refactor: extract shared PDF route handler to eliminate convert route duplication
All checks were successful
Build & Deploy to Staging / Build & Deploy to Staging (push) Successful in 19m19s
All checks were successful
Build & Deploy to Staging / Build & Deploy to Staging (push) Successful in 19m19s
- New src/utils/pdf-handler.ts with handlePdfRoute() helper - Handles: content-type validation, PDF option validation, slot acquire/release, error mapping, response headers - Refactored convert.ts from 388 to 233 lines (40% reduction) - 10 TDD tests for the new helper (RED→GREEN verified) - All 618 tests passing, zero tsc --noEmit errors
This commit is contained in:
parent
54316d45cf
commit
76b2179be9
3 changed files with 265 additions and 179 deletions
|
|
@ -3,33 +3,12 @@ import { renderPdf, renderUrlPdf } from "../services/browser.js";
|
|||
import { markdownToHtml, wrapHtml } from "../services/markdown.js";
|
||||
import dns from "node:dns/promises";
|
||||
import logger from "../services/logger.js";
|
||||
import { errorMessage } from "../utils/errors.js";
|
||||
import { isPrivateIP } from "../utils/network.js";
|
||||
|
||||
import { sanitizeFilename } from "../utils/sanitize.js";
|
||||
import { validatePdfOptions } from "../utils/pdf-options.js";
|
||||
import { handlePdfRoute } from "../utils/pdf-handler.js";
|
||||
|
||||
export const convertRouter = Router();
|
||||
|
||||
interface ConvertBody {
|
||||
html?: string;
|
||||
markdown?: string;
|
||||
css?: string;
|
||||
format?: string;
|
||||
landscape?: boolean;
|
||||
margin?: { top?: string; right?: string; bottom?: string; left?: string };
|
||||
printBackground?: boolean;
|
||||
filename?: string;
|
||||
headerTemplate?: string;
|
||||
footerTemplate?: string;
|
||||
displayHeaderFooter?: boolean;
|
||||
scale?: number;
|
||||
pageRanges?: string;
|
||||
preferCSSPageSize?: boolean;
|
||||
width?: string;
|
||||
height?: string;
|
||||
}
|
||||
|
||||
/**
|
||||
* @openapi
|
||||
* /v1/convert/html:
|
||||
|
|
@ -80,66 +59,18 @@ interface ConvertBody {
|
|||
* description: PDF generation failed
|
||||
*/
|
||||
convertRouter.post("/html", async (req: Request, res: Response) => {
|
||||
let slotAcquired = false;
|
||||
try {
|
||||
// Reject non-JSON content types
|
||||
const ct = req.headers["content-type"] || "";
|
||||
if (!ct.includes("application/json")) {
|
||||
res.status(415).json({ error: "Unsupported Content-Type. Use application/json." });
|
||||
return;
|
||||
}
|
||||
const body: ConvertBody =
|
||||
typeof req.body === "string" ? { html: req.body } : req.body;
|
||||
|
||||
await handlePdfRoute(req, res, async (sanitizedOptions) => {
|
||||
const body = typeof req.body === "string" ? { html: req.body } : req.body;
|
||||
if (!body.html) {
|
||||
res.status(400).json({ error: "Missing 'html' field" });
|
||||
return;
|
||||
return null;
|
||||
}
|
||||
|
||||
// Validate PDF options
|
||||
const validation = validatePdfOptions(body);
|
||||
if (!validation.valid) {
|
||||
res.status(400).json({ error: validation.error });
|
||||
return;
|
||||
}
|
||||
|
||||
// Acquire concurrency slot
|
||||
if (req.acquirePdfSlot) {
|
||||
await req.acquirePdfSlot();
|
||||
slotAcquired = true;
|
||||
}
|
||||
|
||||
// Wrap bare HTML fragments
|
||||
const fullHtml = body.html.includes("<html")
|
||||
? body.html
|
||||
: wrapHtml(body.html, body.css);
|
||||
|
||||
const { pdf, durationMs } = await renderPdf(fullHtml, {
|
||||
...validation.sanitized,
|
||||
});
|
||||
|
||||
const filename = sanitizeFilename(body.filename || "document.pdf");
|
||||
res.setHeader("Content-Type", "application/pdf");
|
||||
res.setHeader("Content-Disposition", `inline; filename="${filename}"`);
|
||||
res.setHeader("X-Render-Time", String(durationMs));
|
||||
res.send(pdf);
|
||||
} catch (err: unknown) {
|
||||
logger.error({ err }, "Convert HTML error");
|
||||
const msg = errorMessage(err);
|
||||
if (msg === "QUEUE_FULL") {
|
||||
res.status(503).json({ error: "Server busy — too many concurrent PDF generations. Please try again in a few seconds." });
|
||||
return;
|
||||
}
|
||||
if (msg === "PDF_TIMEOUT") {
|
||||
res.status(504).json({ error: "PDF generation timed out." });
|
||||
return;
|
||||
}
|
||||
res.status(500).json({ error: "PDF generation failed." });
|
||||
} finally {
|
||||
if (slotAcquired && req.releasePdfSlot) {
|
||||
req.releasePdfSlot();
|
||||
}
|
||||
}
|
||||
const { pdf, durationMs } = await renderPdf(fullHtml, { ...sanitizedOptions });
|
||||
return { pdf, durationMs, filename: sanitizeFilename(body.filename || "document.pdf") };
|
||||
});
|
||||
});
|
||||
|
||||
/**
|
||||
|
|
@ -191,62 +122,16 @@ convertRouter.post("/html", async (req: Request, res: Response) => {
|
|||
* description: PDF generation failed
|
||||
*/
|
||||
convertRouter.post("/markdown", async (req: Request, res: Response) => {
|
||||
let slotAcquired = false;
|
||||
try {
|
||||
// Reject non-JSON content types
|
||||
const ct = req.headers["content-type"] || "";
|
||||
if (!ct.includes("application/json")) {
|
||||
res.status(415).json({ error: "Unsupported Content-Type. Use application/json." });
|
||||
return;
|
||||
}
|
||||
const body: ConvertBody =
|
||||
typeof req.body === "string" ? { markdown: req.body } : req.body;
|
||||
|
||||
await handlePdfRoute(req, res, async (sanitizedOptions) => {
|
||||
const body = typeof req.body === "string" ? { markdown: req.body } : req.body;
|
||||
if (!body.markdown) {
|
||||
res.status(400).json({ error: "Missing 'markdown' field" });
|
||||
return;
|
||||
return null;
|
||||
}
|
||||
|
||||
// Validate PDF options
|
||||
const validation = validatePdfOptions(body);
|
||||
if (!validation.valid) {
|
||||
res.status(400).json({ error: validation.error });
|
||||
return;
|
||||
}
|
||||
|
||||
// Acquire concurrency slot
|
||||
if (req.acquirePdfSlot) {
|
||||
await req.acquirePdfSlot();
|
||||
slotAcquired = true;
|
||||
}
|
||||
|
||||
const html = markdownToHtml(body.markdown, body.css);
|
||||
const { pdf, durationMs } = await renderPdf(html, {
|
||||
...validation.sanitized,
|
||||
});
|
||||
|
||||
const filename = sanitizeFilename(body.filename || "document.pdf");
|
||||
res.setHeader("Content-Type", "application/pdf");
|
||||
res.setHeader("Content-Disposition", `inline; filename="${filename}"`);
|
||||
res.setHeader("X-Render-Time", String(durationMs));
|
||||
res.send(pdf);
|
||||
} catch (err: unknown) {
|
||||
logger.error({ err }, "Convert MD error");
|
||||
const msg = errorMessage(err);
|
||||
if (msg === "QUEUE_FULL") {
|
||||
res.status(503).json({ error: "Server busy — too many concurrent PDF generations. Please try again in a few seconds." });
|
||||
return;
|
||||
}
|
||||
if (msg === "PDF_TIMEOUT") {
|
||||
res.status(504).json({ error: "PDF generation timed out." });
|
||||
return;
|
||||
}
|
||||
res.status(500).json({ error: "PDF generation failed." });
|
||||
} finally {
|
||||
if (slotAcquired && req.releasePdfSlot) {
|
||||
req.releasePdfSlot();
|
||||
}
|
||||
}
|
||||
const { pdf, durationMs } = await renderPdf(html, { ...sanitizedOptions });
|
||||
return { pdf, durationMs, filename: sanitizeFilename(body.filename || "document.pdf") };
|
||||
});
|
||||
});
|
||||
|
||||
/**
|
||||
|
|
@ -303,19 +188,12 @@ convertRouter.post("/markdown", async (req: Request, res: Response) => {
|
|||
* description: PDF generation failed
|
||||
*/
|
||||
convertRouter.post("/url", async (req: Request, res: Response) => {
|
||||
let slotAcquired = false;
|
||||
try {
|
||||
// Reject non-JSON content types
|
||||
const ct = req.headers["content-type"] || "";
|
||||
if (!ct.includes("application/json")) {
|
||||
res.status(415).json({ error: "Unsupported Content-Type. Use application/json." });
|
||||
return;
|
||||
}
|
||||
const body = req.body as { url?: string; format?: string; landscape?: boolean; margin?: string | { top?: string; right?: string; bottom?: string; left?: string }; printBackground?: boolean; waitUntil?: string; filename?: string; headerTemplate?: string; footerTemplate?: string; displayHeaderFooter?: boolean; scale?: number; pageRanges?: string; preferCSSPageSize?: boolean; width?: string; height?: string };
|
||||
await handlePdfRoute(req, res, async (sanitizedOptions) => {
|
||||
const body = req.body as { url?: string; waitUntil?: string; filename?: string };
|
||||
|
||||
if (!body.url) {
|
||||
res.status(400).json({ error: "Missing 'url' field" });
|
||||
return;
|
||||
return null;
|
||||
}
|
||||
|
||||
// URL validation + SSRF protection
|
||||
|
|
@ -324,65 +202,32 @@ convertRouter.post("/url", async (req: Request, res: Response) => {
|
|||
parsed = new URL(body.url);
|
||||
if (!["http:", "https:"].includes(parsed.protocol)) {
|
||||
res.status(400).json({ error: "Only http/https URLs are supported" });
|
||||
return;
|
||||
return null;
|
||||
}
|
||||
} catch {
|
||||
res.status(400).json({ error: "Invalid URL" });
|
||||
return;
|
||||
return null;
|
||||
}
|
||||
|
||||
// DNS lookup to block private/reserved IPs + pin resolution to prevent DNS rebinding
|
||||
// DNS lookup to block private/reserved IPs + pin resolution
|
||||
let resolvedAddress: string;
|
||||
try {
|
||||
const { address } = await dns.lookup(parsed.hostname);
|
||||
if (isPrivateIP(address)) {
|
||||
res.status(400).json({ error: "URL resolves to a private/internal IP address" });
|
||||
return;
|
||||
return null;
|
||||
}
|
||||
resolvedAddress = address;
|
||||
} catch {
|
||||
res.status(400).json({ error: "DNS lookup failed for URL hostname" });
|
||||
return;
|
||||
}
|
||||
|
||||
// Validate PDF options
|
||||
const validation = validatePdfOptions(body);
|
||||
if (!validation.valid) {
|
||||
res.status(400).json({ error: validation.error });
|
||||
return;
|
||||
}
|
||||
|
||||
// Acquire concurrency slot
|
||||
if (req.acquirePdfSlot) {
|
||||
await req.acquirePdfSlot();
|
||||
slotAcquired = true;
|
||||
return null;
|
||||
}
|
||||
|
||||
const { pdf, durationMs } = await renderUrlPdf(body.url, {
|
||||
...validation.sanitized,
|
||||
...sanitizedOptions,
|
||||
hostResolverRules: `MAP ${parsed.hostname} ${resolvedAddress}`,
|
||||
});
|
||||
|
||||
const filename = sanitizeFilename(body.filename || "page.pdf");
|
||||
res.setHeader("Content-Type", "application/pdf");
|
||||
res.setHeader("Content-Disposition", `inline; filename="${filename}"`);
|
||||
res.setHeader("X-Render-Time", String(durationMs));
|
||||
res.send(pdf);
|
||||
} catch (err: unknown) {
|
||||
logger.error({ err }, "Convert URL error");
|
||||
const msg = errorMessage(err);
|
||||
if (msg === "QUEUE_FULL") {
|
||||
res.status(503).json({ error: "Server busy — too many concurrent PDF generations. Please try again in a few seconds." });
|
||||
return;
|
||||
}
|
||||
if (msg === "PDF_TIMEOUT") {
|
||||
res.status(504).json({ error: "PDF generation timed out." });
|
||||
return;
|
||||
}
|
||||
res.status(500).json({ error: "PDF generation failed." });
|
||||
} finally {
|
||||
if (slotAcquired && req.releasePdfSlot) {
|
||||
req.releasePdfSlot();
|
||||
}
|
||||
}
|
||||
return { pdf, durationMs, filename: sanitizeFilename(body.filename || "page.pdf") };
|
||||
});
|
||||
});
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue