28JEs6yPa0NtLd7xe1 AHpp23rBzlGrW264rMroLYlufjLrSJYsOJIpb43ykvZ1yrvlMQ6I6bfVldtbgwdg4YNbc+OWbRow MMa4F2toihnMA9y9ndq3MdGj0RpFCgsqGEII7dJgg4em3CVdOpW+JzTEI9zucF1CsF2O4h+gVjCt esmKNdK6ZTO567ZxxeIPqWqWqMs1NTVD7r/nNp3nAiP6MY5undubdajn6xPOQ2OtO8+cwvJ5w5AB MmPWPE19kyV+/r7yi/Jo0aypWhr9zY8bpNOZqkFKnTq218jbWBXTsfK4usxRcvW8tmcCt6ytPVuP rPZhVUeBhRPLEayCH1GYF6tgG9YIIj/n9X17yVvv75N3dA1h86ZNTPAU3Paw4qIgKh6uYatgiGgb 5y+CUmDN/lGDoIZoRDeYspAACZAACVwaAY8XtVzaodV/VJBaAWvq/X0XHlipy8oKTJJn9ALWLSR+ rkj8IQ9gLXUfvzLqOWkY2qBUp3Fhg2WkdfNmKt6alNoGl1dOTrZxe9VTkZaRmSEdVajBOoOC6MZQ td51aNtG2rVuZdb24UKGSOLbNFAE6UZQEHyBiykEDW7xZbm0IPDg9sNruL8gZiBWMtQVPUotWYgk xras7Cy1mBSqRamlWR+G9hFEgnWASMfhpdHId6g1pXmTKNNeM31Gv2AFLFCXKNxrgeoyjtJI4dDQ YLNPVf+TV+AmX+8Xic9yk0CvQinwCpP4nDrS1mu/BLd5WiTqUfUKH5Svz06Qte6TpFDHlV/oJqkF NWRUM7V2BZYXYGDWXl2pEMiHjx4zYg2u3Kc0+XEXTQaNAutV65YtzJqyQ7oP5n/MiCEyctgQkz/w 1Ok4XZ8ZaSJIIdTtC1ygSLGD6GoE+VyOgnV9Xjqv9dXKN/7GEXpOlPBHm3Vq19Q0LDt0rg7qWPxk ny45uF3Xb3bu2LaUACvbN6x5xA+SNq1anLfvjSIj1NpWX+8ycsBYjPtoFO0YTQYOqzXuUgKRjPyJ SCPkp+flHRqlC+scCpY/1NLgDaz1Q4HLON+cg7b28Lk5p6laWrVsrudwgEnZAo4471AQVYwfI+gD ClIfQfxh3SH2b6vPiFJGChhYKidp25hnFMwJgj8Q2IMCMYmocJz/+Lzg/YNHjhlLd9PGjcw+/IcE SIAESODiCbjpr/mSn+4Xf/wVOeLTTd/Jy0vflPRsXRiud5mwRJV9Z2AxQB7AyJB68vLI52RQsz72 m/l3FRLIUgPllJ8yZVdSgYTqEkt/T3fxdMuT6/wPyWMD2kiIv7d8u26nfH8sTDI8wiUrt0DScgsl Uy2HH430lW71PC7YG+RJ/OCz6RpUUVMiNZoX4gcBHc5e9h04Il/P+FHz7p01ou/xB+9WS11jZx8W +08CJEACJHCVEyhvernKO4zu3aN3+Hjzppeka0RH497M0fQwtqTQSAyda5I9Yx3g8FYD5IPxr1L8 VcOc5qnXruShVkC4B/1UyLjZrGuFflFS6OFnXIOq/8y+WPbl6M+PWjXDZYreEcRHXcxrdD0dUuO4 QmnVoomuy7tFI8Wby/DB/UxEsCuMi2MgARIgARK4ugk4pQXQQpqSpUlhozfrY4scTYqR5KwUTfUS Li1qNTEBIz0adlYLYWnXn3Usn6uOgHqaZcvpfEnVm5TUUGMeltR56n1/W2ikb6i/TQBm67qu/fE5 kpbvKRr7qUIQ2WEKpUtdjYD2Kb9us+p6x5pIgARIgARIgATKEnBqAWg/GCvQweSQU+sfCwmQAAmQ AAmQAAmQQMUEXEYAVjw8vksCJEACJEACJEACJFCWgFOuASw7CL4mARIgARIgARIgARJwnAAFoOOs uCcJkAAJkAAJkAAJuAQBCkCXmEYOggRIgARIgARIgAQcJ0AB6Dgr7kkCJEACJEACJEACLkGAAtAl ppGDIAESIAESIAESIAHHCVAAOs6Ke5IACZAACZAACZCASxCgAHSJaeQgSIAESIAESIAESMBxAhSA jrPiniRAAiRAAiRAAiTgEgQoAF1iGjkIEiABEiABEiABEnCcAAWg46y4JwmQAAmQAAmQAAm4BAEK QJeYRg6CBEiABEiABEiABBwnQAHoOCvuSQIkQAIkQAIkQAIuQYAC0CWmkYMgARIgARIgARIgAccJ UAA6zop7kgAJkAAJkAAJkIBLEKAAdIlp5CBIgARIgARIgARIwHECFICOs+KeJEACJEACJEACJOAS BCgAXWIaOQgSIAESIAESIAEScJwABaDjrLgnCZAACZAACZAACbgEAQpAl5hGDoIESIAESIAESIAE HCdAAeg4K+5JAiRAAiRAAiRAAi5BgALQJaaRgyABEiABEiABEiABxwlQADrOinuSAAmQAAmQAAmQ gEsQoAB0iWnkIEiABEiABEiABEjAcQIUgI6z4p4kQAIkQAIkQAIk4BIEKABdYho5CBIgARIgARIg ARJwnAAFoOOsuCcJkAAJkAAJkAAJuAQBCkCXmEYOggRIgARIgARIgAQcJ0AB6Dgr7kkCJEACJEAC JEACLkGAAtAlppGDIAESIAESIAESIAHHCVAAOs6Ke5IACZAACZAACZCASxCgAHSJaeQgSIAESIAE SIAESMBxAhSAjrPiniRAAiRAAiRAAiTgEgQoAF1iGjkIEiABEiABEiABEnCcAAWg46y4JwmQAAmQ AAmQAAm4BAEKQJeYRg6CBEiABEiABEiABBwnQAHoOCvuSQIkQAIkQAIkQAIuQYAC0CWmkYMgARIg ARIgARIgAccJUAA6zop7kgAJkAAJkAAJkIBLEKAAdIlp5CBIgARIgARIgARIwHECFICOs+KeJEAC JEACJEACJOASBCgAXWIaOQgSIAESIAESIAEScJwABaDjrLgnCZAACZAACZAACbgEAQpAl5hGDoIE SIAESIAESIAEHCdAAeg4K+5JAiRAAiRAAiRAAi5BgALQJaaRgyABEiABEiABEiABxwlQADrOinuS AAmQAAmQAAmQgEsQoAB0iWnkIEiABEiABEiABEjAcQIUgI6z4p4kQAIkQAIkQAIk4BIEKABdYho5 CBIgARIgARIgARJwnAAFoOOsuCcJkAAJkAAJkAAJuAQBCkCXmEYOggRIgARIgARIgAQcJ0AB6Dgr 7kkCJEACJEACJEACLkGAAtAlppGDIAESIAESIAESIAHHCVAAOs6Ke5IACZAACZAACZCASxCgAHSJ aeQgSIAESIAESIAESMBxAhSAjrPiniRAAiRAAiRAAiTgEgQoAF1iGjkIEiABEiABEiABEnCcAAWg 46y4JwmQAAmQAAmQAAm4BAEKQJeYRg6CBEiABEiABEiABBwnQAHoOCvuSQIkQAIkQAIkQAIuQYAC 0CWmkYMgARIgARIgARIgAccJUAA6zop7kgAJkAAJkAAJkIBLEKAAdIlp5CBIgARIgARIgARIwHEC no7v6jx7FhYWCh5WcXNzEzxYSIAESIAESIAESIAERFxGAObn50tubq7guaCgoJwAdHd3Fw8PD/H0 9DQPTj4JkAAJkAAJkAAJXKsE3NRSVmIqc0IK+fk5kpOdJbl5YkQfhlOZtQ/bPDzcpEYND/Hy8nPC EbPLJEACJEACJEACv4ZARma2HDmeIKfikiUtPVMsIRTg5y11aoVI00a1JcDP59c0cdUf69QCMDMj WhLjlkkN7xbi699BYec7ANxNCvLTJT1lhQQGN5bgsOv0GLqHHQDHXUiABEiABEjAqQmkZ2TLtj3R sm13tJw9ly75edAN9hqg0HgLgwJ9pWPrhtKlXZQEBvg69ZjP13mndQGnndst8afm6uSliKdXQ50/ nUBLwp9vtHbv5+bEScKp3ZKXlyFhta7XwxkPY4eHf5IACZAACZCASxGIjk2Upat3S8zJJHF3dzMP T0+PCseYci5Dlq3ZIweOnJKBvdtKs6g6Fe7nzG86perJyjgu8SdnS37uORVuNZQ/hgEF7/jDzc3D 6MUzCUvlbNJaZ55D9p0ESIAESIAESKASAoej42TG/A1y7ERCsfirZHezj4eHu5w4fUZ+WLBB9h46 WdnuTrnN6SyAWPOXFL9Cgz3Sxc0d4k9LYb6u/8vVZ10IeMHipvvmGfFnrRU8m7hCfHwbip9/5AWP 5g4kQAIkQAIkQALOQyD6RKLMXrRZUtOzxN/XR3x8POVcaqYUIGZAh1Fg5z2EMxHvoQT6+0pefoFg veBPi7aIpwrC5o3r2ja6wL9OJwCzss5JTk5yKZdtVvpuyc2O1emwm8XzTo6uAVQBWJB/TvewTXNh QY6kp8WKr1+k8SSf91BuIAESIAESIAEScBoCmVk5snjVTklVwQejT8c2kdK1Q1M5ohbBTTuPqhYo ED9fL7NNE8hJVlauySTSpV1jdfvWlf2HT8qK9XslA/X8skvq1g42wtBpAFTSUacSgAUq03NyslW2 WfrcNrL8/ERdyxevL0q/X8m4iwQk9ododNPjczWNTJ5GBjsVksqGyG0kQAIkQAIkcE0T2K4BHyfj zmr6Nw8VdoVSMzRQHwHm0bp5AyMBIACtkpWDdHIFElQU+JGRla1BIVhmViAJSedk0/YjZk2gtb8z PzuV2oFAM1lryuk89yJBd+lTAVGJ+muoACxX/UVUi1yE7733nmzZskX+8pe/SFRU1EUcfXl2nTlz pnz33Xdy//33y8CBAy9PI6y1HIHMzEyZ+eN8GTKov9SuVbPc9qp4A0mclq74RQ4dOSq33TJWQoKD SlV7LPq4/Dh/oQzo21s6tGtdatuveZGcck4WL1slsSdPmTYH9e8jDSP1y7SSEhefKEuWr5L4xCRp 3ChSBl3fR6PrAoqPOBYTK8t0LFnZ2dK3Vw9p37ZV8Tb8EZ+QKIuWrtAv4bPSSNsaMrC/Hu9v9oEr Z+u2nbJ2w2Z103hKn57dpH0VjrdUR/iCBEjAKQikq+t214FY1Q227kI/ZGbrcrGiEuhfPs2Lv6+3 tdk8ow4IRxUZ+n+h7Dl0Srq0byLBGiXs7MWpgkDy87HGD3f5gBDMtXsUlJqHwkIkgrbfXn5tYLk6 VN2bBNKaRPrXFNSxbNky+eyzzyQxMfHXVFVlxx45ckQWLVokJ06cKFcnEmdDIK5cubLcNr7x6wjk 6C/JdZu26lqTtF9XUSVH48fKgYNH5Md5C2Xb9l3l9ly2arUsWLhMTp46XW7bpb6RnJwib0x9X/Yf PCz169aRlHPn5M13P5YjR6PPW2XimTMy9f1PBCKwXp3asn7jVnn/k+lq0c8xx0Sr+Htbt2fra4jC j7/4SjYoO6skqmj877sf6WfqjNSrW1s2b9shH33+tUn+jn0gRqd9PUOCVQDjR9xHn38lm7fusA7n MwmQwDVIIDEp1VjtEPELDejtVUNqhwdfFInQIH/NB+gthSoC4UI+m5wqcYkpF1XH1bqz01gAIawK CvJ1ArzEx7+9TkaUMlX9qoo+O+uopoNJKnqdL17ekZobENYIiDl3NeemSnbGfv276GeA2vh8/Nuq WTew6D0khq6tVdnuIoK7hjhS0tPT5Yxe2HB3kXr16hUf4uXlJb6+vupO9hJYgRISEiQ0NFQCA9Fe STl79qyc04unt7e31K1rW1iKccbHx5s7mtSvX1/XLaSaB+pHvzIyMoywrFmzpvj5+ZkL4OnTp017 eC8pKUnQL+xfo4YtSOa2226THj16SPPmzU3jaOPUqVPm2BkzZsg//vEPef7556Vfv36CuvArCW2j b8nJyWK1VdJzKd7m4+MjtWvXtt9khDTqgbisVauWLri1/cpKSUkx9eE99C0uLs5w8ff3N22iT2CG 9i5UMpRralq6+GrdQYElVqQUtUyBp4+PtyQmnTGm+9CQkHLVZWRkavJP2/GBdsefU96YTx+t48zZ ZO2npwQHlbaq4bh0TRzqrwuEA7Tv9iU3L0/O6nGYGw+tB8dbwUb2+1l/Z6l7AQLRS7+Y7K136ekZ 5jg/P19JOnNWPHTuQ0LKf3FhzYq3t5e0adVSVq/fKL17djd1oX5YzHbvPSBNm0RpwBSkYklJ1rnI 1l/CIcGBhhe2oN9ghzZRcJ6gbYgqr6JzCe+v37xV82edkxefe1pCi/r06n/eUkvkamnSuBF2KVe2 bd+tljkPefLR+8y4+va6Tl569Q2JjjkhzZs1llk/LZBWLZrJlLsnmmPDwkJl5pz50rplc/3cBKjg 26mfZDd5/OF7zecAx7/y+lRN4hovdVVQzl+4VAb06y1jRw8zx0/7+gdZunKNdOrQrsh9U65LfIME SMDFCcSpyzYnN98Eb8DD165FhAZxXFw6l/p1QqVdq0hZtX6f+T7P07yB8SoAW7hAMIjTCECIElyQ 3N19xT+wpzHHYu2eKkEpSJojebkJemFxVyGRJ94+TcU/5Hr9O1ff85TcrFjJzjyo+8ISiAuhm0b8 dpcaGvmLCGKUwgJYFW1WQPPGBf5ZsmSJvPnmm7Jt2zYVAv5yyy23yFNPPSUBRS4tiIhdu3bJiy++ KFu3bpUOHTrIv/71L2nRooUZx/fffy9vv/22HDp0yAgeiLQnnnjCtPrkk09KTEyM/P3vf5d//vOf Eh0dLY899piMHj3atLF9+3bp2bOnvP7667pgNUvGjRtn6h86dKj897//NQJx5MiRpm2IKYi8//zn P/Lqq6/KoEGD5JVXXpE5c+YYYQlRhrrhHk5LS5NHHnlE8N7vf/97+fe//y2wHvbu3dv0o0EDm4tv 7ty5Zux79+41Ag99x3HgABGHdn7++WcjfiE8//CHP0iXLl3k22+/lZdfflleeOEFOXDggHz11VfS t29fI0CnTZsmH330kanvz3/+s4wYMeK8M7B95x4jDs6eTVGx4mNcgYMH9DX7T//2B6lVu5aeD3my ccs2I0KvV2Fw44gbioXAxq3b5ad5i4zA8/f3k9HDh6rLsbs5fsaseeIfYLtDzJp1m4zwuvmmUdKn RzezffW6jUZswAUKYXTT6OHSvUtHsy0uPkG+/HamHD0WI3W0D/17X6dtVpxjCgccPnJMpn83ywhV 3JmmX++eMmbEUCNwYNGCKzRA+7JMhQy+dG4aNUwG9O9t2rL/J09FZ+eO7WT/gUMq+Pabv7Ed7tDI BvVNpBt4oGCJwvxFy2Tl6nUqAHOkQf26MvGWm6RRowiZu3CJflnmypQ7bzf7Ho89IW998Jn87rEH jMgyb+o/vt4+po/6Y7i4uKlA9a0ka36b1i2M0LPEMESmu66rydcfdfhRc0SZPTjlzuL6ruvaSX6c u1BFbJIRgO3UHdyyRVPTLnby00g+D3cPw8VdP/e3jx8rzezEJ358+KgwLqN7i+vnHyRAAq5PICU1 3Vxv9cvC/PCsUyvIPF/syOvUDDbfVypDzPdpsuYIdIXimKnrKhmpWf9nXMDq3tXI3eKHsfRZnYSp V9PC2G9XIVi2GBex3T6wFqJ+Wxtl9y79ev/+/UY0LV++XLp27SpBaiH629/+ZoSVdYHDhRYCDwIA IgyC64033jAVwd36zDPPGLE0bNgwI1JwPNbpQTjC6rZz50759NNPjaiClRHi8bnnnjP9g/Vs+vTp AhGJCx32nz9/vnz44YcSFRVlLG5oe+rUqWZ/WBEhIrNVULzzzjtGDMLCN3jwYGM5hBizxo66NmzY YNoLCwszFiEItQ8++MD0He7tBx980Kxx7Natm7l4Q7C99dZbxvKIPuJvWDzbtm1rxv3www8b9zPc fXBDf/7553Ly5EljJfvhhx/k8ccfl/Xr10vDhg1N26+99poRo6Wp217Fqwtx2jczpJMKnueeeUxG 3DDIWI8sd1+2ul2xjgxWv2d++5CMHztaFi5eUexOjDl+QmbMmmusRX959im5ceQNevx8XUN3zDSQ qbcVXPHLWqmjc/aHJx+Rrl06yIzZc42V7uDho+pqXSTDhw6SF577nQxR0fnNjNkSo0IJ8/D59O/U MpoqTz12v0y4ZYwc0P2TdL2abQFx6dGcVTfqJ9O+leZNG8uzTz0qkybcIqtVcK5cs97smKc/dpbo OKTQTZ569AFd09Zdvtd+nDgdV7oifWUTO97StXMHWbV2gx4jZl7gJu3fR38s6WvrvIbbdvXajTL5 jlvlz88+qULKV76Z+aP5kmzXppXs3rNfXRw298aefQfVQhgssMbZl86d20uknj8ff/GNWdP34adf 6g+RbBnc3ybC7fe1/oaruFFkhHmZr2Ob9/MStZ4GSMOIBmplTDZC27/I8oidvNUSDJEHiyxKg3p1 JaqhLU0T1uMsWrJS3TJ+KkxrmcXdPbp11nMuRLaopfCzL7+Tg4eP6LkxUC2fTvUVZ8bKf0iABKqG QK7+cIZoQ9Gv0ksuuK5b13ZUkqtWRVcoTvXtaJsATATcajWKH7jC2a/5U/VXZm5Kb4eVEImgS9dh Q2E/yWUqKX45a9YsgQiE1QzWta+//lpgtbNcsqgDF+WJEycaUQdBAzG1ceNGs+apadOmxgr26KOP yh133CF9+vQx++/erelsVDhC4MHaOWTIEFP/hAkTjEUwREUN1uv96U9/Mq5SWODQDvaHcHzppZeM lQ1CEC5lWOog6LAdBS5kvMYzrHrjx483fYYVEu5evA/Biv5DlMFi98c//tEISgg0FIg3uG5h7cTY ISjh0oUYhdUPbcLaBzGLvmJ8mzdvNmsQ0Q+0AZcx+ggLJ96DAH333XeNqIWF9NixY8a1bhos88+G zdvMujNYw+D6g8DBY+nK1cV7tmzeTIYNuV7XitWRvmqF69e7R7GwWqNuUn8VDmEqFk7oujhYAGER hjAyRb8sYDXr16eH1K9XR4YPURGhG5LUnbx2/SZzHFzOJ06eVtdwsM5nrq59i5HTav2L1ffumDBO mkQ1khZNm6hQHGjcs4U6l2XLtp27zS/K8eNGm3ZgwRuqwSLL1I2Kgp8xDbT/GAesdMOGDDB1nTpV XgAikxWi2HurlTI29qREH4+VrTvU5epZQ1q1bKZWPds6O9TbOKqhPHTvXWaeT52O1/MyxLh+M9WS 3Eq5wWW9b79ay7XsO3BYunftWMr9i/chvCIa1DNr/mBl3KmiEdbQUK3rQiUjM8sI5Q3qRr574i16 /vlKLtb2lvmCBXRYCBHcYV/S1Vr48edfyap1G2SisoaAtUqefhZ27t4nu/bsM1/6fr42S661nc8k QALXFoEaGvmrXy2mYA1fds6lCbcs9ZYgMhh1qQLR78nze3acibDTuIAhSvAoKNB0LdnH9TlbOUO0 YVJ8xNu3pT6rC1i3e3iG6gXAdtGF5cPN3Vvdws10X0w+EkEXCm4Fl1+gZlxzgXETzxrh+gg2behO lRZYsSCUOnXqZPZr1KiRsarhBaxsqB+CrFevXmY7LICwEsJdizHArQohhTWAzZo1M9Ywy1WIY/HA Wrg2bdqY42Gtw3FNmjQpfo3tEIsoEIEQfO3btzev8QxRhiAUuNesAlEJ0bp06VLj3sVxERERxpqJ NYNwAaNtuLFh2URB2xC26Du2oU6s6YP1D6Vz586mbbiN4fLGfrD84TgU7Pfxxx/L8ePHTZ/QB4wZ 40UfrfWPEIVYuwihjH2tsZlK7P5JSEost0YQAgmWHxTUD2FoX2rXClNhsNe8hXV9CGKAlQ/7okDM wMJkio4Rx0OookAQoa9glaTHwhq7fNWaomPd1O0YZQQpLH/Yzz7aNzgo0AgcBJCVLbAMBilnrIuz Ss3wUMHaPAhGWMnqqgDFeYQCixiiW233rbSOKHnGl1PN8DAj+BYsXqZu/FTpr+vksEYS82aVEydO qZv6B2P5hXiK03WC2IovRx9d6Ny1Uwez3i6qUUNdPJ0kndrbzkHreDxv1WCT1SrAnn78QSMoU/W8 ef3N92XhkhXGhW2/r/3fsL5++Nl0c24/88TDRmhje7CujQVffMlaBS7vTF2nGWAEuu1dWFrf//hL PT/9te2H1CJosyjCBZ6i/CHMJ0+6VcVwnnz5zUz54uvvjfva+gFk1c1nEiCBa4MAEjm767UTBYEg sXoLOFgFIQwdLfgRGnPiDH6T2or+ERhQPnrY2uxMz04lAN3ddT1fboqkpSzXNV6whNgm0T+4nwSE DtGrmFoStEAs2cQeXuWrIAyRoLAx+jcudSoi8zMkJWmWXkzjzWsIyYDggRo92KX4wq8bzlsgbnCR 2bFjh0yaNMlY57BWDxY6BFPY2i853LoAQ1TgQvfFF1+YqFu4VadMmWJcu0gbg2J/rCVQrOOtZ+t9 qwUIDwSOwCKINXfoF6x0kZGR5kJv7Yf2YYWEWIObFusSIU4bN25sdrHqRx+sv61nvIdHeHi4Wdu3 adMmwZpDrG/Euj8IUAhWCDpYMvEeBB72w3FYP4ixoz6rTusZjWNMeNi/Z/Xb/jlcBaLlrrXeh+XN CpDAhz1BI0bty3EVPbB0oUCIwyp2/z2T9JXtfIAlEEEfKCVSybws1R8Iuto1w2XSbTfrfrC76ReK 1o1UJBCWiFKHSLSCORDckZGRVeE6tLDQYNmxe49mmccCZdt5nAhRqGLIuC3LdORCXGy9FRmoawRf fm2qisFQuU7dosXHKReUlavXq6UuVB65924jxLDWcPHSlcXnHY6Ba/cXFXi1aoZJeFi4VXXxM1K/ 4EcBLKwoiNqtpedRjFoez1cg0CD+IK4h0mCFtQrczCEhQXJYo4hbNLP9yDmmdeG8hqUWBWsuP/jk S4mMqCf3TLrNWEOt47ENaxVvuXGEwI0N0dykcUO1TO41axopAC1SfCaBa4tAWLB/0dq9QrPM5Fhs gkns3K5lpMMgovWYg8dO6feRGpj0KAjJWmGlAzodruwq29Fm5rjKOlVRdyAiLKuM7TIN6431gDiB W9j2sFkG7WvBdpiCrX1wwcVUWsfbLEGoHxedC5Ubb7zRRNS+//77cvvttxsRiDV3sH7hYgNxCAuW dfHFM17jfYwDFjQ8wzJ21113GdcpXmN9XWys7SJqfzyEkyWQ0Df8je14HwX9huXx2WefNW7nBx54 wAhCBIJgLR72RUE/IBIhzmbPnm2CLhCwATEIyxbqqazvqOPOO+80LlysZ4RrGsEfsArefPPNMnz4 cPOAKERgCN7D+kFYSuHOxhpA+37bc0HdKGgfD4ud7d2Sf7t16aQW09Myd8FiHWOirFJBs0ItcgM1 0AMFkbcHDh3RtWm/GLfsKl1Tt37TFuMGxnYEe+xVF+dCzSeHAINNmirktf++I8iXh2LP2byh/9j4 F5o6NmkAyfJf1uqYz8oadQn/S489roIoIqK+Wv9qydca1IGUJrb1ggv11kNpOtflP2Yd27c1gSrI E3g6Ll5dtrvUgrbcrE2srB8VcbHvM9bZQQQhMAXuVfx6hTUR+6B4aWDEObUOIkJ4z74DZgxpmRnF a+0i6tfTcYSbQJf2Wg++7MoWROsiOnjRspWa1iXB8N29b590bN9OA4vS5INPv1SX+pFShyF4BpbC G0fdYCKocRwemeoSxhfrkAH9TX3bd+0xLvUfZs2Vnj26qgC1rT/8Rdc2QuiN1DWfWBdoHY/zHlHa 6OXcBUvluLrA4bpesHi55hFsbSygpTrCFyRAAtcMgVrhQRIS6Ge+B21BYe6ycOVO2bnvuF5nbN+J 54OBtcaHY+Jl7pJten21XbvhKQlSqyLqdYXiNBZAwEaqDM3+oQVf99bjfK/xfkWl4uNg0UGmcEcK LF0IsEAU8Nq1a43LFJa/3/3ud0bYwRIIl6rlvoMohIsTUbIQWffdd58ReogghkUKkbGoC1Y8iB+4 QWFRsywXcMnCgmZFGMPKhvrRDgQBhBXcw7feeqsRlUg9g8ALrOODsES7OB7HYb0hLJCw0mEdI9yu iL49fPiwWUMIiyDqtoQwjkHf0ScUCDkkukZ/161bZyyIL774onEjo3+whGJf5B2EKxfRvIgCRoAH 6oLL2UqHg/HVqVPHWBVRN/oKCyMYWO3jfftSX3PATbptnAZu/KwibqWKaW8VFcPMWjXsh2M7dWhr BMI8jWqF8Bl5w2BBVCkKAgluH3+T/DR/sczVB44fpuv8kC4EBaLJfl0ZxFtgAKxyYgI2bhk7yuTV +2G2RgurFesmTTvSplUL42aYfMcEQRTya2++p78Qw+T6/r1MEIipuMw/EDa/ufM2+er72QJxg3Ov j7psB/SzLRtAv/x8Sta34XxHOhT7dCxWleizrwZMoOD8Gqsiq8TCpi59daP6eNksnIiWhoXvldff UgttuArnPrJSBe26jVtk3JgRhjvWAm7TNYQd2pV3/6INWOl+o5HCc+Yv0gCb5UZUgjGSLyNfH/L3 NW/SWFo2t1nzcI7C7Y4cf1M1X2BOjgp8/Q+fxAk3j5Ge13U1xyanJMsXX32vc5gvXXRN5LjRI8x4 0GZy8jkjxN96/1Oz7hJ1wlI6+Y7xKjzbyn2TJ2pAzhx59Y23zTEd1XV989iRxcejDhYSIIFri0Co WgCbRdWRdVsOSaOImtKyST1ZunqXzFywUXY0itFlKOGaGiZSwkJK0nnhfsG798eq2zdRjqr1D99X 1g9hiMLGDWtJuN5NxBWKm36RlnE2Xd3DSkuLl4ST36r7tsQF7KdpYXz9O+klxZEFnjYXcOrZeVoH EjXbBGFQ6FCpWad38UQ7QgHiCXn3IGQgmlCAE3n/sPYO70H0QKDBJQuBAyGGZ0TmwuoGCx1EIP6G cIPYwt+wJkIE4ngEbuA9iKNgdZdhP9QHwYX3EUkM0QULItqy2rYEpHU8hODkyZNNxC2OQdsIAMGa QFjspk2bZuqFlQ5ubghYtAVhijWH1hgxTitHYEV5AGExQ//wDEGJnIjWMeAF4Yq2MUbsZ9UNsYb8 gRBxaMvqvzm4zD8Zuj4sVfmXzQP42v/eNev5Jt56U1EeQI/iXHX2VaTr8emaR9BbhRZcu1aBdQlp RbDODAV9guUJbl6rP2npGSqcM0yUKkSZfUEaFYgdiEjUgYha5ArEWsKKCoIvYDXDjxu4Qq2C+mG5 s3Ic4rxCXYiUxTlhX+BqxhdU2ZyE1j7oP4QjhCJKpkbspqpLFgEweFjHww2OXICITobLfMpdt1cq oCwOyENoub0RDTxH7zzStHGUdOnU3rSHfzBGMLdZ3vGZsxWwt8Qr3sEYcd6EqwsbLnar2Nzp9ser ANTPEax/EMsoOG/OaGogWBRD1cWOeWQhARK4tgkknk2TaT+s0utFltw4uIteb33lh/kbJSk5TYVc gEyZMFBCg0uWpODewZ98t1JzjJ4xa6/1a8YUiD9/f2+5c1xfTSbtGhZApxOAeXmZcjJ6ukn+DJcu CnIDurlfxKJMDRDJL0jTaxEEo+pfdQ/XbTBRAoKam/qc6Z+DBw/KmDFjjKhCKhgIq/MViAhE237y ySdy7NgxI24gHJGLD2lpYNl09vLGWx+YdWbIC8dycQT27j8k7370uYq3RnLXxPEVCucL1YhE3HvU xd6lY3sVqiX317zQcdxOAiRAApeLwLbdx+THRVs04MxPJo/vZwJBYk+dkZq6li+yfvl1zrMWbpTt u2OKvYK4dur/MnxAB+nesenl6ma11+t0AhCEUpM1yOHEDJvxzlgJYMS8WEOmzTqA9DFBId2ldv0x alFwzAVc7bNUSYOw1sFqBtcfrGZ4vlCB1RARvyiwvsGyaFm3LnTs1b4dAgTuY8uCd7X392rqHxJP w9UK619Fruarqa/sCwmQAAk4SgACbvXG/er+3S1No+pKv+tairkPsJr3QoL8jDfBvq7ZCzfJtt3R RgDC8gf116tbCxnUu61eY0s8E/bHOOPfTikAccu2pNM/y9kz61T+6eQY4XaxAlDnVO/+4RvQXOpF 3KrBAyVrAJxxItlnEiABEiABEiCBiglAyK3ZdEBWrNf8ubq8BkEhTRvWlnHDumukcGlRBwG4dVe0 CkNbKrDr1Oo3sHcbs7yk4tqd892KFyZd5WOBpS687nDx9AqXpISlmtbFll+vyCR4gd6XCMWg0K66 7m+4ir8S//8FDuZmEiABEiABEiABJyMAy11ftfw1igiXVRsP6J2aTmlgGpLQlx8IxCKWEDeOrC19 urWUpo1K55Ytf4RzvuOUFkB71FkZJ9SdeUIXgGsSWWi7CiazeP+i7e56D1Ef33AJDGqsCr9G8Wb+ QQIkQAIkQAIk4NoEkAIm5mSiWTIVpdHBZcvxU0ma+iVXGjWopcujnG9pWNnxnO+10wtAa2C4N5+V Y862YBNqz6b4EC2IB9bHYa0bIjJh2mUhARIgARIgARIggWuRgMsIQGvykLYDaSQgAvG3Jfwg/vDA axYSIAESIAESIAESuJYJuJwAvJYnk2MnARIgARIgARIgAUcIXDhniCO1cB8SIAESIAESIAESIAGn IUAB6DRTxY6SAAmQAAmQAAmQQNUQoACsGo6shQRIgARIgARIgASchgAFoNNMFTtKAiRAAiRAAiRA AlVDgAKwajiyFhIgARIgARIgARJwGgIUgE4zVewoCZAACZAACZAACVQNAQrAquHIWkiABEiABEiA BEjAaQhQADrNVLGjJEACJEACJEACJFA1BCgAq4YjayEBEiABEiABEiABpyFAAeg0U8WOkgAJkAAJ kAAJkEDVEKAArBqOrIUESIAESIAESIAEnIYABaDTTBU7SgIkQAIkQAIkQAJVQ4ACsGo4shYSIAES IAESIAEScBoCFIBOM1XsKAmQAAmQAAmQAAlUDQEKwKrhyFpIgARIgARIgARIwGkIUAA6zVSxoyRA AiRAAiRAAiRQNQQoAKuGI2shARIgARIgARIgAachQAHoNFPFjpIACZAACZAACZBA1RCgAKwajqyF BEiABEiABEiABJyGAAWg00wVO0oCJEACJEACJEACVUOAArBqOLIWEiABEiABEiABEnAaAhSATjNV 7CgJkAAJkAAJkAAJVA0BCsCq4chaSIAESIAESIAESMBpCHjmpiQ6TWfZURIgARIggWuLgJubSEZO vqRk5Yr+yUIC1UqgsFAkKMBNggJF8PfVVvK1Q+ekfMcK3Qol8IyX1MhSO5/1wSnMkTSPDMnwcJM6 eZ7ieXLhJ1fbeNgfEiABEiABEjAEani4y4ZjZ+SbLTHipX+zkEB1EsjOERl7g7uMH+UhWVnV2fKF 24KuS1Hx94NkS34ZEZhfo1BGvtdQGm8PkjxPCEQPcSs4KAvDVsvGMC95IbaOeOZnX2UjuvCYuQcJ kAAJkMA1QsBDRV9OZoakpqaJtycF4DUy7VfNMCEAs7M81PznLoUFV023TEcgAAtU+GWpAMzTZ8vQ h435aq7M08+NpHqJ1EDHPbT/aZLjnSIZ3vpeqq94urnzAwVYLCRAAiRAAlcfAVyj8PDw8NAHr1dX 3wy5do/0tBN3d/1H9NyzV1hXybDdVPjpp0MfpQVgIV6j3/jMoPum/7qfh6d+jjz1PRzDQgIkQAIk QAIkQAIkcE0RoAC8pqabgyUBEiABEiABEiABYxMkBhIgARIgARIgARIggWuJAC2A19Jsc6wkQAIk QAIkQAIkoAQoAHkakAAJkAAJkAAJkMA1RoAC8BqbcA6XBEiABEiABEiABCgAeQ6QAAmQAAmQAAmQ wDVGgALwGptwDpcESIAESIAESIAEKAB5DpAACZAACZAACZDANUaAAvAam3AOlwRIgARIgARIgAT+ P5GFmsqFbJdjAAAAAElFTkSuQmCC --Apple-Mail=_B2A5A624-E9D6-4204-9376-C3C6DD979582-- --Apple-Mail=_E8416D76-63E7-4065-BA11-A39F397F46CE--