Automatisation/browser-cli
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases 5 Wiki Activity

adding better markdown extracting and filter out to not have broken staff, allow that session can list over multiple browsers

Browse Source
This commit is contained in:
Daniel Dolezal
2026-04-12 17:10:19 +02:00
parent 51054422fb
commit 64d804cf32
7 changed files with 899 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files
extension/background.js
+200 -12
View File
@@ -659,6 +659,32 @@ function contentDispatch(funcName, args) {
"li", "main", "nav", "ol", "p", "pre", "section", "table", "tbody", "td", "tfoot",
"th", "thead", "tr", "ul"
]);
const NOISE_SELECTOR = [
"script",
"style",
"noscript",
"template",
"svg",
"canvas",
"iframe",
"dialog",
"button",
"input",
"textarea",
"select",
"option",
"form",
"[hidden]",
"[aria-hidden='true']",
".sr-only",
"[class*='sr-only']",
"[class*='file-tile']",
"form[data-type='unified-composer']",
".composer-btn",
"[data-composer-surface='true']",
"#thread-bottom-container",
"[data-testid*='action-button']",
].join(", ");
function normalizeText(value) {
return value.replace(/\s+/g, " ").trim();
@@ -681,7 +707,7 @@ function contentDispatch(funcName, args) {
}
function escapeMarkdown(text) {
return text.replace(/([\\`*_{}\[\]()#+\-!|>])/g, "\\$1");
return text.replace(/([\\`[\]])/g, "\\$1");
}
function escapeTableCell(text) {
@@ -692,12 +718,55 @@ function contentDispatch(funcName, args) {
return attr || fallback || "";
}
function isNoiseElement(node) {
if (!node || node.nodeType !== Node.ELEMENT_NODE) return false;
const tag = node.tagName.toLowerCase();
if (["script", "style", "noscript", "template", "svg", "canvas", "iframe", "dialog"].includes(tag)) return true;
if (["button", "input", "textarea", "select", "option", "form"].includes(tag)) return true;
if (node.hasAttribute("hidden")) return true;
if ((node.getAttribute("aria-hidden") || "").toLowerCase() === "true") return true;
if (node.matches(".sr-only, [class*='sr-only']")) return true;
if (node.matches("[class*='file-tile'], form[data-type='unified-composer'], .composer-btn, [data-composer-surface='true'], #thread-bottom-container")) return true;
if (node.matches("[data-testid*='action-button']")) return true;
return false;
}
function stripNoise(root) {
const clone = root.cloneNode(true);
clone.querySelectorAll("script, style, noscript, template").forEach(node => node.remove());
clone.querySelectorAll(NOISE_SELECTOR).forEach(node => node.remove());
return clone;
}
function candidateScore(node) {
const text = normalizeText(node.innerText || "");
if (!text) return -Infinity;
const headings = node.querySelectorAll("h1, h2, h3, h4, h5, h6").length;
const paragraphs = node.querySelectorAll("p").length;
const listItems = node.querySelectorAll("li").length;
const tables = node.querySelectorAll("table").length;
const codeBlocks = node.querySelectorAll("pre, code").length;
const images = node.querySelectorAll("img, figure").length;
const mainLike = node.matches("main, article, [role='main']") ? 1 : 0;
const proseBlocks = node.matches(".markdown, .prose, [data-message-author-role='assistant']") ? 1 : 0;
const buttons = node.querySelectorAll("button, input, textarea, select").length;
const forms = node.querySelectorAll("form").length;
const svgs = node.querySelectorAll("svg, canvas").length;
return text.length
+ (mainLike * 4000)
+ (proseBlocks * 5000)
+ (headings * 250)
+ (paragraphs * 60)
+ (listItems * 35)
+ (tables * 80)
+ (codeBlocks * 60)
+ (images * 25)
- (buttons * 120)
- (forms * 200)
- (svgs * 40);
}
function pickRoot() {
if (selector) {
const matched = document.querySelector(selector);
@@ -705,10 +774,12 @@ function contentDispatch(funcName, args) {
return matched;
}
const candidates = Array.from(document.querySelectorAll("main, article, [role='main']"))
const candidates = Array.from(document.querySelectorAll(
"main, article, [role='main'], section, .markdown, .prose, [data-message-author-role]"
))
.filter(node => normalizeText(node.innerText || "").length > 0);
if (!candidates.length) return document.body;
candidates.sort((a, b) => (b.innerText || "").length - (a.innerText || "").length);
candidates.sort((a, b) => candidateScore(b) - candidateScore(a));
return candidates[0];
}
@@ -717,9 +788,9 @@ function contentDispatch(funcName, args) {
return escapeMarkdown(node.textContent || "");
}
if (node.nodeType !== Node.ELEMENT_NODE) return "";
if (isNoiseElement(node)) return "";
const tag = node.tagName.toLowerCase();
if (tag === "script" || tag === "style" || tag === "noscript" || tag === "template") return "";
if (tag === "br") return "\n";
if (tag === "img") {
const src = absoluteUrl(node.getAttribute("src"), node.src);
@@ -762,6 +833,92 @@ function contentDispatch(funcName, args) {
return collapseBlankLines(normalizeInline(Array.from(node.childNodes).map(inlineText).join("")));
}
function preserveNodeText(node) {
if (node.nodeType === Node.TEXT_NODE) {
return node.textContent || "";
}
if (node.nodeType !== Node.ELEMENT_NODE) return "";
const tag = node.tagName.toLowerCase();
if (tag === "br") return "\n";
const parts = [];
for (const child of node.childNodes) {
const rendered = preserveNodeText(child);
if (!rendered) continue;
parts.push(rendered);
}
if (["div", "p", "li"].includes(tag)) {
return `${parts.join("")}\n`;
}
return parts.join("");
}
function repairFlattenedDiagram(text) {
if (text.includes("\n")) return text;
const markerCount = (text.match(/[│▼├└]/g) || []).length;
if (markerCount < 2) return text;
let repaired = text;
repaired = repaired.replace(/\s{2,}([│▼])/g, "\n $1");
repaired = repaired.replace(/([│▼])\s{2,}/g, "$1\n");
repaired = repaired.replace(/([│▼])(?=[^\s\n│▼├└])/g, "$1\n");
repaired = repaired.replace(/(?<=[^\s\n])([├└])/g, "\n$1");
repaired = repaired.replace(/([^\s\n])(\()/g, "$1\n$2");
return repaired
.split("\n")
.map(line => line.replace(/\s+$/, ""))
.filter(line => line.trim())
.join("\n");
}
function convertDashListsToBranches(lines) {
const converted = [];
let index = 0;
while (index < lines.length) {
const match = lines[index].match(/^(\s*)-\s+(.*)$/);
if (!match) {
converted.push(lines[index]);
index += 1;
continue;
}
const indent = match[1];
const items = [];
while (index < lines.length) {
const nextMatch = lines[index].match(new RegExp(`^${indent.replace(/[.*+?^${}()|[\]\\]/g, "\\$&")}-\\s+(.*)$`));
if (!nextMatch) break;
items.push(nextMatch[1]);
index += 1;
}
items.forEach((item, itemIndex) => {
const branch = itemIndex === items.length - 1 ? "└" : "├";
converted.push(`${indent}${branch} ${item}`);
});
}
return converted;
}
function normalizeCodeBlock(text) {
let lines = text.replace(/\r\n?/g, "\n").split("\n").map(line => line.replace(/\s+$/, ""));
while (lines.length && !lines[0].trim()) lines.shift();
while (lines.length && !lines[lines.length - 1].trim()) lines.pop();
const flattened = repairFlattenedDiagram(lines.join("\n"));
lines = flattened ? flattened.split("\n") : [];
lines = lines.map(line => {
const trimmed = line.trim();
if ((trimmed === "│" || trimmed === "▼") && !/^\s+[│▼]\s*$/.test(line)) {
return ` ${trimmed}`;
}
return line;
});
lines = convertDashListsToBranches(lines);
return lines.join("\n");
}
function tableToMarkdown(table) {
const rows = Array.from(table.querySelectorAll("tr"))
.map(row => Array.from(row.children)
@@ -780,10 +937,16 @@ function contentDispatch(funcName, args) {
let headers = normalizedRows[0];
let bodyRows = normalizedRows.slice(1);
const firstRowIsBlank = headers.every(cell => !cell.trim());
if (firstRowIsBlank && normalizedRows.length > 1) {
headers = normalizedRows[1];
bodyRows = normalizedRows.slice(2);
}
const firstRow = table.querySelector("tr");
const thead = table.querySelector("thead");
const firstRowHasTh = firstRow && Array.from(firstRow.children).some(cell => cell.tagName === "TH");
if (!(thead || firstRowHasTh)) {
if (!(thead || firstRowHasTh || firstRowIsBlank)) {
headers = new Array(widths).fill("");
bodyRows = normalizedRows;
}
@@ -818,7 +981,12 @@ function contentDispatch(funcName, args) {
}
const line = collapseBlankLines(normalizeInline(content.join("")));
if (line) items.push(`${indent}${marker}${line}`);
if (line) {
const lineParts = line.split("\n");
items.push(`${indent}${marker}${lineParts[0]}`);
const continuationIndent = `${indent}${" ".repeat(marker.length)}`;
lineParts.slice(1).forEach(part => items.push(`${continuationIndent}${part}`));
}
nested.filter(Boolean).forEach(block => items.push(block));
});
return items.join("\n");
@@ -829,13 +997,21 @@ function contentDispatch(funcName, args) {
return normalizeText(node.textContent || "");
}
if (node.nodeType !== Node.ELEMENT_NODE) return "";
if (isNoiseElement(node)) return "";
const tag = node.tagName.toLowerCase();
if (tag === "script" || tag === "style" || tag === "noscript" || tag === "template") return "";
if (tag === "table") return tableToMarkdown(node);
if (tag === "ul" || tag === "ol") return listToMarkdown(node);
if (node.matches(".cm-editor[data-is-code-block-view='true']")) {
const lines = Array.from(node.querySelectorAll(".cm-line")).map(line => {
const text = preserveNodeText(line);
return text === "\n" ? "" : text.replace(/\n$/, "");
});
const code = normalizeCodeBlock(lines.join("\n"));
return code ? `\`\`\`\n${code}\n\`\`\`` : "";
}
if (tag === "pre") {
const code = node.innerText.replace(/\n$/, "");
const code = normalizeCodeBlock(preserveNodeText(node));
return code ? `\`\`\`\n${code}\n\`\`\`` : "";
}
if (tag === "blockquote") {
@@ -1011,9 +1187,21 @@ async function clientsRenameProfile({ alias }) {
// ── Helpers ───────────────────────────────────────────────────────────────────
async function getActiveTab() {
const [tab] = await chrome.tabs.query({ active: true, lastFocusedWindow: true });
if (!tab) throw new Error("No active tab found");
return tab;
const activeTabs = await chrome.tabs.query({ active: true });
if (!activeTabs.length) throw new Error("No active tab found");
const windows = await chrome.windows.getAll({ populate: false });
const focusedWindowIds = new Set(windows.filter(window => window.focused).map(window => window.id));
const chooseTab = (predicate) => activeTabs.find(predicate);
const byFocusAndScriptable = tab => focusedWindowIds.has(tab.windowId) && isScriptableUrl(tab.url || tab.pendingUrl || "");
const byScriptable = tab => isScriptableUrl(tab.url || tab.pendingUrl || "");
const byFocus = tab => focusedWindowIds.has(tab.windowId);
return chooseTab(byFocusAndScriptable)
|| chooseTab(byScriptable)
|| chooseTab(byFocus)
|| activeTabs[0];
}
async function resolveGroupId(nameOrId) {