Self-Hosted
/
changedetection.io
mirror of https://github.com/dgtlmoon/changedetection.io.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

Puppeteer fetcher, adding disk cache and other fixes (#1563)

Browse Source
puppeteer-fixes
dgtlmoon 2 years ago committed by GitHub
parent 8eee913438
commit 5325918f29
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
3 changed files with 111 additions and 10 deletions
Show Stats Download Patch File Download Diff File

7
.github/workflows/test-only.yml vendored
Unescape View File

@ -30,7 +30,7 @@ jobs:
# Selenium+browserless
docker run --network changedet-network -d --hostname selenium -p 4444:4444 --rm --shm-size="2g" selenium/standalone-chrome-debug:3.141.59
docker run --network changedet-network -d --hostname browserless -e "DEFAULT_LAUNCH_ARGS=[\"--window-size=1920,1080\"]" --rm -p 3000:3000 --shm-size="2g" browserless/chrome:1.53-chrome-stable
docker run --network changedet-network -d --hostname browserless -e "FUNCTION_BUILT_INS=[\"fs\",\"crypto\"]" -e "DEFAULT_LAUNCH_ARGS=[\"--window-size=1920,1080\"]" --rm -p 3000:3000 --shm-size="2g" browserless/chrome:1.53-chrome-stable
- name: Build changedetection.io container for testing
run: |
@ -58,6 +58,11 @@ jobs:
# restock detection via playwright - added name=changedet here so that playwright/browserless can connect to it
docker run --rm --name "changedet" -e "FLASK_SERVER_NAME=changedet" -e "PLAYWRIGHT_DRIVER_URL=ws://browserless:3000" --network changedet-network test-changedetectionio bash -c 'cd changedetectionio;pytest --live-server-port=5004 --live-server-host=0.0.0.0 tests/restock/test_restock.py'
- name: Test with puppeteer fetcher
run: |
docker run --rm -e "USE_EXPERIMENTAL_PUPPETEER_FETCH=yes" -e "PLAYWRIGHT_DRIVER_URL=ws://browserless:3000" --network changedet-network test-changedetectionio bash -c 'cd changedetectionio;pytest tests/fetchers/test_content.py && pytest tests/test_errorhandling.py && pytest tests/visualselector/test_fetch_data.py'
# Browserless would have had -e "FUNCTION_BUILT_INS=[\"fs\",\"crypto\"]" added above
- name: Test proxy interaction
run: |
cd changedetectionio

107
changedetectionio/content_fetcher.py
Unescape View File

@ -291,7 +291,8 @@ class base_html_playwright(Fetcher):
xpath_element_js = self.xpath_element_js.replace('%ELEMENTS%', visualselector_xpath_selectors)
code = f"""module.exports = async ({{ page, context }}) => {{
var {{ url, execute_js, user_agent, extra_wait_ms, req_headers, include_filters, xpath_element_js, screenshot_quality, proxy_username, proxy_password}} = context;
var {{ url, execute_js, user_agent, extra_wait_ms, req_headers, include_filters, xpath_element_js, screenshot_quality, proxy_username, proxy_password, disk_cache_dir}} = context;
await page.setBypassCSP(true)
await page.setExtraHTTPHeaders(req_headers);
@ -313,6 +314,82 @@ class base_html_playwright(Fetcher):
deviceScaleFactor: 1,
}});
// Very primitive disk cache - USE WITH EXTREME CAUTION
// Run browserless container with -e "FUNCTION_BUILT_INS=[\"fs\",\"crypto\"]"
if ( disk_cache_dir ) {{
await page.setRequestInterception(true);
console.log(">>>>>>>>>>>>>>> LOCAL DISK CACHE ENABLED <<<<<<<<<<<<<<<<<<<<<");
const fs = require('fs');
const crypto = require('crypto');
function file_is_expired(file_path) {{
if (!fs.existsSync(dir_path+key)) {{
return true;
}}
var stats = fs.statSync(file_path);
const now_date = new Date();
const expire_seconds = 300;
if ( (now_date/1000) - (stats.mtime.getTime() / 1000) > expire_seconds) {{
console.log("CACHE EXPIRED: "+file_path);
return true;
}}
return false;
}}
page.on('request', async (request) => {{
// if (blockedExtensions.some((str) => req.url().endsWith(str))) return req.abort();
const url = request.url();
const key = crypto.createHash('md5').update(url).digest("hex");
const dir_path = disk_cache_dir + key.slice(0, 1) + '/' + key.slice(1, 2) + '/' + key.slice(2, 3) + '/';
// https://stackoverflow.com/questions/4482686/check-synchronously-if-file-directory-exists-in-node-js
if (fs.existsSync(dir_path+key)) {{
file_is_expired(dir_path+key);
console.log("Cache exists "+dir_path+key+ " - "+url);
const cached_data = fs.readFileSync(dir_path+key);
request.respond({{
status: 200,
//contentType: 'text/html', //@todo
body: cached_data
}});
return;
}}
request.continue();
}});
page.on('response', async (response) => {{
const url = response.url();
// @todo - check response size()
console.log("Cache - Got "+response.request().method()+" - "+url+" - "+response.request().resourceType());
if(response.request().method() != 'GET' || response.request().resourceType() == 'xhr' || response.request().resourceType() == 'document' || response.status() != 200 ) {{
console.log("Skipping- "+url);
return;
}}
const key = crypto.createHash('md5').update(url).digest("hex");
const dir_path = disk_cache_dir + key.slice(0, 1) + '/' + key.slice(1, 2) + '/' + key.slice(2, 3) + '/';
const data = await response.text();
if (!fs.existsSync(dir_path)) {{
fs.mkdirSync(dir_path, {{ recursive: true }})
}}
var expired = false;
if (fs.existsSync(dir_path+key)) {{
if (file_is_expired(dir_path+key)) {{
fs.writeFileSync(dir_path+key, data);
}}
}} else {{
fs.writeFileSync(dir_path+key, data);
}}
}});
}}
const r = await page.goto(url, {{
waitUntil: 'load'
}});
@ -325,7 +402,6 @@ class base_html_playwright(Fetcher):
await page.waitForTimeout(200);
}}
var html = await page.content();
var xpath_data;
var instock_data;
try {{
@ -337,13 +413,27 @@ class base_html_playwright(Fetcher):
// Protocol error (Page.captureScreenshot): Cannot take screenshot with 0 width can come from a proxy auth failure
// Wrap it here (for now)
var b64s;
var b64s = false;
try {{
b64s = await page.screenshot({{ encoding: "base64", fullPage: true, quality: screenshot_quality, type: 'jpeg' }});
}} catch (e) {{
console.log(e);
}}
// May fail on very large pages with 'WARNING: tile memory limits exceeded, some content may not draw'
if (!b64s) {{
// @todo after text extract, we can place some overlay text with red background to say 'croppped'
console.error('ERROR: content-fetcher page was maybe too large for a screenshot, reverting to viewport only screenshot');
try {{
b64s = await page.screenshot({{ encoding: "base64", quality: screenshot_quality, type: 'jpeg' }});
}} catch (e) {{
console.log(e);
}}
}}
var html = await page.content();
return {{
data: {{
'content': html,
@ -375,15 +465,17 @@ class base_html_playwright(Fetcher):
# Actual authentication handled by Puppeteer/node
o = urlparse(self.proxy.get('server'))
proxy_url = urllib.parse.quote(o._replace(netloc="{}:{}".format(o.hostname, o.port)).geturl())
browserless_function_url = f"{browserless_function_url}&--proxy-server={proxy_url}"
browserless_function_url = f"{browserless_function_url}&--proxy-server={proxy_url}&dumpio=true"
try:
amp = '&' if '?' in browserless_function_url else '?'
response = requests.request(
method="POST",
json={
"code": code,
"context": {
'disk_cache_dir': False, # or path to disk cache
'execute_js': self.webdriver_js_execute_code,
'extra_wait_ms': extra_wait_ms,
'include_filters': current_include_filters,
@ -396,9 +488,10 @@ class base_html_playwright(Fetcher):
}
},
# @todo /function needs adding ws:// to http:// rebuild this
url=browserless_function_url+"&--disable-features=AudioServiceOutOfProcess&dumpio=true",
url=browserless_function_url+f"{amp}--disable-features=AudioServiceOutOfProcess&dumpio=true&--disable-remote-fonts",
timeout=wait_browserless_seconds)
# 'ziparchive::addglob() will throw an instance of error instead of resulting in a fatal error if glob support is not available.'
except ReadTimeout:
raise PageUnloadable(url=url, status_code=None, message=f"No response from browserless in {wait_browserless_seconds}s")
except ConnectTimeout:
@ -410,6 +503,10 @@ class base_html_playwright(Fetcher):
x = response.json()
if not x.get('screenshot'):
# https://github.com/puppeteer/puppeteer/blob/v1.0.0/docs/troubleshooting.md#tips
# https://github.com/puppeteer/puppeteer/issues/1834
# https://github.com/puppeteer/puppeteer/issues/1834#issuecomment-381047051
# Check your memory is shared and big enough
raise ScreenshotUnavailable(url=url, status_code=None)
if not x.get('content', '').strip():

1
changedetectionio/res/xpath_element_scraper.js
Unescape View File

@ -17,7 +17,6 @@ try {
// Include the getXpath script directly, easier than fetching
function getxpath(e) {
var n = e;

Write Preview
Loading…
Cancel
Save