<li>Then I cleared my local Ansible fact cache and re-ran the <ahref="https://github.com/ilri/rmg-ansible-public">infrastructure playbooks</a></li>
<li>Open <ahref="https://github.com/ilri/OpenRXV/issues/111">an issue for the value mappings global replacement bug in OpenRXV</a></li>
<li>Advise Peter and Abenet on expected CGSpace budget for 2022</li>
<li>Start a fresh harvesting on AReS (linode20)</li>
</ul>
<h2id="2021-08-02">2021-08-02</h2>
<ul>
<li>Help Udana with OAI validation on CGSpace
<ul>
<li>He was checking the OAI base URL on OpenArchives and I had to verify the results in order to proceed to Step 2</li>
<li>Now it seems to be verified (all green): <ahref="https://www.openarchives.org/Register/ValidateSite?log=R23ZWX85">https://www.openarchives.org/Register/ValidateSite?log=R23ZWX85</a></li>
<li>We are listed in the OpenArchives list of databases conforming to OAI 2.0</li>
</ul>
</li>
</ul>
<h2id="2021-08-03">2021-08-03</h2>
<ul>
<li>Run fresh re-harvest on AReS</li>
</ul>
<h2id="2021-08-05">2021-08-05</h2>
<ul>
<li>Have a quick call with Mishell Portilla from CIP about a journal article that was flagged as being in a predatory journal (Beall’s List)
<ul>
<li>We agreed to unmap it from RTB’s collection for now, and I asked for advice from Peter and Abenet for what to do in the future</li>
</ul>
</li>
<li>A developer from the Alliance asked for access to the CGSpace database so they can make some integration with PowerBI
<ul>
<li>I told them we don’t allow direct database access, and that it would be tricky anyways (that’s what APIs are for!)</li>
</ul>
</li>
<li>I’m curious if there are still any requests coming in to CGSpace from the abusive Russian networks
<ul>
<li>I extracted all the unique IPs that nginx processed in the last week:</li>
<li>Advise IWMI colleagues on best practices for thumbnails</li>
<li>Add a handful of mappings for incorrect countries, regions, and licenses on AReS and start a new harvest
<ul>
<li>I sent a message to Jacquie from WorldFish to ask if I can help her clean up the incorrect countries and regions in their repository, for example:</li>
<li>WorldFish countries: Aegean, Euboea, Caribbean Sea, Caspian Sea, Chilwa Lake, Imo River, Indian Ocean, Indo-pacific</li>
<li>WorldFish regions: Black Sea, Arabian Sea, Caribbean Sea, California Gulf, Mediterranean Sea, North Sea, Red Sea</li>
</ul>
</li>
<li>Looking at the July Solr statistics to find the top IP and user agents, looking for anything strange
<ul>
<li>35.174.144.154 made 11,000 requests last month with the following user agent:</li>
</ul>
</li>
</ul>
<divclass="highlight"><pretabindex="0"style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><codeclass="language-console"data-lang="console"><spanstyle="display:flex;"><span>Mozilla/5.0 (Macintosh; Intel Mac OS X 10_14_3) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/78.0.3904.97 Safari/537.36
</span></span></code></pre></div><ul>
<li>That IP is on Amazon, and from looking at the DSpace logs I don’t see them logging in at all, only scraping… so I will purge hits from that IP</li>
<li>I see 93.158.90.30 is some Swedish IP that also has a normal-looking user agent, but never logs in and requests thousands of XMLUI pages, I will purge their hits too
<ul>
<li>Same deal with 130.255.162.173, which is also in Sweden and makes requests every five seconds or so</li>
<li>Same deal with 93.158.90.91, also in Sweden</li>
</ul>
</li>
<li>3.225.28.105 uses a normal-looking user agent but makes thousands of request to the REST API a few seconds apart</li>
<li>61.143.40.50 is in China and uses this hilarious user agent:</li>
</ul>
<divclass="highlight"><pretabindex="0"style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><codeclass="language-console"data-lang="console"><spanstyle="display:flex;"><span>Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/79.0.3945.{random.randint(0, 9999)} Safari/537.{random.randint(0, 99)}"
</span></span></code></pre></div><ul>
<li>47.252.80.214 is owned by Alibaba in the US and has the same user agent</li>
<li>159.138.131.15 is in Hong Kong and also seems to be a bot because I never see it log in and it downloads 4,300 PDFs over the course of a few hours</li>
<li>95.87.154.12 seems to be a new bot with the following user agent:</li>
<li>129.0.211.251 is in Cameroon and uses a normal-looking user agent, but seems to be a bot of some sort, as it downloaded 900 PDFs over a short period.</li>
<li>217.182.21.193 is on OVH in France and uses a Linux user agent, but never logs in and makes several requests per minute, over 1,000 in a day</li>
<li>103.135.104.139 is in Hong Kong and also seems to be making real requests, but makes way too many to be a human</li>
<li>There are probably more but that’s most of them over 1,000 hits last month, so I will purge them:</li>
</span></span></span><spanstyle="display:flex;"><span><spanstyle="color:#960050;background-color:#1e0010"></span>Total number of bot hits purged: 90485
</span></span></code></pre></div><ul>
<li>Then I purged a few thousand more by user agent:</li>
<li>Then in OpenRefine I merged all null, blank, and en fields into the <code>en_US</code> one for each, removed all spaces, fixed invalid multi-value separators, removed everything other than ISSN/ISBNs themselves
<ul>
<li>In total it was a few thousand metadata entries or so so I had to split the CSV with <code>xsv split</code> in order to process it</li>
<li>I was reminded again how DSpace 6 is very fucking slow when it comes to any database-related operations, as it takes over an hour to process 200 metadata changes…</li>
<li>In total it was 1,195 changes to ISSN and ISBN metadata fields</li>
</ul>
</li>
</ul>
<h2id="2021-08-09">2021-08-09</h2>
<ul>
<li>Extract all unique ISSNs to look up on Sherpa Romeo and Crossref</li>
<li>In OpenRefine I faceted by blank in each column and copied the values from the other, then created a new column to indicate whether the values were the same with this GREL:</li>
<li>Then I exported the list of journals that differ and sent it to Peter for comments and corrections
<ul>
<li>I want to build an updated controlled vocabulary so I can update CGSpace and reconcile our existing metadata against it</li>
</ul>
</li>
<li>Convert my <code>generate-thumbnails.py</code> script to use libvips instead of Graphicsmagick
<ul>
<li>It is faster and uses less memory than GraphicsMagick (and ImageMagick), and produces nice thumbnails from PDFs</li>
<li>One drawback is that libvips uses Poppler instead of Graphicsmagick, which apparently means that it can’t work in CMYK</li>
<li>I tested one item (10568/51999) that uses CMYK and the thumbnail looked OK (closer to the original than GraphicsMagick), so I’m not sure…</li>
<li>Perhaps this is not a problem after all, see this PR from 2019: <ahref="https://github.com/libvips/libvips/pull/1196">https://github.com/libvips/libvips/pull/1196</a></li>
</ul>
</li>
<li>I did some tests of the memory used and time elapsed with libvips, GraphicsMagick, and ImageMagick:</li>
<li>The ImageMagick way is the same as how DSpace does it (first creating an intermediary image, then getting a thumbnail)
<ul>
<li>libvips does use less time and memory… I should do more tests!</li>
<li>I wonder if I can try to use these <ahref="https://github.com/criteo/JVips/blob/master/src/test/java/com/criteo/vips/example/SimpleExample.java">unofficial Java bindings</a> in DSpace</li>
<li>The authors of the JVips project wrote a nice blog post about libvips performance: <ahref="https://medium.com/criteo-engineering/boosting-image-processing-performance-from-imagemagick-to-libvips-268cc3451d55">https://medium.com/criteo-engineering/boosting-image-processing-performance-from-imagemagick-to-libvips-268cc3451d55</a></li>
<li>Ouch, JVips is Java 8 only as far as I can tell… that works now, but it’s a non-starter going forward</li>
</ul>
</li>
</ul>
<h2id="2021-08-11">2021-08-11</h2>
<ul>
<li>Peter got back to me about the journal title cleanup
<ul>
<li>From his corrections it seems an overwhelming majority of his choices match the Sherpa Romeo version of the titles rather than Crossref’s</li>
<li>Anyways, I exported the originals that were the same in Sherpa Romeo and Crossref as well as Peter’s selections for where Sherpa Romeo and Crossref differred:</li>
<li>Now I will create a controlled vocabulary out of this list and reconcile our existing journal title metadata with it in OpenRefine</li>
<li>I exported a list of all the journal titles we have in the <code>cg.journal</code> field:</li>
</ul>
<divclass="highlight"><pretabindex="0"style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><codeclass="language-console"data-lang="console"><spanstyle="display:flex;"><span>localhost/dspace63= > \COPY (SELECT DISTINCT(text_value) AS "cg.journal" FROM metadatavalue WHERE dspace_object_id in (SELECT dspace_object_id FROM item) AND metadata_field_id IN (251)) to /tmp/2021-08-11-journals.csv WITH CSV;
<li>I started looking at reconciling them with reconcile-csv in OpenRefine, but ouch, there are 1,600 journal titles that don’t match, so I’d have to go check many of them manually before selecting a match or fixing them…
<ul>
<li>I think it’s better if I try to write a Python script to fetch the ISSNs for each journal article and update them that way</li>
<li>Or instead of doing it via SQL I could use CSV and parse the values there…</li>
</ul>
</li>
<li>A few more issues:
<ul>
<li>Some ISSNs are non-existent in Sherpa Romeo and Crossref, but appear on issn.org’s web search (their API is invite only)</li>
<li>Some titles are different across all three datasets, for example ISSN 0003-1305:
<ul>
<li><ahref="https://portal.issn.org/resource/ISSN/0003-1305">According to ISSN.org</a> this is “The American statistician”</li>
<li><ahref="https://v2.sherpa.ac.uk/id/publication/20807">According to Sherpa Romeo</a> this is “American Statistician”</li>
<li><ahref="https://search.crossref.org/?q=0003-1305&from_ui=yes&container-title=The+American+Statistician">According to Crossref</a> this is “The American Statistician”</li>
</ul>
</li>
</ul>
</li>
<li>I also realized that our previous controlled vocabulary came from CGSpace’s top 500 journals, so when I replaced it with the generated list earlier today we lost some journals
<ul>
<li>Now I went back and merged the previous with the new, and manually removed duplicates (sigh)</li>
<li>I requested access to the issn.org OAI-PMH API so I can use their registry…</li>
</ul>
</li>
</ul>
<h2id="2021-08-12">2021-08-12</h2>
<ul>
<li>I sent an email to Sherpa Romeo’s help contact to ask about missing ISSNs
<ul>
<li>They pointed me to their <ahref="https://v2.sherpa.ac.uk/romeo/about.html">inclusion criteria</a> and said that missing journals should submit their open access policies to be included</li>
</ul>
</li>
<li>The contact from issn.org got back to me and said I should pay 1,000/year EUR for 100,000 requests to their API… no thanks</li>
<li>Submit a pull request to COUNTER-Robots for the httpx bot (<ahref="https://github.com/atmire/COUNTER-Robots/pull/45">#45</a>)
<ul>
<li>In the mean time I added it to our local ILRI overrides</li>
</ul>
</li>
</ul>
<h2id="2021-08-15">2021-08-15</h2>
<ul>
<li>Start a fresh reindex on AReS</li>
</ul>
<h2id="2021-08-16">2021-08-16</h2>
<ul>
<li>Meeting with Abenet and Peter about CGSpace actions and future
<ul>
<li>We agreed to move three top-level Feed the Future projects into one community, so I created one and moved them:</li>
<li>I made a minor fix to OpenRXV to prefix all image names with <code>docker.io</code> so it works with less changes on podman
<ul>
<li>Docker assumes the <code>docker.io</code> registry by default, but we should be explicit</li>
</ul>
</li>
</ul>
<h2id="2021-08-17">2021-08-17</h2>
<ul>
<li>I made an initial attempt on the policy statements page on DSpace Test
<ul>
<li>It is modeled on Sherpa Romeo’s OpenDOAR policy statements advice</li>
</ul>
</li>
<li>Sit with Moayad and discuss the future of AReS
<ul>
<li>We specifically discussed formalizing the API and documenting its use to allow as an alternative to harvesting directly from CGSpace</li>
<li>We also discussed allowing linking to search results to enable something like “Explore this collection” links on CGSpace collection pages</li>
</ul>
</li>
<li>Lower case all AGROVOC metadata, as I had noticed a few in sentence case:</li>
</ul>
<divclass="highlight"><pretabindex="0"style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><codeclass="language-console"data-lang="console"><spanstyle="display:flex;"><span>dspace=# UPDATE metadatavalue SET text_value=LOWER(text_value) WHERE dspace_object_id IN (SELECT uuid FROM item) AND metadata_field_id=187 AND text_value ~ '[[:upper:]]';
<li>Also update some DOIs using the <code>dx.doi.org</code> format, just to keep things uniform:</li>
</ul>
<divclass="highlight"><pretabindex="0"style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><codeclass="language-console"data-lang="console"><spanstyle="display:flex;"><span>dspace=# UPDATE metadatavalue SET text_value = regexp_replace(text_value, 'https://dx.doi.org', 'https://doi.org') WHERE dspace_object_id IN (SELECT uuid FROM item) AND metadata_field_id = 220 AND text_value LIKE 'https://dx.doi.org%';
<li>Then start a full Discovery re-indexing to update the Feed the Future community item counts that have been stuck at 0 since we moved the three projects to be a subcommunity a few days ago:</li>
<li>I learned how to use the OpenRXV API, which is just a thin wrapper around Elasticsearch:</li>
</ul>
<divclass="highlight"><pretabindex="0"style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;"><codeclass="language-console"data-lang="console"><spanstyle="display:flex;"><span>$ curl -X POST <spanstyle="color:#e6db74">'https://cgspace.cgiar.org/explorer/api/search?scroll=1d'</span><spanstyle="color:#ae81ff">\
</span></span><spanstyle="display:flex;"><span>$ curl -X POST <spanstyle="color:#e6db74">'https://cgspace.cgiar.org/explorer/api/search/scroll/DXF1ZXJ5QW5kRmV0Y2gBAAAAAAAASekWMTRwZ3lEMkVRYUtKZjgyMno4dV9CUQ=='</span>
</span></span></code></pre></div><ul>
<li>This uses the Elasticsearch scroll ID to page through results
<ul>
<li>The second query doesn’t need the request body because it is saved for 1 day as part of the first request</li>
</ul>
</li>
<li>Attempt to re-do my tests with VisualVM from 2019-04
<ul>
<li>I found that I can’t connect to the Tomcat JMX port using SSH forwarding (visualvm gives an error about localhost already being monitored)</li>
<li>Instead, I had to create a SOCKS proxy with SSH (ssh -D 8096), then set that up as a proxy in the VisualVM network settings, and then add the JMX connection</li>
<li>I have to spend more time on this…</li>
</ul>
</li>
<li>I fixed a bug in the Altmetric donuts on OpenRXV
<ul>
<li>We now <ahref="https://github.com/ilri/OpenRXV/pull/113">try to show the donut for the DOI first if it exists, then fall back to the Handle</a></li>
<li>This is working on my local test, but not on the live site… sigh</li>
<li>I started a fresh harvest, maybe it’s something to do with the metadata in Elasticsearch</li>
</ul>
</li>
<li>I improved the quality of the “no thumbnail” placeholder image on AReS: <ahref="https://github.com/ilri/OpenRXV/pull/114">https://github.com/ilri/OpenRXV/pull/114</a></li>
<li>I sent some feedback to some ILRI and CCAFS colleagues about how to use better thumbnails for publications</li>
</ul>
<h2id="2021-08-24">2021-08-24</h2>
<ul>
<li>In the last few days I did a lot of work on OpenRXV
<ul>
<li>I started exploring the Angular 9.0 to 9.1 update</li>
<li>I tested some updates to dependencies for Angular 9 that we somehow missed, like @tinymce/tinymce-angular, @nicky-lenaers/ngx-scroll-to, and @ng-select/ng-select</li>
<li>I changed the default target from ES5 to ES2015 because ES5 was released in 2009 and the only thing we lose by moving to ES2015 is IE11 support</li>
<li>I fixed a handful of issues in the Docker build and deployment process</li>
<li>I started exploring changing the Docker configuration from using volumes to <code>COPY</code> instructions in the <code>Dockerfile</code> because we are having sporadic issues with permissions in containers caused by copying the host’s frontend/backend directories and not being able to write to them</li>
<li>I tested moving from node-sass to sass, as it has been <ahref="https://blog.ninja-squad.com/2019/05/29/angular-cli-8.0/">supported since Angular 8 apparently</a> and will allow us to avoid stupid node-gyp issues</li>
</ul>
</li>
</ul>
<h2id="2021-08-25">2021-08-25</h2>
<ul>
<li>I did a bunch of tests of the OpenRXV Angular 9.1 update and merged it to master (<ahref="https://github.com/ilri/OpenRXV/pull/115">#115</a>)</li>
<li>Last week Maria Garruccio sent me a handful of new ORCID identifiers for Bioversity staff
<ul>
<li>We currently have 1320 unique identifiers, so this adds eleven new ones:</li>
<li>Tag existing items from the Alliance’s new authors with ORCID iDs using <code>add-orcid-identifiers-csv.py</code> (181 new metadata fields added):</li>
<li>Also do more work the past few days on OpenRXV
<ul>
<li>I switched the backend target from ES2017 to ES2019</li>
<li>I did a proof of concept with multi-stage builds and simplifying the Docker configuration</li>
</ul>
</li>
<li>Update the list of ORCID identifiers on CGSpace</li>
<li>Run system updates and reboot CGSpace (linode18)</li>
</ul>
<h2id="2021-08-31">2021-08-31</h2>
<ul>
<li>Yesterday I finished the work to make OpenRXV use a new multi-stage Docker build system and use smarter <code>COPY</code> instructions instead of runtime volumes
<ul>
<li>Today I merged the changes to the master branch and re-deployed AReS on linode20</li>
<li>Because the <code>docker-compose.yml</code> moved to the root the Docker volume prefix changed from <code>docker_</code> to <code>openrxv_</code> so I had to stop the containers and rsync the data from the old volume to the new one in /var/lib/docker</li>
