Two axes, selected explicitly | Explicit non-decisions | Bounded vendor profiles | Contract metadata

Why a selector? | The canonical cases | Encoded unreserved bytes (%41%42) | Encoded reserved bytes (%2F) | Whole-number IPv4 (2130706433) | Diagnostics are facts, not policy | Ports and backslashes | Resolving references | What "whatwg" changes | Which colon starts a scheme | Asking for the standard string instead | Migration notes | What the selector does not govern | Conformance posture

Reading a literal | The leading zero | Short forms and the whole-host number | The trailing dot, and the WHATWG asymmetry | IPv6: where the disagreement moves | The zone travels beside the bits | What raddr models, and what it does not | Classification | Longest prefix match, and why a parent block is not the answer | 0.0.0.0 is two rows, not one | "Not globally reachable" is not "not forwardable" | The N/A rows, and two different reasons for one NA | Reason codes at the classify layer | RFC 6890 is a stale snapshot | Provenance | What none of this decides

Libraries in other ecosystems | Observed behavior on the comparable R packages

Introduction | Safe URL Parsing | Extracting URL Components | Domain and TLD Parsing | Vectorization and Edge Cases | Advanced Host Manipulation with subdomain_levels_to_keep | Getting the whole URL back | Caching | Summary | See also | Acknowledgments

What pslr does differently | Trade-offs

The problem, in one string | Dialects sit on two axes | Measured divergence | For IPv6 the disagreement inverts | raddr is a shower, not a protector | What raddr will never do | Why not just use ipaddress?

Two layers, and they are not the same kind of statement | A helper for the examples below | The parse layer | not_a_number | leading_zero | empty_part | empty_hex | out_of_range | wrong_part_count | trailing_dot | zone_not_permitted | multiple_zones | bad_hextet | empty_group | bad_elision | wrong_group_count | bad_embedded_ipv4 | whitespace | The classify layer | nat64_wk_embedded_not_global — must | sixtofour_embedded_not_global — must | teredo_client_not_global — must | link_local_outside_fe80_64 — must | link_local_reserved_range — must | ipv4_compatible_low_tail — may | nat64_local_layout_unspecified — unspecified | ula_l_bit_unset — unspecified | What a code does not prove

Inspecting a URL | Inspecting what a name resolves to | Fetching through the guard | Following redirects | What the guard cannot protect | Testing against a local server

Introduction | Installation | Quick Start | Screaming Frog crawl exports | Redirect Resolution | Basic redirect handling | Conflicting redirects | Redirect loops | Standalone link resolution | Nofollow Handling | Indexability | noindex pages | robots.txt blocked pages | Weighted Edges | Domain Filtering | HITS hub and authority | SALSA hub and authority | Model Comparison | Comparing two models | Parameter grid search | Distribution metrics | Simulating Changes | Adding links | Removing links | Simulating redirects | Combined changes | Convergence Controls and Damping Stability | pagerank() convergence controls | damping_sensitivity() and pagerank_stability() | Worked example | Topic-Sensitive PageRank | Distinguishing related functions | GA4 Behavioral Transitions | Building transition counts with ga4_page_transitions() | Smoothing sparse transitions with smooth_transitions() | Converting entrances to a teleport vector with ga4_entrance_teleport() | End-to-end example | Function Reference | Acknowledgments

What pslr does | Terminology | Choosing a section | section = "private" fall-through | Unknown-suffix policy | Explicit-membership queries | Unicode and ASCII output | Terminal dots | Extracting and inspecting | Refresh and the active list | Freshness: what is known versus what is merely due | Why "check due" is not "update available" | ETag and Last-Modified versus SHA-256 | The four successful refresh outcomes | Courtesy: no more than daily | Opt-in reminders | Refreshing on a schedule | Snapshots, retention, and pruning | Comparing two snapshots | Multiple lists with engines | Reproducibility | Security and scope notes | See also | Acknowledgments

Reading a sitemap | Validating a sitemap | Discovering a site's sitemaps | Where to go next | Acknowledgments

Introduction | Why punycoder? | The Problem | The Solution | Basic Usage | Domain Encoding and Decoding | Validation and Utilities | Host Normalization | Invalid Input Returns NA | Profile Identity | Unicode Version Selection | Data Analysis Workflows | Bulk Domain Processing | Error Handling | Performance Considerations | Package Options | Integration with Other Packages | Next Steps | Technical Details | See also | Acknowledgments

One string, three hosts | Paper and reality | addr_parse() takes no mode argument | Rejection is not missingness | Why addr_curl() exists alongside the standards | The reality dialects model a measured platform | Classification takes addresses, not strings | Facts, not verdicts | Where to go next

Diagnose, model, ship, re-measure | Loading the two crawls | Two lenses on the same crawl | PageRank does not sum to 1 here | The headline: concentration collapsed | Where the authority went | The mechanism was reclassification, not new links | The full graph barely noticed | Confounds — read this before reusing the method | Running this on your own crawls

The links placement cannot reach | The metric | Why the container is the denominator | Turning it on | It is a downweight, never a drop | The four arguments | Composing with region weighting | What to expect on a real site: author pages will drop | Getting a container column from a crawl | See also

Why presets exist | The four presets | "raw" — the faithful baseline | "declared" — a pin, not a change | "reversed" — the feeder view | "content" — stop letting the navigation rank the site | Layers vs. presets: raw, real, and annotated | Precedence: explicit argument > preset > default | Inspecting and adapting a bundle | Provenance in the transition audit | See also

What TrustRank is | It reuses the personalization path — no new solver | A worked example | Build the seed prior explicitly | Or use the convenience wrapper | Compare against uniform PageRank | Graded trust and a teleport floor | Notes

What robotstxtr does | Crawl-decision entry points | Matching against a supplied body | Vectorized, with per-URL user agents | Rich match metadata | Inspecting the stored body | Fetching robots.txt for a URL | Validating the document itself | The URL you pass is the URL that is matched | Summary

The question this answers | Why it is not just re-reading PageRank | A worked example | Reading the output | Contrast with the forward (authority) view | Where it sits among the reverse-direction tools | Notes