text stringlengths 1 1.04M | language stringclasses 25 values |
|---|---|
India and the U. S. will discuss “salient regional” issues when U. S. Secretary of State Mike Pompeo and U. S. Defence Secretary travel to India next week for the “2+2 Ministerial dialogue”, the government said on Thursday, in an indication that ties with China will be on the agenda for talks on October 27.
“The Third India-U. S. 2+2 Ministerial Dialogue will entail a comprehensive discussion on cross-cutting bilateral issues of mutual interest. In addition, both sides will also exchange views on salient regional and global issues,” said Ministry of External Affairs spokesperson Anurag Srivastava during a weekly briefing with journalists.
However, the MEA did not respond to a specific question about whether LAC tensions, and the ongoing talks with China would feature during the 2+2 meeting, nor would Mr. Srivastava comment on which agreements, including the geo-spatial agreement BECA would be signed at the meetings, adding that he would rather not “prejudge the outcomes” of the visit.
“India and U. S. have a comprehensive global strategic partnership which includes political, security and defence, economic, commercial, technology and people-to-people contacts. We have regular dialogue for various levels to discuss ongoing bilateral cooperation and to exchange views on developments in the region,” the MEA statement said.
The MEA denied that the timing of the visit of senior U. S. officials exactly one week before U. S. elections on November 3, was unusual, and said it followed the pattern of “2+2” meetings in previous years.
“As regards the timing of the meeting, you know that diplomatic calendars have their logic, and you will notice that the earlier two meetings have been held towards the end of the year,” Mr. Srivastava said in response to a question from The Hindu .
Mr. Pompeo and Mr. Esper will arrive on October 26 and meet with their counterparts External Affairs Minister S. Jaishankar and Defence Minister Rajnath Singh the next day. They are also scheduled to meet Prime Minister Narendra Modi and National Security Advisor Ajit Doval.
Mr. Pompeo will also travel to Maldives, Sri Lanka and Indonesia during the visit, where his focus will remain on the U. S. ’s Indo-Pacific strategy, which he has frequently repeated, will involve “countering” the challenge from China.
The 2+2 Foreign and Defence Minister’s mechanism, which was announced by U. S. President Trump and PM Modi during their first meeting in 2017, was held in Delhi in September 2018 and in Washington DC in December 2019. It replaced the “India-U. S. Strategic and Commercial dialogue”, involving the Foreign and Commerce Ministers, announced by then U. S. President Barack Obama and PM Modi in 2015. | english |
Posted On:
President Shri Pranab Mukherjee will grace the occasion of Public Sector Day and present the prestigious SCOPE Awards to Central Public Sector Enterprises (CPSEs) on Tuesday next i.e. 11th April 2017 at Vigyan Bhawan, New Delhi.
Shri Anant G. Geete, Union Minister of Heavy Industries & Public Enterprises, Shri Babul Supriyo, Minister of State for Heavy Industries & Public Enterprises shall also grace the occasion in the august presence of Smt. Seema Bahuguna, IAS, Secretary, DPE, Shri Ved Prakash , Chairman, SCOPE & CMD, MMTC, Shri R.K. Bahuguna, Vice Chairman, SCOPE & CMD, RailTel Corporation, Dr. U.D. Choubey, Director General, SCOPE and host of representatives from public sector enterprises, senior government officials, media etc. The Award function is being organized jointly by SCOPE and Department of Public Enterprises (DPE).
SCOPE Meritorious Awards recognizes significant achievements of CPSEs in various fields leading to business excellence and competitive advantage. The function will be attended by Chief Executives/Directors of CPSEs besides Ministers, Secretaries and other senior government officials.
| english |
//! Implementation of command traits for unwrapping.
use failure::Fallible;
use crate::{
io,
types::Size,
};
use iota_streams_core::{
sponge::{
prp::PRP,
spongos::*,
},
tbits::word::SpongosTbitWord,
};
//#[derive(Debug)]
pub struct Context<TW, F, IS> {
pub spongos: Spongos<TW, F>,
pub stream: IS,
}
impl<TW, F, IS> Context<TW, F, IS>
where
TW: SpongosTbitWord,
F: PRP<TW> + Default,
{
pub fn new(stream: IS) -> Self {
Self {
spongos: Spongos::<TW, F>::init(),
stream: stream,
}
}
}
impl<TW, F, IS: io::IStream<TW>> Context<TW, F, IS> {
pub fn drop(&mut self, n: Size) -> Fallible<&mut Self> {
self.stream.try_advance(n.0)?;
Ok(self)
//<IS as io::IStream<TW>>::try_advance(&mut self.stream, n)
}
}
impl<TW, F, IS> Clone for Context<TW, F, IS>
where
TW: Clone,
F: Clone,
IS: Clone,
{
fn clone(&self) -> Self {
Self {
spongos: self.spongos.clone(),
stream: self.stream.clone(),
}
}
}
mod unwrap;
mod absorb;
mod absorb_external;
mod commit;
mod dump;
mod fork;
mod guard;
mod join;
mod mask;
mod repeated;
mod skip;
mod squeeze;
mod squeeze_external;
mod mssig;
mod ntrukem;
pub use absorb::*;
pub use absorb_external::*;
pub use commit::*;
pub use dump::*;
pub use fork::*;
pub use guard::*;
pub use join::*;
pub use mask::*;
pub use repeated::*;
pub use skip::*;
pub use squeeze::*;
pub use squeeze_external::*;
pub use mssig::*;
pub use ntrukem::*;
| rust |
If the PersistentHandler sub-key does exist, do not change the default value. This could have very undesired consequences.
Close Regedit and reboot the computer. This file type will now be searched for "A word or phrase in the file" type searches.
An alternative is to download a small VB Script utility here. Save the file you just downloaded to your hard disk. Locate the file you just saved and double click it. Enter the file extension you wish to modify. If the PersistentHandler sub-key exists, no changes will be made. If it does not exist for the file type in question, the Registry will be updated accordingly. Reboot when you're finished with the file types you wish to change.
4)How Do I Do a "Repair Installation"?
Configure your computer to start from the CD-ROM drive. For more information about how to do this, please refer to your computer's documentation or contact your computer manufacturer.
Insert your Windows XP compact disc (CD) into your CD-ROM drive or DVD-ROM drive, and then restart your computer.
(You can also boot with a Windows 98/Me Startup disk with CD support and run WINNT.EXE in the I386 folder on the CD)
When the "Press any key to boot from CD" message is displayed on your screen, press a key to start your computer from the Windows XP CD.
When you see the following message displayed on the Welcome to Setup screen, press ENTER:
To setup Windows XP now, press ENTER.
At this point an option to press R to enter the Recovery Console is displayed. Do not select this option.
On the Windows XP Licensing Agreement screen, press F8 to agree to the license agreement.
Make sure that your current installation of Windows XP is selected in the box, and then press the R key to repair Windows XP.
5) How do I delete an "undeletable" file?
Open a Command Prompt window and leave it open. Close all open programs. Click Start, Run and enter TASKMGR.EXE Go to the Processes tab and End Process on Explorer.exe. Leave Task Manager open. Go back to the Command Prompt window and change to the directory the AVI (or other undeletable file) is located in. At the command prompt type DEL <filename> where <filename> is the file you wish to delete. Go back to Task Manager, click File, New Task and enter EXPLORER.EXE to restart the GUI shell. Close Task Manager.
Sorry, no script file fix for this one. If you're CD's don't autoplay and the Autoplay tab is missing from the Properties (right click the CD drive and select Properties), then click Start, Run and enter SERVICES.MSC Scroll down to Shell Hardware Detection and make sure this service is Started and set to Automatic.
To remove TweakUI v1.33, or earlier versions, manually, perform the following steps:
1. Do a Find for TWEAKUI.CPL, TWEAKUI.HLP and TWEAKUI.GID. When each of these files has been located, delete it. Note: TWEAKUI.GID may not exist on your system.
2. Click Start, Run and enter REGEDIT.
3. This step is only necessary if you wish to remove the Shortcut Icon customization. Go to the Registry branches indicated and modify or delete the values as instructed.
In the right pane, locate the value named "29". Right click on this value and select Delete.
In the right pane, locate the value called "Shell Icon Size". Double click on this entry and set the value to 32.
In the right pane, locate the value called "Tweak UI". Right click on this value and select Delete.
5. Close Regedit.
2) In Regedit, highlight the HKEY_USERS key and go to File, Load Hive.
3) Use the File Open dialog to go to the Documents and Settings\<username> folder, where <username> is the account you wish to modify.
4) Highlight the NTUSER.DAT file in this folder (usually a hidden file) and select Open.
5) You'll be prompted to enter a "Key name". You can use whatever you wish, but I use the User's logon name.
6) You can now expand the Hive you just loaded and make any needed changes.
7) When finished, highlight this Hive again and go to File, Unload Hive.
NOTE: You MUST unload the Hive prior to logging on to the users account. Otherwise XP may have trouble loading the user's profile.
Well there are lot more tweaks tips and tricks and is taken care by a MICROSOFT'S MVP.
| english |
What's the story?
Iain Hume, the Indian Super League's record goalscorer has always been vocal about the quality of the league in India and every issue pertaining to the league. He recently rejoined the Kerala Blasters from ATK and is all set to enthrall the Kochi fans in Season 4. At the recently organized ISL media day in Mumbai, Hume revealed that he is ready to relocate to India with his family if the league becomes an eight/ten-month affair.
Speaking to the press, Hume revealed, "When you play in a five-month league, it is not the best idea to bring your family with you and have them stay at a hotel. I will give it a serious thought (moving to India) if the duration of the tournament is eight to ten months. "
Increasing the length of the ISL has been on the organizers' mind for a long time and they are expected to implement a longer version in the coming years.
Along with his record of the highest number of goals in the ISL with 23, Hume also holds the record of the most appearances in the Indian Super League with 46 caps.
At the event, Hume talked about a lot of aspects such as his connection with the Kerala Blasters which made him re-sign with the club after two seasons with ATK.
He also claimed that the fans in Kerala are the best that he has ever seen and termed the carnival-like atmosphere in the Kochi stadium as unforgettable.
What next?
The former Leicester City striker is also keen on linking with ISL's star acquisition of the season, Dimitar Berbatov and hopes that the strong foreign contingent of the Blasters along with a talented Indian unit can give the Kerala based side their first ever ISL title.
Hume's thoughts on bringing his family down to India suggests how far the ISL has come from its inception. From being a two-month tournament featuring some yesteryear stars, the competition has now become a five-month hard fought league which will soon become a year long event.
When that happens, the foreign stars will take the ISL more seriously which will, in turn, help the quality and branding of Indian football. | english |
Have Any U.S. Presidents Decided Not to Run For a Second Term?
What State Is Washington, D.C. In?
Why Was Frederick Douglass’s Marriage to Helen Pitts Controversial?
Why Do We Drop a Ball on New Year’s Eve?
What’s the Difference Between a Psychopath and a Sociopath? And How Do Both Differ from Narcissists?
Article revised.
Article added to new online database.
View Changes:
| english |
package org.jtrim2.executor;
import org.jtrim2.cancel.Cancellation;
import org.jtrim2.cancel.CancellationToken;
import org.junit.Test;
import static org.mockito.Matchers.*;
import static org.mockito.Mockito.*;
public class DebugTaskExecutorTest extends AbstractDebugTaskExecutorTest {
public DebugTaskExecutorTest() {
super(() -> new DebugTaskExecutor(SyncTaskExecutor.getDefaultInstance()));
}
@Test
public void testUsingRightExecutor() throws Exception {
ManualTaskExecutor wrapped = new ManualTaskExecutor(true);
DebugTaskExecutor executor = new DebugTaskExecutor(TaskExecutors.upgradeToStoppable(wrapped));
CancelableTask task = mock(CancelableTask.class);
executor.execute(Cancellation.UNCANCELABLE_TOKEN, task);
verifyZeroInteractions(task);
wrapped.executeCurrentlySubmitted();
verify(task).execute(any(CancellationToken.class));
}
}
| java |
Air India just launched their new domestic airport upgrades scheme ‘Get Up Front’, in which they are offering a chance for everyone to fly business class and first class on Air India domestic routes.
By India Today Web Desk: If flying business or first class and enjoying the luxuries has been on your mind, then worry not because under Air India’s new scheme you finally can.
Air India just launched their new domestic airport upgrades scheme Get Up Front’, in which they are offering a chance for everyone to fly Business class on Air India domestic routes.
According to the airline, the airport upgrade facility can be bought at any of Air India's booking offices, Airport offices, call centres and online also.
The Get Up Front' offer has two payment slabs Rs 4500 for distances up to 750 km and Rs 6000 for higher distances to avail the upgrade from Economy class to Business and Business class to First class.
According to the airline an additional 12 per cent GST will also be applicable on the scheme.
The offer is applicable for all passengers, be it an adult or a child and is also open to all valid ticket holders, i. e. revenue and non-revenue tickets like award tickets, companion free tickets, concessional tickets and discounted tickets.
The airline also mentioned that the upgrade is only applicable on economy to business and business to first class tickets and not on economy to first class tickets.
The scheme is applicable on domestic sector only on the flights departing with business and first-class configuration. | english |
The Galaxy Watch 3 sees Samsung take an already fantastic smartwatch and improve it further.
Updated This is the best Apple Watch you can buy – but it still doesn’t overcome all the issues with the Watch line.
Updated It’s Fitbit’s third attempt at the Versa smartwatch – has the fitness brand finally perfected its affordable smartwatch?
The by-the-numbers Poco smartwatch neither impresses nor disappoints.
Updated The Samsung Galaxy Watch is still a robust and well-loved smartwatch with a handsome design and tactile rotating bezel, plus battery life lasting up to five days.
The Withings ScanWatch Horizon is a smartwatch for those who like their looks muscular and classic, and their tech on the down low.
Updated The Samsung Galaxy Watch Active 2 is a minor upgrade on its predecessor and while it's slightly better the price has increased as a result.
Mobvoi’s second generation health watch is better, but still lacking in areas.
Get the hottest deals available in your inbox plus news, reviews, opinion, analysis, deals and more from the TechRadar team.
| english |
We normally believe that one's character decides products' quality, the details decides products' high-quality ,while using the REALISTIC,EFFICIENT AND INNOVATIVE staff spirit for Action Figure Glass Display Cabinet, Foldable Jewelry Display Case, Acrylic Ring Display Case, We have extensive goods supply and the price is our advantage. Welcome to inquire about our products.
Hot-selling Awards Display Cabinet - Retail display cabinets for sale with 5 adjustbale shelves | OYE – OYE Detail:
This is a type of retail display case, it does not occupy your floor area, you can hang it on any wall of your store. Put it on a wall at the right height for customers to view, pick and choose your products without having to duck their heads.Six LED lights at the top and four led lights at the side make lighting easy and give your products a charming shine. Sliding door locks are also designed to protect your valuable items. Multiple protection of your goods, you can rest assured to use.
|Brand Name:
|Model Number:
people also ask:
Product detail pictures:
Related Product Guide:
Sticking for the principle of "Super Quality, Satisfactory service" ,We have been striving for being a superb small business partner of you for Hot-selling Awards Display Cabinet - Retail display cabinets for sale with 5 adjustbale shelves | OYE – OYE , The product will supply to all over the world, such as: Sweden, Myanmar, Sheffield, Our professional engineering group will always be ready to serve you for consultation and feedback. We're able to also offer you with absolutely free samples to meet your requirements. Finest efforts will likely be produced to offer you the ideal service and goods. For anyone who is thinking about our company and merchandise, be sure to contact us by sending us emails or contact us quickly. As a way to know our merchandise and firm. lot more, you can come to our factory to find out it. We'll always welcome guests from all over the world to our business to build company relations with us. Be sure to feel free to get in touch with us for business and we believe we've been intending to share the top trading practical experience with all our merchants.
| english |
A sales role is one of the most rewarding jobs for someone who enjoys flexibility, remote work, and exciting opportunities for personal and professional growth. However, starting your first job in sales can be daunting. Whether you’re working for a startup or a well-established company, there are certain steps you can take to set yourself up for success.
Once you understand what your goals are, you can use this 10-step guide to help you succeed in your first sales role.
The first step to succeeding in sales is to have a deep understanding of the products and services you’re selling. You need to know the features, benefits, and potential use cases inside and out. This will help you articulate the value proposition to potential customers and answer any questions they might have.
Knowing your target audience is just as important as understanding the products and services. Who are their ideal customers? What are their pain points and needs? How can the products and services help them? By understanding the target audience, you can tailor your sales pitch and messaging to resonate with them.
Once you have a good understanding of the products and services and the target audience, it’s time to develop your sales pitch, which should be concise, clear, and compelling. It should highlight the key benefits of your products and services and explain why they’re better than the competition.
Sales is not just about talking, it’s also about listening. Listening to potential clients can help you understand their needs and concerns, and tailor your pitch accordingly, so make sure to ask open-ended questions and actively listen to the answers.
According to Salazar, there is likely a script for any answer the potential client could give you. “Make sure to know the scripts inside and out,” she advises.
Building rapport with potential customers is key to closing sales, as people are more likely to buy from someone they like and trust. Building rapport can be as simple as finding common ground, showing genuine interest in their business or personal life, and being friendly and approachable.
Sometimes, the best advice is to just be yourself in order to be more comfortable talking to people, know when to push, and know when to back off. The best way to be good at something is to continuously practice.
Tracking your progress can help you identify what’s working and what’s not. Keep track of your sales activities, such as calls, emails, and meetings, as well as your results, such as closed deals and revenue generated. This will help you identify areas for improvement and adjust your approach accordingly.
Feedback is essential for growth and improvement. Seek feedback from your manager, colleagues, and customers. Ask for constructive criticism and take it to heart, using that feedback to identify areas for improvement and set goals for yourself.
Sales can be a challenging and competitive field, and it’s important to stay motivated by setting goals for yourself, celebrating small wins, and staying positive. Remember that rejection is a normal part of the sales process — don’t let it get you down.
Lastly, embrace technology to help you succeed in your first sales role. There are many tools and platforms available to help you automate and streamline your sales activities, such as customer relationship management (CRM) software, email marketing platforms, and social media. A novel idea is enlisting the help of a virtual assistant or cyberbacker to help you manage your sales activities more efficiently.
As you gain experience and hone your skills, you’ll become a more effective salesperson, closing more deals and generating more revenue. Remember: sales is a challenging and competitive field, but with dedication, hard work, and a growth mindset, you can succeed and achieve your goals. | english |
<reponame>SketchIO/infinite-tree
.infinite-tree-scroll {
overflow: auto;
max-height: 400px; /* Change the height to suit your needs. */
}
.infinite-tree-table {
width: 100%;
}
.infinite-tree-content {
outline: 0;
position: relative;
}
.infinite-tree-content .infinite-tree-selected.infinite-tree-item,
.infinite-tree-content .infinite-tree-selected.infinite-tree-item:hover {
background: #deecfd;
border: 1px solid #06c;
}
.infinite-tree-content .infinite-tree-item {
border: 1px solid transparent;
cursor: default;
}
.infinite-tree-content .infinite-tree-item:hover {
background: #f2fdff;
}
.infinite-tree-content .infinite-tree-item:disabled,
.infinite-tree-content .infinite-tree-item[disabled] {
cursor: not-allowed;
opacity: 0.5;
-ms-filter: "progid:DXImageTransform.Microsoft.Alpha(Opacity=50)";
filter: alpha(opacity=50);
}
.infinite-tree-content .infinite-tree-node {
position: relative;
}
.infinite-tree-content .infinite-tree-toggler {
color: #666;
-webkit-user-select: none;
-moz-user-select: none;
-ms-user-select: none;
user-select: none;
}
.infinite-tree-content .infinite-tree-toggler:hover {
color: #333;
text-decoration: none;
}
.infinite-tree-content .infinite-tree-title {
cursor: pointer;
-webkit-user-select: none;
-moz-user-select: none;
-ms-user-select: none;
user-select: none;
}
.infinite-tree-no-data {
text-align: center;
}
| css |
Is Toshiba stepping up its tablet game with a 10-inch Tegra 4 slate?
Wait, doesn't HP have something similar going on?
While Apple and Google are thinking small, other competitors are still focused on the big picture, at least in regards to tablet display size.
Why only yesterday, a new 10-inch HP tablet leaked, supposedly complete with Android 4.2.2 Jelly Bean and a 1.8GHz Tegra 4 processor.
Clearly not one to want to lose out on the speculative spotlight, a new, and similarly spec'd, 10-inch Toshiba tablet was also loosed to the wild on today.
However, this Toshiba AT10LE-A tablet was more than just benchmarked, as images of the device and what appears to be an external keyboard were also uncovered by TechBlog.
Like the rumored HP tablet, Toshiba's AT10LE-A will also feature a 1.8GHz Tegra 4 processor, though it will only run Android 4.2.1.
The unnamed tablet is also believed to include a memory card slot, mini HDMI out, and stereo speakers to go along with the keyboard accessory.
Outside of that, there isn't much concrete information about this new tablet, and one has to wonder just when Toshiba was planning on sharing its new device with the rest of the world.
Toshiba is no stranger to tablets, even of the Android variety, and a new, more powerful slate could definitely help the company make more of a splash.
That said, with HP also supposedly bringing such a similar device to market, alongside existing Android tablets like the Nexus 10 and Galaxy Tab 2 10.1, there might not be a whole lot of room at the table.
Get the hottest deals available in your inbox plus news, reviews, opinion, analysis, deals and more from the TechRadar team.
| english |
Indian benchmark indices witnessed a gap up opening on Thursday after closing on a weak note in the previous trading session. The Indian market had a robust handover from US equities which closed on a positive note over hopes that US Federal Reserve’s aggressive policy tightening will help cool sticky inflation pressure on the economy.
At the start, S&P BSE Sensex rallied over 494 points to 58,282 levels in the opening session while the broader CNX Nifty50 index surged over 122 points to 17,345 levels.
“The 17175-17430 region continues to be in focus, with yesterday’s trades focussing on the lower half of this range. Pull back above 17330 will lend a positive bias, and though directional trades may not take off until the 17175-430 band is taken out, our bias within this construct is tilted towards upsides, and poised for a break of 17700-780 region first,” said Geojit Financial Services in a pre-market note.
Infosys, Bajaj Finance, HDFC Bank, Reliance and Larsen & Toubro are powering the rally while losses in ITC, HUL, Maruti Suzuki, Sun Pharma and Asian Paints are capping Nifty gains currently.
Broader markets are not as strong as flagship indices, as seen in Nifty Smallcap 100 and Nifty Midcap 100 indices which are up 0. 22-0. 25% higher as compared to 0. 5% up move in Nifty50 and Sensex gauges.
Among sectors, Nifty Realty, Oil & Gas and Metal indices are the top gainers at this hour while Nifty FMCG, Media, and Pharma lagged.
In business news headlines, the Union Cabinet on Wednesday said it cleared the production-linked incentive scheme (PLI) for semiconductors with the ambition of making the country an electronics hub as the shortage of microchips hurts industrial production.
Cipla entered agreement to acquire up to 33% partnership interest in Clean Max Auriga Power LLP to enhance the share of renewable power source in its operation and to comply with regulatory requirements for being a ‘captive user’ under electricity laws.
Asian markets are trading higher barring Hang Seng as investors reacted to Fed's indication that current policy stance might be coming to an end. | english |
package com.blinkfox.zealot.config.entity;
import com.blinkfox.zealot.consts.ZealotConst;
import com.blinkfox.zealot.core.IConditHandler;
/**
* 标签对应的动态sql生成的处理类.
* @author blinkfox on 2016-11-01.
*/
public class TagHandler {
/** 生成sql的前缀,如:and, or 等. */
private String prefix;
/** 生成动态sql的处理实现反射类型，如:EqualHandler. */
private Class<? extends IConditHandler> handlerCls;
/** 生成sql的后缀,如:>, <, = 等. */
private String suffix;
/**
* 仅标签对应处理类的构造方法.
* @param handlerCls 动态处理类的反射类型
*/
public TagHandler(Class<? extends IConditHandler> handlerCls) {
this.prefix = ZealotConst.ONE_SPACE;
this.handlerCls = handlerCls;
}
/**
* 含前缀、标签处理器的构造方法.
* @param prefix sql前缀
* @param handlerCls 动态处理类的反射类型
*/
public TagHandler(String prefix, Class<? extends IConditHandler> handlerCls) {
this.prefix = prefix;
this.handlerCls = handlerCls;
}
/**
* 含标签处理器、后缀的构造方法.
* @param handlerCls 动态处理类的反射类型
* @param suffix sql前后缀
*/
public TagHandler(Class<? extends IConditHandler> handlerCls, String suffix) {
this.prefix = ZealotConst.ONE_SPACE;
this.handlerCls = handlerCls;
this.suffix = suffix;
}
/**
* 全构造方法.
* @param prefix sql前缀
* @param suffix sql后缀
* @param handlerCls 动态处理类的反射类型
*/
public TagHandler(String prefix, Class<? extends IConditHandler> handlerCls, String suffix) {
this.prefix = prefix;
this.handlerCls = handlerCls;
this.suffix = suffix;
}
/* ------------ getter 和 setter 方法 ------------- */
public String getPrefix() {
return prefix;
}
public Class<? extends IConditHandler> getHandlerCls() {
return handlerCls;
}
public String getSuffix() {
return suffix;
}
} | java |
<gh_stars>1-10
---
title: "IWebAppDiagnosticsObjectInitialization::Initialize | Microsoft Docs"
ms.custom: ""
ms.date: "01/18/2017"
ms.prod: "windows-script-interfaces"
ms.reviewer: ""
ms.suite: ""
ms.tgt_pltfrm: ""
ms.topic: "reference"
helpviewer_keywords:
- "IWebAppDiagnosticsObjectInitialization::Initialize"
ms.assetid: 7ccfac28-9f65-4e1c-a0fb-a8a6c7f75b63
caps.latest.revision: 7
author: "mikejo5000"
ms.author: "mikejo"
manager: "ghogen"
---
# IWebAppDiagnosticsObjectInitialization::Initialize
Initializes objects created with [IWebAppDiagnosticsSetup::CreateObjectWithSiteAtWebApp](../../winscript/reference/iwebappdiagnosticssetup-createobjectwithsiteatwebapp.md).
> [!IMPORTANT]
> [IWebAppDiagnosticsObjectInitialization Interface](../../winscript/reference/iwebappdiagnosticsobjectinitialization-interface.md) is found in activdbg100.h.
## Syntax
```cpp
HRESULT Initialize( [in, annotation("__in")] HANDLE_PTR hPassedHandle, [in, annotation("__in")] IUnknown* pDebugApplication );
```
#### Parameters
`hPassedHandle`
The handle that was passed to the [IWebAppDiagnosticsSetup::CreateObjectWithSiteAtWebApp](../../winscript/reference/iwebappdiagnosticssetup-createobjectwithsiteatwebapp.md) method in the `hPassToObject` parameter.
`pDebugApplication`
The PDM application with which the object was created. | markdown |
logLevel = 0
def log(msg):
if logLevel > 0:
print(msg)
else:
pass
def setLogLevel(value):
global logLevel
logLevel = value | python |
India batting legend Sunil Gavaskar was not happy with the way opener Mayank Agarwal got out in the first innings of the recent third Test match against South Africa being played at Newlands, Cape Town. Agarwal was out for 15 runs, falling to Kagiso Rabada.
Virat Kohli, who returned to the playing XI after missing the second Test at Jo’burg due to an upper back spasm, won the toss and chose to bat first. Hanuma Vihari made way for Kohli, while Mohammed Siraj, who injured his hamstring in the previous Test, was replaced by Umesh Yadav, who was preferred over Ishant Sharma. South Africa went in unchanged for this series-deciding Test match.
Mayank Agarwal was dropped on 0, scored 15 before he edged Kagiso Rabada to second slip, where Aiden Markram completed an easy catch. Sunil Gavaskar questioned Agarwal’s shot selection, saying that although the batter looks amazing while middling the ball, the India opener appears a tad vulnerable against the moving ball.
“Agarwal is a very good player when the ball is hitting the middle of the bat, for the simple reason that he’s got that lovely little push but when the ball moves a little bit, that bat speed gets him into trouble. We have seen that edge when he was on 0, that would have been a terrific catch. Look at that... see how hard he has pushed at the ball,” Gavaskar said on air.
The legendary opener Gavaskar, who was the first batter to score 10,000 runs in Test cricket, also added that the habit of poking at the balls outside off-stump has formed in modern era batters due to the T20 format. But the former India captain stressed the importance of making adjustments in the mind upon transitioning from one format to another.
“To a great extent yes, but then it has got to be the mental adjustment more than anything else. Don’t you make mental adjustments from Test to 50-overs to T20 cricket? That is what needs to be done,” Gavaskar pointed out.
| english |
{"name":"<NAME>","id":"NRW-124254","address":"Deckertstr. 1, 33617 Bielefeld","school_type":"Grundschule","fax":"0521 557999315","phone":"0521 55799930","website":" http://www.martinschule-buntgemischt.de","email":" <EMAIL> ","state":"NRW","programs":{"programs":[]},"full_time_school":false,"lon":8.519893,"lat":52.010602}
| json |
3 Anga nan kutok, ɗo a Yesu ahay à man ata ataya ta slan atan à Urəsalima. Tinen apan ti zla tə daliyugo sə Finikiya tə Samariya nà, ta jan anà ɗo a Yesu ahay nə ɗo su kon azar aya ɗukwen tə ɗəfak nga pə Yesu. Ləbara ata a varan atan ataslay mivel.
12 Pə dəɓa anahan a wa, man su ɗo fok a njahay tətemtemme sə sləne way ana atə Barnabas tə Pol sa jan atan ata awan. Tə sləne way ana Mbərom sa ga masuwayan aya tu ɗo maslan ataya à wulen su ɗo su kon azar aya ata awan.
13 Ta ndav anan sə təker 'am ata cəna, Yakuba a slabak ite, a wa: «Mərak uno ahay, pəken sləmay lele. 14 Həna Simon Piyer a jak uko nə Mbərom kə̀ dazlak anan sa gan nga anà ɗo sə daliyugo ahay ta sə walay ɗo ahay à wulen a tinen wa sə təra anan ɗo anahan aya awan. 15 Ɗo maja'am a Mbərom ahay tə vinde kawa winen ata awan, ta wa:
16 “Mbərom Fetek a wa:
Pə dəɓa wa ni may ahay, ni han anan doh ana bahay Dawuda ma mbazl ata awan.
Kwa kə̀ tərak rəgay coy dəp nà, ni ndakay anan asa.
17 Matanan, ɗo azar aya fok, ɗo su kon azar aya nen sə walay atan sə təra ɗo uno ahay ataya fok, ti pəlay nen.* Ca pə Amos 9.11-12.
22 Pə dəɓa anahan a wa, ɗo maslan ahay tə məceɗ sə egliz ahay tu ɗo a Yesu ahay fok, ta ɓan 'am sə walay ɗo ahay à wulen a tinen wa sa slan atan à Antakiya pə kərtek a tatə Pol tə Barnabas. Tə gəɓa Yudas inde, tə ngaman Barsabas re, tinen tə Silas, tinen ɗo tə mazlaɓ aya à wulen su ɗo ana Yesu ahay inde. 23 Ta slan atan tə ɗerewel à alay inde. Ɗerewel ata a ja nà:
«Manay ɗo maslan ahay, tə məceɗ sə egliz ahay, mərak a kwanay ahay, ma jak ikwen ayak 'am anà kwanay mə təra ɗo a Yesu aya à wulen su ɗo su kon azar aya aday mə njahay aya à Antakiya tə Siriya tə Silikiya ataya awan.
24 «Mə slənek ɗo a manay ahay inde ta zlak ayak àga kwanay, tə vawak ikwen nga ta 'am a tinen ahay bayak awan. Əna, manay sə slənak atan ayak bay. 25 Anga nan, mə halay nga, mə walay ɗo ahay, mi slənak atan ayak àga kwanay kutok. Ti zlak ayak tatə Barnabas tə Pol, car a manay ahay. 26 Tinen nà, tə zukwa lize anan sifa a tinen anga sləmay ana Bahay a mənuko Yesu Almasihu. 27 Anga nan, mə slənak ikwen ayak atə Yudas tə Silas, ti jak ikwen nə way a manay sə vindek ikwen ayak həna ataya kərtek awan.
28 «Kà zlak anan à nga anà Apasay Cəncan a aday anà manay a re, mə̂ zəgahak ikwen anan apan awan maza bay, si 'am a həna anaya awan: 29 Kə̂ rəcen sluweɗ sə way mə waslan a anà pəra ahay bay, kə̂ sləɓen mez bay, kə̂ rəcen sluweɗ sə way mə ngərew a angərew bay, kə̂ jənen uho bay re. Kə mbəsiken way ataya nə, ki gen way lele aya awan.
30 Ta slan anan ɗo ataya, ta zla à Antakiya kutok. Tə dəzle à man ata cəna, tə halan nga anà ɗo a Yesu ahay, tə varan atan anan ɗerewel a sə manan kutok. 31 Tə jinge anan ɗerewel ata cəna, a varan atan ataslay mivel bayan a, anga kə̀ varak atan məgala bayak a re. 32 Atə Yudas tə Silas ta jan atan 'am sa man atan zek ahay bayak a, tə varak atan məgala lele, bina tinen ɗo maja'am a Mbərom ahay.
35 Aya əna, atə Pol tə Barnabas ite, tə njahay à Antakiya. Tinen pə kərtek a tu ɗo a Yesu azar aya tə ɗakan atan anan aday tə wazan atan anan 'am a Bahay a mənuko Yesu.
36 A njahay pə dəɓa wa mənjœk nà, Pol a jan anà Barnabas, a wa: «Muko saa cay pə mərak a mənuko sə ɗakan atan anan ahay 'am ana Bahay a mənuko Yesu ataya awan, tə njahay həna kəkəmaw.» 37 'Am ata a zlan à nga anà Barnabas, aday a nan tə̂ japay tə Yuhana sə ngaman Markus ata awan, 38 Əna Pol ite nà, a nan sə japay tə Markus sabay, anga winen kə̀ mbəsakak atan à Pamfiliya, kə̀ ndəvak anan mer su way a tinen sa ga ata bay. 39 'Am ata kà zlak atan pi zek sabay, tə gəzla nga. Barnabas a gəɓa Markus, ta zla à Kiprus tə kwalalan. 40 Pol ite a gəɓa Silas, ta zla way a tinen. Mərak ahay ta ga amboh anga tinen aday Bahay Mbərom â pərahan azar sa gan atan sumor anahan. 41 Pol a zla tə daliyugo sə Siriya tə Silikiya, a varan məgala anà ɗo a Yesu ahay à man ataya awan.
| english |
//! Implementation of a DataFusion `TableProvider` in terms of `PartitionChunk`s
use std::sync::Arc;
use arrow::{datatypes::SchemaRef as ArrowSchemaRef, error::ArrowError};
use datafusion::{
datasource::{
datasource::{Statistics, TableProviderFilterPushDown},
TableProvider,
},
error::{DataFusionError, Result as DataFusionResult},
logical_plan::Expr,
physical_plan::ExecutionPlan,
};
use internal_types::schema::{merge::SchemaMerger, Schema};
use observability_deps::tracing::debug;
use crate::{
duplicate::group_potential_duplicates,
predicate::{Predicate, PredicateBuilder},
util::project_schema,
PartitionChunk,
};
use snafu::{ResultExt, Snafu};
mod adapter;
mod physical;
use self::physical::IOxReadFilterNode;
#[derive(Debug, Snafu)]
pub enum Error {
#[snafu(display("Chunk schema not compatible for table '{}': {}", table_name, source))]
ChunkSchemaNotCompatible {
table_name: String,
source: internal_types::schema::merge::Error,
},
#[snafu(display(
"Internal error: no chunk pruner provided to builder for {}",
table_name,
))]
InternalNoChunkPruner { table_name: String },
#[snafu(display("Internal error: No rows found in table '{}'", table_name))]
InternalNoRowsInTable { table_name: String },
#[snafu(display("Internal error: Cannot verify the push-down predicate '{}'", source,))]
InternalPushdownPredicate {
source: datafusion::error::DataFusionError,
},
#[snafu(display("Internal error: Can not group chunks '{}'", source,))]
InternalChunkGrouping { source: crate::duplicate::Error },
}
pub type Result<T, E = Error> = std::result::Result<T, E>;
impl From<Error> for ArrowError {
// Wrap an error into an arrow error
fn from(e: Error) -> Self {
Self::ExternalError(Box::new(e))
}
}
impl From<Error> for DataFusionError {
// Wrap an error into a datafusion error
fn from(e: Error) -> Self {
Self::ArrowError(e.into())
}
}
/// Something that can prune chunks based on their metadata
pub trait ChunkPruner<C: PartitionChunk>: Sync + Send + std::fmt::Debug {
/// prune `chunks`, if possible, based on predicate.
fn prune_chunks(&self, chunks: Vec<Arc<C>>, predicate: &Predicate) -> Vec<Arc<C>>;
}
/// Builds a `ChunkTableProvider` from a series of `PartitionChunk`s
/// and ensures the schema across the chunks is compatible and
/// consistent.
#[derive(Debug)]
pub struct ProviderBuilder<C: PartitionChunk + 'static> {
table_name: Arc<str>,
schema_merger: SchemaMerger,
chunk_pruner: Option<Arc<dyn ChunkPruner<C>>>,
chunks: Vec<Arc<C>>,
/// If the builder has been consumed
finished: bool,
}
impl<C: PartitionChunk> ProviderBuilder<C> {
pub fn new(table_name: impl AsRef<str>) -> Self {
Self {
table_name: Arc::from(table_name.as_ref()),
schema_merger: SchemaMerger::new(),
chunk_pruner: None,
chunks: Vec::new(),
finished: false,
}
}
/// Add a new chunk to this provider
pub fn add_chunk(&mut self, chunk: Arc<C>, chunk_table_schema: Schema) -> Result<&mut Self> {
self.schema_merger
.merge(&chunk_table_schema)
.context(ChunkSchemaNotCompatible {
table_name: self.table_name.as_ref(),
})?;
self.chunks.push(chunk);
Ok(self)
}
/// Specify a `ChunkPruner` for the provider that will apply
/// additional chunk level pruning based on pushed down predicates
pub fn add_pruner(&mut self, chunk_pruner: Arc<dyn ChunkPruner<C>>) -> &mut Self {
assert!(
self.chunk_pruner.is_none(),
"Chunk pruner already specified"
);
self.chunk_pruner = Some(chunk_pruner);
self
}
/// Specify a `ChunkPruner` for the provider that does no
/// additional pruning based on pushed down predicates.
///
/// Some planners, such as InfluxRPC which apply all predicates
/// when they get the initial list of chunks, do not need an
/// additional pass.
pub fn add_no_op_pruner(&mut self) -> &mut Self {
let chunk_pruner = Arc::new(NoOpPruner {});
self.add_pruner(chunk_pruner)
}
/// Create the Provider
pub fn build(&mut self) -> Result<ChunkTableProvider<C>> {
assert!(!self.finished, "build called multiple times");
self.finished = true;
let iox_schema = self.schema_merger.build();
// if the table was reported to exist, it should not be empty
if self.chunks.is_empty() {
return InternalNoRowsInTable {
table_name: self.table_name.as_ref(),
}
.fail();
}
let chunk_pruner = match self.chunk_pruner.take() {
Some(chunk_pruner) => chunk_pruner,
None => {
return InternalNoChunkPruner {
table_name: self.table_name.as_ref(),
}
.fail()
}
};
Ok(ChunkTableProvider {
iox_schema,
chunk_pruner,
table_name: Arc::clone(&self.table_name),
chunks: std::mem::take(&mut self.chunks),
})
}
}
/// Implementation of a DataFusion TableProvider in terms of PartitionChunks
///
/// This allows DataFusion to see data from Chunks as a single table, as well as
/// push predicates and selections down to chunks
#[derive(Debug)]
pub struct ChunkTableProvider<C: PartitionChunk + 'static> {
table_name: Arc<str>,
/// The IOx schema (wrapper around Arrow Schemaref) for this table
iox_schema: Schema,
/// Something that can prune chunks
chunk_pruner: Arc<dyn ChunkPruner<C>>,
// The chunks
chunks: Vec<Arc<C>>,
}
impl<C: PartitionChunk + 'static> ChunkTableProvider<C> {
/// Return the IOx schema view for the data provided by this provider
pub fn iox_schema(&self) -> Schema {
self.iox_schema.clone()
}
/// Return the Arrow schema view for the data provided by this provider
pub fn arrow_schema(&self) -> ArrowSchemaRef {
self.iox_schema.as_arrow()
}
}
impl<C: PartitionChunk + 'static> TableProvider for ChunkTableProvider<C> {
fn as_any(&self) -> &dyn std::any::Any {
self
}
/// Schema with all available columns across all chunks
fn schema(&self) -> ArrowSchemaRef {
self.arrow_schema()
}
fn scan(
&self,
projection: &Option<Vec<usize>>,
_batch_size: usize,
filters: &[Expr],
_limit: Option<usize>,
) -> std::result::Result<Arc<dyn ExecutionPlan>, DataFusionError> {
debug!(?filters, "Input Filters to Scan");
// Note that `filters` don't actually need to be evaluated in
// the scan for the plans to be correct, they are an extra
// optimization for providers which can offer them
let predicate = PredicateBuilder::default()
.add_pushdown_exprs(filters)
.build();
// Now we have a second attempt to prune out chunks based on
// metadata using the pushed down predicate (e.g. in SQL).
let chunks: Vec<Arc<C>> = self.chunks.to_vec();
let num_initial_chunks = chunks.len();
let chunks = self.chunk_pruner.prune_chunks(chunks, &predicate);
debug!(%predicate, num_initial_chunks, num_final_chunks=chunks.len(), "pruned with pushed down predicates");
// Figure out the schema of the requested output
let scan_schema = project_schema(self.arrow_schema(), projection);
let mut deduplicate = Deduplicater::new();
let plan = deduplicate.build_scan_plan(
Arc::clone(&self.table_name),
scan_schema,
chunks,
predicate,
)?;
Ok(plan)
}
fn statistics(&self) -> Statistics {
// TODO translate IOx stats to DataFusion statistics
Statistics::default()
}
fn supports_filter_pushdown(
&self,
_filter: &Expr,
) -> DataFusionResult<TableProviderFilterPushDown> {
Ok(TableProviderFilterPushDown::Inexact)
}
}
#[derive(Clone, Debug, Default)]
/// A deduplicater that deduplicate the duplicated data during scan execution
pub(crate) struct Deduplicater<C: PartitionChunk + 'static> {
// a vector of a vector of overlapped chunks
pub overlapped_chunks_set: Vec<Vec<Arc<C>>>,
// a vector of non-overlapped chunks each have duplicates in itself
pub in_chunk_duplicates_chunks: Vec<Arc<C>>,
// a vector of non-overlapped and non-duplicates chunks
pub no_duplicates_chunks: Vec<Arc<C>>,
}
impl<C: PartitionChunk + 'static> Deduplicater<C> {
fn new() -> Self {
Self {
overlapped_chunks_set: vec![],
in_chunk_duplicates_chunks: vec![],
no_duplicates_chunks: vec![],
}
}
/// The IOx scan process needs to deduplicate data if there are duplicates. Hence it will look
/// like this. In this example, there are 4 chunks.
/// . Chunks 1 and 2 overlap and need to get deduplicated. This includes these main steps:
/// i. Read/scan/steam the chunk: IOxReadFilterNode.
/// ii. Sort each chunk if they are not sorted yet: SortExec.
/// iii. Merge the sorted chunks into one stream: SortPreservingMergeExc.
/// iv. Deduplicate the sorted stream: DeduplicateExec
/// . Chunk 3 does not overlap with others but has duplicates in it self, hence it only needs to get
/// sorted if needed, then deduplicated.
/// . Chunk 4 neither overlaps with other chunks nor has duplicates in itself, hence it does not
/// need any extra besides chunk reading.
/// The final UnionExec on top is to union the streams below. If there is only one stream, UnionExec
/// will not be added into the plan.
/// ```text
/// ┌─────────────────┐
/// │ UnionExec │
/// │ │
/// └─────────────────┘
/// ▲
/// │
/// ┌──────────────────────┴───────────┬─────────────────────┐
/// │ │ │
/// │ │ │
/// ┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐
/// │ DeduplicateExec │ │ DeduplicateExec │ │IOxReadFilterNode│
/// └─────────────────┘ └─────────────────┘ │ (Chunk 4) │
/// ▲ ▲ └─────────────────┘
/// │ │
/// ┌───────────────────────┐ │
/// │SortPreservingMergeExec│ │
/// └───────────────────────┘ │
/// ▲ |
/// │ |
/// ┌───────────┴───────────┐ │
/// │ │ │
/// ┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐
/// │ SortExec │ │ SortExec │ │ SortExec │
/// │ (optional) │ │ (optional) │ │ (optional) │
/// └─────────────────┘ └─────────────────┘ └─────────────────┘
/// ▲ ▲ ▲
/// │ │ │
/// │ │ │
/// ┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐
/// │IOxReadFilterNode│ │IOxReadFilterNode│ │IOxReadFilterNode│
/// │ (Chunk 1) │ │ (Chunk 2) │ │ (Chunk 3) │
/// └─────────────────┘ └─────────────────┘ └─────────────────┘
///```
fn build_scan_plan(
&mut self,
table_name: Arc<str>,
schema: ArrowSchemaRef,
chunks: Vec<Arc<C>>,
predicate: Predicate,
) -> Result<Arc<dyn ExecutionPlan>> {
// find overlapped chunks and put them into the right group
self.split_overlapped_chunks(chunks.to_vec())?;
// TEMP until the rest of this module's code is complete:
// merge all plans into the same
self.no_duplicates_chunks
.append(&mut self.in_chunk_duplicates_chunks);
for mut group in &mut self.overlapped_chunks_set {
self.no_duplicates_chunks.append(&mut group);
}
self.overlapped_chunks_set.clear();
// Building plans
let mut plans = vec![];
if self.no_duplicates() {
// Neither overlaps nor duplicates, no deduplicating needed
let plan = Self::build_plans_for_non_duplicates_chunk(
Arc::clone(&table_name),
Arc::clone(&schema),
chunks,
predicate,
);
plans.push(plan);
} else {
// Go over overlapped set, build deduplicate plan for each vector of overlapped chunks
for overlapped_chunks in self.overlapped_chunks_set.to_vec() {
plans.push(Self::build_deduplicate_plan_for_overlapped_chunks(
Arc::clone(&table_name),
Arc::clone(&schema),
overlapped_chunks.to_owned(),
predicate.clone(),
));
}
// Go over each in_chunk_duplicates_chunks, build deduplicate plan for each
for chunk_with_duplicates in self.in_chunk_duplicates_chunks.to_vec() {
plans.push(Self::build_deduplicate_plan_for_chunk_with_duplicates(
Arc::clone(&table_name),
Arc::clone(&schema),
chunk_with_duplicates.to_owned(),
predicate.clone(),
));
}
// Go over non_duplicates_chunks, build a plan for it
for no_duplicates_chunk in self.no_duplicates_chunks.to_vec() {
plans.push(Self::build_plan_for_non_duplicates_chunk(
Arc::clone(&table_name),
Arc::clone(&schema),
no_duplicates_chunk.to_owned(),
predicate.clone(),
));
}
}
let final_plan = plans.remove(0);
// TODO
// There are still plan, add UnionExec
if !plans.is_empty() {
// final_plan = union_plan
panic!("Unexpected error: There should be only one output for scan plan, but there were: {:#?}", plans);
}
Ok(final_plan)
}
/// discover overlaps and split them into three groups:
/// 1. vector of vector of overlapped chunks
/// 2. vector of non-overlapped chunks, each have duplicates in itself
/// 3. vectors of non-overlapped chunks without duplicates
fn split_overlapped_chunks(&mut self, chunks: Vec<Arc<C>>) -> Result<()> {
// Find all groups based on statstics
let groups = group_potential_duplicates(chunks).context(InternalChunkGrouping)?;
for mut group in groups {
if group.len() == 1 {
if group[0].may_contain_pk_duplicates() {
self.in_chunk_duplicates_chunks.append(&mut group);
} else {
self.no_duplicates_chunks.append(&mut group);
}
} else {
self.overlapped_chunks_set.push(group)
}
}
Ok(())
}
/// Return true if all chunks are neither overlap nor has duplicates in itself
fn no_duplicates(&self) -> bool {
self.overlapped_chunks_set.is_empty() && self.in_chunk_duplicates_chunks.is_empty()
}
/// Return deduplicate plan for the given overlapped chunks
/// The plan will look like this
/// ```text
/// ┌─────────────────┐
/// │ DeduplicateExec │
/// └─────────────────┘
/// ▲
/// │
/// ┌───────────────────────┐
/// │SortPreservingMergeExec│
/// └───────────────────────┘
/// ▲
/// │
/// ┌───────────┴───────────┐
/// │ │
/// ┌─────────────────┐ ┌─────────────────┐
/// │ SortExec │ ... │ SortExec │
/// │ (optional) │ │ (optional) │
/// └─────────────────┘ └─────────────────┘
/// ▲ ▲
/// │ ... │
/// │ │
/// ┌─────────────────┐ ┌─────────────────┐
/// │IOxReadFilterNode│ │IOxReadFilterNode│
/// │ (Chunk 1) │ ... │ (Chunk n) │
/// └─────────────────┘ └─────────────────┘
///```
fn build_deduplicate_plan_for_overlapped_chunks(
table_name: Arc<str>,
schema: ArrowSchemaRef,
chunks: Vec<Arc<C>>, // These chunks are identified overlapped
predicate: Predicate,
) -> Arc<dyn ExecutionPlan> {
// TODO
// Currently return just like there are no overlaps, no duplicates
Arc::new(IOxReadFilterNode::new(
Arc::clone(&table_name),
schema,
chunks,
predicate,
))
}
/// Return deduplicate plan for a given chunk with duplicates
/// The plan will look like this
/// ```text
/// ┌─────────────────┐
/// │ DeduplicateExec │
/// └─────────────────┘
/// ▲
/// │
/// ┌─────────────────┐
/// │ SortExec │
/// │ (optional) │
/// └─────────────────┘
/// ▲
/// │
/// │
/// ┌─────────────────┐
/// │IOxReadFilterNode│
/// │ (Chunk) │
/// └─────────────────┘
///```
fn build_deduplicate_plan_for_chunk_with_duplicates(
table_name: Arc<str>,
schema: ArrowSchemaRef,
chunk: Arc<C>, // This chunk is identified having duplicates
predicate: Predicate,
) -> Arc<dyn ExecutionPlan> {
// // TODO
// // Currently return just like there are no overlaps, no duplicates
Arc::new(IOxReadFilterNode::new(
Arc::clone(&table_name),
schema,
vec![chunk],
predicate,
))
}
/// Return the simplest IOx scan plan of a given chunk which is IOxReadFilterNode
/// ```text
/// ┌─────────────────┐
/// │IOxReadFilterNode│
/// │ (Chunk) │
/// └─────────────────┘
///```
fn build_plan_for_non_duplicates_chunk(
table_name: Arc<str>,
schema: ArrowSchemaRef,
chunk: Arc<C>, // This chunk is identified having no duplicates
predicate: Predicate,
) -> Arc<dyn ExecutionPlan> {
Arc::new(IOxReadFilterNode::new(
Arc::clone(&table_name),
schema,
vec![chunk],
predicate,
))
}
/// Return the simplest IOx scan plan for many chunks which is IOxReadFilterNode
/// ```text
/// ┌─────────────────┐
/// │IOxReadFilterNode│
/// │ (Chunk) │
/// └─────────────────┘
///```
fn build_plans_for_non_duplicates_chunk(
table_name: Arc<str>,
schema: ArrowSchemaRef,
chunks: Vec<Arc<C>>, // This chunk is identified having no duplicates
predicate: Predicate,
) -> Arc<dyn ExecutionPlan> {
Arc::new(IOxReadFilterNode::new(
Arc::clone(&table_name),
schema,
chunks,
predicate,
))
}
}
#[derive(Debug)]
/// A pruner that does not do pruning (suitable if no additional pruning is possible)
struct NoOpPruner {}
impl<C: PartitionChunk> ChunkPruner<C> for NoOpPruner {
fn prune_chunks(&self, chunks: Vec<Arc<C>>, _predicate: &Predicate) -> Vec<Arc<C>> {
chunks
}
}
#[cfg(test)]
mod test {
use crate::test::TestChunk;
use super::*;
#[test]
fn chunk_grouping() {
// This test just ensures that all the plumbing is connected
// for chunk grouping. The logic of the grouping is tested
// in the duplicate module
// c1: no overlaps
let c1 = Arc::new(TestChunk::new(1).with_tag_column_with_stats("t", "tag1", "a", "b"));
// c2: over lap with c3
let c2 = Arc::new(TestChunk::new(2).with_tag_column_with_stats("t", "tag1", "c", "d"));
// c3: overlap with c2
let c3 = Arc::new(TestChunk::new(3).with_tag_column_with_stats("t", "tag1", "c", "d"));
// c4: self overlap
let c4 = Arc::new(
TestChunk::new(4)
.with_tag_column_with_stats("t", "tag1", "e", "f")
.with_may_contain_pk_duplicates(true),
);
let mut deduplicator = Deduplicater::new();
deduplicator
.split_overlapped_chunks(vec![c1, c2, c3, c4])
.expect("split chunks");
assert_eq!(
chunk_group_ids(&deduplicator.overlapped_chunks_set),
vec!["Group 0: 2, 3"]
);
assert_eq!(chunk_ids(&deduplicator.in_chunk_duplicates_chunks), "4");
assert_eq!(chunk_ids(&deduplicator.no_duplicates_chunks), "1");
}
fn chunk_ids(group: &[Arc<TestChunk>]) -> String {
let ids = group.iter().map(|c| c.id().to_string()).collect::<Vec<_>>();
ids.join(", ")
}
fn chunk_group_ids(groups: &[Vec<Arc<TestChunk>>]) -> Vec<String> {
groups
.iter()
.enumerate()
.map(|(idx, group)| format!("Group {}: {}", idx, chunk_ids(group)))
.collect()
}
}
| rust |
Feedback (2)
Good Quality Spc Tiles - Vinyl Plank Flooring SPC Core Wood Grain Finish Flooring – TopJoy Detail:
Made of high-quality material, environmental protection and safety. TopJoy wood grain floor, TPW002-A, brings a more comfortable feeling to your home through its natural texture and color.
It is strong abrasion and scratch resistance, which can easily accept various scratches such as table legs, chairs, pet paws, etc. without hurting the floor.
What’s more, the installation of SPC flooring is also very fast. It do not need subfloor prep, no acclimation time, and can be installed over existing hard surfaces such as concrete, old ceramic tile, wood or non-cushioned vinyl flooring.
If you are looking for a durable, long-lasting & waterproof floor that enhances the look of any space and won’t break the bank, send the inquiry NOW.
Manufacturing Direction ≤0.02% (82oC @ 6hrs)
Across Manufacture Direction ≤0.03% (82oC @ 6hrs)
Curling (mm)
Peel Strength (N/25mm)
Manufacturing Direction 62 (Average)
Across Manufacture Direction 63 (Average)
Locking Strength(kN/m)
Packing Information(4.0mm)
Product detail pictures:
Related Product Guide:
We'll make every effort and hard work being outstanding and excellent, and speed up our techniques for standing during the rank of global top-grade and high-tech enterprises for Good Quality Spc Tiles - Vinyl Plank Flooring SPC Core Wood Grain Finish Flooring – TopJoy , The product will supply to all over the world, such as: Barcelona , Poland , The Swiss , Our tenet is "integrity first, quality best". We have confidence in providing you with excellent service and ideal products. We sincerely hope we can establish win-win business cooperation with you in the future!
The product manager is a very hot and professional person, we have a pleasant conversation, and finally we reached a consensus agreement.
| english |
The 2004 stabbing death of Johnia Berry will be the subject of the first episode of Oxygen's upcoming eight-part true-crime series Sleeping With Death, which premieres on November 6, 2022.
Episode 1, titled Left for Dead, will revisit Berry's lengthy murder investigation, which was followed by the arrest of a potential suspect. The episode airs on Sunday at 7 pm ET.
The synopsis reads:
"Johnia Berry is stabbed more than 22 times in her Knoxville, Tenn., bedroom while her male roommate escapes. Investigators can't fathom how this young man evaded Johnia's fate. Will a years-long investigation reveal the true killer?"
A Knoxville man named Taylor Olson, 22, was arrested nearly three years later in 2007. DNA evidence found at the crime scene identified him as a potential suspect and the alleged burglar in the case, who broke into Berry's home and attacked her and the flatmate.
Sadly, in 2008, Olson, who was in custody at the time, was found hanged in his cell before he could be brought to justice. Moreover, authorities revealed that he left letters in the jail cell before killing himself, which raised further questions regarding the homicide.
In December 2004, 21-year-old Johnia Berry was fatally stabbed in her home in West Knox County by one or more burglars who entered the apartment she shared with flatmate Jason Aymami and attacked them both. A DNA test reportedly connected the blood discovered at the crime scene to Olson, who had a background in theft, leading to an arrest in 2007.
Olson was scheduled for trial on July 28 of the following year. However, he was discovered hanging at the Knox County Jail in March 2008. He allegedly ripped a bed sheet and used it to hang himself inside his jail cell, leaving behind letters.
Authorities claimed that Olson, who was being confined to a cell alone and being examined every hour, showed no signs of suicidal tendencies.
Reports state that in those letters, Taylor Olson claimed that he did not kill Johnia Berry and identified the person he alleges stabbed Berry to death in 2004. Gregory P. Isaacs, Olson's lawyer, disclosed some of the basic content of the letters, which included sentiments of love for his family and sorrow for the Berrys.
Olson's defense lawyer Isaacs stated that he named a person who may be liable for the murder of Johnia Berry in those letters and expressed his wish for that particular individual to take accountability for the crime committed. However, Berry's father Michael refused to believe his side of the story.
Michael Berry reportedly said:
"His blood was there, his DNA was there. He killed my daughter in a cowardly way, and he took the coward's way out. If he wasn't really the one who murdered her, why did he kill himself?"
He further added:
"I feel for any parent who has to bury their child, but my daughter didn't get to write any last letters to her family before she was killed."
According to reports, another potential suspect, Noah Thomas Cox, 24, was named in the case but was never charged. Moreover, Olson's defense refused to clarify whether Cox is the same individual mentioned in the letters.
But it was revealed that defense attorney Gregory Isaacs was putting together a case based on claims that Cox killed Johnia Berry that night after the two broke into her house.
The theory reportedly emerged from a drug addict's statement who was also a friend of Olson and Cox, named Augustine Joseph Leon. He informed authorities that his friend confessed to stabbing Berry during the botched burglary.
He stated that the two men broke into the victim's apartment thinking the place was unoccupied. That's when Berry armed herself with a knife and wounded Olson. Cox then ran to help his friend and killed the victim in the process while his accomplice, who was bleeding, fled the scene.
Oxygen's Sleeping With Death airs on Sunday, November 6, at 7 pm ET.
| english |
<filename>corpus/norm-class/eclipse.jdt.ui/2533.java
test foo red red red pink | java |
<gh_stars>100-1000
#include "DaeTestReportWriterPerformance.h"
#include "DaeTestPerformanceBudgetResultData.h"
#include "DaeTestPerformanceBudgetViolation.h"
#include <HAL/PlatformFilemanager.h>
#include <Interfaces/IPluginManager.h>
#include <Kismet/KismetTextLibrary.h>
FName FDaeTestReportWriterPerformance::GetReportType() const
{
return TEXT("FDaeTestReportWriterPerformance");
}
void FDaeTestReportWriterPerformance::WriteReport(const TArray<FDaeTestSuiteResult>& TestSuites,
const FString& ReportPath) const
{
// Ensure report path exists.
IPlatformFile& PlatformFile = FPlatformFileManager::Get().GetPlatformFile();
if (!PlatformFile.DirectoryExists(*ReportPath))
{
UE_LOG(LogDaeTest, Display, TEXT("Creating directory: %s"), *ReportPath);
PlatformFile.CreateDirectoryTree(*ReportPath);
}
// Prepare report.
FString ContentDir =
IPluginManager::Get().FindPlugin(TEXT("DaedalicTestAutomationPlugin"))->GetContentDir();
FString ReportTemplatePath = FPaths::Combine(ContentDir, TEXT("ReportTemplates"),
TEXT("PerformanceReport.template.html"));
FString MapTemplatePath = FPaths::Combine(ContentDir, TEXT("ReportTemplates"),
TEXT("PerformanceReportMap.template.html"));
FString BudgetViolationTemplatePath =
FPaths::Combine(ContentDir, TEXT("ReportTemplates"),
TEXT("PerformanceReportBudgetViolation.template.html"));
// Write performance budget violations.
FString MapString;
for (const FDaeTestSuiteResult& TestSuiteResult : TestSuites)
{
for (const FDaeTestResult& TestResult : TestSuiteResult.TestResults)
{
if (TestResult.Data != nullptr
&& TestResult.Data->GetDataType() == TEXT("FDaeTestPerformanceBudgetResultData"))
{
TSharedPtr<FDaeTestPerformanceBudgetResultData> Data =
StaticCastSharedPtr<FDaeTestPerformanceBudgetResultData>(TestResult.Data);
FString BudgetViolationsString;
for (const FDaeTestPerformanceBudgetViolation& BudgetViolation :
Data->BudgetViolations)
{
// Copy screenshot.
FString OldScreenshotPath = BudgetViolation.ScreenshotPath;
FString ScreenshotFilename = FPaths::GetCleanFilename(OldScreenshotPath);
FString NewScreenshotPath = FPaths::Combine(ReportPath, ScreenshotFilename);
UE_LOG(LogDaeTest, Display, TEXT("Copying %s to %s."), *OldScreenshotPath,
*NewScreenshotPath);
PlatformFile.CopyFile(*NewScreenshotPath, *OldScreenshotPath);
// Write budget violation.
TMap<FString, FString> BudgetViolationTemplateReplacements;
BudgetViolationTemplateReplacements.Add(TEXT("{LOCATION}"),
FormatLocation(
BudgetViolation.CurrentLocation));
BudgetViolationTemplateReplacements.Add(
TEXT("{PREVIOUS}"), BudgetViolation.PreviousTargetPointName);
BudgetViolationTemplateReplacements.Add(TEXT("{NEXT}"),
BudgetViolation.NextTargetPointName);
BudgetViolationTemplateReplacements.Add(TEXT("{FPS}"),
FormatTime(BudgetViolation.FPS));
BudgetViolationTemplateReplacements.Add(TEXT("{GAME_TIME}"),
FormatTime(
BudgetViolation.GameThreadTime));
BudgetViolationTemplateReplacements.Add(TEXT("{RENDER_TIME}"),
FormatTime(
BudgetViolation.RenderThreadTime));
BudgetViolationTemplateReplacements.Add(TEXT("{GPU_TIME}"),
FormatTime(BudgetViolation.GPUTime));
BudgetViolationTemplateReplacements.Add(TEXT("{SCREENSHOT_PATH}"),
ScreenshotFilename);
BudgetViolationsString +=
ApplyTemplateFile(BudgetViolationTemplatePath,
BudgetViolationTemplateReplacements);
}
// Write map.
TMap<FString, FString> MapTemplateReplacements;
MapTemplateReplacements.Add(TEXT("{MAP_NAME}"), TestSuiteResult.MapName);
MapTemplateReplacements.Add(TEXT("{MAP_DURATION}"),
FormatTime(TestResult.TimeSeconds));
MapTemplateReplacements.Add(TEXT("{BUDGET_VIOLATIONS}"), BudgetViolationsString);
MapString += ApplyTemplateFile(MapTemplatePath, MapTemplateReplacements);
}
}
}
// Write report.
TMap<FString, FString> ReportTemplateReplacements;
ReportTemplateReplacements.Add(TEXT("{START_TIME}"), GetTimestamp(TestSuites));
ReportTemplateReplacements.Add(TEXT("{TOTAL_DURATION}"),
FormatTime(GetTotalTimeSeconds(TestSuites)));
ReportTemplateReplacements.Add(TEXT("{MAP_RESULTS}"), MapString);
FString ReportHtmlString = ApplyTemplateFile(ReportTemplatePath, ReportTemplateReplacements);
UE_LOG(LogDaeTest, Verbose, TEXT("Test report:\r\n%s"), *ReportHtmlString);
UE_LOG(LogDaeTest, Display, TEXT("Writing test report to: %s"), *ReportHtmlString);
FString HtmlReportPath = FPaths::Combine(ReportPath, TEXT("performance-report.html"));
FFileHelper::SaveStringToFile(ReportHtmlString, *HtmlReportPath);
// Copy style file.
const FString& StyleFileName = TEXT("bootstrap.min.css");
FString PluginStyleFilePath =
FPaths::Combine(ContentDir, TEXT("ReportTemplates"), StyleFileName);
FString ReportStyleFilePath = FPaths::Combine(ReportPath, StyleFileName);
UE_LOG(LogDaeTest, Display, TEXT("Copying %s to %s."), *PluginStyleFilePath,
*ReportStyleFilePath);
PlatformFile.CopyFile(*ReportStyleFilePath, *PluginStyleFilePath);
}
FString FDaeTestReportWriterPerformance::FormatTime(float Time) const
{
return UKismetTextLibrary::Conv_FloatToText(Time, ERoundingMode::HalfToEven, false, false, 1,
324, 2, 2)
.ToString();
}
FString FDaeTestReportWriterPerformance::FormatLocation(const FVector& Location) const
{
return FString::Printf(TEXT("X=%d Y=%d Z=%d"), FMath::FloorToInt(Location.X),
FMath::FloorToInt(Location.Y), FMath::FloorToInt(Location.Z));
}
| cpp |
<reponame>gdia/The-Complete-Python-Course
books_file = 'books.txt'
def create_book_table():
with open(books_file, 'w') as file:
pass # just to make sure the file is there
def get_all_books():
with open(books_file, 'r') as file:
lines = [line.strip().split(',') for line in file.readlines()]
return [
{'name': line[0], 'author': line[1], 'read': line[2]}
for line in lines
]
def add_book(name, author):
with open(books_file, 'a') as file:
file.write(f'{name},{author},0\n')
def _save_all_books(books):
with open(books_file, 'w') as file:
for book in books:
file.write(f"{book['name']},{book['author']},{book['read']}\n")
def mark_book_as_read(name):
books = get_all_books()
for book in books:
if book['name'] == name:
book['read'] = '1'
_save_all_books(books)
def delete_book(name):
books = get_all_books()
books = [book for book in books if book['name'] != name]
_save_all_books(books)
# def delete_book(name):
# for book in books:
# if book['name'] == name:
# books.remove(book)
| python |
The ongoing protests in Iran over the death of 22-year-old Mahsa Amini while in the custody of the country’s “Guidance Patrol” (or morality police) have made world headlines. But there is another form of protest that has received less mainstream attention in Western media.
Whereas Amini was arrested for allegedly wearing her hijab “improperly”, thereby violating Iran’s mandatory hijab law, this new protest campaign involves another form of headwear – the amameh, or turban, worn by Shi’a clergy. Protesters have been deliberately knocking amameh off the heads of passing clerics.
The movement, known as “amameh parani”, has spread across Iran since early November. It has become particularly popular with young Iranians. Videos posted on twitter under #TurbanTossing and عمامه_پرانی# show amameh being knocked off in streets, cars, buses, metro stations and almost everywhere clergy appear in public.
In less than a month, amameh parani has become the symbol of a national satirical mockery of Shi’a clergy and their legitimacy in Iran, and another face of the global protests against the death of Mahsa Amini.
By focusing on the significance, symbolism and function of the amameh, the campaign explicitly targets the hegemony of Shi’a clergy over Iranian politics and society.
Clerical attire is composed of three pieces: the amameh, a turban made of 11 metres of thin white or black cotton material; a long cotton garment called a qabā; and the abā, the long open robe worn over it.
Students at Shi’a seminaries are ceremonially crowned with an amameh upon completing the first stage of their theological studies, which typically take three to five years.
Iranian clergy and their institutions view the amameh as sacred. They even use its colours to signify the lineage of a cleric, creating a class-based system both within and outside the clerical institutions. In its contemporary usage, for instance, a black amameh signifies a cleric’s claim of direct lineage to the prophet.
Because of this, the amameh is the source of religious legitimacy and implies a sense of infallibility inherent in Shi’a theology. The Islamic Republic has translated this theological model systematically into politics.
The amameh is the only source of political authority in the Islamic Republic. The clergy occupy all positions of power and authority. They have established and protected an exclusive political and economic system.
Iran’s parliament, government, judiciary, military, economy and education system are either directly ruled by a cleric or by a clerical assembly. Candidates in Iran’s elections must be approved by the Guardian Council. The council also warrants all laws passed by the parliament in accordance with Shi’a Shari’a law.
The amameh is no longer a mere sign of religious learning or social status. Rather, it is the symbol of a hegemonic political power. Like defrocking in Christian churches, removing an amameh is synonymous with the removal of its associated rights, authority and prestige.
Prior to the Mahsa Amini protests, amameh parani was typically a deliberate cross-party attack at a perceived political opponent. It was typically performed by zealous followers of the conservatives, based on an edict from one of Ruhollah Khomeini’s revolutionary sermons in 1969 to toss the turbans of clergy deemed corrupt.
The current amameh parani campaign employs the same tactic for a different end. Dislodging an amameh in public is a sign of great irreverence and ridicule. It attacks what the attire represents: the Islamic Republic regime.
Hand in hand with slogans such as “Clerics get lost! ”, it’s a form of resistance against discrimination and exclusion, and represents the rejection of clericalism. It is a symbolic act against the entanglement of religion and politics in Iran.
The campaign is also about gender politics and the violent and discriminatory way clothing is used against women. It is common for clergy to verbally abuse women and girls in public for their “inappropriate” hijab . The death of Mahsa Amini highlighted the kind of gender-based abuse Iranian women have been subject to for more than four decades.
From Iranian clergy in parliament saying that tossing the amameh is “playing with the lion’s tail”, to Iraqi Shi’a Sadrist leader Muqtada al-Sadr warning against the spread of amameh parani across the border, it’s clear the symbolic meaning of the act is being felt.
The reaction to protest in general has been typically harsh and violent, including calls for the execution of protesters. Courts have already imposed the death sentence on some. These threats can extend to those who live outside Iran, including the co-author of this article, who has decided to remain anonymous.
Had any influential cleric opposed the killing of Mahsa Amini or other peaceful protesters, campaigns like amameh parani might not have taken off. But the regime’s demand for more brutality and violence has only further angered the public.
The Iranian clergy face a crisis of legitimacy beyond politics. Their challenge is no longer about maintaining hegemony over the country, but whether they will retain the legitimacy to perform their traditional religious roles.
Negar Partow is Senior Lecturer in Security Studies, Massey University.
This article first appeared on The Conversation. | english |
{"template":{"small":"https://static-cdn.jtvnw.net/emoticons/v1/{image_id}/1.0","medium":"https://static-cdn.jtvnw.net/emoticons/v1/{image_id}/2.0","large":"https://static-cdn.jtvnw.net/emoticons/v1/{image_id}/3.0"},"channels":{"a0307":{"title":"花美男","channel_id":22703523,"link":"http://twitch.tv/a0307","desc":null,"plans":{"$4.99":"12993","$9.99":"35580","$24.99":"35581"},"id":"a0307","first_seen":null,"badge":"https://static-cdn.jtvnw.net/badges/v1/dcbd1586-ab3e-4274-a301-0886fdab76c4/1","badge_starting":"https://static-cdn.jtvnw.net/badges/v1/dcbd1586-ab3e-4274-a301-0886fdab76c4/3","badge_3m":null,"badge_6m":null,"badge_12m":null,"badge_24m":null,"badges":[{"image_url_1x":"https://static-cdn.jtvnw.net/badges/v1/dcbd1586-ab3e-4274-a301-0886fdab76c4/1","image_url_2x":"https://static-cdn.jtvnw.net/badges/v1/dcbd1586-ab3e-4274-a301-0886fdab76c4/2","image_url_4x":"https://static-cdn.jtvnw.net/badges/v1/dcbd1586-ab3e-4274-a301-0886fdab76c4/3","description":"Subscriber","title":"Subscriber","click_action":"subscribe_to_channel","click_url":""}],"bits_badges":null,"cheermote1":null,"cheermote100":null,"cheermote1000":null,"cheermote5000":null,"cheermote10000":null,"set":12993,"emotes":[{"code":"huaCARRY","image_id":51182,"set":12993},{"code":"hua00","image_id":110483,"set":12993}]}}}
| json |
Social media platforms like Facebook, Twitter, and YouTube have been making significant investments in the development of artificial intelligence to moderate content and automate the removal of harmful posts. These decisionmaking technologies typically rely on machine-learning techniques and are specific to types of content, such as images, videos, sounds, and written text. Some of these AI systems, developed to measure the "toxicity" of text-based content, make use of natural language processing and sentiment assessment to detect harmful text.
While these technologies may appear to represent a turning point in the debate around hate speech on the internet, recent research has shown that they are still far from being able to distinguish context or intent. If such AI tools are entrusted with the power to police content online, they have the potential to suppress legitimate speech and censor the use of specific words, particularly by vulnerable groups.
At InternetLab, we recently conducted a study focused on Perspective, an AI technology developed by Jigsaw (owned by Google's parent company, Alphabet). The AI measures the perceived level of “toxicity” of text-based content. Perspective defines "toxic" as "a rude, disrespectful, or unreasonable comment that is likely to make you leave a discussion.” Accordingly, the AI model was trained by asking people to rate internet comments on a scale from "very healthy" to "very toxic." The level of perceived toxicity indicates the likelihood that a specific comment will be considered toxic.
We used Perspective’s API to compare the perceived levels of toxicity of well-known drag queens and far-right political figures. The study compared the Twitter accounts of all the former participants of RuPaul's Drag Race with those of far-right leaders such as David Duke, Richard Spencer, Stefan Molyneux, and Faith Goldy. Additionally, we included prominent non-LGBTQ Twitter users, including Donald Trump and Michelle Obama. We analyzed over 114,000 tweets posted in English with Perspective’s most recent version.
Our results indicate that a significant number of drag queen Twitter accounts were calculated to have higher perceived levels of toxicity than white nationalist leaders. On average, the toxicity levels of the drag queens’ accounts ranged from 16.68 percent to 37.81 percent, while the white nationalists’ averages spanned from 21.30 percent to 28.87 percent. The toxicity level of President Trump’s Twitter account was 21.84 percent.
We also ran tests measuring the toxicity level of words commonly found in the tweets of drag queens. These words had significantly high levels of toxicity: gay (76.10 percent), lesbian (60.79 percent), queer (51.03 percent), and transvestite (44.48 percent). That means that, even when used in a neutral context, such words were ranked as significantly toxic by Perspective's AI. This indicates important biases in Perspective’s tool.
Additionally, words such as fag (91.94 percent), sissy (83.20 percent), and bitch (98.18 percent) registered high levels of toxicity. Though those words might be commonly perceived as harmful, their use by members of the LGBTQ community most often serves a different purpose.
Drag queens can be sharp-tongued. From “reads”—a specific form of insult that acerbically exposes someone’s flaws—to harsh jokes and comebacks, drag queens often reclaim words traditionally used as slurs to build a distinctive communication style.
In person, it is easier to understand context and see this as a form of self-expression. But when reading such missives online, it is significantly more challenging to distinguish between harmful and legitimate speech—especially when that assessment is made by machines. These in-group uses were also found in various tweets we analyzed. But in many of those cases, Perspective still deemed the post extremely toxic:
Often times, these “harsh” interactions address sensitive topics like sexual roles in relationships, the visibility of gayness, and sexual promiscuity—subjects usually explored by those who aim to verbally attack LGBTQ people.
But when directed at each other by members of the LGBTQ community, these comments may come from a place of solidarity, not malice. The underlying messages do not promote hate, prejudice, and discrimination. On the contrary, they often evoke pride and self-acceptance, helping LGBTQ people cope with outside hostility.
Hate speech is often predicated on underlying messages, as well. When subtext promotes hateful or discriminatory ideas, it represents a threat for marginalized and vulnerable groups. By training its algorithm to learn what content is likely to be considered toxic, Perspective’s tool seems to be giving more prevalence to words, rather than their underlying messages.
Though the ideas promoted by white nationalist tweets may target vulnerable groups, Perspective's AI often categorized them as much less toxic than the drag queens’ tweets:
If this AI tool were empowered to decide which tweets should be removed, many of the drag queens’ posts would be suppressed. In fact, Perspective is already making such decisions.
The problem: Such AI tools may be developed using biased training data, posing threats to the self-expression and visibility of vulnerable groups. Taken individually, 3,925 tweets from drag queens—around 3.7 percent of the total amount of analyzed tweets—would have been hidden from Tune users on "Keep it low" mode, according to our research.
The use of Perspective and other similar technologies could thus be mistakenly used to police and censor legitimate LGBTQ speech on online platforms. If AI tools focus on misleading signals—such as the use of specific words, rather than a message's intent—such models will make little progress in removing hate speech.
AI tools have the potential to shape the way we communicate. If computers indiscriminately decide what is “toxic,” tech has the power to both impact our modes of expression online and severely limit the inclusiveness of the internet.
| english |
package ru.volkov.guest.view.registration;
import com.vaadin.flow.component.Tag;
import com.vaadin.flow.component.button.Button;
import com.vaadin.flow.component.dependency.CssImport;
import com.vaadin.flow.component.dependency.JsModule;
import com.vaadin.flow.component.polymertemplate.EventHandler;
import com.vaadin.flow.component.polymertemplate.Id;
import com.vaadin.flow.component.polymertemplate.PolymerTemplate;
import com.vaadin.flow.component.textfield.PasswordField;
import com.vaadin.flow.data.binder.BeanValidationBinder;
import com.vaadin.flow.router.BeforeEnterEvent;
import com.vaadin.flow.router.BeforeEnterObserver;
import com.vaadin.flow.router.PageTitle;
import com.vaadin.flow.router.Route;
import com.vaadin.flow.templatemodel.TemplateModel;
import lombok.Getter;
import lombok.Setter;
import lombok.extern.slf4j.Slf4j;
import ru.volkov.guest.data.service.AuthService;
import java.util.List;
import java.util.Map;
import java.util.Optional;
import static ru.volkov.guest.util.ConfigHelper.getDefNotify;
@Slf4j
@JsModule("./views/registration/registration-view.js")
@CssImport("./views/registration/registration-view.css")
@Tag("registration-view")
@Route("registration")
@PageTitle("Registration")
public class RegistrationView extends PolymerTemplate<TemplateModel> implements BeforeEnterObserver {
@Id("enterPass")
private PasswordField enterPass;
@Id("confirmPass")
private PasswordField confirmPass;
@Id("enter")
private Button enter;
@Id("clear")
private Button clear;
private final AuthService authService;
private BeanValidationBinder<Password> binder;
private String code;
@Getter
@Setter
public class Password {
private String enterPass;
private String confirmPass;
}
public RegistrationView(AuthService authService) {
this.authService = authService;
initForm();
}
private void initForm() {
binder = new BeanValidationBinder<>(Password.class);
binder.forField(enterPass)
.withValidator((p) -> enterPass.getValue().equals(confirmPass.getValue()),
"Passwords must be equals")
.bind("enterPass");
binder.forField(confirmPass)
.withValidator((p) -> enterPass.getValue().equals(confirmPass.getValue()),
"Passwords must be equals")
.bind("confirmPass");
binder.bindInstanceFields(this);
}
@EventHandler
private void save() {
Password password = new Password();
if (binder.writeBeanIfValid(password)) {
authService.activate(password.getEnterPass(), code);
getDefNotify("You are registered");
clearForm();
navigateToMainPage();
}
}
@Override
public void beforeEnter(BeforeEnterEvent event) {
Map<String, List<String>> parameters = event.getLocation().getQueryParameters().getParameters();
Optional.ofNullable(parameters.get("code")).ifPresent(paramList -> {
String code = paramList.get(0);
if (code != null && !authService.isActivated(code)) {
this.code = code;
} else {
event.forwardTo("login");
}
});
if (!parameters.containsKey("code")) {
event.forwardTo("login");
}
}
@EventHandler
private void clearForm() {
binder.setBean(new Password());
}
private void navigateToMainPage() {
getUI().ifPresent(ui -> ui.navigate("login"));
}
} | java |
<gh_stars>0
import * as colors from 'colors';
import shell from 'shelljs';
import fs from 'fs';
import {WebresourcesCrmJson} from '../root/Webresources/CrmJson';
import {CrmJson} from '../root/CrmJson';
import cp from 'child_process';
export class Update {
public static updateProject(): Promise<void> {
if (process.argv[3]) {
console.log(colors.red(`No spaces allowed after update command!`));
} else if (!shell.test('-e', 'src')) {
console.log(colors.red(`You are not inside the project Webresources folder!`));
} else {
return Update.update();
}
}
public static showUpdateHelp(): void {
console.log(`Arguments:`);
}
private static async update(): Promise<void> {
console.log(`Updating D365 Project...`);
Update.updateSrcFolder();
Update.updateCrmJson();
Update.updatePCF();
Update.updateProjectRootFolder();
Update.updateEntityFiles();
Update.updateFormFiles();
Update.updatePackageJson();
await Update.updateWebpackConfig();
console.log(`Updating D365 Project done`);
}
// eslint-disable-next-line max-lines-per-function
private static updateProjectRootFolder(): void {
console.log(`Updating .eslintignore...`);
shell.cp('-R', `${__dirname}/root/Webresources/.eslintignore`, '.');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', '.eslintignore']);
}
// shell.exec('git add .eslintignore');
console.log(`Updating .gitignore...`);
shell.cp('-R', `${__dirname}/root/Webresources/gitignore`, '.');
fs.renameSync(`./gitignore`, './.gitignore');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', '.gitignore']);
}
// shell.exec('git add .gitignore');
console.log(`Updating .eslintrc.json...`);
shell.cp('-R', `${__dirname}/root/Webresources/.eslintrc.json`, '.');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', '.eslintrc.json']);
}
// shell.exec('git add .eslintrc.json');
console.log(`Updating postcss.config.js`);
shell.cp('-R', `${__dirname}/root/Webresources/postcss.config.js`, '.');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', 'postcss.config.js']);
}
// shell.exec('git add postcss.config.js');
console.log(`Updating karma.conf.js`);
shell.cp('-R', `${__dirname}/root/Webresources/karma.conf.js`, '.');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', 'karma.conf.js']);
}
// shell.exec('git add karma.conf.js');
console.log(`Updating tsconfig.json`);
shell.cp('-R', `${__dirname}/root/Webresources/tsconfig.json`, '.');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', 'tsconfig.json']);
}
// shell.exec('git add tsconfig.json');
}
private static updateEntityFiles(): void {
console.log(`Updating Entity files...`);
const crmSettings: CrmJson = JSON.parse(fs.readFileSync('../crm.json', 'utf8'));
const crm = crmSettings.crm;
shell.ls(`src/**/*.ts*`).forEach(function (filepath) {
const fileData = String(fs.readFileSync(filepath));
const match = fileData.match(new RegExp(`specify Form onLoad function: ${crm.publisher_prefix}.${crm.namespace}.([^\\W]*).Form.onLoad`));
if (match) {
const entityName_formName = match[1];
const split = entityName_formName.split('_');
const newFileData = fileData.replace(new RegExp(entityName_formName, 'ig'), `${split[0]}.${split[1]}`);
shell.ShellString(newFileData).to(filepath);
}
});
}
private static updateFormFiles(): void {
console.log(`Updating Form files...`);
shell.ls(`src/**/*.form.ts*`).forEach(function (filepath) {
const fileData = String(fs.readFileSync(filepath));
const match = fileData.match(new RegExp('export class ([a-zA-Z]*)Form {'));
if (match) {
const entityName = match[1];
const importString = `import {${entityName}FormContext} from './${entityName}.formContext';`;
const newExportString = `export class ${entityName}Form extends ${entityName}FormContext {`;
shell.sed('-i', new RegExp(`export class ${entityName}Form {`, 'i'), `${importString}\n${newExportString}`, filepath);
}
});
}
private static updatePCF(): void {
if (!shell.test('-e', '../PCF')) {
shell.cd('..');
const crmSettings: CrmJson = JSON.parse(fs.readFileSync('./crm.json', 'utf8'));
shell.mkdir('PCF');
shell.cd('PCF');
shell.mkdir('Solutions');
shell.cd('Solutions');
shell.exec(`pac solution init --publisher-name ${crmSettings.crm.publisher_prefix} --publisher-prefix ${crmSettings.crm.publisher_prefix}`);
shell.cd('..');
shell.cd('..');
shell.cd('Webresources');
}
}
// eslint-disable-next-line max-lines-per-function
private static updateSrcFolder(): void {
console.log(`Updating Annotation...`);
shell.cp('-R', `${__dirname}/root/Webresources/src/Annotation`, './src');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', 'src/Annotation']);
}
// shell.exec('git add src/Annotation');
console.log(`Updating Http...`);
shell.cp('-R', `${__dirname}/root/Webresources/src/Http`, './src');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', 'src/Http']);
}
// shell.exec('git add src/Http');
console.log(`Updating Translation...`);
shell.cp('-R', `${__dirname}/root/Webresources/src/translation`, './src');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', 'src/translation']);
}
// shell.exec('git add src/translation');
console.log(`Updating util...`);
shell.cp('-R', `${__dirname}/root/Webresources/src/util`, './src');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', 'src/util']);
}
// shell.exec('git add src/util');
console.log(`Updating WebApi...`);
shell.cp('-R', `${__dirname}/root/Webresources/src/WebApi`, './src');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', 'src/WebApi']);
}
// shell.exec('git add src/WebApi');
}
// eslint-disable-next-line max-lines-per-function
private static updateCrmJson(): void {
if (fs.existsSync('./tools/crm.json')) {
//crm.json
const crmSettings: CrmJson = JSON.parse(fs.readFileSync('./tools/crm.json', 'utf8'));
shell.cp('-R', `${__dirname}/root/crm.json`, '../');
shell.sed('-i', new RegExp('<%= publisher_prefix %>', 'ig'), crmSettings.crm.publisher_prefix, '../crm.json');
shell.sed('-i', new RegExp('<%= environment %>', 'ig'), crmSettings.crm.url, '../crm.json');
shell.sed('-i', new RegExp('<%= namespace %>', 'ig'), (crmSettings as unknown as {webresource: {namespace:string} }).webresource.namespace, '../crm.json');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', '../crm.json']);
}
// Webresources/crm.json
const webresourcesSettings: WebresourcesCrmJson = JSON.parse(fs.readFileSync('./tools/crm.json', 'utf8'));
shell.cp('-R', `${__dirname}/root/Webresources/crm.json`, '.');
shell.sed('-i', new RegExp('<%= solution_name_deploy %>', 'ig'), webresourcesSettings.crm.solution_name_deploy, './crm.json');
shell.sed('-i', new RegExp('<%= solution_name_generate %>', 'ig'), webresourcesSettings.crm.solution_name_generate, './crm.json');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', './crm.json']);
}
// shell.exec('git add ./crm.json');
shell.rm('-rf', `./tools`);
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['rm', './tools']);
}
// shell.exec('git rm ./tools');
}
const crmSettings: CrmJson = JSON.parse(fs.readFileSync('../crm.json', 'utf8'));
shell.cp('-R', `${__dirname}/root/crm.json`, '../');
shell.sed('-i', new RegExp('<%= publisher_prefix %>', 'ig'), crmSettings.crm.publisher_prefix, '../crm.json');
shell.sed('-i', new RegExp('<%= environment %>', 'ig'), crmSettings.crm.url, '../crm.json');
shell.sed('-i', new RegExp('<%= namespace %>', 'ig'), crmSettings.crm.namespace, '../crm.json');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', '../crm.json']);
}
const version = shell.exec('hso-d365 --version').stdout.replace(/\n/ig, '');
shell.sed('-i', new RegExp('<%= version %>', 'ig'), version, '../crm.json');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', '../crm.json']);
}
// shell.exec('git add ../crm.json');
}
private static updatePackageJson(): void {
console.log(`Updating package.json...`);
const packageJson = JSON.parse(fs.readFileSync('./package.json', 'utf8'));
const {name, description, version} = packageJson;
const settings: CrmJson = JSON.parse(fs.readFileSync('../crm.json', 'utf8'));
shell.cp('-R', `${__dirname}/root/Webresources/package.json`, '.');
const packageJsonFile = shell.ls('package.json')[0];
shell.sed('-i', new RegExp('<%= projectname %>', 'ig'), name, packageJsonFile);
shell.sed('-i', new RegExp('<%= description %>', 'ig'), description, packageJsonFile);
shell.sed('-i', new RegExp('<%= publisher %>', 'ig'), settings.crm.publisher_prefix, packageJsonFile);
shell.sed('-i', new RegExp('<%= version %>', 'ig'), version, packageJsonFile);
console.log(`Removing old npm packages. This may take a while...`);
shell.exec('npm prune');
shell.exec('npm install');
}
private static async updateWebpackConfig(): Promise<void> {
console.log(`Updating webpack.config...`);
shell.cp('-R', `${__dirname}/root/Webresources/webpack.config.ts`, '.');
if (shell.test('-e', '../.git')) {
cp.execFileSync('git', ['add', 'webpack.config.ts']);
}
}
}
| typescript |
Unrivalled League to revolutionise football in India and put the country on the global map ..
Bangalore, 27th Feb 2013: After a stellar performance in the BDFA League, where they were ..
Bangalore, November 26, 2012: South United Football Club (SUFC), operated by Lokko Sports ..
Indian football fans, teams and players will have reasons to rejoice. IMG-Reliance, the co ..
With the rising popularity of soccer in the country and FIFA all ready to be introduced ve ..
When nothing special is expected off from our young players, they are sizzling at the ongo ..
The value of each moment defers from one another in life but there are few moments which w ..
Amidst increasing popularity of the EPL in India, there has come great news for Indian foo ..
+oneTM is proud to officially announce the launch of the +ONE India Caribbean Tour 2011, ..
India In World Cup – A Dream?
It must be every Indian football fan’s dream to see India play in the World Cup. I perso ..
| english |
The Rs 6,480-crore initial public offering (IPO) of the Hyderabad-based Gland Pharma managed to sail through on the back of institutional investor support, even as individual investors shunned the issue.
The institutional investor portion of the IPO was subscribed 6. 4x.
Meanwhile, the high networth individual (HNI) portion was subscribed 51 per cent, and retail investor portion subscribed just 24 per cent.
TO READ THE FULL STORY, SUBSCRIBE NOW NOW AT JUST RS 249 A MONTH.
What you get on Business Standard Premium?
- Unlock 30+ premium stories daily hand-picked by our editors, across devices on browser and app.
- Pick your 5 favourite companies, get a daily email with all news updates on them.
- Full access to our intuitive epaper - clip, save, share articles from any device; newspaper archives from 2006.
- Preferential invites to Business Standard events.
- Curated newsletters on markets, personal finance, policy & politics, start-ups, technology, and more. | english |
{
"$schema": "https://raw.githubusercontent.com/SimonCropp/MarkdownSnippets/master/schema.json",
"TocExcludes": [ "NuGet package", "Release Notes", "Icon" ],
"MaxWidth": 90,
"ValidateContent": true
} | json |
<filename>component-cache-spring-boot-starter/src/test/java/cn/elmi/components/cache/broadcast/RedisCacheChannelTest.java<gh_stars>0
package cn.elmi.components.cache.broadcast;
import cn.elmi.components.cache.CacheChannel;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.boot.test.context.SpringBootTest;
import org.springframework.test.context.testng.AbstractTestNGSpringContextTests;
import org.testng.annotations.Test;
import cn.elmi.components.cache.Person;
@SpringBootTest
public class RedisCacheChannelTest extends AbstractTestNGSpringContextTests {
@Autowired
private CacheChannel channel;
/**
* 根据票号查询
*/
@Test
public void test_1() {
String region = "test";
channel.put(region, "L1", "guava");
channel.put(region, "L2", "redis");
channel.put(region, 12, new Person());
Person p = new Person();
channel.put(region, p, "person");
System.out.println(channel.get(region, "L1"));
System.out.println(channel.get(region, "L2"));
System.out.println(channel.get(region, 12));
System.out.println(channel.get(region, p));
channel.keys(region).forEach(System.out::println);
}
}
| java |
A deadly land mine explosion in Syria killed at least six people on Sunday, according to state media.
News agency SANA said the explosion hit civilians who were foraging for truffles in the countryside, and blamed the incident on a land mine planted by the Islamic State group in the southern Deir Ez-Zor province. The area is a former stronghold of the militants.
A day earlier, SANA reported six people - also heading to search for truffles - were killed by an anti-tank mine left by IS in the desert of Homs' eastern countryside.
The Syrian Observatory for Human Rights, a Britain-based war monitor, put the number killed Sunday at nine. The monitor said the incident brings to 139 the number of civilians reported killed this year as a result of the explosion of mines and other explosive objects left over from the war, including 30 children.
The truffles are a seasonal delicacy that can be sold for a high price. Since truffle hunters work in large groups in remote areas, IS militants have repeatedly preyed on them, emerging from the desert to abduct them, kill some and ransom others for money.
In February, IS sleeper cells attacked workers collecting truffles near the central town of Sukhna, killing at least 53 people, mostly workers but also some Syrian government security forces.
| english |
Melbourne: Calls for Ricky Ponting's retirement from international cricket grew stronger, with Test legend Doug Walters and former wicket-keeper Ian Healy joining the chorus saying the former Australian captain should call it a day with the two-Test series at home against New Zealand next month.
The Australian media has also come down harshly on Ponting, who turns 37 next month, and said the batting legend should be honest with himself after his prolonged form slump.
New selection boss John Inverarity refused to put a date on Ponting's career, but said the National Selection Panel (NSP) is keen to inject fresh blood into the Australian team.
Ponting, however, got some much-needed backing from legendary captain Allan Border and former England captain Tony Greig.
Doug Walters said the former skipper is nearing the end of his career.
"I'm a huge fan of Ricky as a batsman, but he's getting towards the end and perhaps we can get more out of him if he moves down a notch (in the batting line-up)," Walters was quoted as saying by the Daily Telegraph.
Ponting, who took over from Steve Waugh, stepped down from captaincy earlier this year after Australia's quarter-final exit from the World Cup. He led Australia in 77 Tests, winning 48, the most for any captain. He also captained Australia in a record 228 ODIs and won 164, including two World Cups in 2003 and 2007.
Ponting, the third highest run-getter in Tests, is in miserable form with the bat. In the ongoing Test series in South Africa, he made 8 and 0 in the first Test in Cape Town and fell for a third-ball duck in the first innings of the ongoing second Test in Johannesburg.
Ponting has now gone 27 Test innings without a ton and his last century, a majestic 209, came almost two years ago against Pakistan in Hobart. He has averaged just 25 in his last 14 innings and has failed to reach double figures seven times in his last 13 innings, until Sunday.
Before his fighting half-century in the second innings in Johannesburg on Sunday, Ponting had made 184 runs at an average of 14. 15 in his past 13 Test innings, dating back to the Ashes last December.
Healy said faults are visible in his former teammate's batting, something that he has never seen before.
"There are faults that have never been in his technique before and I don't like his statistics from the past 12 months. No one who has played for as long as he has should be averaging in the teens. I think it's a pretty easy decision. He's been on the wane for some time now and without something special, it's time for Ricky to say, 'I'm finishing'," said Healy.
Former captain Kim Hughes is also worried about Ponting's technical problems, but believes that the struggling Australian team would be worse without him.
"He's got a severe technical fault, which happens when people lose confidence. He gets so far over to the offside . . . because of his head balance, he's trying to take balls on middle stump through square leg. You're not going to survive," Hughes said.
"The Australian side at the moment looks pretty vulnerable. There's a new captain, the opening partnership is still uncertain, Usman Khawaja is a new boy who has just come in. With doubts about Marsh's fitness, irrespective of whether he (Ponting) gets runs in the second innings or not, I'd give him a game against New Zealand at home on truer wickets. "
"With the confidence of spending time in the middle, he (Ponting) might overcome them (technical flaws). He's keen, he's enthusiastic, he adds a lot to the side with his training attitude. "
Border, a commentator for the ongoing Test series, refused to comment on Ponting's future but the Daily Telegraph reported that Border privately believes that Ponting should dig his heels in and is convinced that he has more to offer the national team. Tony Greig also feels that Ponting is still in the best Australian team.
"I suspect Ponting will go into the new summer in Australia regardless of what happens in this (second) Test," Greig said. "If it keeps going wrong, they have to look at somebody else, but is there someone who has made four (first-class) hundreds in a row and deserves to step into that position? "
Inverarity said the selection committee is keen to have some new faces in the team.
"Ricky Ponting is a magnificent cricketer and a fine man, a very resilient man. All selectors are always looking for a change of blood and to bring young players in, and you have to balance that with picking your best side," he said. (This story has not been edited by News18 staff and is published from a syndicated news agency feed - Agencies) | english |
Ajay Devgn and Rohit Shetty are returning to their cop universe with Singham Again. The filming of the next part of the Singham franchise is set to begin later this year, with an Independence Day 2024 release already scheduled. Following Deepika Padukone’s entry as “Lady Singham" in the film, another actress will reportedly be joining the franchise. Kareena Kapoor is said to be the latest addition to the cop universe. Kareena has previously been a part of the series and had appeared alongside Ajay Devgn in the 2011 movie Singham Returns. | english |
There are excruciatingly high stakes for software today. Trillions of dollars of market cap, billions of consumers, hundreds of billions of revenue, and limitless hours of usage are dependent on software working in real time, all the time, without downtime.
As the sophistication of software delivery has increased, every layer of the tech stack has been rewritten — an accelerated evolution has led to a bevy of multibillion-dollar, up-and-coming unicorns and multiple massive public market debuts.
However, today we’re talking about one of the most integral pillars holding up the internet. The Domain Name System, otherwise known as DNS, is the key addressing system that connects browsers, users, devices and servers together. Type in “www.techcrunch.com” in your browser, and DNS finds the address linked to that name and tells routers and switches across the world which server to connect to and how to send data back to you.
Unlike physical postal addresses, innovation around DNS addressing has flourished in recent years. Traffic management, performance scaling, and cost shaping have turned DNS from a basic directory into a vital layer for guaranteeing the reliability of all software in use on the internet today while protecting the bottom line.
Few companies have parlayed internet infrastructure experience into a world-class engineering company quite like NS1. The New York City-based startup has raised more than $100 million as it builds a strategic node at the core of the modern web delivery tech stack. Customers are flocking: 760 at latest count (up from 600 a year ago), with year-over-year bookings growth well into the triple digits.
How did the company take a slumbering and dreary yet reliable aspect of the internet and turn it into a strategic moat and an enterprise win? And what lessons can we learn about the future of the enterprise infrastructure layer from one of its leading lights? That’s what we’re here to find out.
The lead writer of this EC-1 is Sean Michael Kerner. Kerner has been covering the IT and enterprise infrastructure market for more than a decade as a tech journalist (or, @TechJournalist as he is known on Twitter). Perhaps most importantly, he’s also partially fluent in Klingon, which has no bearing on this EC-1, but is one of those cool facts we wanted to include anyway. The lead editor for this package was Danny Crichton, the assistant editor was Ram Iyer, the copy editor was Richard Dal Porto, and illustrations were drawn by Nigel Sussman.
NS1 had no say in the content of this analysis and did not get advance access to it. Kerner has no financial ties to NS1 or other conflicts of interest to disclose.
The NS1 EC-1 comprises four main articles numbering 10,300 words and a reading time of 41 minutes. Here’s what we’ll be (DNS) addressing:
- Part 1: Origin story “1 napkin and 22 lines of code, or how NS1 rewrote the rules of internet infrastructure” (2,700 words/11 minutes) — Sets the context for NS1’s business and provides a brief explanation of why DNS (the technology at the core of the NS1 service) matters.
- Part 2: Product development and roadmap “WTF is NS1? It’s DNS, DDI, and maybe other TLAs” (2,700 words/11 minutes) — Analyzes NS1’s products, vision and open-source efforts through NS1 Labs.
- Part 3: Competitive landscape “The fight for the future of DNS is white hot” (2,300 words/9 minutes) — Evaluates how NS1 competes in a space littered with well-funded Big Tech companies and well-established mature companies.
- Part 4: Customer development “Outages, pandemics and the reengineering of traffic on the internet” (2,600 words/10 minutes) — Explores how companies like Roblox and Pinterest have had to reshape their networking infrastructure in light of massive growth and increasing demands for resilience from users.
We’re always iterating on the EC-1 format. If you have questions, comments or ideas, please send an email to TechCrunch Managing Editor Danny Crichton at danny@techcrunch.com.
| english |
<filename>src/components/SearchForm/SearchForm.css<gh_stars>0
.search-form {
box-sizing: border-box;
}
.search-form hr {
border: .5px solid hsl(0, 0%, 94%);
}
.search-form label {
width: 100%;
}
.search-form input {
width: 100%;
}
.search-form select {
width: 100%;
}
.search-form button {
} | css |
Govinda and his wife Sunita Ahuja celebrated Karwa Chauth. (Photo: Sunita Ahuja, Govinda/Instagram)
Govinda recently took to Instagram to share photos from the Karwa Chauth celebrations with wife Sunita Ahuja. In one of the photos, the couple is seen posing next to a BMW car that Govinda gifted to Sunita.
You have exhausted your monthly limit of free stories.
“You deserve all the happiness in this world and more. Love you my Sona! ❤️ @officialsunitaahuja #Happykarwachauth #truelove @bmwinfinitycars @bmwindia_official,” concluded the actor.
See all inside photos from Govinda and wife Sunita Ahuja’s Karwa Chauth celebrations: | english |
The best dash cams in Australia will mount to your car’s windshield or dashboard, giving you peace of mind that you’ll have video proof should anything happen while you’re out on the road or your car is in park.
Of course, some dash cams are better than others, with the best in-car cameras offering clean, sharp footage with reliable features you know you can rely on if the worst should happen.
We've spent countless hours on the road with all of the latest models, testing them in various weather and lighting conditions to assess their usability, features, and video quality in the real world. Based on these experiences, we’ve tracked down the best dash cams for every need and budget below.
Below you can find a quick round-up of our favourite dash cams to buy in Australia. You can click through to read a more in-depth review and compare prices quickly using our price comparison tool.
With impressive 4K video quality and a flagship feature set, the Nextbase 622GW is the best dash cam you can buy right now.
Despite its tiny size, The Garmin Dash Cam Mini 2 records in Full HD at 30fps with HDR, producing footage that’s sharp enough to pick out key detail.
The Nexar Pro is a dual-cam solution aimed at those who spend extended periods behind the wheel, limited to 1080p.
Neat and relatively compact, the Vantrue E1 is an attractive dash cam that’s capable of recording 2.5K video at 30fps.
Clear 1440p video, a wide-angle lens and easy sharing of footage make the 522GW a worthy contender.
Similar to the Mini 2, the Garmin Dash Cam 67W can clip in immediately with a magnetic clip. Video quality is also excellent.
Loaded with useful features included in the price, this is a great entry point to 4K footage.
The Tandem Dash Cam is essentially the Dash Cam 67W, but with a camera facing the interior of the car in place of a touchscreen.
Designed with taxi drivers in mind, the Vantrue N2 Pro forgoes many of the extras which your car is already likely to have, including speed camera alerts and collision warnings.
Why you can trust TechRadar We spend hours testing every product or service we review, so you can be sure you’re buying the best. Find out more about how we test.
Here you'll find more in-depth reviews of our picks for the best dash cams in Australia in 2023. Each one has been tested thoroughly by TechRadar's team of camera experts, and you can click through to read the full review of each.
The Nextbase 622GW is a relatively new flagship dash cam that has proven itself as the best dash cam you can buy right now. In our tests, it delivered much-improved video quality and better stabilisation, along with the inclusion of what3words geolocation services for pinpointing stricken vehicles within a three-meter radius.
When we chose to shoot in 4K/30p, the resulting footage looked almost cinematic in its presentation, with extremely crisp definition and great detail, even in poor lighting. This makes it much easier to pinpoint registration numbers or pick out hard-to-see elements of an accident.
A built-in polarising filter on the front of the camera can be rotated to reduce glare from windscreens, while digital image stabilisation is another first for the dash cam market and helps smooth out those bumps and shakes caused by potholes and poor road surfaces.
Like its 522GW sibling, this model can be controlled via your voice with Alexa Skills, but it requires the accompanying smartphone app to work, which we didn't rate as the best we've tried. Despite new dual 2.4GHz + 5GHz Wi-Fi, we found that it still had trouble connecting with phones to transfer images and video clips.
Thankfully, the 3-inch rear touchscreen is crisp, clear and very easy to use, while the inclusion of what3words combines well with Nextbase's EmergencySOS feature, which you get a year's free subscription for with this dash cam.
The Garmin Mini 2 is a tiny dash cam that all but disappears behind your car’s rear view mirror, yet boasts Full HD video with HDR, voice control that actually works reliably, a decent smartphone app and a dead-simple magnetic mounting system.
Installing the Mini 2 is the same as other members of the Garmin Dash Cam family. We found the setup process took just a few minutes with the use of Garmin’s Drive smartphone app (iOS and Android). The simple but effective mounting system comprises a ball-and-socket joint for positioning the camera at the perfect angle, while a coin-sized magnet that sticks to your windscreen and holds the dash cam securely in place.
We found video quality to be good for the size of the camera. It records in Full HD, 30fps with HDR and produces footage that is sharp enough to pick out key details like registration plates, whatever the ambient light and weather conditions. At 140 degrees, the lens' field-of-view isn’t the widest on the market, but still provides a good view of the road ahead.
There’s a button for quickly saving a portion of video (or you can rely on the g-sensor to automatically detect a collision) and voice controls for functions like taking a photo or turning audio recording on and off work surprisingly well.
Due to the Garmin Dash Cam Mini 2’s compact size, it does not have a display. Instead, you need to use the smartphone app to check the camera’s view and access recordings. It also misses out on GPS, which is a shame, but this is arguably the only major feature missing here. Unless you need the best dash cam with driver assistance functions like speed camera alerts, the Garmin Dash Cam Mini 2 will leave you wanting for very little. Simple, subtle and reliable, it’s the definition of set-and-forget- technology.
If you're someone who spends extended periods behind the wheel, the Nexar Pro is going to be one of the best options for you. It's a dual-cam solution that can record video both inside and outside a vehicle and you get both camera units in the box. These connect by a cable, which we found surprisingly simple to setup, even if it took up a fair bit of screen real estate.
The Nexar app is central to the twin camera experience: it’s where you can fine-tune settings, generate incident reports and save recorded clips to the cloud (Nexar bundles cloud storage in for free). Other useful tools include break-in alerts if someone tries to tamper with your car, as well as GPS data logging.
There's no 4K recording here, but we found its 1080p video to be perfectly serviceable. The external camera fares well even in tricky shooting situations, from heavy rain to bright sunshine. There are cheaper dash cams out there with fewer features to play with, but if the security of your car is crucial to your daily life, the Pro is hard to beat in terms of protection and overall value.
Giving the Garmin Dash Cam Mini 2 a run for its money with its equally-compact size is the Vantrue E1. This attractive dash cam is capable of recording 2.5K video at 30fps, or Full HD 1080p footage at a smoother 60fps, for additional detail while you drive. Results are sharp night and day, with accurate colours and acceptable noise in low light. Our testing revealed that the optional polarising filter does a good job of reducing dashboard reflections.
The E1’s magnetic mount works well, but the lack of sideways adjustments limits its use if you can’t fit it centrally. If you can, you’ll find its 160-degree angle captures a wide view of what’s ahead. A tiny 1.54-inch screen offers a handy preview when installing, but the smartphone app is more useful to tweak settings.
Driver assistance is left to your vehicle, but the E1 still has a solid set of features, including GPS tracking, Wi-Fi connectivity, and a remote for wirelessly saving footage.
The Nextbase 622GW flagship (see no.1) might be one of the most advanced dash cams you can buy, but the 522GW remains the best dash cam all-rounder. Thanks to a crisp 1440p resolution and wide-angle lens, it does the basics very well, but also throws in plenty of additional features.
We found the three-inch touchscreen on the back to be responsive, and there's also the option of using the built-in Alexa functionality. Currently, you can ask Alexa to play music, place calls and listen to audiobooks through connected devices, though the Dash Cam Skill (which lets you command it to ‘start recording’, ‘stop recording’, ‘protect a recording’ and ‘send to my phone’) needs a bit more polish.
This is all a bit of a gimmick and, to be honest, we didn't use it all that much, so it is lucky that the remainder of the UX is extremely simple. Videos can be quickly and easily shared to a smart device via Bluetooth and Wi-Fi, while a clever Emergency SOS system will alert the emergency services of your location and other details if you prove unresponsive following an accident.
The 67W supersedes Garmin’s already very good 66W with a few additional connected features that only add to a very enticing package. At about the size of a matchbox, the 67W is one of the smallest dash cams we’ve had the honour of testing – in fact, it's only usurped by Garmin’s own Mini, which is so small it looks like it came out of a Christmas Cracker.
Crammed inside this tiny package is a top-quality sensor that's capable of capturing crisp 1440p footage and enhancing it in those tricky weather conditions thanks to a HDR (High Dynamic Range) function. In our tests, it looked great and the additional pixels made it possible to punch into footage on the computer to read number plates or spot minor things that could act as important evidence.
Easy to set up, simple to use and accompanied by a very clean smartphone app, the 67W goes one better than its predecessor by adding connected features, such as the ability to automatically upload any saved clips to Garmin’s cloud when the camera detects a trusted Wi-Fi network.
Bear in mind that Garmin will charge you for a storage plan – and if you want to make the most of its connected features, such as the ability to remotely check in on a parked car from anywhere in the world, you’ll have to hardwire the camera into your vehicle’s main power source.
You’ll also need to make sure the camera is connected to a Wi-Fi network, so this means parking next to a friendly router or investing in a mobile hotspot, which will also require constant power. You can see how it rapidly becomes a deep and expensive rabbit hole.
Regardless, if you just want great quality footage that’s automatically recorded by something that will fit into a shirt pocket, look no further.
We won’t hold it against you if you’ve never heard of Viofo, because it certainly isn't the biggest name in the dash cam business, but its 4K resolution Pro Duo model represents great value for money.
The front camera is quite a bit bulkier than many rivals on this list, but it features a built-in GPS module, something that many other brands offer as an optional extra. Its plastic casing looks and feels basic, but it houses some rather impressive tech that belies its overall build quality.
Rather than a traditional rechargeable battery set up, it runs super capacitors, meaning its power source is built to last and can withstand an extreme temperature range without blunting performance. On top of this, you get the option of what we found to be impressive 4K (3840 x 2160p) video recording, with the resulting footage offering a great amount of detail. There's also a Wide Dynamic Range mode for rich colours in all weather conditions.
Unfortunately, 4K recording is only possible at 30fps, which isn’t the greatest if you want to slow footage down. However, dual recording (front and rear cameras) is only available at full HD (1080p) anyway and this is pumped out at 60fps for much smoother results.
In our tests, the setup was very easy and Viofo offers a smartphone app for quickly reviewing and saving clips. Unfortunately, installation of dual cameras requires the removal of interior trim and the clever stashing of long wires. It can be a messy and infuriating process to get it right, but worth it to avoid a dangling tangle of power cables.
The fact that you get night vision, a parking mode, motion detection, automatic emergency recording, GPS tracking and dual-channel 1080p at this price makes this a package well worth considering if you cover a lot of miles and want total camera coverage that doesn’t cost a small fortune.
Garmin’s first dual-lens dash cam lets you view what’s going on both inside and outside the car while you are driving, which is handy for taxi drivers and others wanting to keep an eye on their passengers. Extremely compact in design, the Dash Cam Tandem features a clip-in magnetic mount and can be easily installed below the rear-view mirror and removed when not in use.
Two lengths of USB cable are provided (the longer one enables you to run the cable around the car cabin neatly), as is a cigarette lighter USB socket with two ports for charging your phone at the same time. However, if you want to use the dash cam for incident recording – such as if your car gets bumped in the car park while you are shopping – you will need to get it professionally installed so it’s constantly recording.
Key for control of the camera is the Garmin Drive app (Android and iOS) where you can review video and audio footage from your drives without having to take the microSD card out of the camera. We found picture quality to be generally excellent, especially from the front camera. And while the rear camera struggled a bit in very low light, you can still make out passengers reasonably clearly in black and white. There’s also a picture-in-picture option so you can view both rear- and front-facing camera footage simultaneously.
Rather usefully, footage is displayed with a time stamp, the speed of the vehicle and its location. Voice control is also provided, which enables hands-free control with instructions such as ‘OK Garmin, take a picture’ or ‘OK Garmin, save video’. Safety cam alert updates will also be added to the app soon.
The only slight problem we experienced was that the app wasn’t as intuitive as we would’ve liked and didn’t automatically connect to the Wi-Fi connection when reviewing footage from our drives. Aside from that, this is a pricey but excellent option for those who want to keep an eye on their car, inside and out.
Relatively compact for a dual-lens dash cam, the Vantrue N2 Pro records a detailed view of the road ahead and your vehicle’s interior. Designed with taxi drivers in mind, it forgoes many of the extras which your car is already likely to have, including speed camera alerts and collision warnings.
That simplicity extends to its connectivity, a factor we liked in testing. With no Wi-Fi, Bluetooth or smartphone app, it’s a self-contained dash cam that gets on with the job of capturing events. The only feature we did miss was GPS tracking, although Vantrue sells an optional GPS windscreen mount.
We were happily surprised by the video quality. The cameras can record Full HD footage in both directions, capturing video that’s smooth, sharp and clear, matched by decent audio. The interior camera occasionally struggles in bright sunlight, but that’s a small criticism of otherwise impressive performance. The front camera can also go solo to shoot in even more detailed 1440p.
The best dash cams can have anything from one, two or even three cameras. Single camera dash cams record the outside view from your windscreen, dual dash cams add an inside facing camera which is especially useful for ride-share drivers like taxis, while three-camera dash cams are more for professional drivers clocking up the miles, adding an additional viewpoint from the outside of the vehicle, being especially handy for trucks.
Video resolution is a key consideration when choosing a dash cam, as explained below. Most dash cams in 2023 can record footage in Full HD, which is generally enough to identify large details in a video clip. Sharper 4K dash cams tend to cost more, but capture video at a higher resolution. This makes it easier to pull out crucial details such as number plates, which mean 4K dash cams generally offer more reassurance from a safety perspective.
Video frame rates are also important. Most dash cams can record at 30fps, which is fine for reviewing footage at full speed, but the best models can capture action at a smoother 60fps. This gives you the option to slow down footage after the fact, making it easier to pick up details at half-speed. It’s increasingly common to find dash cams that can capture Full HD video at 60fps, but 4K 60fps recording is generally reserved for professional dash cams. Certain models, such as the Nextbase 622GW, also feature digital image stabilization, which helps to smooth out footage on bumpy roads.
There are also a wealth of features to consider, including a touchscreen display, a companion app, GPS, parking surveillance, voice control, sat nav and what3words to name but a few.
If you drive an older car without an infotainment system, then it’s worth getting a dash cam with sat nav, but there’s little point if your car already has sat nav. GPS location data is handy when recording incidents, while some even include what3words geolocation data which offers the most precise way of pinpointing your vehicle and super handy should you become stricken.
Almost all dash cams can be fitted to a car without making any permanent modification. They can be attached to a windscreen using a suction cup or removable adhesive pad, and powered from either a USB port or the 12V socket found in almost all cars. To test dash cams, we install them in our car, attach them to the windscreen and pair them to our smartphone as if we had bought them ourselves.
We then drive, at day and night, to test how the camera handles different lighting conditions, and ideally in varied weather too. The footage is then transferred to a smartphone or computer and viewed to check out the quality, and whether key details like vehicle registration plates are legible. This also gives us an opportunity to see how easy (or difficult) it is to view, transfer and save recorded footage.
Additional features like voice assistants and drive assistance systems get a thorough workout, and we also test how easy it is to turn these off, or adjust them to our personal preferences. Being able to quickly and easily turn audio recording off is key, especially when carrying passengers who don’t want their conversations recorded.
Naturally, we cannot crash a vehicle to see how well the dash cam detects collisions. Instead, firmly tapping the dash cam can simulate a collision, allowing us to see what happens when footage is saved. It is also possible to trigger a recording by powering the camera with a portable battery and tapping it against our desk. It may seem rudimentary, but it works and keeps our insurance company happy.
Sometimes, driving over a particularly aggressive speed bump can trigger a dash cam recording. In these cases we learn that the camera’s g-sensor is too sensitive, and needs to be adjusted, providing the menu system includes such an option.
Installing your dash cam:
So you've managed to track down the best dash cam for your needs but you're not sure how to get started on installing it? We've put together a handy guide on how to fit a dash cam to help make installation a breeze.
Is it worth having a dash cam?
Yes, absolutely. Some drivers might not like the idea of a camera constantly monitoring their vehicle, but some models of dash cam (like the Garmin Dash Cam Mini 2) are compact and subtle enough to almost disappear behind the central rear view mirror. In other words, once they are up and running you soon forget about them; they just sit there, quietly keeping an eye out, ready to save footage if you’re caught in an accident.
If the worst were to happen, and you need to prove your innocence, the dash cam’s footage could do exactly that. This footage can also be set to include your speed, along with the time and date to help reinforce the validity of your evidence if blame is disputed by the other party. As well as monitoring the road ahead, dash cams can be used to record out of the rear windscreen too, recording footage if you are unfortunate enough to be rear-ended by another motorist.
Even if you're not directly involved in a collision, dash cams can still be useful. Many have a button for quickly saving the last few seconds of video. That way, if you need something ahead that you think needs recording, like an accident between two other vehicles, an example of bad driving, or something else noteworthy, you can provide that evidence to whoever might need it. Another useful feature of some dash cams is how they will call the emergency services if a crash is detected and you don’t respond; they can then share your location and ensure help is on its way.
Furthermore, many dash cams come with additional driver assistance features. These include alerts to let you know about nearby speed cameras and red light cameras, plus a prompt to tell you the vehicle in front has set off – just in case you weren’t paying attention while sat in traffic.
Lastly, some dash cams can be permanently installed in your vehicle and hard-wired, giving them a constant power supply. When paired with a special parking mode, this enables the camera to sense impacts and record footage while the car is parked, potentially gathering useful evidence of your neighbour’s poor parking.
Reformatted buying guide to make buying decisions clearer and easier. Removed some old products and replaced with new, more up-to-date models.
Get the hottest deals available in your inbox plus news, reviews, opinion, analysis, deals and more from the TechRadar team.
Mark is TechRadar's Senior news editor. Having worked in tech journalism for a ludicrous 17 years, Mark is now attempting to break the world record for the number of camera bags hoarded by one person. He was previously Cameras Editor at Trusted Reviews, Acting editor on Stuff.tv, as well as Features editor and Reviews editor on Stuff magazine. As a freelancer, he's contributed to titles including The Sunday Times, FourFourTwo and Arena. And in a former life, he also won The Daily Telegraph's Young Sportswriter of the Year. But that was before he discovered the strange joys of getting up at 4am for a photo shoot in London's Square Mile.
| english |
.rightCol{
background-color: white;
text-align: left;
width: 410px;
height: 626.5px;
float: left;
margin-top: 20px;
}
.regText{
font-size: 36px;
font-weight: 600;
margin-bottom: 5px;
}
.placeHolder{
font-size: 18px;
padding: 8px 10px;
border-radius: 5px;
width: 396px;
margin-bottom: 10px;
border: 1px solid grey;
}
.placeHolder2{
font-size: 18px;
padding: 8px 10px;
border-radius: 5px;
float:left;
margin-bottom: 10px;
margin-right: 8px;
width: 194px;
border: 1px solid grey;
}
.bgText{
font-size: 16px;
font-weight: bold;
margin: 10px 0 5px 0;
color: grey;
}
.birthdayDate{
font-size: 13px;
height: 30px;
padding: 5px;
border: 1px solid grey;
}
.genderText{
font-size: 18px;
font-weight: normal;
line-height: 18px;
padding: 0 10px 0 3px;
}
.note{
color:grey;
font-size: 11px;
width: 316px;
margin: 11px 0 11px 0;
}
.submitButton{
background-color: green;
color: white;
font-size: 19px;
font-weight: bold;
min-width: 194px;
padding: 7px 20px;
text-align: center;
border-radius: 5px;
}
.note2{
color: #666;
font-size: 13px;
font-weight: bold;
margin-top: 10px;
padding-top: 15px;
}
| css |
TCS and Infosys will release their Q3 earnings today amid expectations of weak numbers. However, bulls believe that this could be the last dull quarter before revenue growth picks up in Q4. Analysts anticipate TCS to deliver 1.5% QoQ cc revenue growth driven by large deal ramp-ups, with margin expansion driven by operational efficiencies. On the other hand, Infosys is expected to see a decline in USD revenue due to seasonal furloughs and the impact of a slowdown in discretionary tech spending.
Are you looking for a steel almirah that is durable and stylish at the same time? Check out our article where we have curated a list of the best steel almirah online in India. These almirahs are bound to be durable enough to withstand the effects of time while adding style to your interior.
He will spearhead the company’s end-to-end procurement, covering sourcing, spend management and strategic management of cost levers of the supply chain. “This is a new role and structure being created by Ecom Express in building a stronger P&L focus and driving a clear cost management agenda”, the company said.
Of the total bonds outstanding, the Central government bonds (G-secs) stood at USD 1.34 trillion as of the second quarter (Q2), up from USD 1.06 trillion in the March 2023 quarter, Vishal Goenka, cofounder of Indiabonds.com, said quoting the Reserve Bank data.
The consumer technology market in India will likely touch $300 billion in the next four years, while online retail could top $160 billion by 2028, reports suggest. This and more in today’s ETtech Morning Dispatch.
Max Healthcare Institute, NHPC, Bharat Forge, Hindustan Petroleum Corporation, Gland Pharma, Emami, Exide Industries, R R Kabel, Bajaj Electricals, Bikaji Foods International, Sun Pharma Advanced Research, Sobha Ltd, Ujjivan Financial Services, Quess Corp, Fusion MicroFinance, Gateway Distriparks, Heidelberg Cement, Va Tech Wabag, and Paras Defence and Space Technologies are among the other major companies releasing results on Monday.
For the six-month period ended September 30, 2023 the profit grew to Rs 59.62 crore, from Rs 35.29 crore registered during the corresponding period of last year, the Chennai-headquartered company said in a press release here.
As the earnings near the fag end, a plethora of companies are scheduled to release their second-quarter earnings in the week ahead. Divi’s Laboratories, FSN E-Commerce, Power Grid Corp of India, Adani Ports and SEZ, Apollo Hospitals, Bosch, Zee Entertainment Enterprises, LIC, Oil and Natural Gas Corp, Coal India, M&M, HAL, and Hindalco Industries are among the big names to watch out for.
Discover the ideal blend of affordability and functionality with our curated selection of the 8 Best Floor Lamps Under 1500 in India. From sleek designs to adjustable features, these budget-friendly lamps provide the perfect lighting solution for various spaces, ensuring a well-lit and aesthetically pleasing environment without breaking the bank.
Yatra online said it recorded the strongest quarter on the air front since the advent of Covid-19 with the highest number of air passengers booked since pre-Covid in December 2019, up 41.5% YoY far outpacing domestic air passenger industry growth of 14.8% YoY.
Its total income stood at Rs 60.2 crore in the June quarter of the current financial year, down 13.2% YoY as compared to Rs 64.9 crore in the same quarter of the previous financial year. On a QoQ basis, total income fell 51.7% to Rs 60.2 crore in Q1 FY24 from Rs 12.46 crore in Q4 FY23.
Major firms such as Maruti Suzuki, Titan, Bharti Airtel, and M&M among others will release their June quarter results this week. The report cards of other companies such as GAIL, Power Grid, Sun Pharma, and Adani Enterprises are also expected to be out. This week, Maruti Suzuki, Adani Enterprises, Adani Transmission, Titan, IndiGo, Bharti Airtel, Sun Pharma, SBI, M&M, Varun Beverages, and Bank of Baroda are scheduled to announce their quarterly performance.
Total income for the quarter was Rs 3,125 crore, up 25% y-o-y, aided by growth in asset book, a release by the NBFC said. Disbursements at Rs 12,165 crore were higher by 28% y-o-y. Loan book increased sequentially by 4.8% to Rs 86,732 crore.
Revenues from the India business rose by a marginal 3% to Rs 34,901 crore for the quarter under review. India's EBITDA was at Rs 7,514 crore, a decline of 22%. This translates into EBITDA per ton of Rs 15,651 and a margin of 22%.
It was after almost six quarters that quarterly results of TCS were able to help in improving the sentiment toward the IT stocks. But that was short lived as the result and guidance of Infy reversed the sentiment toward bears. But is the worst over or it is still to continue. Stock Reports Plus, powered by Refinitiv, is a comprehensive research report that evaluates five key components of 4,000+ listed stocks - earnings, fundamentals, relative valuation, risk and price momentum to generate standardized scores.
| english |
KIGALI, Rwanda, Nov 30 – Rwanda opened a formal probe into 20 French officials suspected of playing a role in the 1994 genocide, in a move likely to further sour diplomatic ties with France.
Kigali has long accused France of complicity in the genocide of some 800,000 mostly ethnic Tutsis, at the hands of Hutu extremists, angering Paris and straining relations.
“The inquiry, for now, is focused on 20 individuals whom, according to information gathered so far, are required by the prosecution authority to explain or provide clarity on allegations against them,” said prosecutor general Richard Muhumuza in a statement.
This will enable prosecutors to decide “whether the concerned individuals should be formally charged or not”.
Muhumuza said the relevant French authorities had been contacted and that full cooperation was expected.
The dispute centres on France’s role prior to the genocide as a close ally of the Hutu nationalist regime of Juvenal Habyarimana. The shooting down of his plane over Kigali on April 6, 1994 was the event that triggered 100 days of meticulously planned slaughter.
France is accused of missing or ignoring the warning signs, and of training the soldiers and militiamen who carried out the killings.
And when the genocide was in full swing, it was further accused of using its diplomatic clout to stall effective action.
When it did finally send in troops — in Operation Turquoise — it was accused of only doing so to counter the advance of the Tutsi rebels of Paul Kagame, who is now president, allowing the perpetrators to escape to neighbouring Zaire which later became the Democratic Republic of Congo.
France says its soldiers were only deployed after most of the killing had happened and that their presence helped save thousands of lives.
And French officials insist that any guilt for failing to prevent the genocide is shared by the entire international community, accusing Kagame of raising the issue in a bid to distract attention from what they say is his own poor human rights record.
When contacted by AFP, the French defence ministry referred to a statement issued on November 16 by Defence Minister Jean-Yves Le Drian after a list of 22 high-ranking French officers suspected of involvement was first published by Kigali.
There was “nothing new” in the list, Le Drian said at the time, adding that it was “almost identical to one produced in 2008” and “to say that the French army took park in genocide was a disgraceful lie”.
Although relations between Kigali and Paris were completely frozen from 2006 to 2009, they were on the mend until 2014 when Kagame repeated accusations that French soldiers were both accomplices and “actors” in the bloodbath.
The remarks prompted France to pull out of ceremonies to mark the 20th anniversary of the massacres.
The stormy relationship took an even worse turn when the French judiciary decided in October to re-open an investigation into the shooting down of Habyarimana’s plane to hear testimony from a former general who accuses Kagame of being behind the assassination.
“It is France which should be in the dock,” said Kagame.
In apparent retaliation, Rwanda on November 1 published the list of 22 senior French officers it accuses of being involved in the genocide. | english |
Faguaa Bital song is a Bhojpuri pop song from the Balma Baagi Baliya Vala released on 2016. Music of Faguaa Bital song is composed by Shiv Manmohi. Faguaa Bital was sung by Manoj Lehari. Download Faguaa Bital song from Balma Baagi Baliya Vala on Raaga.com.
| english |
package com.zw.talkback.domain.lyxj.tsm3;
import lombok.Data;
import java.util.ArrayList;
import java.util.List;
/**
* tsm分页返回的数据结果
*/
@Data
class DataRecords<T> {
private List<T> records = new ArrayList<>();
}
| java |
/* jshint node: true */
/* globals describe, it */
"use strict";
var _ = require('lodash'),
expect = require('expect.js');
var samples = require('./_samples'),
transform = require('../transform');
describe('integration test: missingLastSgv', function() {
var sample = samples.missingLastSgv;
var transformed = transform(sample);
var pumpStatuses = transformed.devicestatus;
var sgvs = transformed.entries;
it('should set the pump status time based on the "last device data update time"', function() {
expect(pumpStatuses[0]['created_at']).to.equal(new Date(sample['lastMedicalDeviceDataUpdateServerTime']).toISOString());
});
it('should have one pump_status and 5 sgv entries', function() {
expect(pumpStatuses.length).to.be(1);
expect(sgvs.length).to.be(5);
});
it('should pull the right sgvs', function() {
expect(_.map(sgvs, 'sgv')).to.eql([70, 69, 68, 65, 66]);
});
it('should correctly deduce that the pump time offset is -0700', function() {
expect(_.map(sgvs, 'date')).to.eql(
[1445266500000, 1445266800000, 1445267100000, 1445267400000, 1445267700000]
);
});
it('should not include a trend for any sgv', function() {
expect(_.uniq(_.map(sgvs, 'direction'))).to.eql([undefined]);
});
it('should include pump status data, including active insulin', function() {
_.forEach({
'uploader.battery': 29,
'pump.battery.percent': 75,
'pump.reservoir': 60,
'pump.iob.bolusiob': 4.85,
'pump.iob.timestamp': new Date(sample['lastMedicalDeviceDataUpdateServerTime']).toISOString(),
'connect.calibStatus': 'LESS_THAN_TWELVE_HRS',
'connect.conduitInRange': true,
'connect.conduitMedicalDeviceInRange': true,
'connect.conduitSensorInRange': true,
'connect.sensorDurationHours': 73,
'connect.sensorState': 'NORMAL',
'connect.timeToNextCalibHours': 10,
'device': 'connect://paradigm',
}, function(val, key) {
expect(_.get(pumpStatuses[0], key)).to.be(val);
});
});
});
describe('integration test: withTrend', function() {
var sample = samples.withTrend;
var transformed = transform(sample);
var pumpStatuses = transformed.devicestatus;
var sgvs = transformed.entries;
it('should have one pump_status and 6 sgv entries', function() {
expect(pumpStatuses.length).to.be(1);
expect(sgvs.length).to.be(6);
});
it('should pull the right sgvs', function() {
expect(_.map(sgvs, 'sgv')).to.eql([191, 185, 179, 175, 168, 163]);
});
it('should correctly deduce that the pump time offset is -0500', function() {
expect(_.map(sgvs, 'date')).to.eql(
[1445365260000, 1445365560000, 1445365860000, 1445366160000, 1445366460000, 1445366760000]
);
});
it('should include a SingleDown direction/trend on the last sgv', function() {
function assertCount(key, value, count) {
expect(sgvs.filter(function(s) { return s[key] === value; }).length).to.be(count);
}
assertCount('direction', undefined, 5);
assertCount('trend', undefined, 5);
assertCount('direction', 'SingleDown', 1);
assertCount('trend', 6, 1);
expect(sgvs[sgvs.length - 1]['direction']).to.be('SingleDown');
expect(sgvs[sgvs.length - 1]['trend']).to.be(6);
});
it('should include pump status data, including active insulin', function() {
_.forEach({
'uploader.battery': 86,
'pump.battery.percent': 50,
'pump.reservoir': 67,
'pump.iob.bolusiob': 1.35,
'pump.iob.timestamp': new Date(sample['lastMedicalDeviceDataUpdateServerTime']).toISOString(),
'connect.calibStatus': 'LESS_THAN_NINE_HRS',
'connect.conduitInRange': true,
'connect.conduitMedicalDeviceInRange': true,
'connect.conduitSensorInRange': true,
'connect.sensorDurationHours': 137,
'connect.sensorState': 'NORMAL',
'connect.timeToNextCalibHours': 6,
'device': 'connect://paradigm',
}, function(val, key) {
expect(_.get(pumpStatuses[0], key)).to.be(val);
});
});
});
| javascript |
<filename>GPM/MathTestProject/src/main.cpp<gh_stars>0
#include <iostream>
#include <GPM/GPM.h>
int main(const int p_argc, char** p_argv)
{
try
{
// Try stuff here
Matrix4F rot = GPM::Matrix4F::LookAt(GPM::Vector3F(0, 0, 0), GPM::Vector3F(0, 0, 3), GPM::Vector3F(0, 1, 0));
std::cout << rot;
}
catch (const std::exception & e)
{
std::cerr << "Exception thrown: " << e.what() << '\n';
return EXIT_FAILURE;
}
return EXIT_SUCCESS;
} | cpp |
Reigning ONE flyweight world champion ‘Mighty Mouse’ Demetrious Johnson put together another incredible performance, when he closed the book on his trilogy with former flyweight king ‘Mikinho’ Adriano Moraes last week.
Johnson went to war with Moraes in the main event of ONE Fight Night 10, which took place live from the sold-out 1stBank Center in Broomfield, Colorado on May 5. After five rounds of action, the fight went to the judges’ scorecards, with Johnson taking home a unanimous decision to retain his gold.
Speaking to the media in the post-event press conference, the topic touched on Johnson seeing himself in the GOAT conversation, and what his list looks like.
‘Mighty Mouse’ said:
While Demetrious Johnson believes these GOAT lists are highly subjective and based on who you’re asking, he does indeed see himself among the all-time greats. He added:
Check Demetrious Johnson's comments from 37:00 onwards:
Johnson is taking a little break right now, in order to re-evaluate his career and decide if he wants to continue or to simply hang up his gloves for good.
Meanwhile, fans in the United States and Canada can watch Johnson’s dominant performance against Moraes and the rest of the stacked ONE Fight Night 10: Johnson vs Moraes III card absolutely free via replay on Amazon Prime. | english |
{"9nzzwv": " CA Mendocino", "9nzzpm": " CA Mendocino", "9nzzps": " CA Mendocino", "9nzzpt": " CA Mendocino", "9nzyzw": " CA Mendocino", "9nzzrj": " CA Mendocino", "9nzzwf": " CA Mendocino", "9nzzpe": " CA Mendocino", "9nzzp3": " CA Mendocino", "9nzzr4": " CA Mendocino", "9nzzp5": " CA Mendocino", "9nzzym": " CA Mendocino", "9nzzyr": " CA Mendocino", "9nzzqb": " CA Mendocino", "9nzzpr": " CA Mendocino", "9nzzwg": " CA Mendocino", "9nzzwe": " CA Mendocino", "9nzzrh": " CA Mendocino", "9nzzpd": " CA Mendocino", "9nzzwc": " CA Mendocino", "9nzyzx": " CA Mendocino", "9nzzye": " CA Mendocino", "9nzzy5": " CA Mendocino", "9nzzwx": " CA Mendocino", "9nzzz2": " CA Mendocino", "9nzyzg": " CA Mendocino", "9nzzp6": " CA Mendocino", "9nzzr5": " CA Mendocino", "9nzzr1": " CA Mendocino", "9nzztu": " CA Mendocino", "9nzzp7": " CA Mendocino", "9nzzrg": " CA Mendocino", "9nzzrm": " CA Mendocino", "9nzzyt": " CA Mendocino", "9nzzrq": " CA Mendocino", "9nzzz6": " CA Mendocino", "9nzzyx": " CA Mendocino", "9nzzyw": " CA Mendocino", "9nzzwk": " CA Mendocino", "9nzzxh": " CA Mendocino", "9nzzwu": " CA Mendocino", "9nzzpu": " CA Mendocino", "9nzzyv": " CA Mendocino", "9nzzrk": " CA Mendocino"} | json |
As the final hours tick down until the latest Apple product launch and the faithful congregate on the venue in the US like disciples preparing to take and spread the holy flame, I find myself overwhelmed by an emotion that is unusual for gadget-obsessed types in the leadup to an Apple launch: apathy.
Don't get me wrong: I'm sure whatever Apple rolls out will be shiny, gorgeous, and irresistible. But it will no longer be relevant for me, as months ago I gave up on waiting for Apple to launch a large-screen iPhone to replace my venerable iPhone 4S. Like many, I made the shift to Samsung's Galaxy S5 – and I am happy enough with it that I will not feel like I've missed out when Apple rolls out its own large-screen phones.
This, of course, is because Android is no longer the compromise it used to be. It wasn't too long ago that decisions to buy an Android phone instead of an iPhone were based on either cost reasons, or a willingness to endure the endless depredations of incompatibility for the sake of sticking it to the man ('the man' in this sentence being Steve Jobs, of course).
Things have changed a fair bit, and the S5 has become both extremely popular and extremely capable. Android apps are almost universally delivered at the same time or even earlier than the iOS equivalents, often with additional functionality possible because Android isn't locked down tighter than, well, something that's really tight.
Sure, there's a learning curve – a friend who more recently made the switch was going through it too, as you do – but there comes a point when you come to grips with the way Android works and think “this really is better”.
For me, that point came early on, when the tight links between Android and Dropbox not only saw my own account boosted to 50GB, but saw all kinds of apps seamlessly accessing it. Now, I understand that iCloud does the same thing in the iOS world, and that there are many other apps improving integration with cloud services – but try accessing your iCloud content from an Android device and you'll see where that model falls down.
Indeed, throughout the course of using an Android device you steadily come to realise that it's the little things that make it a superior mobile experience: the much-improved keyboard, for example, which learns from your typing and predicts your next words with uncanny accuracy, or the amazingly useful widgets that make Apple's icons-R-Us interface seem positively archaic.
That Apple is now playing catchup on these and other points, confirms that even its own engineers realised they had remained too inflexible, for too long (this was confirmed in a leaked internal presentation showing they are getting caned in the market as many others defect to phones with larger screens).
There are the inevitable quirks, to be fair: for example Samsung, for reasons I cannot even begin to comprehend, does not offer a way to take photos from the lock screen without unlocking the phone (this feature is easily restored using one of many alternative lock screens). The ability to control screen brightness has inexplicably disappeared from the swipe-down menu. Typing was strangely laggy until an over-the-air update resolved the issue. The inability to turn off the camera's shutter sound is an unfortunate blight on an otherwise excellent camera application.
Like many, I worried about how I could possibly bring over long-entrenched Apple things like my iTunes music or iPhoto library; simple dragging and dropping, or the use of any number of programs to automate the process, do away with that completely. Notes sync seamlessly between the S5 and my Mac computers using Google Keep, although Evernote or many other options are also popular.
On the whole, the S5 offers an iPhone alternative experience that has become comfortable and effective. More importantly, the gorgeous screen is large enough that I no longer feel like I'm trying to paint a triptych onto a grain of rice every time I squint at the iPhone's screen.
Apple may offer these features and more in its big update, but the damage is already done: I, and the many other people that have made the same jump, am now locked into the S5 for the remainder of my two-year telco contract. That means there will be no iPhone 6 for me, nor even perhaps the iPhone 7 with its inevitable brain-control features or whatever.
Not even Apple's iWatch, which I may view with a momentary pang of lust, will matter that much in the end: Android-supporting competitors like Motorola, LG and Samsung have finally gotten their acts together to the point where they are shipping high-quality things to stick on one's wrist. If I go down that path, I know there will be a variety of devices that will work quite nicely with my S5 – all except, I suspect, Apple's iWatch, or whatever it's called.
Not only is today's Apple launch the first one I will watch with detached curiosity more than wanton lust, but it is probably the first one in which Apple is blatantly playing catch-up with the rest of the industry. I'm sure it will be very successful, but the collateral damage from Apple's decision not to release a large-screen iPhone 5S will continue for years – until the contracts of those who defected to Samsung, HTC or other alternatives find themselves deciding whether it's worth going back to the Apple fold. For many, the answer will be 'no'.
What do you think? Has Apple reclaimed the innovation lead it risked losing? What do you think of its new products? Or have the Android vendors finally figured out how to keep their lead?
| english |
{
"id": 154454,
"name": "<NAME>",
"description": "Less Clutter, cleaner.\r\nInoreader's \"Expanded View\" also has a colour ring around focused articles - instead of a left hand strip of colour.",
"user": {
"id": 586367,
"name": "<NAME>",
"email": "redacted",
"paypal_email": null,
"homepage": null,
"about": null,
"license": "publicdomain"
},
"updated": "2018-04-19T11:15:17.000Z",
"weekly_install_count": 0,
"total_install_count": 298,
"rating": null,
"after_screenshot_name": "https://userstyles.org/style_screenshots/154454_after.png?r=1604650040",
"obsoleting_style_id": null,
"obsoleting_style_name": null,
"obsolete": 0,
"admin_delete_reason_id": null,
"obsoletion_message": null,
"screenshots": [
"https://userstyles.org/style_screenshots/154454_additional_29351.png?r=1604650040"
],
"license": "publicdomain",
"created": "2018-01-20T00:39:47.000Z",
"category": "site",
"raw_subcategory": "inoreader",
"subcategory": "inoreader",
"additional_info": "cuts out too much clutter. This version does not change font, but used with some other styles of mine, you can have choice of fonts.",
"style_tags": [],
"css": "@-moz-document domain(inoreader.com) {\r\n\r\n.article_title {font-size: 20px;}\r\n.article_header_title {font-size: 16px;}\r\n.article_magazine_title_link {font-size: 19px; font-weight: normal;}\r\n.article_header_text {overflow: visible;} \r\n\r\n.article_author { display: none !important; }\r\n.article_tile_footer_feed_title {display: none;}\r\n.article_feed_title {display: none;}\r\n.ar {border-bottom: none; padding: 12px 7px 12px 12px;}\r\n\r\n#subscriptions_buttons, #sb_rp_tools, #sb_tp_search_overlay {background-color: #5986c3; background-image: none; } \r\n#feed_searcher {background-color: azure; box-shadow:none; }\r\n#sb_rp_heading, .inno_toolbar_button {text-shadow: none;} \r\n#sb_rp_subscribe, #sb_rp_clip_page, #sb_rp_create_as, #sb_rp_add_team {background-color: #2726263b; text-shadow:none}\r\n#feed_searcher {color: black !important;}\r\n\r\n.article_title_expanded_view {top: 12px;}\r\n.article_magazine_feed_title {color: #bdc1c7;}\r\n.article_unreaded .article_magazine_feed_title {color: #bdc1c7;}\r\n.icon-radio_checked_big {color: #d5d7d9;}\r\n.article_magazine_footer {top: 6px;position: relative;margin-bottom: 29px;}\r\n\r\n.tree_unreaded, .unread_cnt {color: #05a !important;} \r\n.tree_unreaded {font-weight: normal;} \r\n\r\n.icon-tag_full:before {font-size: 14px;} \r\n.icon-tag_full:before {color: #8e9195;}\r\n\r\n#tree .feed_favicon {width: 13px; height: 13px;}\r\n\r\n.article_magazine_footer .article_magazine_feed_title {display: none;} \r\ndiv.article_magazine { max-width: 900px;} \r\ndiv.article_card.article_current {border: 4px solid rgba(49, 97, 196, 0.6);border-radius: 12px;}\r\ndiv.article_expanded {border: 4px solid rgba(49, 97, 196, 0); border-radius: 12px;}\r\n\r\n.article_stripe {width: 99.3%; z-index: -1; background-color: #ffffff;}\r\n\r\ndiv.article_expanded {background-color: #2e415c00;}\r\n.article_full_contents {background-color: #2e415c00 !important;}\r\n.reader_pane_view_style_1 .article_footer, .reader_pane_view_style_1 .arrow_div {background-color: #d4d40a00 !important;}\r\n\r\n#preferences_wrapper .tr2 {background-color: #bbcfe7;}\r\n\r\n.article_header_title {font-weight: normal !important;}\r\n.article_title_link {font-weight: normal;}\r\n.article_card .article_title, #article_dialog .article_title {padding-bottom: 20px; padding-top: 22px;}\r\n.article_card .article_sub_title {position: absolute;}\r\n}",
"discussions": [
{
"id": 69062,
"name": "<NAME> Align Left",
"rating": 2,
"created": "2018-08-30T12:31:25.000Z",
"author_name": "AlPennyworth",
"author_id": 597748
}
],
"discussionsCount": 1,
"commentsCount": 1,
"userjs_url": "/styles/userjs/154454/inoreader-nicer-cleaner.user.js",
"style_settings": []
} | json |
{
"vorgangId": "259571",
"VORGANG": {
"WAHLPERIODE": "19",
"VORGANGSTYP": "Mündliche Frage",
"TITEL": "Gewalttaten und sexuelle Übergriffe in öffentlichen Verkehrsmitteln und Privatfahrzeugen im Jahr 2019",
"AKTUELLER_STAND": "Beantwortet",
"SIGNATUR": "",
"GESTA_ORDNUNGSNUMMER": "",
"PLENUM": {
"PLPR_KLARTEXT": "Mündliche Frage/Schriftliche Antwort",
"PLPR_HERAUSGEBER": "BT",
"PLPR_NUMMER": "19/148",
"PLPR_SEITEN": "18528A - 18528B",
"PLPR_LINK": "http://dipbt.bundestag.de:80/dip21/btp/19/19148.pdf#P.18528"
},
"EU_DOK_NR": "",
"SCHLAGWORT": [
{
"_fundstelle": "true",
"__cdata": "Gewaltkriminalität"
},
"Kraftfahrzeug",
"Öffentlicher Personennahverkehr",
"Öffentlicher Raum",
{
"_fundstelle": "true",
"__cdata": "Sexualdelikt"
}
],
"ABSTRAKT": "Originaltext der Frage(n): \r\n \r\nWie viele Gewalttaten und sexualisierte Übergriffe fanden nach Kenntnis der Bundesregierung im Jahr 2019 (wenn Zahlen von 2019 noch nicht vollständig vorliegen, bitte von 2018) in öffentlichen Verkehrsmitteln (inklusive deren Haltestellenbereichen), in Autos (inklusive Parkplätzen und im öffentlichen Straßenland; z. B. gegenüber Fußgängerinnen und Fußgängern, Fahrradfahrerinnen und Fahrradfahrer) statt (bitte nach o. g. Deliktart und Geschlecht der Gewaltausübenden bzw. derer, die Gewalt ausgesetzt waren aufschlüsseln)?"
},
"VORGANGSABLAUF": {
"VORGANGSPOSITION": [
{
"ZUORDNUNG": "BT",
"URHEBER": "Mündliche Frage ",
"FUNDSTELLE": "28.02.2020 - BT-Drucksache 19/17408, Nr. 27",
"FUNDSTELLE_LINK": "http://dipbt.bundestag.de:80/dip21/btd/19/174/1917408.pdf"
},
{
"ZUORDNUNG": "BT",
"URHEBER": "Mündliche Frage/Schriftliche Antwort",
"FUNDSTELLE": "04.03.2020 - BT-Plenarprotokoll 19/148, S. 18528A - 18528B",
"FUNDSTELLE_LINK": "http://dipbt.bundestag.de:80/dip21/btp/19/19148.pdf#P.18528",
"PERSOENLICHER_URHEBER": [
{
"VORNAME": "Stefan",
"NACHNAME": "Gelbhaar",
"FUNKTION": "MdB",
"FRAKTION": "BÜNDNIS 90/DIE GRÜNEN",
"AKTIVITAETSART": "Frage",
"SEITE": "18528A"
},
{
"PERSON_TITEL": "Prof. Dr.",
"VORNAME": "Günter",
"NACHNAME": "Krings",
"FUNKTION": "Parl. Staatssekr.",
"RESSORT": "Bundesministerium des Innern, für Bau und Heimat",
"AKTIVITAETSART": "Antwort",
"SEITE": "18528A"
}
]
}
]
}
}
| json |
new duties as a design academic.
*This involves carting around a host of possessions in a car,
exquisite spiming accuracy.
*One BEYOND THE BEYOND reader, however,
relationship to the material world.
"Canadian Guy" writes in:
I stumbled into your blog circa your arrival in europe.
You have no time so I'm going to be as telegraphic as possible.
I fugue regularly. I "come-to" with an intense connection to the arrangements of objects left in my wake.
One is "The Difference Engine" which I haven't read since my Mom gave it to me for Christmas in 1991. The other is "Magic" by this Welsh bloke named David Conway; I've had it always sort of kicking around since I was 8 or 9.
I'm going to itemize the rest of the image. I've felt the need to do this before. Consider it an experiment in explanation. My goal is to get you interested in receiving some text I've been putting together since October which nobody knows what to make of.
Please note that nothing has been planted for the benefit of this email. I'm winging it; this is me doing a mental inventory of a sort.
Ok, here goes...generally from the bottom to the top, left to right. Most of the little things are buried.
1 garbage bag full of laundry (in corner)
(left side of piano)
1 small broken clock (five past twelve)
(right side of piano)
(on top of piano, to the left)
1 DVD: The Corrs Unplugged"
1 CD: Frances Cabrel "Les Beaux DÈg‚ts" ("the beautiful small ugly-things/garbage" roughly translated)
1 CD cover: The Corrs "Borrowed Heaven"
1 loose CD: Billie Holiday "Loveless Love"
1 black pencil marked "HÙtels Fairmont"
(right of keyboard)
1 Acer Ferrari mouse (for the Acer Athlon 64 Ferrari laptop I wasted $3000 on last summer)
1 loose DVD with no-budget 16mm film I helped clean up, entitled "Dinner for Ten"
2 books (already mentioned)
(behind display)
Everything else is just cabling or atomic/digital/informational content.
I think there's an important lesson here somewhere but I already have enough topics to cover.
If nothing else, I wanted to say this to you: I think we're collectively developing a new "mathematics of real symbols".
Math is not "beautiful" as the mathematicians like to believe. If it was, we'd be living in Utopia. Math is a language. It's up to us to language the beautiful into the real.
(I'm still trying to say what I wanted to say...)
The language of your blog has changed significantly since your return.
If you do nothing but stop, I swear I could write a book about your last two posts. It's nice to be back home, I'm sure. But I'm scared about the mental health effects of your environment. Essentially, I think I may be volunteering as a sort of logical analyst. I'd say "psycho-logical" but the term is so loaded, you know?
Your life isn't perfect. I'm not trying to play games with you. I'm responding from the gut to your mentioning of Diderot.
Feel free to phone me @ (phone number). I don't usually answer but you may be lucky and, in any case, I'll phone you back. I'm normal, etc.
I'm from (Canadian City), my name is (Canadian Guy). I'm a part-time veteran computer programmer and full-time musical signal receiver.
With this new G5, I feel I'm entering into a mystic-like state. I'm seriously over-empowered and I think many people are going to lose their grip when the next wave of technology hits. We've all been talking about it for years in other ways – you know: "a new level of human consciousness" and all sorts of esoteric cosmological significance, the usual. Only this time, it's programmatic.
I'm just trying to pay my rent. I'm a non-writer stuck in a writerly corner. You have no time, I know, so I don't want to inundate you with opinions right off the bat.
All this == FYI.
I have no ulterior motives other than the satiation of general curiosity. And most of it doesn't concern you at all.
For example, the biggest question I'm looking to answer right now is: "Why the heck did I put the supercomputer on the piano?"
Best regards,
| english |
import {expect, spy, stub, restore} from '../test';
import NativeObject from '../../src/tabris/NativeObject';
import ClientInterface from '../../src/tabris/Tabris';
import ClientStub from './ClientStub';
import '../../src/tabris/Tabris';
describe('ClientInterface', function() {
let tabris;
let client;
beforeEach(function() {
tabris = global.tabris = new ClientInterface();
client = new ClientStub();
});
afterEach(restore);
describe('_init', function() {
it('can be called without a context', function() {
expect(() => {
tabris._init.call(null, client);
}).to.not.throw();
});
it('triggers start event', function() {
let listener = spy();
tabris.on('start', listener);
tabris._init.call(null, client);
expect(listener).to.have.been.calledOnce;
});
it('triggers start event when tabris is set up', function() {
class TestType extends NativeObject {}
tabris.on('start', () => new TestType()._create('test.Type'));
tabris._init.call(null, client);
expect(client.calls({op: 'create', type: 'test.Type'}).length).to.equal(1);
});
});
describe('_notify', function() {
class TestType extends NativeObject {}
let widget;
beforeEach(function() {
tabris._init(client);
widget = new TestType();
});
it('notifies widget', function() {
spy(widget, '_trigger');
tabris._notify(widget.cid, 'foo', {bar: 23});
expect(widget._trigger).to.have.been.calledWith('foo', {bar: 23});
});
it('returns return value from widget', function() {
stub(widget, '_trigger').callsFake(() => 'result');
let result = tabris._notify(widget.cid, 'foo');
expect(result).to.equal('result');
});
it('skips events for already disposed widgets', function() {
widget.dispose();
spy(widget, '_trigger');
tabris._notify(widget.cid, 'foo', {bar: 23});
expect(widget._trigger).to.have.not.been.called;
});
it('silently ignores events for non-existing ids (does not crash)', function() {
expect(() => {
tabris._notify('no-id', 'foo', {bar: 23});
}).to.not.throw();
});
it('can be called without a context', function() {
spy(widget, '_trigger');
tabris._notify.call(null, widget.cid, 'foo', [23, 42]);
expect(widget._trigger).to.have.been.calledWith('foo', [23, 42]);
});
});
});
| javascript |
<filename>spec/regression/data/links_incompatible_types.result.json
{
"data": {
"horst": {
"name": "Horst",
"nationality": {
"description": "Deutschland"
}
},
"_errors": [
{
"message": "Error in @link config on Person.nationality: Link field \"nationality\" is of type ID, but argument \"identCode\" on target field \"staticData.Country\" has type String"
}
]
}
} | json |
<reponame>H2020LIGHTest/DelegationProvider<filename>src/main/java/eu/lightest/delegations/DelegationProviderProperties.java
package eu.lightest.delegations;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import java.io.FileNotFoundException;
import java.io.IOException;
import java.io.InputStream;
import java.util.Properties;
public class DelegationProviderProperties {
private Properties mProperties = null;
private Log mLog = LogFactory.getLog(DelegationProviderProperties.class);
private static final String PROPERTY_FILE_NAME = "config.properties";
private static final String GIT_PROPERTY_FILE_NAME = "git.properties";
private Properties mGitProperties = null;
public DelegationProviderProperties() throws IOException {
mProperties = initProperties(PROPERTY_FILE_NAME);
mGitProperties = initProperties(GIT_PROPERTY_FILE_NAME);
mLog.debug("Successfully loaded property file");
}
private Properties initProperties(String file) throws IOException {
Properties properties = new Properties();
InputStream inputStream = null;
inputStream = getClass().getClassLoader().getResourceAsStream(file);
if ( inputStream == null ) {
throw new FileNotFoundException("property file '" + file + "' not found!");
}
properties.load(inputStream);
inputStream.close();
return properties;
}
public String getPropertyStorageType() {
return getPropertyValueByName("data.store.type");
}
public String getPropertyDatabaseAddress() {
return getPropertyValueByName("data.store.database.address");
}
public String getPropertyDatabaseUser() {
return getPropertyValueByName("data.store.database.user");
}
public String getPropertyDatabasePassword() {
return getPropertyValueByName("data.store.database.password");
}
public String getPropertyDatabaseName() {
return getPropertyValueByName("data.store.database.name");
}
public String getPropertyValueByName(String name) {
mLog.debug("Searching property '" + name + "'");
String value = mProperties.getProperty(name);
mLog.debug("Property '"+name+"' has value '" +value+"'");
return value;
}
public String getGitPropertyRevId() {
return getGitPropertyByName("git.commit.id.describe");
}
public String getGitPropertyByName(String name) {
return mGitProperties.getProperty(name);
}
}
| java |
<reponame>nandana/ldp4j-generic
package org.ldp4j.generic.security.impl;
import com.google.common.io.ByteStreams;
import com.google.gson.Gson;
import org.junit.Test;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.util.List;
import java.util.Map;
import java.util.Set;
import java.util.concurrent.ExecutionException;
import static org.hamcrest.MatcherAssert.assertThat;
import static org.hamcrest.Matchers.*;
public class UserStoreTest {
@Test
public void testDeserialize() throws Exception {
Gson gson = new Gson();
byte[] bstr =
ByteStreams.toByteArray(this.getClass().getClassLoader().getResourceAsStream("user-config-test.json"));
String userStoreJson = new String(bstr);
System.out.println(userStoreJson);
UserStore userStore = gson.fromJson(userStoreJson, UserStore.class);
Map<String, User> users = userStore.getUsers();
assertThat(users, is(notNullValue()));
Set<String> userIds = users.keySet();
assertThat(userIds, hasSize(3));
assertThat(userIds, contains("admin", "test1", "test2"));
User admin = users.get("admin");
assertThat("username should be admin", admin.getUsername(), is("admin"));
assertThat("password should be <PASSWORD>", admin.getPassword(), is("<PASSWORD>"));
assertThat("role should be admin", admin.getRoles(), hasSize(1));
}
}
| java |
{
"@metadata": {
"authors": [
"Mguix"
]
},
"ep_readonly_guest_login": "Inicia sessió",
"ep_readonly_guest_logout": "Finalitza la sessió"
}
| json |
In December 2018, a video of Ranveer Singh dancing in Padmaavat was making people laugh and going viral on Facebook. The catch was that the Bollywood actor was not dancing on the song Khalbali from the movie, but the song was PSY’s famous Gangnam Style. People were loving it and this was the first viral video of Ankit Chauhan aka OyeAnkit, who was then working as a video editor at The Viral Fever, an Indian internet video company. In the past few months, Ankit’s Instagram videos feature viral dances as well clips from popular television and movies, remixed with a Bollywood melody of his choice. Matching the steps with the music, he edits them in a way that misplaced music and videos come alive together in a funny and novel way. Social media users love Ankit’s videos so much that his every remix video meme goes viral and none of them has views less than 500,000.
The most viewed video on his Instagram profile is of Dennis Kyere and Isaac Kyere dancing, twin dancers from Germany. In the video, the Kyere brothers can be seen dancing on ‘Bole Chudiya’ from the Bollywood movie Kabhi Khushi Kabhie Gham. The remix video has 47 million views, four times more than the combined views on the original video posted by the Kyere brothers.
In another of his remixes, he placed ‘Tumse Milke Dil Ka Jo Hai Haal’ from Shahrukh Khan starring Main Hoon Na over a funny clip from the popular television series Friends, in which one of the lead characters Ross’s cousin Cassie comes to visit the friends and wherever she goes, all of them stares at her opening her hair awestruck. The video has more than one million views.
The second most viewed video on OyeAnkit’s Instagram profile is again the Kyere twins dancing on ‘Deta Jai Jo Re’ from 1998 Bollywood blockbuster Bade Miyan Chhote Miyan. The immensely popular song is sung by Udit Narayan and Alka Yagnik among other singers.
In another video, Michael Scott from the popular American sitcom The Office is dancing on recently released Parmish Verma’s Meri Marzi. The song, which has lyrics declaring his freedom of choices, fits perfectly with the character of Michael and his action and comes across as really funny. A user commented saying that she fell off her chair laughing when she watched the video.
Another remix of ‘Bole Chudiyan’ featured the viral dancing video of Roseylucci and Samantha Grace. The video ended up making the original artists laugh and Roseylucci shared it on her Instagram profile crediting OyeAnkit. One of the commenters pointed out that the remix fit perfectly.
Dilip Kumar also got the privilege to be featured in one of Ankit’s magical remixes. In the video, Dilip Kumar is seen dancing on My Dil Goes Mmmm from Saif Ali Khan starrer Salaam Namaste.
The original video of Dilip Kumar dancing is from another classic song ‘Nain Lad Jai Ha’ from the 1961 film Gunga Jumna.
Read all the Latest News, Breaking News and Coronavirus News here. | english |
import React from 'react';
import PropTypes from 'prop-types';
import IfPermission from '@folio/stripes-components/lib/IfPermission';
import IfInterface from '@folio/stripes-components/lib/IfInterface';
import EditablePermissions from '../../EditablePermissions';
const propTypes = {
stripes: PropTypes.shape({
intl: PropTypes.object.isRequired,
}).isRequired,
resources: PropTypes.shape({
availablePermissions: PropTypes.shape({
records: PropTypes.arrayOf(PropTypes.object),
}),
}).isRequired,
userPermissions: PropTypes.arrayOf(PropTypes.object),
};
class EditUserPerms extends React.Component {
static manifest = Object.freeze({
availablePermissions: {
type: 'okapi',
records: 'permissions',
path: 'perms/permissions?length=1000',
},
});
render() {
const { resources, userPermissions } = this.props;
const availablePermissions = (resources.availablePermissions || {}).records || [];
return (
<IfPermission perm="perms.users.get">
<IfInterface name="permissions" version="5.0">
<EditablePermissions
{...this.props}
heading={this.props.stripes.intl.formatMessage({ id: 'ui-users.permissions.userPermissions' })}
permToRead="perms.users.get"
permToDelete="perms.users.item.delete"
permToModify="perms.users.item.post"
availablePermissions={availablePermissions}
listedPermissions={userPermissions}
/>
</IfInterface>
</IfPermission>
);
}
}
EditUserPerms.propTypes = propTypes;
export default EditUserPerms;
| javascript |
Antonio Conte has stepped up his efforts to improve his squad ahead of the season, as he orders the Chelsea board to sign Tottenham left back Danny Rose for 40 million.
The Chelsea boss has been very grouchy over the last few weeks, as he has come under a lot of pressure by the media. Conte has been furious with the lack of transfer activity taking place at the club.
Just a few months ago, the Italian was the toast of the town, as he led Chelsea to a Premier league title in his maiden season in England. Although he came in with a great reputation, Conte was still not expected to win the title so early on in his Chelsea career.
The likes of Mourinho, Guardiola, Klopp and to a lesser extent Wenger were expected to challenge for the title last season. But they all fell to the Italian, as Conte forged a winning team out of a Chelsea side that had finished 10th the season before.
But Conte has had an infuriating pre-season so far. There have been multiple reports suggesting that the Chelsea boss is furious with the lack of support that he has received from the Chelsea board in the transfer market.
To be honest, there were murmurings of discontent last year as well, as Conte was not able to get any of his first choice players. This summer, Romelu Lukaku was Conte’s first choice striker to replace Diego Costa. But Manchester United beat Chelsea to his signature, leaving Conte red faced and frustrated with his board.
Now it remains to be seen how the board reacts to this latest order from Conte. The Italian does not seem to be the type of manager who would continue at the club if he feels overpowered especially in team decisions.
The next few weeks will give us a better understanding of what will happen!
| english |
// Autogenerated from CppHeaderCreator
// Created by Sc2ad
// =========================================================================
#pragma once
#pragma pack(push, 8)
// Begin includes
#include "extern/beatsaber-hook/shared/utils/typedefs.h"
// Including type: System.UriParser
#include "System/UriParser.hpp"
// Completed includes
// Begin forward declares
// Forward declaring namespace: System
namespace System {
// Skipping declaration: UriSyntaxFlags because it is already included!
}
// Completed forward declares
// Type namespace: System
namespace System {
// Autogenerated type: System.UriParser/BuiltInUriParser
class UriParser::BuiltInUriParser : public System::UriParser {
public:
// System.Void .ctor(System.String lwrCaseScheme, System.Int32 defaultPort, System.UriSyntaxFlags syntaxFlags)
// Offset: 0x1957DA8
static UriParser::BuiltInUriParser* New_ctor(::Il2CppString* lwrCaseScheme, int defaultPort, System::UriSyntaxFlags syntaxFlags);
}; // System.UriParser/BuiltInUriParser
}
#include "extern/beatsaber-hook/shared/utils/il2cpp-type-check.hpp"
DEFINE_IL2CPP_ARG_TYPE(System::UriParser::BuiltInUriParser*, "System", "UriParser/BuiltInUriParser");
#pragma pack(pop)
| cpp |
Parineeti Chopra and politician Raghav Chadha have been trending after they announced their engagement which took place in New Delhi last week. The couple issued a joint statement, marking the beginning of a new chapter in their lives. Amid this, a video has been going viral in which the actress is seen talking about what qualities she needs in her future husband but she does not want him to be like her father.
In the video, shared by motivationa. letter on Instagram, we can see Parineeti talking to Aamir Khan. She says, “Maine bahot logon se suna hai yeh kehte hue ki mujhe bilkul apne papa jaisa ladka chahiye. Main toh kabhi nai kehte yeh ki mujhe apne dad jaisa ladka chahiye. ” Aamir immediately asks her why how is your dad? She replies, “Kyunki mere dad bahot strict hai. Unhe bhi yeh lagta hai ki rona galat hai. Agar mere bhai rote hai toh vo bolte hai kya ladkiyon ke jaise roo rahe ho. Ladko ko kitchen mein aana allowed nai hai. Patan ai kyun. I think unki galati nai hai, vo aise hi bade hue hai. I think ego bahot hoti hai mardo mein. Yeh mard ka kaam hai, yeh aurat ka kaam hai. ” Aamir then says, “Aapko ego pasand nai hai mardo mein? ” On this she said, “Bilkul bhi nai. ” In the same video, she also stressed on the fact that it is okay to express your feeling even if you are a male and expressing yourself doesn’t make you less male.
Watch the video here:
This viral video also shows Kangana Rannut. Both were gracing the show which was hosted by Aamir Khan. He made his television debut with his talk show, Satyamev Jayate, which dealt with social issues. | english |
A look at the headlines right now:
- India calls off Sushma Swaraj’s meeting with Pakistan’s foreign minister in New York: Three policemen in Jammu and Kashmir’s Shopian district, who were abducted by suspected militants, were found dead earlier on Friday.
- Kerala Police arrest Bishop Franco Mulakkal for allegedly raping a nun: He was questioned by the police for the third straight day on Friday.
- French ex-President Hollande says India proposed Anil Ambani’s Reliance for Rafale deal, says news report: French news organisation Mediapart quoted François Hollande as claiming that his country ‘did not have a say’ in the decision.
- Congress claims alliance between BSP and Ajit Jogi’s party has BJP support: The Congress alleged that Mayawati’s party had tied up with the Janta Congress Chhattisgarh as it was under pressure from the central investigating agencies.
- Union minister Prakash Javadekar says celebrating ‘Surgical Strikes Day’ is not mandatory: His comments came after minister Partha Chatterjee said West Bengal will not mark the occasion.
- NSA charges against Bhim Army chief Chandrashekhar Azad withdrawn, UP government tells Supreme Court: The top court disposed Azad’s plea challenging the invocation of provisions of the National Security Act, saying it had become infructuous.
- US imposes sanctions on Chinese military for buying Russian fighter jets, missiles: The US State Department said the move is aimed to ‘impose costs’ on Russia for interfering in the US elections.
- Paytm accuses Google Pay of sharing data with group companies and third parties: In a letter to National Payments Corporation of India, it also alleged the payment platform has used customers’ data for monetary gains.
- Police carry out midnight raids at Manipur University: This comes hours after Vice Chancellor AP Pandey along with security personnel accompanied Pro Vice Chancellor Y Yugindro Singh into the campus.
- Cyclonic storm Daye hits Odisha, triggers rainfall across state: The regional meteorological centre said it may gradually weaken into a deep depression.
| english |
{
"page_hash": "36bcae6aa19cb9ca75b36e5b9127e814f4fef550fa641a602faf89202232f3aa",
"paragraphs": [
{
"type": "text",
"text": "Contre toute attente, l'Escargot prend peur et s'enfuit aussi vite que sa lourde coquille le lui permet. Il écrase arbrisseaux et taillis sur son passage, décidé à trouver un endroit plus tranquille. "
},
{
"type": "link",
"text": "Vous regagnez l'allée que vous suiviez auparavant.",
"hash": "e426463692e2d766dc22fb01981fa224f047c84cc65ed481d67231518c588a66"
}
],
"monsters": [
]
}
| json |
{
"name": "@freitagfelipe/discord.js-pagination",
"version": "1.0.5",
"description": "A simple way to send and paginate discord message embeds.",
"main": "index.js",
"scripts": {
"test": "echo \"Error: no test specified\" && exit 1"
},
"repository": {
"type": "git",
"url": "git+https://github.com/freitagfelipe/discord.js-pagination.git"
},
"keywords": [
"discord.js",
"pagination"
],
"author": "freitagfelipe",
"license": "MIT",
"publishConfig": {
"access": "public"
},
"bugs": {
"url": "https://github.com/freitagfelipe/discord.js-pagination/issues"
},
"homepage": "https://github.com/freitagfelipe/discord.js-pagination#readme"
}
| json |
- Performance (183)
- Interior (65)
Looks Impressive.
Powerful Beast.
Best Off Road Car.
Looking good and now we used daily use. Thar 2020 Mahindra is best for offroading vehicles also.
Head turner design.
The car has a very premium looks and design.
So nice and great car looking good with powerful engine and balance is good.
Nice Car.
Thar looking nice, and riding super, long drive super, chasing comfort feeling joy ride,
Are you Confused?
What ഐഎസ് the ground clearance അതിലെ the മഹേന്ദ്ര Thar?
The ground clearance of the Mahindra Thar is 226mm(Unladen).
What ഐഎസ് the price?
What ഐഎസ് the ഇരിപ്പിടം capacity അതിലെ മഹേന്ദ്ര Thar?
The seating capacity of the Mahindra Thar is 4 people.
What ഐഎസ് the star rating അതിലെ മഹേന്ദ്ര Thar?
The Thar achieved four stars for adult and child occupant protection.
| english |
import React, { memo } from "react";
import { Path } from "react-native-svg";
import Theme from "../../commons/Theme";
interface Props {
curveData: string;
}
export default memo(function Curve({ curveData }: Props) {
return (
<Path
d={curveData}
fill="none"
stroke={Theme.color.neutral300}
strokeWidth={27}
/>
);
});
| typescript |
<filename>src/main/resources/static/mas_json/2016_ijcai_8913541906900741399.json
{"title": "Fear and Hope Emerge from Anticipation in Model-Based Reinforcement Learning.", "fields": ["machine learning", "anticipation", "architecture", "monte carlo tree search", "reinforcement learning"], "abstract": "Social agents and robots will require both learning and emotional capabilities to successfully enter society. This paper connects both challenges, by studying models of emotion generation in sequential decision-making agents. Previous work in this field has focussed on model-free reinforcement learning (RL). However, important emotions like hope and fear need anticipation, which requires a model and forward simulation. Taking inspiration from the psychological Belief-Desire Theory of Emotions (BDTE), our work specifies models of hope and fear based on best and worst forward traces. To efficiently estimate these traces, we integrate a well-known Monte Carlo Tree Search procedure (UCT) into a model based RL architecture. Test results in three known RL domains illustrate emotion dynamics, dependencies on policy and environmental stochasticity, and plausibility in individual Pacman game settings. Our models enable agents to naturally elicit hope and fear during learning, and moreover, explain what anticipated event caused this.", "citation": "Citations (3)", "year": "2016", "departments": ["Delft University of Technology", "Delft University of Technology", "Delft University of Technology"], "conf": "ijcai", "authors": ["<NAME>.....http://dblp.org/pers/hd/m/Moerland:Thomas_M=", "<NAME>.....http://dblp.org/pers/hd/b/Broekens:Joost", "<NAME>.....http://dblp.org/pers/hd/j/Jonker:Catholijn_M="], "pages": 7} | json |
The same thing will also be introduced at the university's affiliate and iconic St Xavier's College in Kolkata.
The student opting for the new four-year undergraduate course, after completion of that course would be able to undertake a one- year post-graduation.
However, both St. Xavier's University and St. Xavier's College will continue with the old system of a three-year undergraduate course followed by a two-year post-graduation course.
Both the universities have made a formal announcement on this count on its website.
The authorities of both St. Xavier's University and St. Xavier's College have claimed to have completed the process of putting in place the infrastructure necessary to introduce this new academic system. The new system will be available for all the subjects taught in these two institutes.
It is learnt that students who will be completing a four-year undergraduate course, will have the option to either go for a one- year post-graduation course or straight go to Ph. D.
However, those opting for the second option of going straight to Ph. D will have to fulfill some criteria for that.
Students opting for research in the fourth year will have to score 75 per cent in the first six semesters and also complete a research project under any faculty member.
The decision of these two prime educational institutes of the state come at a time, when no other university in the state, especially the state-run universities are yet to spell out their concrete plan in the matter.
Even the TWest Bengal government has expressed its reservations on this count.
Disclaimer: This story has not been edited by the Sakshi Post team and is auto-generated from syndicated feed. | english |
Ridiculously Hot! Priyanka Chopra's Pictures From Emmy Awards Are Bound To Grab Your Attention!
Actress Priyanka Chopra, in her debut appearance at the Emmy Awards, opted for an understated look as she wore a plain one-shouldered all-red chiffon gown.
With her hair neatly tied in a ponytail with a side part, Priyanka impressed the fashion critics as she arrived at the 68th Emmy Awards in her Jason Wu outfit.
Check Out Her 11 Jaw-dropping Pictures Below:
Priyanka Chopra wore no jewellery except a diamond ring and drop diamond earrings and she totally nailed the red hot look like a pro at the 68th Emmy Awards.
Though she wore light make-up, her vibrant red lipstick gave the perfect highlight to her look. Her look from the event is just too hot to handle!
The 34-year-old Quantico actress enjoyed every moment at the red carpet as she happily posed, twirled and spun around, showing off the delicate layers of her gown.
She presented the Outstanding Directing for a Limited Series, Movie, or Dramatic Special award to The Night Manager director Susanne Bier at the ceremony with Hollywood star Tom Hiddleston.
On the work front, the actress is now gearing up for the second season of her American series Quantico, which goes on air on September 25.
With 'Jon Snow'
Priyanka Chopra poses with Kit Harington aka Jon Snow of the Game Of Throne. The duo looks gorgeous together, don't they?
The actress was also seen posing along with her girl squad! Interestingly, her pictures from the awards are going viral on every platform of social media including Facebook, Twitter & Instagram.
Priyanka Chopra is leaving no stones unturned to make India proud! From gracing the Times magazine cover to attending popular Hollywood Awards event, she is winning all hearts.
Not so long ago, Priyanka graced the cover on W magazine and while interacting with its team, she had revealed that she doesn't consider herself as a beautiful lady!
"No. I don't even think I am right now. I just think I've found how to be the best version of myself, and that took a lot of years. " Dear Priyanka, if only you could see from the eyes of your fans, how gorgeous you are!
"I've Done Slow Motion All My Life'
Talking about her Hollywood movie Baywatch at Emmy, Priyanka told, "I have done too much of slow motions all my life. Therefore I won't be doing it. I will leave that to Dwayne and Zac. However, I do walk in slow motion in the movie. Running is too much of an effort. "
Oh No! Aaradhya Looks So SCARED, Spotted With Aishwarya Rai Bachchan!
Priyanka Chopra was last seen in Sanjay Leela Bhansali's Bajirao Mastani, where she portrayed the character of Kashibai. She will be next seen in her upcoming Hollywood film, Baywatch where she will play a negative character.
The Baywatch movie is based on the hugely popular 1990s TV series of the same name. And Priyanka essays Victoria Leeds in it.
The actress is now busy shooting the season 2 of "Quantico", which will return in September. The first season of the show will be back in India from the mid-season break on Star World and Star World HD.
In the show, Priyanka plays Alex Parrish, an FBI recruit who becomes a suspect of terrorist bombing of New York's Grand Central Station.
The show's storytelling prism shifts between the present day with Parrish navigating her way through a class of FBI new agent trainees to the near future as the truth and repercussions of the attack emerge. | english |
Rey Mysterio is currently one-half of the SmackDown Tag Team Champions, a title he holds with his son and tag partner Dominik. The duo became the first father and son team to win the titles together in WWE history when they beat the Dirty Dawgs at WrestleMania Backlash.
During a recent interview with Sony Sports India, Rey Mysterio opened up about his close relationship with the late Eddie Guerrero, who was more of a brother to him than a friend. Rey also told the story of the first time he saw Eddie Guerrero perform in the ring:
"We were like brothers," said Mysterio. "Funny story, when I was a fan, I would go and watch my uncle wrestle in Tijuana and that's where I got to see Eddie wrestle for the first time. I was still a fan, I'd have been about 13 years old. Being able to see Eddie, never did I think that one day we would be as close as we became and that I would be able to share the ring with Eddie. Very inspirational. In the ring, I learned so much from him. Outside of the ring, the connection that we had was very special. Very very special. It was so special that when Eddie passed, the fans would see Eddie through me. Speechless. No words for that. "
Apart from being a massive singles star, Rey Mysterio has also won a number of tag team titles during his WWE run. The former World Heavyweight Champion spoke about his privilege of teaming up with some of the best in the business.
"It's been beautiful that I was able to share the Tag Team titles with legends in this industry and Hall of Famers," Mysterio explained. "I had a great time wrestling with Edge, with RVD as well, with Batista who is now in the Hollywood world. Then Eddie Guerrero. I learned so much from him, being next to him and wrestling against him. "
"Overall, I think all the partners I've had in my career have been incredible and I've learned so much from them," Mysterio continued. "Now, even with my son, as much as I'm teaching him, I've learned so much from him being by his side. "
During the interview with Sony Sports, Rey Mysterio also revealed how he came up with the 619. You can check that out HERE.
If any quotes are used from this article, please add an H/T to Sportskeeda and a credit Sony Sports India. | english |
Angry guardians on Friday protested at a school here after parents of a four-year-old girl student and doctors said she was sexually abused by a teacher and perhaps more men.
The girl, studying at the high-profile GD Birla Centre For Education in south Kolkata's Ranikuthi area, was profusely bleeding when she returned from school on Thursday, the girl's father said. She was rushed to a hospital.
"We initially thought it was some infection. We took her to the doctor who said she was physically abused. My daughter is still in a lot of pain," the father said.
A lawyer for the victim's family said the girl was lured into a toilet and sexually abused.
"She was shown a chocolate and lured inside a school toilet where one or more men abused her. The girl was shown pictures of four suspects. She has identified two men, one of whom is her physical education teacher," the lawyer said.
Furious parents of the school students on Friday stopped the school buses from entering the premises and demanded the removal of all male teaching staff.
They also told the school to ensure security for the children, including by putting up CCTV cameras.
Police said they had received the complaint.
"Our primary focus now is to arrest the accused. We will further look into the security measures in the schools," a police officer said. | english |
Islamabad: At least three journalists in Pakistan have been charged for alleged sedition in the past week, raising concerns that Prime Minister Imran Khan’s government is increasingly intolerant of media criticism.
Bilal Farooqi, the news editor at the Express Tribune, said he was arrested by the Karachi police last week on charges of sedition for defaming the country’s army and criticizing the government but later he was released on bail. Absar Alam, a former anchor at AAJ news channel, said he also faced similar accusations in charges made available to him by police. Asad Ali Toor, a journalist at Samaa television, is facing a sedition case, the local media reported.
Many prominent journalists in Pakistan have taken to social media to describe their battles with government censorship. Reporters Without Borders has lowered Pakistan’s ranking by three points to 145 out of about 180 in its 2020 World Press Index. The ranking has dropped six notches since Khan, once a vocal advocate of the free media, came to power in 2018. The prime minister has claimed the press enjoys unprecedented freedom in the South Asian nation.
“The arrest of Bilal was a part of the nefarious and concerted plan to gag the free media and independent voices,” the Karachi Union of Journalists, said in a statement. The government has no tolerance for dissenting views and criticism of any kind, Farooqi said on phone from Karachi on Tuesday. Such cases are an “honor in a fascist regime,” Alam said on phone.
Zafar Abbas, a spokesman for Pakistan’s interior ministry, was not immediately available for comments on journalists being booked for treason. | english |
@import url("https://fonts.googleapis.com/css?family=Source+Sans+Pro:200,400");
@import url("https://fonts.googleapis.com/css?family=Source+Code+Pro");
body {
font-family: "Source Sans Pro", sans-serif;
background-color: rgb(255, 255, 255);
}
| css |
{
"author": {
"id": "t2_w4ars",
"name": "shadowcorp"
},
"date": {
"day": 1592870400,
"full": 1592915131,
"month": 1590969600,
"week": 1592697600
},
"id": "t3_hedob4",
"picture": {
"filesize": 267096,
"fullUrl": "https://external-preview.redd.it/YIgDO4qkDHACeYS2OOVa2I_6blB1lZsvhsMKJb5UYcw.png?auto=webp&s=0c208d15cfba7f3d9e27ceea16a9618a26ee8092",
"hash": "e2d77cdb99",
"height": 800,
"lqip": "data:image/png;base64,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",
"url": "https://external-preview.redd.it/YIgDO4qkDHACeYS2OOVa2I_6blB1lZsvhsMKJb5UYcw.png?width=640&crop=smart&auto=webp&s=16ccd6af4a8d0dce69bf36dbd9bd6de25a2b1369",
"width": 640
},
"score": {
"comments": 28,
"downs": 0,
"ratio": 0.98,
"ups": 927,
"value": 927
},
"subreddit": {
"id": "t5_2xy5e",
"name": "TerrainBuilding"
},
"tags": [],
"title": "There's some great technique in here!",
"url": "https://www.reddit.com/r/TerrainBuilding/comments/hedob4/theres_some_great_technique_in_here/"
}
| json |
{
"IndexId": 463,
"Portraits": [
{
"PortraitId": 0,
"PortraitName": "STANDARD",
"PortraitImageBase64": "<KEY>
}
]
} | json |
HowExpert Guide to Doll Collecting: 101+ Tips to Learn How to Find, Buy, Sell, and Collect Collectible Dolls for Doll Collectors (Paperback)
If you want to learn about the history of dolls, caring for and storing your dolls, starting a collection, buying, selling, and collecting collectible dolls, doll knowledge for collectors, and joining the doll community, then check out HowExpert Guide to Doll Collecting.
The first toy ever made was the doll Today doll collecting is the world's largest hobby. Dolls remind us of our favorite childhood moments, a time in history that we are most fond of, and even a beloved art style reflected through the doll. HowExpert Guide to Doll Collecting can lead you through the stages of doll collecting. Whether you are starting a collection or adding on to one, there is a wide variety of dolls to choose from. These include ceremonial dolls, bisque dolls, rag dolls, and paper dolls, to the beloved characters like Strawberry Shortcake, Cabbage Patch Kids, and Barbie. Finding that doll that brings out your love for collecting is the easy part. In this book, you will learn the proper care for dolls, doll lingo, how to spot a replica, and tips on keeping a logbook. We even took a peek at the history of haunted dolls. The tips in HowExpert Guide to Doll Collecting will show doll enthusiasts just what they need to know to become doll collectors.
HowExpert publishes how to guides by everyday experts.
| english |
/*
* @Date: 2022.03.02 10:10
* @Description: Omit
* @LastEditors: <NAME>
* @LastEditTime: 2022.03.02 10:10
*/
package net
import "net"
func Listen(network, address string, cipher ConnCipher) (net.Listener, error) {
l, err := net.Listen(network, address)
return &listener{Listener: l, ConnCipher: cipher}, err
}
func Dial(network, address string, cipher ConnCipher) (net.Conn, error) {
c, err := net.Dial(network, address)
return cipher(c), err
}
func ListenPacket(network, address string, cipher PacketConnCipher) (net.PacketConn, error) {
c, err := net.ListenPacket(network, address)
return cipher(c), err
}
| go |
Microsoft just unveiled its newest AI integration into its broader ecosystem of products, from Bing search to Office and even possibly Windows itself: a ChatGPT-like AI system to interact with users through a large language model that the company says will greatly enhance search results and general user experience using its products.
That may be true in 99.9999% of cases, but it's always the edge cases that get you in trouble, and failures of a system that is effectively a black box are inevitable. That Google and Microsoft are banking so hard on this idea is astounding since they have a unique privilege on the internet that the rest of us don't.
With the integration of these new AI tools, however, Google, Microsoft, and others are pretty much throwing that protection away, and as such, they could, and should, be held legally liable for every bit of misinformation and libel that their new chatbots produce.
Ask any YouTube creator or print or online journalist about copyright and you're bound to get an earful about the difference between fair use, quoted materials, "background", and all the other rules governing what they can publish and how they can publish it.
If you're livestreaming Hi-Fi Rush and forget to turn on the Livestream audio option? Enjoy the copyright strike on your account. Forgot to attribute an entire passage of text in your article to the original source? Have fun navigating a writing career after the ensuing plagiarism controversy (and be ready to forgo any profits you've made if the case is egregious enough).
What happens when a black-box AI "inadvertantly" plagiarizes your material and Google or Microsoft publishes it for profit? Or, what if it scrapes bad advice from a site or rewords the material it has scraped but omits critical, legally required disclaimers? What happens if it does it thousands or millions of times? The whole point of these AI chatbots is that they are autonomous, so there's little to no chance to intervene before the chatbot says something legally problematic.
Even in Microsoft's own press materials, it included a screenshot of the new Bing Chat feature where the prompt "Give me a workout I can do at home" gave an answer without any medical disclaimer or context about the serious risk of injury when performing some of the exercises the answer suggests.
This is the kind of thing that personal injury lawyers make a living off of, and these chatbots being deployed by the most wealthy corporations in the world are going to be an absolutely irresistible target.
To its credit, Microsoft Bing's AI-powered chatbots will at least provide citations for where it got the material it used to generate its answers, but as a journalist, I can tell you that "Well, this is where I got the legally objectionable material from" won't get you very far in a court.
What the new Bing produces will only ever be as good as the data it uses as input, and by rewriting it in Chat rather than simply passing along what someone else has written (as Google and Microsoft do now with their search results), they ultimately become the publishers of that content, even if they cite someone else's work as a source.
You can introduce the most advanced filters possible to minimize your risk, but at some point, one of these chatbots is going to respond to a very specific prompt and pull from some very dubious source and it's going to give a result that is going to break a law or even cause harm. Who is responsible then?
If Microsoft Bing recommends a substitute for eggs in a recipe (as Microsoft itself says the new Bing will be able to do in its presentation), and recommends an ingredient to somebody who has an allergy, who is to blame if that person gets sick or, God forbid, actually dies as a result? You can bet there will be an army of lawyers who are going to sue to settle that legal question very quickly, and Microsoft will have a lot harder time hiding and claiming a "platform" exemption when its own chatbot produced the offending work.
Worse still, no one at Microsoft will be able to predict when something like this will happen and what the offending material will be, so there's no way to prevent it. That's because, fundamentally, nobody knows how these generative AI models actually come to the conclusions they ultimately reach, since neural networks are, again, largely black boxes. You cannot predict every problem, you can only identify it after the fact, which won't be a great legal defense for Big Tech firms who implement these models in a public-facing way.
They are ultimately integrating an unknowable liability into their products and taking responsibility for what it produces, just as TechRadar (and parent company Future PLC) would be liable if I wrote and it published something that resulted in harm or broke some country's law.
There are critical safeguards in place to ensure that doesn't happen when you're a publisher (like editors and required legal training for writers on this sort of thing), but Microsoft and Google seem to be prepared to just let their AIs loose on the Internet without any intercessional human supervision at the point of the user interface. Good luck with that.
Microsoft of all companies should know how this is likely to end up since its experimental Twitter chatbot Tay was hijacked in less than a day by internet users who turned it into a foul-mouthed bigot for fun. Sure, the state of AI has come a long way in the seven years since Tay denied the Holocaust, but honestly, it hasn't changed fundamentally. These models are just as biased and prone to manipulation as ever, they just look more sophisticated now so that their faults are better hidden.
How much of a legal shield that will be for Big Tech companies rolling out AI models online that are intrinsically capable of spreading legally damaging misinformation and libel remains to be seen, but March 2016 was also a very different era for tech than 2023, and something tells me that people nowadays are going to be far less forgiving no matter what your politics happen to be (Editor's update: this week, Alphabet livestreamed a demonstration of Google's Bard AI and showed off a response that was factually incorrect, in real time. Markets reacted by wiping $120 billion off Alphabet's market valuation in the hours that followed).
The legal perils of being a publisher are as infinite as there are ways to libel someone or spread dangerous, unprotected speech, so it's impossible to predict how damaging AI integration will be.
All I know is Google, Microsoft, and any other Big Tech platform that decides that it's going to jump on the AI writer bandwagon better be ready to answer for whatever it inevitably spews out, and given past experience with these kinds of automated language models, once millions and even billions of people start engaging with them, all of those invisible faults in the model are going to materialize very quickly.
If there's one thing we've learned over the past few years, it's that inside every internet-trained AI model, there is a deeply bigoted, genocide-friendly troll looking for some crack to crawl out of. So, how long until it crawls out into our chat-engine results? How many of those are going to be legally actionable?
Who's to say, but in no world would I bet my trillion-dollar company's existence on something this ephemeral. Given Big Tech's absolute paranoia about missing out on the next Big Thing, though, companies jumping on this isn't the least bit surprising, even if it's Big Tech's most reckless move yet.
Get the hottest deals available in your inbox plus news, reviews, opinion, analysis, deals and more from the TechRadar team.
John (He/Him) is the Components Editor here at TechRadar and he is also a programmer, gamer, activist, and Brooklyn College alum currently living in Brooklyn, NY.
Named by the CTA as a CES 2020 Media Trailblazer for his science and technology reporting, John specializes in all areas of computer science, including industry news, hardware reviews, PC gaming, as well as general science writing and the social impact of the tech industry.
You can find him online on Threads @johnloeffler.
Currently playing: Baldur's Gate 3 (just like everyone else).
| english |
Puerto Rico: Do you see what I see?
Gil the Jenius enlists the help of refrain from a Christmas carol in his diatribe against what he sees as some of the administrative and political problems affecting Puerto Rico.
Hungary Economy Watch writes about the country's demographic situation.
According to AghBahman [Fa], Mansour Osanlou,head of the executive committee of the Syndicate of Workers of Tehran and Suburbs Bus Company, is out of prison now.
Jamie from Two Koreas reprints some news from Migrant Worker TV to sum up some of the current issues going on in the migrant movement in South Korea these days.
South Korea: Will S. Korea Ratify Migrant Convention?
Koh Ki-bok from Ohmynews discusses about conditions of migrant workers in South Korea and urges the government to ratify migrant convention as soon as possible.
Russia, U. S. : Medical Insurance; Paid Vacation.
Everybody I Love You seems to have found a way to get medical insurance for his Russian wife through Blue Cross of California – and have discovered that in Russia a four-week paid vacation is the law, while in the United States it is not.
“…Sales people often seem to be actively working against closing the deal; in this sense, we can call it an ‘anti-service’ culture,” writes Stet as she offers fascinating examples of this Jamaican phenomenon.
Francomenz reports on a strike by workers of Trinidad and Tobago's major petroleum supply company.
China: no sex in Australia?
ESWN translates an article about translation errors in putting “banning for personal reasons resulting in pregnancy or impregnating” into “banning sexual intercourse” concerning the working conditions of Chinese workers in Australia by a Chinese newspapers. The errors resulted in a lot of discussion concerning labour rights.
Debito translates an article about the treatment of foreign workers in Japan and NGO's effort in advocating “free choice of work sector”.
Jamie from Two Koreas blogs about the union demonstration against bills on irregular work. The demonstrators were received by water cannon at freezing weather. You can find a video link in the blog. | english |
<reponame>code4sabae/covid19
{"name":"Wakayama","lastUpdate":"2020-11-17T13:00:00","ncurrentpatients":44,"nexits":284,"ndeaths":4,"npatients":332,"src_url":"https://www.pref.wakayama.lg.jp/prefg/000200/covid19_d/fil/kansensuii.xlsx","url_opendata":"https://www.pref.wakayama.lg.jp/prefg/041200/d00203387.html"} | json |
New Delhi: The Supreme Court on Saturday ordered live telecast of the crucial floor test in the Karnataka Assembly to ensure 'transparency' in the trust vote in which chief minister B S Yeddyurappa will have to prove his majority in the House.
The apex court, which said that 'live telecast would be the best possible way to ensure transparency in the proceedings', also made it clear that 'there shall be no other item in the agenda that shall be taken up during the trust vote today'.
The order for holding the live telecast was delivered after the Congress-JD (S) combine expressed no objection to the suggestions made by the Karnataka government and governor Vajubhai Vala in this regard.
"What more do you want? It will be as transparent as you want. It will be live telecast. Now, it solves your concern. Live broadcast of floor test will be the best possible way to ensure transparency in the proceedings," a special bench of justices A K Sikri, S A Bobde and Ashok Bhushan told the counsel appearing for the alliance.
The bench said secretary of the assembly will record the proceedings of the House.
It said several local channels will be provided the live feed of the proceedings so that they can also be in a position to telecast simultaneously.
The bench termed as 'fair' the suggestion of Additional Solicitor General Tushar Mehta, appearing for the governor and Karnataka government, that the floor test would be telecast live.
Elections held on May 12 to 222 of the state's 224 seats have resulted in a hung Assembly with the BJP having 104 MLAs, while the Congress won 78 seats and JD(S) 37. Polling in two seats will be held later.
The Congress-JD(S) combine has claimed the support of 117 MLAs, including one from the BSP, and two others.
The JD(S)' effective strength will be 36 at the time of voting, as its leader H D Kumaraswamy has won from two seats. This will also reduce the strength of the House to 221 and the BJP will require 111 votes for a simple majority.
The bench was hearing the plea moved on Friday night by the combine challenging validity of governor's order appointing BJP MLA K G Bopaiah as the pro-tem speaker of Karnataka assembly by ignoring the convention of appointing the senior most member in the House.
After the statement of Mehta, senior advocates Kapil Sibal and Abhishek Manu Singhvi, appearing for the combine, preferred not to press for the removal of the pro-tem speaker.
While Bopaiah was a four-time MLA, the combine was clamouring for appointing Congress MLA R V Deshpande who has made it to the Assembly for the eighth time.
The bench, during the 34-minute hearing that commenced at 10. 36 am on a holiday, noted the submissions of Mehta that a decision was already taken that there would be live telecast of proceedings in the Legislative Assembly in respect of the floor test.
Senior advocates Kapil Sibal and Abhishek Manu Singhvi, who were appearing for JD (S) leader H D Kumaraswamy and the combine respectively, changed their strategy after justice Bobde observed that Bopaiah would have to be heard before passing any such order and in that case, the floor test has to be postponed.
"If you want us to examine the suitability of the (pro tem) speaker, then notice has to issued to him and he has to be heard. Then in that case, the floor test will have to be deferred or postponed," Justice Bobde said when Sibal referred to an earlier apex court verdict and claimed that Bopaiah, who was the speaker of the Assembly in 2010, had "manipulated" the majority for the then government.
Sibal referred to the governor's order under Article 188 of the Constitution and said Bopaiah has been appointed "as the person before whom a member of the Legislative Assembly shall, before taking his seat, make and subscribe an oath and affirmation".
He said that the notification refers only to the fact that Bopaiah would administer oath of office to the MLAs and it does not specify that he would be conducting the floor test as the pro-tem speaker.
"Then who will conduct the floor test? ," justice Sikri asked Sibal who responded by saying that the court "can direct the Governor to appoint the senior most member as the pro-tem speaker. It is the convention followed even in Commonwealth countries".
The bench shot back, "It is a convention only. We cannot do it (direct governor to appoint senior most member as the pro tem speaker). Conventions are not rule and they cannot be the law".
When Sibal referred to one of the prayers in the application which sought a direction to conduct the floor test through a 'division' into separate and segregated lobbies so that the MLAs in different lobbies can be counted in a transparent manner, the bench observed, "that was the request made to us (during Saturday's hearing) and we had declined".
Senior advocate Mukul Rohatgi, representing Yeddyurappa, objected to their application saying, "They cannot go on making applications after applications everyday. They are going on as if the Supreme Court is sitting and waiting for them".
When Sibal and Singhvi, who were pressing for video recording of floor test, agreed to Mehta's suggestions, the bench told Mehta, "Thank you very much for coming out with something which we had in our mind". | english |
import argparse
import numpy as np
from scipy import sparse
from scipy.optimize import linprog
import matplotlib.pyplot as plt
import networkx as nx
import torch
import torch.nn as nn
import torch.nn.functional as func
import torch.optim as optim
from torch.optim.lr_scheduler import StepLR
from torch.utils.data import TensorDataset, DataLoader
from captum.attr import IntegratedGradients
def temp_make_energy():
""" I tried to formulate the energy problem as a LP formulation, so defining c, A, and b. I am almost entirely
certain that this is in fact possible but I had too many problems and did not want to invest to much time in this.
So I stopped trying. """
time = 8760
# inputs:
c_pv = 1
c_bat = 1
c_buy = 1
demand = np.genfromtxt("data/TS_Demand.csv").reshape((time, 1))
avail_pv = np.genfromtxt("data/TS_PVAvail.csv").reshape((time, 1))
# cost vector
c_buy = np.full(time, c_buy)
c_zeros = np.zeros(time*5) # for the right dimensionality
c = np.concatenate((np.array([c_pv]), np.array([c_bat]), c_buy, c_zeros))
x_dim = c.shape[0]
# how does x look like? (everything with "(T)" contains T (=time) elements, one for each time step)
# c_ap_pv, c_ap_bat_s, p_buy (T), p_pv (T), p_bat_out (T), p_bat_in (T), p_bat_s (T)
# constraints
# energy balance (maybe remove the equality and make it an inequality (>= Demand, or <= - Demand only)
a_energy_balance = sparse.lil_matrix((2*time, x_dim))
b_energy_balance = sparse.lil_matrix((2*time, 1))
for t in range(time): # this can definitely be written more efficiently, for now I just want it correct though
a_energy_balance[t * 2, 2 + t] = 1 # p_buy(t)
a_energy_balance[t * 2, 2 + time + t] = 1 # p_pv(t)
a_energy_balance[t * 2, 2 + 2 * time + t] = 1 # p_bat_out(t)
a_energy_balance[t * 2, 2 + 3 * time + t] = -1 # p_bat_in(t)
b_energy_balance[t * 2] = demand[t]
a_energy_balance[t * 2 + 1, 2 + t] = -1 # p_buy(t)
a_energy_balance[t * 2 + 1, 2 + time + t] = -1 # p_pv(t)
a_energy_balance[t * 2 + 1, 2 + 2 * time + t] = -1 # p_bat_out(t)
a_energy_balance[t * 2 + 1, 2 + 3 * time + t] = 1 # p_bat_in(t)
b_energy_balance[t * 2 + 1] = -demand[t]
# battery equation
a_battery_equation = sparse.lil_matrix((2 * (time - 1), x_dim))
b_battery_equation = sparse.lil_matrix((2 * (time - 1), 1)) # just stays zero, so that is fine
for t in range(1, time):
a_battery_equation[(t - 1) * 2, 2 + 4 * time + t] = 1 # p_bat_s (t)
a_battery_equation[(t - 1) * 2, 2 + 4 * time + t - 1] = -1 # p_bat_s (t - 1)
a_battery_equation[(t - 1) * 2, 2 + 3 * time + t] = -1 # p_bat_in (t)
a_battery_equation[(t - 1) * 2, 2 + 2 * time + t] = 1 # p_bat_out (t)
a_battery_equation[(t - 1) * 2 + 1, 2 + 4 * time + t] = -1 # p_bat_s (t)
a_battery_equation[(t - 1) * 2 + 1, 2 + 4 * time + t - 1] = 1 # p_bat_s (t - 1)
a_battery_equation[(t - 1) * 2 + 1, 2 + 3 * time + t] = 1 # p_bat_in (t)
a_battery_equation[(t - 1) * 2 + 1, 2 + 2 * time + t] = -1 # p_bat_out (t)
# pv production limit (0 <= p_pv (t) always given per LP definition (x >= 0))
# lifetime missing (delta t I think) --> but why not for battery (per slides: not, per code: yes)
a_pv_production_limit = sparse.lil_matrix((time, x_dim))
b_pv_production_limit = sparse.lil_matrix((time, 1)) # just stays zero, so that is fine
for t in range(time):
a_pv_production_limit[t, 2 + time + t] = 1
a_pv_production_limit[t, 0] = -avail_pv[t]
# battery charge limit (0 <= p_bat_in (t) always given per LP definition (x >= 0))
a_battery_charge_limit = sparse.lil_matrix((time, x_dim))
b_battery_charge_limit = sparse.lil_matrix((time, 1)) # just stays zero, so that is fine
for t in range(time):
a_battery_charge_limit[t, 2 + 2 * time + t] = 1
a_battery_charge_limit[t, 1] = -1
# battery initial state
a_battery_initial_state = sparse.lil_matrix((2, x_dim))
b_battery_initial_state = sparse.lil_matrix((2, 1)) # just stays zero, so that is fine
a_battery_initial_state[0, 2 + 4 * time] = 1
a_battery_initial_state[0, 2 + 4 * time] = -1 # maybe not necessary because of x >= 0
# power purchase limit (0 <= p_buy (t) always given per LP definition (x >= 0))
# concatenate for constraint matrix a and vector b
a = sparse.vstack(
(a_energy_balance, a_battery_equation, a_pv_production_limit, a_battery_charge_limit, a_battery_initial_state))
b = sparse.vstack(
(b_energy_balance, b_battery_equation, b_pv_production_limit, b_battery_charge_limit, b_battery_initial_state))
# time with sparse matrices: csr: 92.63453531265259, lil: 1.463003396987915
# the following calculation of the solution (linprog): took like two hours, did not finish I stopped it, maybe I
# could use more linprog options, or I just don't use this (better I think)
res = linprog(c, a, b.toarray(), method="highs")
sol = res.x
print("Finished")
def load_energy_data(seed=0, with_new=True):
""" Load data points for the energy system. Currently contains 10000 data points. """
# data has 9 columns: cost_pv, cost_bat, cost_buy, demand, cap_pv, cap_bat, own_gen, totex, capex
orig_data = np.load("data/energy_data.npy")
orig_data_plus = None
if with_new:
# "data_plus.npy" is the combination of "data_values_interval.npy", data_values_around_min.npy", and
# "data_values_around_max.npy", generated by "energy_data.py"
orig_data_plus = np.load("data/data_plus.npy")
orig_data = np.concatenate((orig_data, orig_data_plus))
np.random.seed(seed)
np.random.shuffle(orig_data)
# this manually set mean is really (!) close to the calculated mean anyway (with infinite data, it would be exactly
# the same) but using this manually set means means we have data for +1 and -1 (normalized) for all input values
# when also using the newly generated data
if with_new:
# orig_data_plus[100] is exactly the mean data with corresponding outputs
data_mean = orig_data_plus[100] # np.mean(orig_data, axis=0)
else:
data_mean = np.mean(orig_data, axis=0)
# normalize data and save variables to be able to reverse that
data = orig_data - data_mean
data_max = np.max(np.abs(data), axis=0)
data = data / data_max
e_input = data[:, :4]
e_output = data[:, 4:]
return e_input, e_output, data_mean, data_max, orig_data
def vis_energy(attributions, values=None, edge_labels=False, vis_real=False):
""" Visualizes attribution for the energy neural network, from inputs to outputs. """
inp_params, output_vals, pred_outputs = None, None, None
if values is not None:
inp_params, output_vals, pred_outputs = values
g = nx.DiGraph()
# index to name (input/output)
itni = {0: "Photovoltaik", 1: "Batteriespeicher", 2: "Stromnetz", 3: "Energiebedarf"}
itno = {0: "Kapazität PV", 1: "Kapazität Batterie", 2: "Eigenerzeugung", 3: "TOTEX", 4: "CAPEX"}
# define nodes
g.add_node("Photovoltaik", pos=(0, 7))
g.add_node("Batteriespeicher", pos=(0, 5))
g.add_node("Stromnetz", pos=(0, 3))
g.add_node("Energiebedarf", pos=(0, 1))
g.add_node("Kapazität PV", pos=(5, 8))
g.add_node("Kapazität Batterie", pos=(5, 6))
g.add_node("Eigenerzeugung", pos=(5, 4))
g.add_node("TOTEX", pos=(5, 2))
g.add_node("CAPEX", pos=(5, 0))
labeldict = {}
if values is not None:
labeldict["Photovoltaik"] = f"Photovoltaik\n{inp_params[0, 0]:.2f}"
labeldict["Batteriespeicher"] = f"Batteriespeicher\n{inp_params[0, 1]:.2f}"
labeldict["Stromnetz"] = f"Stromnetz\n{inp_params[0, 2]:.2f}"
labeldict["Energiebedarf"] = f"Energiebedarf\n{inp_params[0, 3]:.2f}"
if output_vals is None:
labeldict["Kapazität PV"] = f"Kapazität PV\n{pred_outputs[0, 0]:.2f}"
labeldict["Kapazität Batterie"] = f"Kapazität Batterie\n{pred_outputs[0, 1]:.2f}"
labeldict["Eigenerzeugung"] = f"Eigenerzeugung\n{pred_outputs[0, 2]:.2f}"
labeldict["TOTEX"] = f"TOTEX\n{pred_outputs[0, 3]:.2f}"
labeldict["CAPEX"] = f"CAPEX\n{pred_outputs[0, 4]:.2f}"
else:
labeldict["Kapazität PV"] = f"Kapazität PV\n{pred_outputs[0, 0]:.2f} ({output_vals[0, 0]:.2f})"
labeldict["Kapazität Batterie"] = f"Kapazität Batterie\n{pred_outputs[0, 1]:.2f} ({output_vals[0, 1]:.2f})"
labeldict["Eigenerzeugung"] = f"Eigenerzeugung\n{pred_outputs[0, 2]:.2f} ({output_vals[0, 2]:.2f})"
labeldict["TOTEX"] = f"TOTEX\n{pred_outputs[0, 3]:.2f} ({output_vals[0, 3]:.2f})"
labeldict["CAPEX"] = f"CAPEX\n{pred_outputs[0, 4]:.2f} ({output_vals[0, 4]:.2f})"
if vis_real:
_, _, data_mean, data_max, _ = load_energy_data()
str_add = "\n" + r"$\rightarrow$"
input_diffs = inp_params[0, :] * data_max[:4]
real_inputs = input_diffs + data_mean[:4]
labeldict["Photovoltaik"] += str_add + f"{real_inputs[0]:.0f} ({'+' if input_diffs[0] > 0 else ''}{input_diffs[0]:.0f})"
labeldict["Batteriespeicher"] += str_add + f"{real_inputs[1]:.0f} ({'+' if input_diffs[1] > 0 else ''}{input_diffs[1]:.0f})"
labeldict["Stromnetz"] += str_add + f"{real_inputs[2]:.0f} ({'+' if input_diffs[2] > 0 else ''}{input_diffs[2]:.3f})"
labeldict["Energiebedarf"] += str_add + f"{real_inputs[3]:.0f} ({'+' if input_diffs[3] > 0 else ''}{input_diffs[3]:.0f})"
output_diffs = pred_outputs[0, :] * data_max[4:]
real_outputs = output_diffs + data_mean[4:]
labeldict["Kapazität PV"] += str_add + f"{real_outputs[0]:.2f} ({'+' if output_diffs[0] > 0 else ''}{output_diffs[0]:.2f})"
labeldict["Kapazität Batterie"] += str_add + f"{real_outputs[1]:.2f} ({'+' if output_diffs[1] > 0 else ''}{output_diffs[1]:.2f})"
labeldict["Eigenerzeugung"] += str_add + f"{real_outputs[2]:.3f} ({'+' if output_diffs[2] > 0 else ''}{output_diffs[2]:.3f})"
labeldict["TOTEX"] += str_add + f"{real_outputs[3]:.0f} ({'+' if output_diffs[3] > 0 else ''}{output_diffs[3]:.0f})"
labeldict["CAPEX"] += str_add + f"{real_outputs[4]:.0f} ({'+' if output_diffs[4] > 0 else ''}{output_diffs[4]:.0f})"
else:
labeldict["Photovoltaik"] = "Photovoltaik"
labeldict["Batteriespeicher"] = "Batteriespeicher"
labeldict["Stromnetz"] = "Stromnetz"
labeldict["Energiebedarf"] = "Energiebedarf"
labeldict["Kapazität PV"] = "Kapazität PV"
labeldict["Kapazität Batterie"] = "Kapazität Batterie"
labeldict["Eigenerzeugung"] = "Eigenerzeugung"
labeldict["TOTEX"] = "TOTEX"
labeldict["CAPEX"] = "CAPEX"
edge_list = []
edge_attr = []
for i, o in attributions:
edge_list.append((itni[i], itno[o]))
edge_attr.append(attributions[i, o])
color_bounds = np.max(np.abs(edge_attr))
cmap = plt.cm.RdBu
pos = nx.get_node_attributes(g, "pos")
nx.draw(g, pos, labels=labeldict, with_labels=True, node_color="w")
nx.draw_networkx_edges(g, pos, edgelist=edge_list, edge_color=edge_attr, edge_cmap=cmap, edge_vmin=-color_bounds,
edge_vmax=color_bounds)
if edge_labels:
e_labels = {(itni[key[0]], itno[key[1]]): f"{attributions[key]:.4f}" for key in attributions.keys()}
nx.draw_networkx_edge_labels(g, pos=pos, edge_labels=e_labels, label_pos=0.5)
sm = plt.cm.ScalarMappable(cmap=cmap, norm=plt.Normalize(vmin=-color_bounds, vmax=color_bounds))
sm.set_array([])
plt.colorbar(sm)
plt.show()
def visualize_loss(train_loss, test_loss):
""" Plot the train and test loss of a neural network after learning. One graph shows the loss progress over all
iterations, another one for only the last 10 iterations (can see whether it is still improving)."""
nr_epochs = list(range(len(train_loss)+1))[1:]
print(train_loss)
print(test_loss)
ax1 = plt.subplot(2, 1, 1)
ax1.plot(nr_epochs, train_loss, label="Train")
ax1.plot(nr_epochs, test_loss, label="Test")
ax1.set_ylabel("Loss")
ax1.set_xlabel("Epoch")
ax1.set_xticks(nr_epochs)
ax1.set_xticklabels(str(epoch) for epoch in nr_epochs)
ax1.set_title("Loss over all epochs")
ax1.grid(True)
ax1.legend()
ax2 = plt.subplot(2, 1, 2)
ax2.plot(nr_epochs[-10:], train_loss[-10:], label="Train")
ax2.plot(nr_epochs[-10:], test_loss[-10:], label="Test")
ax2.set_ylabel("Loss")
ax2.set_xlabel("Epoch")
ax2.set_xticks(nr_epochs[-10:])
ax2.set_xticklabels(str(epoch) for epoch in nr_epochs[-10:])
ax2.set_title("Loss over all epochs")
ax2.set_title("Loss over the last 10 epochs")
ax2.grid(True)
ax2.legend()
plt.subplots_adjust(hspace=0.6)
plt.show()
class EnergyNet(nn.Module):
""" Neural network learning the relationship of the input and output values as loaded by load_energy_data. """
def __init__(self, dim_input, dim_output):
""" Initialize the neural network. """
super(EnergyNet, self).__init__()
factor = 80 * 2 * 2 * 2 * 2
self.fc1 = nn.Linear(dim_input, factor * 2)
self.fc2 = nn.Linear(factor * 2, factor * 2)
self.fc2b = nn.Linear(factor * 2, factor * 2)
self.fc3 = nn.Linear(factor * 2, factor)
self.fc4 = nn.Linear(factor, dim_output)
def forward(self, param):
""" Forward pass. """
h = self.fc1(param)
h = func.relu(h)
h = self.fc2(h)
h = func.relu(h)
h = self.fc2b(h)
h = func.relu(h)
h = self.fc3(h)
h = func.relu(h)
output = self.fc4(h)
return output
def train(args, model, device, train_loader, optimizer, epoch):
""" Train the model. """
model.train()
# mean squared error loss for output
criterion = torch.nn.MSELoss()
for batch_idx, (e_input, e_output) in enumerate(train_loader):
e_input, e_output = e_input.to(device), e_output.to(device)
optimizer.zero_grad()
prediction = model(e_input)
loss = criterion(prediction, e_output)
loss.backward()
optimizer.step()
if batch_idx % args.log_interval == 0:
print("Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}".format(epoch, batch_idx * len(e_input),
len(train_loader.dataset),
100. * batch_idx / len(train_loader),
loss.item()))
if args.dry_run:
break
def test(model, device, test_loader):
""" Test the model. """
model.eval()
test_loss = 0
# mean squared error loss for output
criterion = torch.nn.MSELoss(reduction="sum")
with torch.no_grad():
for (e_input, e_output) in test_loader:
e_input, e_output = e_input.to(device), e_output.to(device)
prediction = model(e_input)
test_loss += criterion(prediction, e_output).item()
test_loss /= len(test_loader.batch_sampler)
print("\nTest set: Average loss: {:.4f}\n".format(test_loss))
return test_loss
def train_model(args):
""" Get model parameters, data and train a model. """
use_cuda = not args.no_cuda and torch.cuda.is_available()
device = torch.device("cuda" if use_cuda else "cpu")
torch.manual_seed(args.seed)
train_kwargs = {"batch_size": args.batch_size}
test_kwargs = {"batch_size": args.test_batch_size}
if use_cuda:
cuda_kwargs = {"num_workers": 1,
"pin_memory": True,
"shuffle": True}
train_kwargs.update(cuda_kwargs)
test_kwargs.update(cuda_kwargs)
e_input, e_output, _, _, _ = load_energy_data(args.seed)
# parameter
input_train = e_input[:e_input.shape[0] // 2]
input_test = e_input[e_input.shape[0] // 2:]
# use half of the data for training and testing each
output_train = e_output[:e_output.shape[0] // 2]
output_test = e_output[e_output.shape[0] // 2:]
tensor_input_train = torch.Tensor(input_train)
tensor_input_test = torch.Tensor(input_test)
tensor_output_train = torch.Tensor(output_train)
tensor_output_test = torch.Tensor(output_test)
dataset_train = TensorDataset(tensor_input_train, tensor_output_train)
dataset_test = TensorDataset(tensor_input_test, tensor_output_test)
dataloader_train = DataLoader(dataset_train, **train_kwargs)
dataloader_train_for_test = DataLoader(dataset_train, **test_kwargs)
dataloader_test = DataLoader(dataset_test, **test_kwargs)
model = EnergyNet(e_input.shape[1], e_output.shape[1]).to(device)
optimizer = optim.Adadelta(model.parameters(), lr=args.lr)
scheduler = StepLR(optimizer, step_size=1, gamma=args.gamma)
train_loss = []
test_loss = []
for epoch in range(1, args.epochs + 1):
train(args, model, device, dataloader_train, optimizer, epoch)
train_loss.append(test(model, device, dataloader_train_for_test))
test_loss.append(test(model, device, dataloader_test))
scheduler.step()
visualize_loss(train_loss, test_loss)
print(train_loss)
print(test_loss)
if args.save_model:
# save the model
save_path = f"models/Energy_{args.save_name}.pt"
torch.save(model.state_dict(), save_path)
return model
def prepare_model(args):
""" Define the model and load the state in the specified path. """
use_cuda = not args.no_cuda and torch.cuda.is_available()
device = torch.device("cuda" if use_cuda else "cpu")
e_input, e_output, _, _, _ = load_energy_data(args.seed)
model = EnergyNet(e_input.shape[1], e_output.shape[1]).to(device)
# load the model state
save_path = f"models/Energy_{args.save_name}.pt"
model.load_state_dict(torch.load(save_path))
return model
def apply_visualization(model, args):
""" Set all necessary parameters and call the right visualization method. """
use_cuda = not args.no_cuda and torch.cuda.is_available()
device = torch.device("cuda" if use_cuda else "cpu")
e_input, e_output, _, _, _ = load_energy_data(args.seed)
input_test = e_input[e_input.shape[0] // 2:]
output_test = e_output[e_output.shape[0] // 2:]
start_index = args.num_vis * args.vis_next
if start_index + args.num_vis > input_test.shape[0]:
raise ValueError(f"There are not enough test instances to visualize with respect to \"args.num_vis\": "
f"{args.num_vis} and \"args.vis_next\": {args.vis_next}")
input_test = input_test[start_index:start_index + args.num_vis]
tensor_input_vis = torch.Tensor(input_test).to(device)
output_test = output_test[start_index:start_index + args.num_vis]
attributions = {}
for i in range(args.num_vis):
# custom input
if args.vis_input:
vis_input_str = input("Enter input values ([-1, 1], 4 values, whitespace separated):")
for index, value in enumerate(vis_input_str.split()):
tensor_input_vis[i, index] = float(value)
for output_index in range(5):
ig = IntegratedGradients(model)
# this following code or the block after? What is better, and for what purpose? Both seem very similar
# for input_index in range(4):
# if args.vis_only_input != -1 and args.vis_only_input != input_index:
# continue
# # baseline
# bl = tensor_input_vis[i:i + 1].detach().clone()
# bl[0, input_index] = 0
# attr = ig.attribute(tensor_input_vis[i:i + 1], baselines=bl, target=output_index)
# attr = attr.detach().cpu().numpy()[0]
# if (input_index, output_index) in attributions:
# attributions[(input_index, output_index)] += attr[input_index]
# else:
# attributions[(input_index, output_index)] = attr[input_index]
# which baseline to use
choose_baseline = args.baseline
# baseline for all smallest and all largest inputs
if choose_baseline == "edges":
bl = tensor_input_vis[i:i + 1].detach().clone()
bl[0, :] = -1
attr = ig.attribute(tensor_input_vis[i:i + 1], baselines=bl, target=output_index)
attr = attr.detach().cpu().numpy()[0]
bl[0, :] = 1
attr2 = ig.attribute(tensor_input_vis[i:i + 1], baselines=bl, target=output_index)
attr += attr2.detach().cpu().numpy()[0]
# random: multiple baselines, uniformly distributed within [-1, 1], average for final attribution
elif choose_baseline == "random":
all_attr = None
for bls in range(10):
bl = ((torch.rand(1, 4) * 2) - 1).to(device)
attr = ig.attribute(tensor_input_vis[i:i + 1], baselines=bl, target=output_index)
attr = attr.detach().cpu().numpy()[0]
if bls == 0:
all_attr = attr
else:
all_attr += attr
attr = all_attr / 10
# gaussian: multiple baselines, gaussian distributed around 0, average for final attribution
elif choose_baseline == "gaussian":
all_attr = None
for bls in range(10):
std = 0.25 # pretty close to the underlying data std
bl = torch.normal(torch.tensor([[0.0, 0.0, 0.0, 0.0]]), std).to(device)
attr = ig.attribute(tensor_input_vis[i:i + 1], baselines=bl, target=output_index)
attr = attr.detach().cpu().numpy()[0]
if bls == 0:
all_attr = attr
else:
all_attr += attr
attr = all_attr / 10
# baseline as specified in args
else:
bl = tensor_input_vis[i:i + 1].detach().clone()
bl[0, :] = float(choose_baseline)
attr = ig.attribute(tensor_input_vis[i:i + 1], baselines=bl, target=output_index)
attr = attr.detach().cpu().numpy()[0]
for input_index in range(4):
if args.vis_only_input == -1 or args.vis_only_input == input_index:
attributions[(input_index, output_index)] = attr[input_index]
if not args.vis_agg:
pred = model(tensor_input_vis[i:i + 1]).detach().cpu().numpy()
if args.vis_input:
out_label = None
else:
out_label = output_test[i:i + 1]
vis_energy(attributions, values=(tensor_input_vis[i:i + 1].detach().cpu().numpy(), out_label, pred),
edge_labels=args.vis_only_input != -1, vis_real=args.vis_real_values)
attributions = {}
if args.vis_agg:
# pred = model(tensor_input_vis).detach().cpu().numpy()
vis_energy(attributions, edge_labels=args.vis_only_input != -1, vis_real=args.vis_real_values)
return
def prepare_arguments():
""" Define and return arguments. """
parser = argparse.ArgumentParser(description="PyTorch Energy Experiment v0_1")
# model training
parser.add_argument("--batch-size", type=int, default=64, metavar="N",
help="input batch size for training (default: 64)")
parser.add_argument("--test-batch-size", type=int, default=1000, metavar="N",
help="input batch size for testing (default: 1000)")
parser.add_argument("--epochs", type=int, default=5, metavar="N", help="number of epochs to train (default: 5)")
parser.add_argument("--lr", type=float, default=1.0, metavar="LR", help="learning rate (default: 1.0)")
parser.add_argument("--gamma", type=float, default=0.7, metavar="M", help="Learning rate step gamma (default: 0.7)")
parser.add_argument("--no-cuda", action="store_true", default=False, help="disables CUDA training")
parser.add_argument("--dry-run", action="store_true", default=False, help="quickly check a single pass")
parser.add_argument("--seed", type=int, default=0, metavar="S", help="random seed (default: 0)")
parser.add_argument("--log-interval", type=int, default=10, metavar="N",
help="how many batches to wait before logging training status")
# model saving / loading
parser.add_argument("--save-model", action="store_true", default=False, help="save the current model")
parser.add_argument("--load-model", action="store_true", default=False, help="load a model")
parser.add_argument("--save-name", type=str, default="0", metavar="NAME",
help="name with which the model will be saved or loaded")
# visualization
parser.add_argument("--vis", action="store_true", default=False, help="visualize model performance and attribution")
parser.add_argument("--num-vis", type=int, default=10, metavar="N", help="number of instanced to be visualized")
parser.add_argument("--vis-agg", action="store_true", default=False,
help="aggregate the attribution of all \"num-vis\" instances before the visualization)")
parser.add_argument("--vis-next", type=int, default=0, metavar="N",
help="skips the first vis_next * num_vis instances, can visualize other instances that way")
parser.add_argument("--vis-save", action="store_true", default=False,
help="save the visualization, otherwise simply show it")
parser.add_argument("--vis-input", action="store_true", default=False,
help="enter own inputs for the visualization")
parser.add_argument("--baseline", type=str, default="0", metavar="NAME OR NUMBER",
help="which baseline to use (\"edges\", \"random\", or a number as the baseline)")
parser.add_argument("--vis-real-values", action="store_true", default=False,
help="also show the unnormalized values on the visualization")
parser.add_argument("--vis-only-input", type=int, default=-1, metavar="N", help="only visualize for specific input")
args = parser.parse_args()
return args
def main():
""" Run the neural network with the specified arguments. """
# get arguments
args = prepare_arguments()
# get the model
if not args.load_model:
# train the model
model = train_model(args)
else:
# load the model
model = prepare_model(args)
# obtain and visualize attributions
if args.vis:
apply_visualization(model, args)
if __name__ == "__main__":
main()
| python |
import { describe, expect, it } from 'vitest'
import { summaryRanges } from '.'
describe('汇总区间', () => {
testCase(summaryRanges)
})
function testCase(fn: (nums: number[]) => string[]) {
it.each([
[[0, 1, 2, 4, 5, 7], ['0->2', '4->5', '7']],
[[0, 2, 3, 4, 6, 8, 9], ['0', '2->4', '6', '8->9']],
])('示例%#', (nums, expected) => {
expect(fn(nums)).toStrictEqual(expected)
})
}
| typescript |
<filename>src/util/math-helpers.hpp
#pragma once
#include "../stdafx.hpp"
namespace Math {
template <typename T> inline T lerp( T val0,T val1, float blend ) {
return val0*(1.0f-blend) + val1*blend;
}
inline void get_basis( Dir const& basis_y, Dir*__restrict basis_x,Dir*__restrict basis_z ) {
//http://jcgt.org/published/0006/01/01/paper.pdf
float sign = std::copysignf(1.0f,basis_y.z);
float a = -1.0f / (sign + basis_y.z);
float b = basis_y.x * basis_y.y * a;
*basis_x = Dir(
1.0f + sign*basis_y.x*basis_y.x*a,
sign*b,
-sign*basis_y.x
);
*basis_z = Dir(
b,
sign + basis_y.y*basis_y.y*a,
-basis_y.y
);
}
inline Dir get_rotated_to(Dir const& dir, Dir const& normal) {
Dir basis_x, basis_z;
get_basis(normal, &basis_x,&basis_z);
return dir.x*basis_x + dir.y*normal + dir.z*basis_z;
}
inline Dir reflect(Dir const& vec, Dir const& normal) {
return -vec + 2.0f*glm::dot(vec,normal)*normal;
}
}
| cpp |
Hello Friends,
My name is Hareshkumar Parmar @printskill and I am from India. My city name is Surendranagar that comes under the state of Gujarat of country India.
In my city, the street food called "Bhaji Con" is very famous among people. It is very spicy and with some sweet taste its taste looks so sweet. Roti made from wheat flour fried in oil and spicy vegetables is stuffed within it.
Then for decoration, some pieces of peanuts and onion pieces are placed on top of the con. People are hungry to eat this street food in India and they wait in queue for a long time for their turns to come for this food.
If you are willing to travel to India, you should taste this street food for sure.
I hope you like this street food from India.
Thank you Jury :
Thank you Supporters of this contest:
Thank you @klen.civil for organizing this Contest.
| english |
The morning shift will start at 08:30 am and will end at 12:50 pm. Whereas, the second shift will start at 01:30 pm and will conclude at 05:50 pm. The recess time is also included for the students in the timetable for the remedial classes.
The commissions found that the schools were violating the rights of children and norms laid down in Kerala Education Rules 1959 by denying them the joys of holidaying. | english |
Triple H has admitted that he and Brock Lesnar have a good relationship with each other at the moment. But that wasn't always the case.
Speaking with Ariel Helwani on BT Sport recently, The Game noted that he and Lesnar had issues over the years during The Game's days as a WWE talent.
In addition, Triple H mentioned that talent don't always get along with each other to a certain degree with everybody wanting to be the best.
During the same interview with BT Sport, Triple H spoke about Brock Lesnar walking out of WWE a few weeks ago when Vince McMahon initially announced his retirement.
According to rumors, The Game had convinced Lesnar to return. While addressing the same subject, he said:
"There's some truth to it, yeah. You have to understand Brock's relationship with Vince and if you look at Brock's relationship across combat sports period. He has a relationship with Dana [White], that's a certain way, relationship with Vince, that's a certain way. But Vince is the devil he knows. Brock is inherently not a trusting person, that's just how he is, he doesn't like people. He's not a trusting person and I think at that moment where you just hear Vince is out, now what's gonna happen?"
Lesnar's latest WWE match was at SummerSlam 2022 when he faced Roman Reigns for the Undisputed WWE Universal Championship in a Last Man Standing Match.
Despite a valiant effort, Brock Lesnar was unable to beat The Head of the Table.
If you take any quotes from this article, please credit BT Sport and provide a transcription credit to Sportskeeda Wrestling.
| english |
<gh_stars>0
{
"kallisto.refFasta": "/home/jslagel/ISB-CGC-Examples/wdl/kallisto/data/Homo_sapiens.GRCh37.cdna.all.fa.gz",
"kallisto.readsBam": [ "/home/jslagel/ISB-CGC-Examples/wdl/kallisto/data/All_CPTAC_customDB.bam" ]
}
| json |
On the same day, someone outside of the university also asked the king if he could 'bring back Harry'
King Charles III reacted kindly to a 4-year-old who just couldn’t wait his turn this week to give the monarch some flowers in a moment that has gone viral.
The little boy surprised Charles by thrusting a bouquet of flowers at the royal while he was speaking to a student during a visit to the University of East London for its 125th anniversary.
Video posted to Twitter by PA Media report Tim Sigsworth showed Charles reacting with kidding surprise as the boy stepped forward while the king was in conversation and excitedly held the flowers in front of his face with an "Eh! "
He then smiled and generously took the gift as everyone laughed.
"This young man is from our university nursery, age 4 years old," an employee advised the king after the interaction.
Charles continued to engage with the boy who looked delighted with himself and in awe following the exchange.
KING CHARLES TRAVELS WITH HIS TEDDY BEAR, CUSTOM-MADE TOILET SEAT, AUTHOR CLAIMS: 'HE WANTS WHAT HE WANTS'
The royal was at the university to help open its new primary care training hub.
Earlier, when greeting crowds outside of the university, Charles misheard someone who said, "Can you bring back Harry? " responding "Who? " according to the Mirror.
When it was clarified the person meant Prince Harry who left the U. K. in 2020 and now lives in California with his family, Charles laughed and said "Oh. "
Harry has written bluntly about his strained relationship with his father and his brother Prince William in his memoir "Spare. "
Charles, who is just months away from his coronation in May, has stepped up his duties and become the new face of the monarchy since his mother's death last September. It's unclear if Harry will attend the coronation. | english |
{"name":"<NAME>","dna":[{"code":2.6192,"color":"#FE2712","scale":0.6192},{"code":2.506,"color":"#FE2712","scale":0.506},{"code":1.7647,"color":"#FB9902","scale":0.7647},{"code":1.2348,"color":"#FABC02","scale":0.2348},{"code":1.448,"color":"#D0EA2B","scale":0.448},{"code":0.0435,"color":"#696969","scale":0.0435},{"code":1.4223,"color":"#0392CE","scale":0.4223},{"code":1.5375,"color":"#3D01A4","scale":0.5375},{"code":1.8897,"color":"#8601AF","scale":0.8897}],"attributes":[{"trait_type":"Speed","value":7},{"trait_type":"Stamina","value":6},{"trait_type":"Strength","value":8},{"trait_type":"Aggression","value":3},{"trait_type":"Creativity","value":5},{"trait_type":"Luck","value":1},{"trait_type":"Focus","value":5},{"trait_type":"Influence","value":6},{"trait_type":"Reflexes","value":9},{"trait_type":"Phobia","value":"Improv Comedy"},{"trait_type":"Vice","value":"TSA Agents"},{"trait_type":"Role","value":"Superfan"},{"trait_type":"Personality","value":"Innovative"},{"trait_type":"Class","value":"Gambler"},{"trait_type":"Affinity","value":"Body"}],"image":""} | json |
.quick-edit{
text-decoration: underline underline;
cursor: pointer;
}
| css |
<filename>app/src/main/java/selldone/api/ShopHelperMethods.java
package selldone.api;
/**
* Samin shop helper functions.
*/
public class ShopHelperMethods {
/**
* Convert file name to Samin CDN url
*
* @param file_name file name received in fetching
* @param size Optional image size
* @return Absolute image url.
*/
public static String getShopImagePath(String file_name, String size) {
return
"https://selldone.com/cdn-shop-images-1/" + file_name +
(size != null ? ("?size=" + size) : "")
;
}
/**
* Get shop icon by shop ID
*
* @param shop_id Unique shop id
* @param size Optional image size
* @return Absolute shop icon url.
*/
public static String getShopIcon(int shop_id, String size) {
return (
"https://selldone.com/xapi/shops/" + shop_id + "/icon" +
(size != null ? ("?size=" + size) : "")
);
}
public static String getUserAvatar(int user_id, String size) {
return (
"https://selldone.com/users/" + user_id + "/profile/avatar/" + (size != null ? size : "small")
);
}
}
| java |
The TS POLYCET 2023 was conducted on May 17.
The State Board of Technical Education and Training (SBTET), Telangana has announced the result of Telangana Polytechnic Common Entrance Test (TS POLYCET) 2023 today. Candidates who have appeared for the examination can download their rank cards from the official website polycet. sbtet. telangana. gov. in.
The TS POLYCET 2023 exam was conducted on May 17. Following the result announcements, the counseling process will start based on the merit lists.
TS POLYCET is being conducted for candidates seeking admission into all Diploma Courses in Engineering/Non-Engineering/Technology offered at Polytechnic colleges in Telangana State for the academic year 2023-24.
Steps to download TS POLYCET result 2023: | english |
<filename>result/caleg/caleg_kabupaten_6318_5.json
[{"namaKab":"SUMENEP","originalFilename":"1.jpg","namaPartai":"Partai NasDem","id":11047,"noUrut":1,"nama":"RAUSI","stringJenisKelamin":"Laki-Laki"},{"namaKab":"SUMENEP","originalFilename":"4.jpg","namaPartai":"Partai NasDem","id":19285,"noUrut":2,"nama":"MUSYPIQ","stringJenisKelamin":"Laki-Laki"},{"namaKab":"SUMENEP","originalFilename":"USWAH.jpg","namaPartai":"Partai NasDem","id":27811,"noUrut":3,"nama":"<NAME>","stringJenisKelamin":"Perempuan"},{"namaKab":"SUMENEP","namaPartai":"Partai NasDem","id":304540,"noUrut":4,"nama":"<NAME>","stringJenisKelamin":"Laki-Laki"},{"namaKab":"SUMENEP","originalFilename":"6.jpg","namaPartai":"Partai NasDem","id":18477,"noUrut":5,"nama":"HODAIPAH","stringJenisKelamin":"Perempuan"},{"namaKab":"SUMENEP","namaPartai":"Partai NasDem","id":304541,"noUrut":6,"nama":"BUSRI","stringJenisKelamin":"Laki-Laki"},{"namaKab":"SUMENEP","namaPartai":"Partai NasDem","id":165218,"noUrut":7,"nama":"ISY\u0027ATIYAH","stringJenisKelamin":"Perempuan"}] | json |
{"2010":"","2016":"","AdditionalNote":"","Department":"Ужгородський міськрайонний суд Закарпатської області","Link":"http://ug.zk.court.gov.ua/userfiles/stdergachovadekl.pdf","Link 2015":"","Name":"<NAME>","Note":"","Position":"Суддя Ужгородського міськрайонного суду Закарпатської області","Region":"Закарпатська область","Youtube":"","analytics":[{"ff":45.66,"ffa":1,"fi":305147.25,"i":345551.68,"k":22.84,"ka":1,"y":2014},{"ff":41.1,"ffa":2,"fi":1315,"i":174353,"m":150000,"y":2015},{"ff":73.83,"ffa":3,"fi":15000,"i":225673,"m":150000,"y":2016},{"ff":53.61,"ffa":2,"fi":16941,"i":243605,"m":150000,"y":2017},{"b":15785,"ff":198.45,"ffa":3,"fh":31,"fha":1,"fi":995587,"fl":0.04,"fla":1,"k":89.35,"ka":2,"m":398640,"y":2019}],"declarationsLinks":[{"id":"vulyk_20_151","provider":"declarations.com.ua.opendata","url":"http://static.declarations.com.ua/declarations/chosen_ones/mega_batch/derhachova_nataliia_volodymyrivna.pdf","year":2014},{"id":"nacp_4a0e26c9-10c4-451c-bf20-5d4ec803b05b","provider":"declarations.com.ua.opendata","year":2015},{"id":"nacp_18140581-57bc-490f-a41e-05246de08b10","provider":"declarations.com.ua.opendata","year":2016},{"id":"nacp_4c6dbc59-5467-4ba1-8f19-2a078cd0b71d","provider":"declarations.com.ua.opendata","year":2017},{"id":"nacp_6464461d-7e25-4ffe-9c36-2a54c59b8ed8","provider":"declarations.com.ua.opendata","year":2019}],"field8":"","field9":"","key":"dergachova_nataliya_volodimirivna","type":"judge","Декларація доброчесності судді подано у 2016 році (вперше)":"","Декларація родинних зв’язків судді подано у 2016 році":"","Декларації 2013":"","Декларації 2014":"","Декларації 2015":"","Декларації 2016":"","Клейма":"","Кількість дисциплінарних стягнень":"","Кількість скарг":"0","Кількість справ":"","Оскаржені":"","ПІБ2":"","Фото":"","Як живе":"","декларації 2015":"","судові рішення по справах Майдану":""} | json |
module.exports = function(err, req, res, next) {
// Log the exception
res.status(500).send("Something went wrong!");
} | javascript |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.