frenzy-ios/ReaderTests/ReaderTests.swift

//
//  ReaderTests.swift
//  ReaderTests
//
//  Created by Shadowfacts on 10/29/21.
//

import XCTest
import Reader
//import SwiftSoup

class ReaderTests: XCTestCase {

    override func setUpWithError() throws {
        // Put setup code here. This method is called before the invocation of each test method in the class.
    }

    override func tearDownWithError() throws {
        // Put teardown code here. This method is called after the invocation of each test method in the class.
    }

    func testExample() throws {
        // This is an example of a functional test case.
        // Use XCTAssert and related functions to verify your tests produce the correct results.
    }
    
    // HTML parsing comparison conducted on iPhone 12 Pro
    // SwiftSoup: approx 0.53 sec
    // lol-html: approx 0.003 sec

    // note: when testing this, make sure to set the Reader scheme to build in release mode for fair comparison
//    func testSwiftSoupPerformance() {
//        self.measure {
//            for i in 0..<100 {
//                if i % 10 == 0 {
//                    print(i)
//                }
//                let doc = try! SwiftSoup.parseBodyFragment(html)
//                let excerpt = try! doc.text()
//            }
//        }
//    }
    
    func testLolHtmlPerformance() {
        self.measure {
            for i in 0..<100 {
                if i % 10 == 0 {
                    print(i)
                }
                let excerpt = ExcerptGenerator.excerpt(from: html)
            }
        }
    }
    
    let html = """
    <div class="page" id="readability-page-1"><div data-reactid="221"><div data-reactid="222"><p><u>On Tuesday, Republicans</u> on the House Committee on Oversight and Reform <a href="https://republicans-oversight.house.gov/wp-content/uploads/2022/01/Letter-Re.-Feb-1-Emails-011122.pdf">released a letter </a>that
     paints a damning picture of U.S. government officials wrestling with
    whether the novel coronavirus may have leaked out of a lab they were
    funding, acknowledging that it may have, and then keeping the discussion
     from spilling out into public view.</p><p>The letter, signed by James Comer, R-Ky., and Jim Jordan, R-Ohio, was followed by pages of notes on <a href="https://s3.documentcloud.org/documents/20793561/leopold-nih-foia-anthony-fauci-emails.pdf">emails</a> that were first obtained through the Freedom of Information Act by <a href="https://www.buzzfeednews.com/article/nataliebettendorf/fauci-emails-covid-response">BuzzFeed News</a> and the <a href="https://www.washingtonpost.com/politics/interactive/2021/tony-fauci-emails/">Washington Post</a>,
     but were heavily redacted when published in June 2021. The redacted
    emails included the agenda for a February 1, 2020, telephone conference
    between National Institute of Allergy and Infectious Diseases director
    Anthony Fauci; his then-boss, former National Institutes of Health
    director Francis Collins; and several of the world’s leading
    virologists. The communications contained extensive notes summarizing
    what was said during the call, but their substance was hidden at the
    time.</p><p>Oversight Committee staff were able to view the full emails
    “in camera,” which means they could physically look at them and take
    notes but couldn’t take copies with them. The information released
    Tuesday for the first time reveals the content of notes taken on the
    February 1 call.</p></div><a data-reactid="223" href="https://beta.documentcloud.org/documents/21177759-house-oversight-letter-and-email-transcriptions"><img data-reactid="224" src="data:image/jpeg;base64,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
     that call, virologists Michael Farzan and Robert Garry told Fauci and
    Collins the virus might have leaked from the Wuhan lab. It might have
    been genetically engineered, the transcription of Garry’s notes
    suggests, but this now seems unlikely. Another possibility, put forward
    by Farzan, was that it could have been evolved in the lab through a
    process known as serial passage.</p><p>“The email is out-of-context,”
    Garry wrote Wednesday in an email to The Intercept. “This was one email
    among many I was sharing with my colleagues.”</p></div><div data-reactid="283"><p>The
     two methods represent two different ideas behind the so-called lab-leak
     hypothesis. The one that gained notoriety early in the pandemic is
    genetic engineering, where scientists insert and delete nucleotides in
    the virus’s genetic code, in this case viral RNA, to turn it into
    something new. This version forms the basis of accusations that the
    virus was intentionally created as a bioweapon — which practically every
     credible scientist has dismissed as an illogical conspiracy, but was
    quickly embraced by former President Donald Trump and much of the
    American right wing, souring scientists, liberals, and the mainstream on
     the possibility of lab origin. The less lurid but seemingly more
    plausible version is the idea of evolution through serial passage, in
    which scientists allow a virus to jump between host species or cell
    cultures, spurring new mutations.</p><p>The day before the call, Scripps
     Research infectious disease expert Kristian Andersen had warned Fauci
    that the virus may have been engineered in a lab, noting that he and
    several other high-profile scientists “all find the genome inconsistent
    with expectations from evolutionary theory.” The scientists agreed to
    have a conference call the next day. “It was a very productive
    back-and-forth conversation where some on the call felt it could
    possibly be an engineered virus,” Fauci told Alison Young, <a href="https://www.usatoday.com/story/opinion/2021/06/17/covid-19-fauci-lab-leaks-wuhan-china-origins/7737494002/">writing for USA Today</a>, in June 2021.</p><p>Not long after the call, Andersen was the lead author on a <a href="https://www.nature.com/articles/s41591-020-0820-9">paper in Nature Medicine</a>
     titled “The Proximal Origin of SARS-CoV-2.” The paper proposed “two
    scenarios that can plausibly explain the origin of SARS-CoV-2: (i)
    natural selection in an animal host before zoonotic transfer; and (ii)
    natural selection in humans following zoonotic transfer.” For the
    scientists and pundits who sought to discount the emerging lab-leak
    hypothesis, it offered the authoritative proof they needed. The paper
    has since been accessed more than 5.6 million times, with over 2,000
    citations.</p><p>The authors acknowledged a third scenario, “selection
    during passage,” but they discussed it briefly and presented it as by
    far the least plausible. The newly released notes from the call,
    however, suggest that the scientists Fauci consulted initially
    considered that possibility to be much more serious than the paper let
    on.</p></div><div data-reactid="285"><p>On February 2, Jeremy Farrar, an
     infectious disease expert and the director of Wellcome, sent around
    notes, including to Fauci and Collins, summarizing what some of the
    scientists had said on the call. Farzan, a Scripps professor who studied
     the spike protein on the 2003 SARS virus, “is bothered by the furin
    site and has a hard time explain that as an event outside the lab
    (though, there are possible ways in nature, but highly unlikely),”
    Farrar’s note reads, referring to a spike protein feature that aids
    interaction with furin, a common enzyme in human lung cells. Farzan
    didn’t think the site was the product of “directed engineering,” but
    found that the changes would be “highly compatible with the idea of
    continued passage of the virus in tissue culture.”</p><p>According to
    the transcribed notes, Garry, a professor at the Tulane University
    School of Medicine, said on the call that he had aligned the SARS-CoV-2
    genome with that of RaTG13, a 96-percent similar virus isolated from
    bats at the Wuhan Institute of Virology that was long regarded as the
    new virus’s closest known relative — though a closer one <a href="https://www.nature.com/articles/d41586-021-02596-2">has since been identified</a>.
     Garry found that the spike proteins of RaTG13 and SARS-CoV-2, which
    makes the latter so infectious, were nearly identical. The key
    distinction was in the ability of the new virus’s spike protein to
    interact with furin, which Garry found too perfect to make natural
    sense. “I just can’t figure out how this gets accomplished in nature,”
    he said.</p><p>“My initial impression and that of others about the
    [furin cleavage site] was wrong. I changed my mind with new
    information/new data,” Garry wrote to The Intercept. “That’s how science
     works. No one was trying to mislead the public. What was in the
    Proximal Origins paper was our best analysis — it’s held up extremely
    well.”</p></div><div data-reactid="287"><p>As they discussed what to
    present to the public, the scientists determined that questions of
    potential lab origin might prove more trouble than they’re worth. “Given
     the evidence presented and the discussions around it, I would conclude
    that a follow-up discussion on the possible origin of 2019-nCoV would be
     of much interest,” wrote Ron Fouchier, a virologist at the Erasmus MC
    Center for Viroscience in the Netherlands, on February 2. Years earlier,
     Fouchier’s gain-of-function research had brought the discipline under
    fire for <a href="https://www.science.org/content/article/exclusive-controversial-experiments-make-bird-flu-more-risky-poised-resume">a 2011 experiment</a>
     in which he infected ferrets in adjacent cages with the avian influenza
     virus, allowing it to become airborne and infect mammals. “However,
    further debate about such accusations would unnecessarily distract top
    researchers from their active duties and do unnecessary harm to science
    in general and science in China in particular,” Fouchier wrote.</p><p>Farzan, Fauci, and Fouchier did not immediately respond to The Intercept’s requests for comment.</p><p>Several
     of the scientists on the email chain ended up co-authoring the Nature
    Medicine paper with Andersen and Garry. In a February 4 email, which
    House Republicans presented as a response to a first copy of the draft,
    Fauci wrote: “?? Serial passage in ACE2-transgenic mice.”</p><p>The
    early draft has not been made public, so we don’t know what, exactly,
    sparked Fauci’s reaction. But his words, which refer to the process of
    passaging a virus in “humanized” laboratory mice — or mice that have
    been genetically modified to express receptors for human ACE2, an enzyme
     that occurs in the lungs — do not appear in the published paper.</p><p>“Neither
     Drs. Fauci or Collins edited our Proximal Origins paper in any way. The
     major feedback we got from the Feb 1 teleconference was: 1. Don’t try
    to write a paper at all — it’s unnecessary or 2. If you do write it
    don’t mention a lab origin as that will just add fuel to the
    conspiracists,” Garry wrote on Wednesday.</p><p>When the paper appeared
    in Nature Medicine on March 17, 2020, it noted near the end that in
    order for the novel coronavirus to have emerged in a lab via serial
    passage, scientists would have to conduct those experiments using a
    relative with very high genetic similarity, but there was no evidence
    that such experiments had been done. The authors added, “Subsequent
    generation of a polybasic cleavage site,” which lets the virus process
    furin, “would have then required repeated passage in cell culture or
    animals with ACE2 receptors similar to those of humans, but such work
    has also not previously been described.”</p><p>Though the paper was
    publicly embraced by the scientific community and the mainstream media,
    Collins worried that its impact wasn’t sufficient. “Wondering if there
    is something NIH can do to help put down this very destructive
    conspiracy,” Collins wrote on April 16, 2020, in reference to a Fox News
     segment on the lab-leak theory. “I hoped the Nature Medicine article on
     the genomic sequence of SARS-CoV-2 would settle this. But probably
    didn’t get much visibility. Anything more we can do?”</p><p>“I would not do anything about this right now,” Fauci replied. “It is a shiny object that will go away in times.”</p></div></div></div>
    """

}