GoogleScholarScraper::ScrapeDoc doesn't work

[GerHobbelt]

The first line in there:

HtmlNodeCollection NoAltElements_outer = doc.DocumentNode.SelectNodes("//*[@class='gs_r']");

always delivers NULL.

Nett result: Google Scholar Scrape activity will never deliver anything.

Analysis / Resolution

https://stackoverflow.com/questions/13771083/html-agility-pack-get-all-elements-by-class

vs. Google Scholar spitting out HTML like this:

              <div class="gs_r gs_or gs_scl" data-cid="ngzxngyKFX0J" data-did="ngzxngyKFX0J" data-lid="" data-rp="0">
                <div class="gs_ggs gs_fl">
                  <div class="gs_ggsd">
                    <div class="gs_or_ggsm" ontouchstart="gs_evt_dsp(event)" tabindex="-1"><a href="https://commons.emich.edu/cgi/viewcontent.cgi?article=1193&amp;context=loexquarterly" data-clk="hl=nl&amp;sa=T&amp;oi=gga&amp;ct=gga&amp;cd=0&amp;d=9013262016062753950&amp;ei=NkirXeGUFIjTmQGml5nIAg&amp;scisig=AAGBfm2DYn4rw9QE9Orpp4GAaHPhnaBB8w&amp;nossl=1" data-clk-atid="ngzxngyKFX0J"><span class=gs_ctg2>[PDF]</span> emich.edu</a></div>
                  </div>
                </div>
                <div class="gs_ri">
                  <h3 class="gs_rt" ontouchstart="gs_evt_dsp(event)"><span class="gs_ctc"><span class="gs_ct1">[PDF]</span><span class="gs_ct2">[PDF]</span></span> <a id="ngzxngyKFX0J" href="https://commons.emich.edu/cgi/viewcontent.cgi?article=1193&amp;context=loexquarterly" data-clk="hl=nl&amp;sa=T&amp;oi=ggp&amp;ct=res&amp;cd=0&amp;d=9013262016062753950&amp;ei=NkirXeGUFIjTmQGml5nIAg&amp;scisig=AAGBfm2DYn4rw9QE9Orpp4GAaHPhnaBB8w&amp;nossl=1" data-clk-atid="ngzxngyKFX0J">TechMatters:&quot; <b>Qiqqa</b>&quot; than you can say Reference Management: <b>A </b>Tool to Organize <b>the </b>Research Process</a></h3>
                  <div class="gs_a">K Graham - LOEX Quarterly, 2013 - commons.emich.edu</div>
                  <div class="gs_rs">… however, they are well documented in <b>the</b> <b>manual</b> that is included in <b>the</b> “guest� library that come<br>
                    with initial download of <b>the</b> software … Whether searching for <b>a</b> means to organize your own research<br>
                    process or seeking <b>a</b> tool that you can recommend to students, <b>Qiqqa</b> is most … 
                  </div>

(after pulling the HTML through an online formatter which very probably b0rked the Unicode in there...)

shows that the solution can be had from the last SO answer:

HtmlNodeCollection NoAltElements_outer = doc.DocumentNode.SelectNodes("//*[contains(@class,'gs_r')]");

[GerHobbelt]

Solution reference: https://stackoverflow.com/questions/13771083/html-agility-pack-get-all-elements-by-class#answer-14087707

[GerHobbelt]

As .NET only supports XPath 1.0 the tokenize(...) solution doesn't fly.

Binging on SO:

• https://stackoverflow.com/questions/1390568/how-can-i-match-on-an-attribute-that-contains-a-certain-string -- using concatenate whitespace around class value and then 'fake' word boundary via contains with surrounding spaces: this is what we'll use, basically
• https://stackoverflow.com/questions/17352340/tokenize-or-split-string-using-xsl-in-visual-studio
• https://stackoverflow.com/questions/53229668/xpath-normalize-space-with-contains

[GerHobbelt]

Fixed in dev repo for v82 release: https://github.com/GerHobbelt/qiqqa-open-source/tree/v82-build