Dependency Updates:
- Bump Brave to 1.67.123
- Update puppeteer-core to latest, fixes possible crash when loading
current browser with old profiles
- Tests: simplifies extra hops test to avoid complex pages that could
lead to timeout
- update to wabac.js 2.19.0 to use new html rewriting support in
wabac.js 2.19.0
- update to browsertrix-behaviors to 0.6.1 to fix instagram behavior
- bump to 1.2.0-beta.3
- parse URL username/password, store in 'auth' field in seed, or pass in 'auth' field directly (from yaml config)
- add 'Authorization' header with base64 encoded basic auth via setExtraHTTPHeaders()
- tests: add test for crawling with auth using http-server using local docs build (now build docs as part of CI)
- docs: add HTTP Auth to YAML config section
---------
Co-authored-by: Ed Summers <ehs@pobox.com>
fixes#587
The proxy env vars PROXY_HOST and PROXY_PORT were being ignored, as they
were hardcoded to obsolete values in the Dockerfile.
Proxy settings can now be set, in order of precedence via:
- --proxyServer cli flag
- PROXY_SERVER env var
- PROXY_HOST and PROXY_PORT env vars, which set an HTTP proxy server
only (for backwards compatibility with 0.12.x)
The --proxyServer / PROXY_SERVER settings are passed to the browser via
the --proxy-server flag.
AsyncFetcher / direct fetch also supports HTTP and SOCKS5 proxying.
Supported proxies are: HTTP no auth, SOCKS5 no auth, SOCKS5 with auth
(supported in Brave, but not Chrome!)
---------
Co-authored-by: Vinzenz Sinapius <Vinzenz.Sinapius@gmail.com>
Co-authored-by: Tessa Walsh <tessa@bitarchivist.net>
The 0.6.0 release of Browsertrix Behaviors /
webrecorder/browsertrix-behaviors#70 introduces support for site-specific behaviors to implement an `awaitPageLoad()` function which allows for waiting for specific resources on the page load.
- This PR just adds a call to this function directly after page load.
- Factors out into an `awaitPageLoad()` method used in both crawler and replaycrawler to support the same wait in QA Mode
- This is to support custom loading wait time for Instagram (other sites in the future)
Using latest puppeteer-core to keep up with latest browsers, mostly
minor syntax changes
Due to change in puppeteer hiding the executionContextId, need to create
a frameId->executionContextId mapping and track it ourselves to support
the custom evaluateWithCLI() function
Initial (beta) support for QA/replay crawling!
- Supports running a crawl over a given WACZ / list of WACZ (multi WACZ) input, hosted in ReplayWeb.page
- Runs local http server with full-page, ui-less ReplayWeb.page embed
- ReplayWeb.page release version configured in the Dockerfile, pinned ui.js and sw.js fetched directly from cdnjs
Can be deployed with `webrecorder/browsertrix-crawler qa` entrypoint.
- Requires `--qaSource`, pointing to WACZ or multi-WACZ json that will be replay/QAd
- Also supports `--qaRedisKey` where QA comparison data will be pushed, if specified.
- Supports `--qaDebugImageDiff` for outputting crawl / replay/ diff
images.
- If using --writePagesToRedis, a `comparison` key is added to existing page data where:
```
comparison: {
screenshotMatch?: number;
textMatch?: number;
resourceCounts: {
crawlGood?: number;
crawlBad?: number;
replayGood?: number;
replayBad?: number;
};
};
```
- bump version to 1.1.0-beta.2
Adds a new SAX-based sitemap parser, inspired by:
https://www.npmjs.com/package/sitemap-stream-parser
Supports:
- recursively parsing sitemap indexes, using p-queue to process N at a
time (currently 5)
- `fromDate` and `toDate` filter dates, to only include URLs between the given
dates, filtering nested sitemap lists included
- async parsing, continue parsing in the background after 100 URLs
- timeout for initial fetch / first 100 URLs set to 30 seconds to avoid
slowing down the crawl
- save/load state integration: mark if sitemaps have already been parsed
in redis, serialize to save state, to avoid reparsing again. (Will
reparse if parsing did not fully finish)
- Aware of `pageLimit`, don't add URLs pass the page limit, interrupt
further parsing when at limit.
- robots.txt `sitemap:` parsing, check URL extension and mime type
- automatic detection of sitemaps for a seed URL if no sitemap url provided - first check robots.txt,
then /sitemap.xml
- tests: test for full sitemap autodetect, sitemap with limit, and sitemap from specific URL.
Fixes#496
---------
Co-authored-by: Tessa Walsh <tessa@bitarchivist.net>
Requires webrecorder/browsertrix-behaviors#69 / browsertrix-behaviors
0.5.3, which will add support for behaviors to add links.
Simplify adding links by simply adding the links directly, instead of
batching to 500 links. Errors are already being logged in queueing a new
URL fails.
This adds prettier to the repo, and sets up the pre-commit hook to
auto-format as well as lint.
Also updates ignores files to exclude crawls, test-crawls, scratch, dist as needed.
Follows #424. Converts the upcoming 1.0.0 branch based on native browser-based traffic capture and recording to TypeScript. Fixes#426
---------
Co-authored-by: Tessa Walsh <tessa@bitarchivist.net>
Co-authored-by: emma <hi@emma.cafe>
Major refactoring of Browsertrix Crawler to native capture network traffic to WARC files
via the Chrome Debug Protocol (CDP). Allows for more flexibility and accuracy when dealing
with HTTP/2.x sites and avoids a MITM proxy. Addresses #343
Changes include:
- Recorder class for capture CDP network traffic for each page.
- Handling requests from service workers via matching active frames, skipping unrelated requests outside the page (from background pages, etc..)
- WARC writing support via TS-based warcio.js library.
- Generates single WARC file per worker (still need to add size rollover).
- Request interception via Fetch.requestPaused
- Rule-based rewriting response support (via wabac.js), using Fetch.getResponseBody() / Fetch.fulfillRequest()
- Streaming responses via three methods: inline response fetch via Fetch.takeResponseBodyAsStream,
async loading via browser network stack with Network.loadNetworkResource() and node-based async fetch
via fetch()
- Direct async fetch() capture of non-HTML URLs
- Awaiting for all requests to finish before moving on to next page, upto page timeout.
- Experimental: generate CDXJ on-the-fly as WARC is being written (not yet in use).
- removed pywb, using cdxj-indexer for --generateCDX option.
* behavior logging tweaks, add netIdle
* fix shouldIncludeFrame() check: was actually erroring out and never accepting any iframes!
now used not only for link extraction but also to run() behaviors
* add logging if iframe check fails
* Dockerfile: add commented out line to use local behaviors.js
* bump behaviors to 0.5.2
treat non-regexes as strings and pass to RegExp constructor
tests: add additional scope parsing tests for different types passed in as exclusions
update yargs
bump to 0.10.4
* support loading custom behaviors from a specified directory via --customBehaviors
* call load() for each behavior incrementally, then call selectMainBehavior() (available in browsertrix-behaviors 0.5.1)
* tests: add tests for multiple custom behaviors
---------
Co-authored-by: Ilya Kreymer <ikreymer@gmail.com>
* base: update to chrome 112
headless: switch to using new headless mode available in 112 which is more in sync with headful mode
viewport: use fixed viewport matching screen dimensions for headless and headful mode (if GEOMETRY is set)
profiles: fix catching new window message, reopening page in current window
versions: bump to pywb 2.7.4, update puppeteer-core to (20.2.1)
bump to 0.10.0-beta.4
* profile: force reopen in current window only for headless mode (currently breaks otherwise), remove logging messages
- reduced memory usage, avoids memory leak issues caused by using playwright (see #298)
- browser: split Browser into Browser and BaseBrowser
- browser: puppeteer-specific functions added to Browser for additional flexibility if need to change again later
- browser: use defaultArgs from playwright
- browser: attempt to recover if initial target is gone
- logging: add debug logging from process.memoryUsage() after every page
- request interception: use priorities for cooperative request interception
- request interception: move to setupPage() to run once per page, enable if any of blockrules, adblockrules or originOverrides are used
- request interception: fix originOverrides enabled check, fix to work with catch-all request interception
- default args: set --waitUntil back to 'load,networkidle2'
- Update README with changes for puppeteer
- tests: fix extra hops depth test to ensure more than one page crawled
---------
Co-authored-by: Tessa Walsh <tessa@bitarchivist.net>
* various loading improvements to avoid pages getting 'stuck' + load state tracking
- add PageState object, store loadstate (0 to 4) as well as other per-page-state properties on defined object.
- set loadState to 0 (failed) by default
- set loadState to 1 (content-loaded) on 'domcontentloaded' event
- if page.goto() finishes, set to loadState to 2 'full-page-load'.
- if page.goto() times out, if no domcontentloaded either, fail immediately. if domcontentloaded reached, extract links, but don't run behaviors
- page considered 'finished' if it got to at least loadState 2 'full-pageload', even if behaviors timed out
- pages: log 'loadState' as part of pages.jsonl
- improve frame detection: detect if frame actually not from a frame tag (eg. OBJECT) tag, and skip as well
- screencaster: try screencasting every frame for now instead of every other frame, for smoother screencasting
- deps: behaviors: bump to browsertrix-behaviors 0.5.0-beta.0 release (includes autoscroll improvements)
- workers ids: just use 0, 1, ... n-1 worker indexes, send numeric index as part of screencast messages
- worker: only keeps track of crash state to recreate page, decouple crash and page failed/succeeded state
- screencaster: allow reusing caster slots with fixed ids
- interrupt timedCrawlPage() wait if 'crash' event happens
- crawler: pageFinished() callback when page finishes
- worker: add workerIdle callback, call screencaster.stopById() and send 'close' message when worker is empty
* Migrate from Puppeteer to Playwright!
- use playwright persistent browser context to support profiles
- move on-new-page setup actions to worker
- fix screencaster, init only one per page object, associate with worker-id
- fix device emulation: load on startup, also replace '-' with space for more friendly command-line usage
- port additional chromium setup options
- create / detach cdp per page for each new page, screencaster just uses existing cdp
- fix evaluateWithCLI to call CDP command directly
- workers directly during WorkerPool - await not necessary
* State / Worker Refactor (#252)
* refactoring state:
- use RedisCrawlState, defaulting to local redis, remove MemoryCrawlState and BaseState
- remove 'real' accessors / draining queue - no longer neede without puppeteer-cluster
- switch to sorted set for crawl queue, set depth + extraHops as score, (fixes#150)
- override console.error to avoid logging ioredis errors (fixes#244)
- add MAX_DEPTH as const for extraHops
- fix immediate exit on second interrupt
* worker/state refactor:
- remove job object from puppeteer-cluster
- rename shift() -> nextFromQueue()
- condense crawl mgmt logic to crawlPageInWorker: init page, mark pages as finished/failed, close page on failure, etc...
- screencaster: don't screencast about:blank pages
* more worker queue refactor:
- remove p-queue
- initialize PageWorkers which run in its own loop to process pages, until no pending pages, no queued pages
- add setupPage(), teardownPage() to crawler, called from worker
- await runWorkers() promise which runs all workers until completion
- remove: p-queue, node-fetch, update README (no longer using any puppeteer-cluster base code)
- bump to 0.9.0-beta.1
* use existing data object for per-page context, instead of adding things to page (will be more clear with typescript transition)
* more fixes for playwright:
- fix profile creation
- browser: add newWindowPageWithCDP() to create new page + cdp in new window, use with timeout
- crawler: various fixes, including for html check
- logging: addition logging for screencaster, new window, etc...
- remove unused packages
---------
Co-authored-by: Tessa Walsh <tessa@bitarchivist.net>
* This commit removes puppeteer-cluster as a dependency in favor of
a simpler concurrency implementation, using p-queue to limit
concurrency to the number of available workers. As part of the
refactor, the custom window concurrency model in windowconcur.js
is removed and its logic implemented in the new Worker class's
initPage method.
* Remove concurrency models, always use new tab
* logging improvements: include worker-id in logs, use 'worker' context
- logging: log info string / version as first line
- logging: improve logging of error stack traces
- interruption: support interrupting crawl directly with 'interrupt' check which stops the job queue
- interruption: don't repair if interrupting, wait for queue to be idle
- log text extraction
- init order: ensure wb-manager init called first, then logs created
- logging: adjust info->debug logging
- Log no jobs available as debug
* tests: bail on first failure
* iframe filtering:
- fix filtering for about:blank iframes, support non-async shouldProcessFrame()
- filter iframes both for behaviors and for link extraction
- add 5-second timeout to link extraction, to avoid link extraction holding up crawl!
- cache filtered frames
* healthcheck/worker reuse:
- refactor healthchecker into separate class
- increment healthchecker (if provided) if new page load fails
- remove expermeintal repair functionality for now
- add healthcheck
* deps: bump puppeteer-core to 17.1.2
- bump to 0.9.0-beta.0
--------
Co-authored-by: Ilya Kreymer <ikreymer@gmail.com>
- Ensure page is included in all logging details
- Update logging messages to be a single string, with variables added in the details
- Always wait for all pending wait requests to finish (unless counter <0)
- Don't set puppeteer-cluster timeout (prep for removing puppeeteer-cluster)
- Add behaviorTimeout to running behaviors in crawler, in addition to in behaviors themselves.
- Add logging for behavior start, finish and timeout
- Move writeStats() logging to beginning of each page as well as at the end, to avoid confusion about pending pages.
- For events from frames, use frameUrl along with current page
- deps: bump browsertrix-behaviors to 0.4.2
- version: bump to 0.8.1
* profiles: use vnc for automatic profile creation (fixes#194):
- add x11vnc and serve via vnc when not headless, keep existing screencast for headless mode
- use @novnc/novnc to serve vnc JS library
- add novnc_lite.html to serve the content from an iframe
- optimization: don't show initial blank page / don't wait for initial page in puppeteer
* more vnc work:
- set position of browser at 0,0, avoid needing offset to fit
- add /vncpass endpoint to query vnc password (for use with browsertrix-cloud)
- remove websockify, x11vnc now supports ws connections directly!
- vnc_lite: support reconnecting ws if gracefully disconnected
* x11vnc cleanup: just pass password via cmdline to simplify setup
* make interactive profile creation default, automated enabled only if --automated or --username / --password flags are specified
README updates:
- mention new VNC-based streaming
- mention new --automated flag, move automated info below interactive
* README: adjust auto-login example to use mastodon example instead of twitter, which works more consistently
* switch base image to chrome/chromium 105 with node 18.x
* convert all source to esm for node 18.x, remove unneeded node-fetch dependency
* ci: use node 18.x, update to latest actions
* tests: convert to esm, run with --experimental-vm-modules
* tests: set higher default timeout (90s) for all tests
* tests: rename driver test fixture to .mjs for loading in jest
* bump to 0.8.0
- netIdleWait better defaults: if not set, set to 15 seconds for page/page-spa scope, otherwise to 2 seconds
- default behaviors: include autoscroll in default behavior as well
- restart: if crawl already done, don't attempt to crawl further. if 'waitOnDone' set, wait for signal before exiting.
- bump to puppeteer-core 17.1.2
- bump to 0.7.0-beta.4
* new window: use cdp instead of window.open
* new window tweaks: add reuseCount, use browser.target() instead of opening a new blank page
* rename NewWindowPage -> ReuseWindowConcurrency, move to windowconcur.js
potential fix for #156
* browser repair:
- when using window-concurrency, attempt to repair / relaunch browser if cdp errors occur
- mark pages as failed and don't reuse if page error or cdp errors occur
- screencaster: clear previous targets if screencasting when repairing browser
* bump version to 0.7.0-beta.3
improved logging of pywb + redis:
- if 'logging' includes 'pywb', log pywb and redis output, to pywb.log and redis.log
- otherwise, just ignore (don't print to stdout as that's too confusing)
- print if wb-manager fails, likely due to existing collection
waitUntil: default to just 'load' to avoid potential infinite loop, separate --netIdle can configure idle wait
dependency: update to latest puppeteer-core (16.1.0)
- add --netIdleWait option, default to 10 seconds - necessary for some sites that start fetching immediately after page load
- add openssl.conf to allow pywb to avoid 'unsafe legacy renegotiation disabled' from openssl
- update to browsertrix-behaviors 0.3.2
- update current url for screencasting of page before page load starts
bump to 0.7.0-beta.2
- Add optional health check via `--healthCheckPort`. If set, runs a server on designated port that returns 200 if healthcheck succeeds (num of consecutive failed page loads < 2*num workers), or 503 if fails. Useful for k8s health check
- Add crawl size limit (in bytes), via `--sizeLimit`. Crawl exits (and state optionally saved) when size limit is exceeded.
- Add crawl total time limit (in seconds), via `--timeLimit`. Crawl exists (and state optionally saved) when total running time is exceeded.
- Add option to overwrite existing collection. If `--overwrite` is included, any existing data for specified collection is deleted.
- S3 Storage refactor, simplify, don't add additional paths by default.
- Add interpolateFilename as generic utility, supported in filename and STORE_PATH env value.
- wacz save: reenable wacz validation after save.
- Profiles: support /navigate endpoint, return origins from /ping, prevent opening new tabs.
- bump to 0.6.0-beta.1
- update pywb to 2.6.7, fix possible error cdx indexing ever via --generateCDX
- update wacz to 0.4.6, ensure wacz file is closed and better and more error-resilient text extraction
- update browsertrix-behaviors to 0.3.0, support for telegram behavior
- bump version to 0.5.1
* update to browsertrix-behaviors 0.2.5 to support improved autoscroll
- add evaluateWithCLI() to support evaluate() with 'getEventListeners()' and other devtools command-line api functions, to allow autoscroll behavior to check if it should exit out early
- inject behaviors into interactive loader to allow testing
- fix signal handler if state not inited yet
- dependencies: update puppeteer-cluster to latest, update pywb to 2.6.5
- wacz: update to py-wacz 0.4.1, avoid reading full file into memory to compute hashes
state: fix pending state, account for puppeteer-cluster popping/pushing jobs from queue:
* puppeteer-cluster: add custom 'start()' callback to indicate task actually starting
* new semantics: add pending urls in pending state immediately, remove if readded to queue, add 'started' when actaully started
minio: use fPutObject to support parallel uploading, compute hash and size separately (for now)
dependencies: update to latest minio
error checking:
* print number of WARCs found, exit with error if 0
* ensure wacz creation succeeds, exit with error code if not
* validate wacz after creation, exit with error code if validation fails
bump to 0.5.0-beta.3
* extra hops depth: add support for --extraHops option, which expands the inclusion scope to go N 'extra hops' beyond the existing scope. fixes most common use case in #83
* update README with info on `extraHops`, add tests for extraHops
* dependency fix: use pywb 2.6.3, warcio 1.5.0
* bump to 0.5.0-beta.2
- support uploading WACZ to s3-compatible storage (via minio client)
- config storage loaded from env vars, enabled when WACZ output is used.
- support pinging either or an http or a redis key-based webhook,
- webhook: include 'completed' bool to indicate if fully completed crawl or partial (eg. interrupted via signal)
- consolidate redis init to redis.js
- support upload filename with custom variables: can interpolate current timestamp (@ts), hostname (@hostname) and user provided id (@crawlId)
- README: add docs for s3 storage, remove unused args
- update to pywb 2.6.2, browsertrix-behaviors 0.2.4
* fix to `limit` option, ensure limit check uses shared state
* bump version to 0.5.0-beta.1
* save state work:
- support interrupting and saving crawl
- support loading crawl state (frontier queue, pending, done) from YAML
- support scope check when loading to apply new scoping rules when restarting crawl
- failed urls added to done as failed, can be retried if crawl is stopped and restarted
- save state to crawls/crawl-<ts>-<id>.yaml when interrupted
- --saveState option controls when crawl state is saved, default to partial/when interrupted, also always, never.
- support in-memory or redis based crawl state, using fork of puppeteer-cluster
- --redisStore used to enable redis-based state
* signals/crawl interruption:
- crawl state set to drain/not provide any more urls to crawl
- graceful stop of crawl in response to sigint/sigterm
- initial sigint/sigterm waits for graceful end of current pages, second terminates immediately
- initial sigabrt followed by sigterm terminates immediately
- puppeteer disable handleSIGTERM, handleSIGHUP, handleSIGINT
* redis state support:
- use lua scripts for atomic move from queue -> pending, and pending -> done
- pending key expiry set to page timeout
- add numPending() and numSeen() to support better puppeteer-cluster semantics for early termination
- drainMax returns the numPending() + numSeen() to work with cluster stats
* arg improvements:
- add --crawlId param, also settable via CRAWL_ID env var, defaulting to os.hostname() (used for redis key and crawl state file)
- support setting cmdline args via env var CRAWL_ARGS
- use 'choices' in args when possible
* build update:
- switch base browser image to new webrecorder/browsertrix-browser-base, simple image with .deb files only for amd64 and arm64 builds
- use setuptools<58.0
* misc crawl/scoping rule fixes:
- scoping rules fix when external is used with scopeType
state:
- limit: ensure no urls, including initial seeds, are added past the limit
- signals: fix immediate shutdown on second signal
- tests: add scope test for default scope + excludes
* py-wacz update
- add 'seed': true to pages that are seeds for optimized wacz creation, keeping non-seeds separate (supported via wacz 0.3.2)
- pywb: use latest pywb branch for improved twitter video capture
* update to latest browsertrix-behaviors
* fix setuptools dependency #88
* update README for 0.5.0 beta
* blockrules improvements:
- add await to continue/abort to catch errors, each called only in one place.
- avoid adding multiple interception handlers for same page to avoid 'request already handled' errors
- disallow blocking full pages via blockRules (should be handled via scope exclusion) and print warning
* setup: ensure the 'cwd' for the crawl output exists on startup, in case a custom cwd was set.
* scopeType rename:
- rename 'page' -> page-spa to indicate support for hashtag / single-page-app intended usage
- rename 'none' -> page to indicate default single-page-only crawl
- messaging: adjust error message displaying valid scopeTypes
* README: Add additional examples for scope rules, update scopeType param, explain different between scope rules vs block rules, to better address confusion as per #80
bump to 0.4.4
