Repository navigation
html: deep pages, text in other scripts, every element of the standard, and a byte order mark - #177
Merged
Merged
Conversation
…d, and a byte order mark Four new settings for the html format, each off by default, so files made with the earlier settings keep their bytes. - nesting puts the content inside up to 100 000 div elements, the label outside them. lxml drops the content of a page about 250 deep, Chromium flattens past 511, and jsdom, parse5 and DOMPurify can run out of stack writing out a page from about 2 000 deep. - characters=international adds paragraphs in Polish, Russian, Japanese and Arabic (right to left), emoji, a letter in two code points and entities. - elements=every adds a block using every element of the HTML standard but script, all 22 kinds of input, and pictures and a sound written into the page as data. The Nu HTML Checker finds nothing wrong with it. - bom puts a UTF-8 byte order mark in front, from the shared text encoding axis. Content sniffing as the WHATWG describes it then reads the page as plain text. The structural check is told every setting and checks each one from the bytes. Four golden cases pin the new paths. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
Important Review skippedAuto incremental reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the ⚙️ Run configuration
You can disable this status message by setting the Use the checkbox below for a quick retry:
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info
📜 Recent review details
📝 Walkthrough
Merge Risk: ⚪ Minimal · up to
|
| Check name | Status | Explanation | Resolution |
|---|---|---|---|
| Safe File Parsing |  |
The new HTML checker can crash or hang on malformed settings or input. In internal/oracle/strict.py:700-703, settings.get("nesting") is accepted after .isdigit() and passed to int() without a … |
Bound and validate nesting before conversion. Require a string matching a bounded decimal form (for example [0-9]{1,6}), enforce 0 <= value <= 100000, and convert inside guarded handling; report invalid values with fail() instead of… |
-
Check name Status Explanation Description Check Check skipped - CodeRabbit’s high-level summary is enabled. Title check The title clearly identifies the main user-facing HTML changes: deep nesting, international text, broad element coverage, and UTF-8 BOM support. It is specific and suitable for release notes. Linked Issues check Check skipped because no linked issues were found for this pull request. Out of Scope Changes check Check skipped because no linked issues were found for this pull request. Tests For Changed Behavior The PR adds non-UI HTML runtime behavior and adds direct coverage for it. internal/guard/htmlcatalogue_test.gochecks elements, input types, international text, ASCII output, and BOM placement. `int…No Secrets Or Debug Leftovers The pull-request diff adds no CLAUDE.md, CLAUDE.local.md, AGENTS.md, .claude/, or .env files. Added-line scans found no credentials, API keys, private URLs, email addresses, IPs, local absolute paths,… No Hardcoded Ui Styling The pull request does not change GUI implementation code. Its only internal/guichanges are English and Polish locale JSON entries for labels and messages. No XAML, Slint, Fyne, Tkinter, or WPF code…No Obvious Performance Problems No clear performance problem was introduced. The HTML generator performs linear-size string construction and record filling, including bounded 100,000-level nesting. The strict checker uses linear sca… Desktop Robustness No desktop robustness failure is introduced. The changed HTML generator writes only to its supplied io.Writer and passes context cancellation to core.FillRecords. The embedded media uses data: URIs, w… System Changes Are Reversible PASS. The pull request changes HTML generation, validation, tests, documentation, translations, and static pages. The implementation uses local byte construction and embedded data:resources; it doe…Clear User-Facing Text PASS. The PR adds user-facing HTML setting descriptions, localized labels, format-table entries, notes, and validation messages. The GUI composes non-choice controls with generated allowed values and … No Resource Leaks No resource leak is introduced. The changed Go code generates bounded HTML buffers and passes the existing context-aware core.FillRecords; it adds no handlers, timers, goroutines, external handles, …Scope, Duplication And Docs No scope, duplication, documentation, or breaking-change violation found. The PR is cohesive: HTML settings, validation, tests, localization, and generated site pages. README and CHANGELOG document th…
Full details: Safe File Parsing
-
Explanation
The new HTML checker can crash or hang on malformed settings or input. In
internal/oracle/strict.py:700-703,settings.get("nesting")is accepted after.isdigit()and passed toint()without a bound or exception handling. A 5000-digit value raisesValueErrorin Python 3.11. Inhtml_settingsline 744,re.findall(r'<input[^>]*\btype="([a-z-]+)"', text)scans to the next>for every<input; malformed input with repeated<inputand no>shows quadratic behavior (about 4.2 seconds for 8,000 markers in an isolated reproduction). The newelements=everypath activates this code before the existing tag scan.Resolution
Bound and validate
nestingbefore conversion. Require a string matching a bounded decimal form (for example[0-9]{1,6}), enforce0 <= value <= 100000, and convert inside guarded handling; report invalid values withfail()instead of allowingint()to raise. Replace the unboundedre.findallinput-type expression with a single-pass tag/attribute scanner orhtml.parser.HTMLParsercallback that cannot rescan an unterminated tag. Add checker tests for a 5000-digit nesting setting and repeated unterminated<inputmarkers, and require a normalFAILresult.
- Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.
Comment @coderabbitai help to get the list of available commands.
coderabbitai
Bot
added
bug
…eads nesting safely The parity guard classifies html.bom, characters, elements and nesting as reachable from the window. The structural check accepts only ASCII digits up to 100000 for nesting before converting it, and asks for the page skeleton with one literal pattern. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Four new settings for the
htmlformat. Each is off by default, so files made with the earlier settings keep their bytes (the golden values are unchanged).nesting- the content of a page inside up to 100 000divelements, each inside the last, with the label outside them. Readers part ways early: lxml drops everything but the label from a page about 250 deep, Chromium keeps 511 levels and moves the deeper ones out, and jsdom, parse5 and DOMPurify can run out of stack writing out a page from about 2 000 deep. The Nu HTML Checker calls a page past 513 elements deep an error.characters=international- paragraphs in Polish, Russian, Japanese and Arabic (right to left), emoji with a ZWJ sequence, a letter written as two code points, and named and numbered entities.elements=every- a block that uses every element of the HTML standard butscriptat least once (114 of the 115 names in its index), all 22 kinds ofinput, and pictures and a sound written into the page asdata:, so opening it fetches nothing. The Nu HTML Checker finds nothing wrong with it. A fragment has no head, so it leaves out the eight elements only a whole page holds, and the manifest carries a note saying so.bom- a UTF-8 byte order mark in front, from the shared text encoding axis. Chromium andfilestill see HTML. Content sniffing as the WHATWG describes it, which Go'snet/httpfollows, reads it as plain text.Around it
htmlfilesplit intosettings.go,page.go,blocks.go(bytes unchanged),catalogue.goandcharacters.go. The bytes that close the levels, the body and the document stay in the last record. The minimum is summed from lengths, so a deep page keeps 600 KB out of every plan.strict.pycheck_htmlis told all five settings and checks each from the bytes. It no longer takes<header>for<head>.Checked by
htmlsettings_test.go,htmlcatalogue_test.go), four new golden values.🤖 Generated with Claude Code
Summary by CodeRabbit