This commit is contained in:
2026-07-15 17:55:56 -05:00
parent 2925f8aa35
commit 741b38b4f6
7 changed files with 737 additions and 221 deletions

View File

@@ -1,12 +1,12 @@
let state;
let groups = [];
let selectedIndex = -1;
let currentPdfUrl = "";
let pdfViewer = null;
const people = document.querySelector("#people");
const fields = document.querySelector("#fields");
const pdf = document.querySelector("#pdf");
const viewer = document.querySelector(".viewer");
const pdfContainer = document.querySelector("#pdfViewer");
const pdfMessage = document.querySelector("#pdfMessage");
const status = document.querySelector("#status");
const search = document.querySelector("#search");
@@ -23,6 +23,7 @@ const fieldDefs = [
null,
["Address", "address"],
["State", "state"],
null,
["Businesses from Notes", "notes_business_raw"],
["Types of Businesses", "types_of_business_raw"],
["Background Experience", "background_experience"],
@@ -111,30 +112,34 @@ function renderList() {
}
async function loadPdf(index) {
if (currentPdfUrl) {
URL.revokeObjectURL(currentPdfUrl);
currentPdfUrl = "";
if (pdfViewer) {
pdfViewer.destroy();
}
pdf.removeAttribute("src");
pdfViewer = new PdfViewer(pdfContainer);
viewer.classList.remove("loaded");
pdfMessage.hidden = false;
pdfMessage.textContent = "Loading PDF\u2026";
const startTime = performance.now();
try {
const response = await fetch(`/api/pdf?index=${index}`, {
cache: "no-store",
const response = await fetch("/api/pdf/prepare", {
method: "POST",
headers: { "content-type": "application/json" },
body: JSON.stringify({ index }),
});
if (!response.ok) {
const body = await response.json().catch(() => ({}));
throw new Error(
body.error || `PDF request failed with HTTP ${response.status}.`,
);
throw new Error(body.error || `Prepare failed: HTTP ${response.status}`);
}
const blob = await response.blob();
currentPdfUrl = URL.createObjectURL(blob);
pdf.src = currentPdfUrl;
const data = await response.json();
const prepareMs = Math.round(performance.now() - startTime);
console.log(
`[BizMatch QC] PDF prepare: ${prepareMs}ms, disk cache ${data.cacheStatus}`,
);
viewer.classList.add("loaded");
pdfMessage.hidden = true;
await pdfViewer.load(data.url, data.size);
} catch (error) {
const message = error instanceof Error ? error.message : String(error);
pdfMessage.textContent = `Cannot open PDF: ${message}`;
@@ -222,7 +227,6 @@ saveSettings.onclick = async (event) => {
event.preventDefault();
saveSettings.disabled = true;
settingsError.hidden = true;
try {
const response = await fetch("/api/config", {
method: "POST",
@@ -247,31 +251,6 @@ saveSettings.onclick = async (event) => {
}
};
let resizeTimer;
globalThis.addEventListener("resize", () => {
clearTimeout(resizeTimer);
resizeTimer = setTimeout(() => {
fetch("/api/window-size", {
method: "POST",
headers: { "content-type": "application/json" },
body: JSON.stringify({
width: globalThis.innerWidth,
height: globalThis.innerHeight,
}),
});
}, 300);
});
globalThis.addEventListener("beforeunload", () => {
navigator.sendBeacon(
"/api/window-size",
JSON.stringify({
width: globalThis.innerWidth,
height: globalThis.innerHeight,
}),
);
});
async function load() {
try {
const response = await fetch("/api/state", { cache: "no-store" });

View File

@@ -5,6 +5,12 @@
<meta name="viewport" content="width=device-width,initial-scale=1">
<title>BizMatch QC</title>
<link rel="stylesheet" href="/styles.css">
<script type="module">
import * as pdfjsLib from "/pdfjs/pdf.min.mjs";
pdfjsLib.GlobalWorkerOptions.workerSrc = "/pdfjs/pdf.worker.min.mjs";
globalThis.pdfjsLib = pdfjsLib;
</script>
<script type="module" src="/pdf_viewer.js"></script>
</head>
<body>
<header>
@@ -22,7 +28,7 @@
<div id="fields"></div>
</section>
<section class="viewer">
<iframe id="pdf" title="PDF document"></iframe>
<div id="pdfViewer"></div>
<div id="pdfMessage">Select a document</div>
</section>
</main>
@@ -30,15 +36,15 @@
<form method="dialog">
<h2>Settings</h2>
<label class="checkbox-label">
<input type="checkbox" id="useAnonData">
Use anonymized sample data
</label>
<input type="checkbox" id="useAnonData">
Use anonymized sample data
</label>
<label>buyers_vision.json
<input id="jsonPath" autocomplete="off">
</label>
<input id="jsonPath" autocomplete="off">
</label>
<label>PDF base directory
<input id="pdfBase" autocomplete="off">
</label>
<input id="pdfBase" autocomplete="off">
</label>
<div id="settingsError" class="dialog-error" hidden></div>
<div class="actions">
<button value="cancel">Cancel</button>

287
web/pdf_viewer.js Normal file
View File

@@ -0,0 +1,287 @@
// PDF.js viewer with memory LRU cache and lazy page rendering.
// Depends on global pdfjsLib (loaded via module import in index.html).
// Uses pdfjs-dist 6.1.200.
const MAX_CACHED_DOCS = 3;
const MAX_CACHED_BYTES = 150 * 1024 * 1024;
const PRELOAD_PAGES_AHEAD = 2;
const PRELOAD_PAGES_BEHIND = 1;
const MAX_RENDER_SCALE = 2.0;
class PdfViewer {
constructor(container) {
this.container = container;
this.lru = [];
this.currentEntry = null;
this.renderTasks = [];
this.observer = null;
this.currentDocId = 0;
this.loadingDiv = null;
this.errorDiv = null;
this.pagesDiv = null;
this._createDOM();
}
_createDOM() {
this.container.innerHTML = `
<div class="pdf-v-loading" hidden>Loading PDF\u2026</div>
<div class="pdf-v-error" hidden></div>
<div class="pdf-v-pages"></div>
`;
this.loadingDiv = this.container.querySelector(".pdf-v-loading");
this.errorDiv = this.container.querySelector(".pdf-v-error");
this.pagesDiv = this.container.querySelector(".pdf-v-pages");
}
_showError(msg) {
this.loadingDiv.hidden = true;
this.errorDiv.textContent = msg;
this.errorDiv.hidden = false;
}
_showLoading() {
this.errorDiv.hidden = true;
this.loadingDiv.hidden = false;
}
_hideLoading() {
this.loadingDiv.hidden = true;
}
async load(url, byteSize) {
const docId = ++this.currentDocId;
this._showLoading();
this._cancelAllRenders();
if (this.observer) this.observer.disconnect();
// Check memory LRU
const cached = this.lru.find((e) => e.url === url);
if (cached) {
cached.lastUsed = Date.now();
this.lru = this.lru.filter((e) => e !== cached);
this.lru.push(cached);
this.currentEntry = cached;
this._renderDoc(docId, cached);
console.log("[BizMatch QC] PDF memory cache hit");
return;
}
const startLoad = performance.now();
try {
const task = pdfjsLib.getDocument({ url, rangeChunkSize: 65536 });
const doc = await task.promise;
const loadMs = Math.round(performance.now() - startLoad);
console.log(
`[BizMatch QC] PDF.js document loaded: ${loadMs}ms`,
);
const pageCount = doc.numPages;
const entry = {
url,
doc,
byteSize: byteSize || 0,
pageCount,
lastUsed: Date.now(),
};
this._evictIfNeeded(entry.byteSize);
this.lru.push(entry);
if (docId !== this.currentDocId) {
doc.destroy();
return;
}
this.currentEntry = entry;
console.log("[BizMatch QC] PDF memory cache miss");
this._renderDoc(docId, entry, startLoad);
} catch (err) {
if (docId !== this.currentDocId) return;
this._showError(
`Cannot open PDF: ${err instanceof Error ? err.message : String(err)}`,
);
}
}
_evictIfNeeded(incomingBytes) {
let totalBytes = incomingBytes;
for (const e of this.lru) totalBytes += e.byteSize;
while (
this.lru.length > 0 &&
(this.lru.length >= MAX_CACHED_DOCS || totalBytes > MAX_CACHED_BYTES)
) {
const victim = this.lru.shift();
totalBytes -= victim.byteSize;
try {
victim.doc.destroy();
} catch { /* ok */ }
console.log(
"[BizMatch QC] PDF memory cache evicted: " +
victim.url.split("/").pop(),
);
}
}
_cancelAllRenders() {
for (const t of this.renderTasks) {
t.cancelled = true;
}
this.renderTasks = [];
}
async _renderDoc(docId, entry, startLoad = 0) {
this.pagesDiv.innerHTML = "";
this._hideLoading();
const pageCount = entry.pageCount;
const placeholderStyle = "pdf-v-placeholder";
for (let i = 1; i <= pageCount; i++) {
const ph = document.createElement("div");
ph.className = placeholderStyle;
ph.dataset.page = String(i);
// Pre-allocate aspect ratio from first page
this.pagesDiv.appendChild(ph);
}
// Get first page dimensions for placeholder sizing
try {
const firstPage = await entry.doc.getPage(1);
const vp = firstPage.getViewport({ scale: 1 });
const ratio = vp.height / vp.width;
const placeholders = this.pagesDiv.querySelectorAll(
`.${placeholderStyle}`,
);
for (const ph of placeholders) {
// Will be sized in CSS via aspect-ratio; set data attrs for CSS
ph.dataset.ratio = String(ratio);
}
} catch { /* ignore */ }
// Render first page immediately
this._renderPage(docId, entry, 1);
// Set up lazy rendering
this.observer = new IntersectionObserver(
(entries) => {
for (const e of entries) {
if (e.isIntersecting) {
const pageNum = parseInt(e.target.dataset.page);
if (pageNum) {
this._renderPage(docId, entry, pageNum);
// Pre-render nearby pages
for (
let p = pageNum - PRELOAD_PAGES_BEHIND;
p <= pageNum + PRELOAD_PAGES_AHEAD;
p++
) {
if (p >= 1 && p <= pageCount && p !== pageNum) {
this._renderPage(docId, entry, p);
}
}
}
}
}
},
{ root: this.pagesDiv, rootMargin: "200px" },
);
const placeholders = this.pagesDiv.querySelectorAll(`.${placeholderStyle}`);
for (const ph of placeholders) {
this.observer.observe(ph);
}
if (startLoad) {
const readyMs = Math.round(performance.now() - startLoad);
console.log(`[BizMatch QC] PDF ready: ${readyMs}ms`);
}
}
async _renderPage(docId, entry, pageNum) {
// Don't re-render if already rendered
const existing = this.pagesDiv.querySelector(
`.pdf-v-page[data-page="${pageNum}"]`,
);
if (existing) return;
if (docId !== this.currentDocId) return;
// Don't re-render if cancelled
const existingTask = this.renderTasks.find((t) => t.pageNum === pageNum);
if (existingTask) return;
const placeholder = this.pagesDiv.querySelector(
`.pdf-v-placeholder[data-page="${pageNum}"]`,
);
if (!placeholder) return;
const renderTask = { pageNum, cancelled: false, promise: null };
this.renderTasks.push(renderTask);
const startRender = performance.now();
try {
const page = await entry.doc.getPage(pageNum);
if (renderTask.cancelled || docId !== this.currentDocId) return;
const containerWidth = this.pagesDiv.clientWidth || 600;
const vp1 = page.getViewport({ scale: 1 });
const fitScale = containerWidth / vp1.width;
const dpr = Math.min(globalThis.devicePixelRatio || 1, MAX_RENDER_SCALE);
const scale = Math.min(fitScale * dpr, MAX_RENDER_SCALE);
const viewport = page.getViewport({ scale: fitScale });
const pageDiv = document.createElement("div");
pageDiv.className = "pdf-v-page";
pageDiv.dataset.page = String(pageNum);
pageDiv.style.width = `${viewport.width}px`;
pageDiv.style.height = `${viewport.height}px`;
const canvas = document.createElement("canvas");
canvas.style.width = `${viewport.width}px`;
canvas.style.height = `${viewport.height}px`;
canvas.width = Math.floor(viewport.width * (scale / fitScale));
canvas.height = Math.floor(viewport.height * (scale / fitScale));
pageDiv.appendChild(canvas);
placeholder.replaceWith(pageDiv);
const ctx = canvas.getContext("2d");
if (ctx && !renderTask.cancelled && docId === this.currentDocId) {
await page.render({ canvasContext: ctx, viewport }).promise;
if (pageNum === 1) {
const renderMs = Math.round(performance.now() - startRender);
console.log(
`[BizMatch QC] First page rendered: ${renderMs}ms`,
);
}
}
} catch (err) {
if (renderTask.cancelled) return;
placeholder.textContent = `Error rendering page ${pageNum}: ${
errorMsg(err)
}`;
} finally {
this.renderTasks = this.renderTasks.filter((t) => t !== renderTask);
}
}
destroy() {
this._cancelAllRenders();
if (this.observer) this.observer.disconnect();
this._evictAll();
}
_evictAll() {
for (const entry of this.lru) {
try {
entry.doc.destroy();
} catch { /* ok */ }
}
this.lru = [];
this.currentEntry = null;
}
}
function errorMsg(err) {
return err instanceof Error ? err.message : String(err);
}
// Export global
globalThis.PdfViewer = PdfViewer;

View File

@@ -3,7 +3,7 @@
}
body {
margin: 0;
font: 14px system-ui,sans-serif;
font: 14px system-ui, sans-serif;
color: #202124;
}
header {
@@ -29,7 +29,7 @@ header span {
main {
height: calc(100vh - 52px);
display: grid;
grid-template-columns: 330px 430px minmax(500px,1fr);
grid-template-columns: 330px 430px minmax(500px, 1fr);
}
aside,
.details {
@@ -77,12 +77,14 @@ aside,
.viewer {
position: relative;
background: #555;
overflow: hidden;
}
.viewer iframe {
.viewer #pdfViewer {
width: 100%;
height: 100%;
border: 0;
background: white;
overflow-y: auto;
overflow-x: hidden;
background: #525659;
}
.viewer #pdfMessage {
position: absolute;
@@ -91,12 +93,13 @@ aside,
place-items: center;
color: white;
pointer-events: none;
z-index: 2;
}
.viewer.loaded #pdfMessage {
display: none;
}
dialog {
width: min(720px,90vw);
width: min(720px, 90vw);
}
dialog label {
display: block;
@@ -185,3 +188,39 @@ hr.field-sep {
width: auto;
margin-top: 0;
}
/* PDF.js viewer */
.pdf-v-loading {
padding: 40px 20px;
color: #aaa;
text-align: center;
font-size: 15px;
}
.pdf-v-error {
padding: 30px 20px;
color: #f88;
text-align: center;
white-space: pre-wrap;
}
.pdf-v-pages {
display: flex;
flex-direction: column;
align-items: center;
gap: 8px;
padding: 10px 0;
}
.pdf-v-page {
background: white;
box-shadow: 0 1px 4px rgba(0, 0, 0, 0.3);
flex-shrink: 0;
}
.pdf-v-placeholder {
background: rgba(255, 255, 255, 0.05);
flex-shrink: 0;
}
.pdf-v-page canvas,
.pdf-v-placeholder canvas {
display: block;
width: 100%;
height: 100%;
}