Dataset Information
===================

.. This page relies on raw HTML/JavaScript to fetch and render dataset details
   from the GitLab and NCBI APIs. Content will be empty or broken in non-HTML
   output formats (PDF, LaTeX, man pages) and requires internet access to display.

.. raw:: html

   <style>
     .ds-links {
       display: flex;
       flex-wrap: wrap;
       gap: 0.5rem;
       margin-bottom: 1.5rem;
     }
     .ds-link {
       display: inline-block;
       padding: 0.3rem 0.8rem;
       border-radius: 4px;
       font-size: 0.9rem;
       font-weight: 600;
       text-decoration: none;
     }
     .ds-link--pubmed {
       background: #dbeafe;
       color: #1e40af;
     }
     .ds-link--pubmed:hover { background: #bfdbfe; }
     .ds-link--geo {
       background: #dcfce7;
       color: #166534;
     }
     .ds-link--geo:hover { background: #bbf7d0; }
     .ds-abstract {
       border: 1px solid #e5e7eb;
       border-radius: 0.5rem;
       overflow: hidden;
       margin-bottom: 2rem;
     }
     .ds-abstract__header {
       background: #f9fafb;
       border-bottom: 1px solid #e5e7eb;
       padding: 0.6rem 1rem;
       font-weight: 700;
       font-size: 0.95rem;
     }
     .ds-abstract__citation {
       padding: 0.8rem 1rem 0;
       color: #374151;
       font-size: 0.9rem;
     }
     .ds-abstract__title {
       font-weight: 600;
       margin-bottom: 0.2rem;
     }
     .ds-abstract__authors {
       color: #6b7280;
       margin-bottom: 0.2rem;
       font-size: 0.85rem;
     }
     .ds-abstract__journal {
       color: #6b7280;
       font-size: 0.85rem;
     }
     .ds-abstract__text {
       padding: 0.8rem 1rem;
       color: #1f2937;
       font-size: 0.92rem;
       line-height: 1.65;
       white-space: pre-wrap;
       margin: 0;
       font-family: inherit;
     }
     .ds-readme__header {
       font-size: 1.1rem;
       font-weight: 700;
       margin: 1.5rem 0 0.75rem;
       padding-bottom: 0.4rem;
       border-bottom: 1px solid #e5e7eb;
     }
     #dataset-status {
       color: #4b5563;
       font-size: 0.95rem;
     }
   </style>

   <div id="dataset-status">Loading...</div>
   <div id="dataset-content"></div>

   <script>
   (function () {
     var params = new URLSearchParams(window.location.search);
     var datasetName = params.get("dataset");
     var statusEl = document.getElementById("dataset-status");
     var contentEl = document.getElementById("dataset-content");

     if (!datasetName) {
       statusEl.textContent = "No dataset selected. Return to the Datasets page.";
       return;
     }

     // Update the page <h1> to the dataset name
     var h1 = document.querySelector("h1");
     if (h1) h1.textContent = datasetName;

     var apiBase = "https://gitlab.com/api/v4";
     var projectPath = encodeURIComponent(
       "reproducible-analyses-framework-and-tools/modules/dataset-prep/" + datasetName
     );

     function escapeHtml(v) {
       return String(v).replace(/[&<>"']/g, function (c) {
         return { "&": "&amp;", "<": "&lt;", ">": "&gt;", '"': "&quot;", "'": "&#39;" }[c];
       });
     }

     // Extract PubMed and GEO URLs from README text
     function extractLinks(text) {
       var pubmedMatch = text.match(
         /https?:\/\/(?:www\.)?(?:pubmed\.ncbi\.nlm\.nih\.gov|ncbi\.nlm\.nih\.gov\/pubmed)\/(\d+)[^\s\)\]"']*/
       );
       var geoMatch = text.match(
         /https?:\/\/www\.ncbi\.nlm\.nih\.gov\/geo\/[^\s\)\]"']*/
       );
       return {
         pubmed: pubmedMatch ? { url: pubmedMatch[0].replace(/[.,;]+$/, ""), pmid: pubmedMatch[1] } : null,
         geo: geoMatch ? geoMatch[0].replace(/[.,;]+$/, "") : null
       };
     }

     // Fetch publication metadata from NCBI XML API
     function fetchPubMedInfo(pmid) {
       var url = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi" +
                 "?db=pubmed&id=" + encodeURIComponent(pmid) + "&retmode=xml";
       return fetch(url)
         .then(function (r) {
           if (!r.ok) throw new Error("HTTP " + r.status);
           return r.text();
         })
         .then(function (xml) {
           var doc = new DOMParser().parseFromString(xml, "text/xml");

           var titleEl = doc.querySelector("ArticleTitle");
           var title = titleEl ? titleEl.textContent : "";

           var abstractParts = Array.from(doc.querySelectorAll("AbstractText")).map(function (el) {
             var label = el.getAttribute("Label");
             return (label ? label + ": " : "") + el.textContent;
           });
           var abstract = abstractParts.join("\n\n");

           var authors = Array.from(doc.querySelectorAll("Author")).slice(0, 6).map(function (a) {
             var last = a.querySelector("LastName");
             var init = a.querySelector("Initials");
             return (last ? last.textContent : "") + (init ? " " + init.textContent : "");
           });
           var authorStr = authors.join(", ") + (doc.querySelectorAll("Author").length > 6 ? ", et al." : "");

           var journalEl = doc.querySelector("Journal Title");
           var journal = journalEl ? journalEl.textContent : "";

           var yearEl = doc.querySelector("PubDate Year");
           var year = yearEl ? yearEl.textContent : "";

           return { title: title, abstract: abstract, authors: authorStr, journal: journal, year: year };
         });
     }

     function loadScript(src) {
       return new Promise(function (resolve, reject) {
         if (window.marked) return resolve();
         var s = document.createElement("script");
         s.src = src;
         s.onload = resolve;
         s.onerror = function () { reject(new Error("Failed to load " + src)); };
         document.head.appendChild(s);
       });
     }

     statusEl.textContent = "Loading dataset information from GitLab...";

     // 1. Fetch project info
     fetch(apiBase + "/projects/" + projectPath, { headers: { Accept: "application/json" } })
       .then(function (r) {
         if (!r.ok) throw new Error("Dataset not found (HTTP " + r.status + ")");
         return r.json();
       })
       // 2. Fetch README
       .then(function (project) {
         var branch = encodeURIComponent(project.default_branch || "main");
         return fetch(
           apiBase + "/projects/" + encodeURIComponent(project.id) +
           "/repository/files/README.md/raw?ref=" + branch,
           { headers: { Accept: "text/plain" } }
         ).then(function (r) {
           if (!r.ok) throw new Error("README not found (HTTP " + r.status + ")");
           return r.text();
         });
       })
       // 3. Extract links, fetch PubMed info and marked.js in parallel
       .then(function (readmeText) {
         var links = extractLinks(readmeText);
         var pubmedPromise = links.pubmed
           ? fetchPubMedInfo(links.pubmed.pmid).catch(function () { return null; })
           : Promise.resolve(null);

         return Promise.all([
           Promise.resolve(readmeText),
           Promise.resolve(links),
           pubmedPromise,
           loadScript("https://cdn.jsdelivr.net/npm/marked/marked.min.js")
         ]);
       })
       // 4. Render
       .then(function (results) {
         var readmeText = results[0];
         var links      = results[1];
         var pubmed     = results[2];

         var html = "";

         // Publication link badges
         if (links.pubmed || links.geo) {
           html += '<div class="ds-links">';
           if (links.pubmed) {
             html += '<a class="ds-link ds-link--pubmed" href="' + escapeHtml(links.pubmed.url) +
                     '" target="_blank" rel="noopener">PubMed</a>';
           }
           if (links.geo) {
             html += '<a class="ds-link ds-link--geo" href="' + escapeHtml(links.geo) +
                     '" target="_blank" rel="noopener">GEO</a>';
           }
           html += "</div>";
         }

         // Abstract box
         if (pubmed && pubmed.abstract) {
           html += '<div class="ds-abstract">';
           html += '<div class="ds-abstract__header">Abstract</div>';
           html += '<div class="ds-abstract__citation">';
           if (pubmed.title) {
             html += '<div class="ds-abstract__title">' + escapeHtml(pubmed.title) + "</div>";
           }
           if (pubmed.authors) {
             html += '<div class="ds-abstract__authors">' + escapeHtml(pubmed.authors) + "</div>";
           }
           if (pubmed.journal || pubmed.year) {
             html += '<div class="ds-abstract__journal">' +
                     escapeHtml([pubmed.journal, pubmed.year].filter(Boolean).join(", ")) +
                     "</div>";
           }
           html += "</div>";
           html += '<pre class="ds-abstract__text">' + escapeHtml(pubmed.abstract) + "</pre>";
           html += "</div>";
         }

         // Full README
         html += '<div class="ds-readme__header">README</div>';
         html += '<div id="ds-readme-content">' + window.marked.parse(readmeText) + "</div>";

         statusEl.style.display = "none";
         contentEl.innerHTML = html;
       })
       .catch(function (err) {
         statusEl.textContent = "Unable to load dataset: " + String(err && err.message || err);
       });
   })();
   </script>