class

Analyzer::Crystal::Amber

Inherits Analyzer::Crystal::CrystalEngine < FileScanEngine < Analyzer < FileHelper < Reference < Object

Constants

AMBER_MARKER_RE = /require\s+"amber"|Amber::|^\s*routes\s+:\w+/m

get "/path" is the same line in Amber, Kemal and Grip — the DSLs are indistinguishable one line at a time, and this analyzer is handed every .cr file in the scan. Without a file-level gate it read Kemal's, Grip's and Lucky's routes as Amber ones (22 phantom endpoints across the Crystal fixture tree), and swallowed their public_folder declarations on the way through.

A file that declares Amber routes belongs to an Amber app, so it names the framework: require "amber" in the fixtures and in an app's entry point, Amber::Server.configure in a real config/routes.cr, or the routes :web do pipeline block, which is Amber's DSL and nobody else's.

CONTROLLER_ROUTE_RE = /(?:^|[^.\w])(get|post|put|delete|patch|head|options|ws)\s*(?:\(\s*)?['"](.+?)['"]\s*,\s*\w+,\s*:\w+/

Amber prefers verb "/path", Controller, :action. One combined PCRE2 scan replaces the previous 8 sequential controller-form scans; on miss, the shared simple-route helper covers the controller-less fallback (another single scan). Was 16 unguarded .scan calls per source line.

NAMESPACE_PATTERN = /^(\s*)namespace\s+["']([^"']*)["']\s+do\b/

namespace "/admin" do … end — Amber's DSL scope macro.

RESOURCE_ROUTES = [{"GET", "", :index}, {"GET", "/new", :new}, {"POST", "", :create}, {"GET", "/:id", :show}, {"GET", "/:id/edit", :edit}, {"PUT", "/:id", :update}, {"PATCH", "/:id", :update}, {"DELETE", "/:id", :destroy}]

Amber's resources macro expands to the seven RESTful routes below (update is registered for both PUT and PATCH). resource (singular) shares the same action set; the :id segment is auto-detected as a path param downstream, so we only emit the verb + URL + action here.

RESOURCES_PATTERN = /^\s*resources\s+["']([^"']+)["']\s*,\s*([A-Za-z_]\w*(?:::[A-Za-z_]\w*)*)\s*(.*)$/

resources "/posts", PostController[, only: …][, except: …].

ROUTES_SCOPE_PATTERN = /^(\s*)routes\s+:\w+(?:\s*,\s*["']([^"']*)["'])?\s+do\b/

routes :web, "/admin" do … end — the optional second argument is a path scope that prefixes every route declared inside the block.

SERVE_STATIC_DISABLED_RE = Regex.union("serve_static false", "serve_static(false)")

serve_static false / serve_static(false) — checked on every source line, so the two fixed OR-ed String#includes? substring scans are precompiled into a single regex union (one PCRE2 JIT scan per line instead of two naive substring scans).

Class methods

tech_name
Source

Instance methods

analyze
Source
analyze_file(path : String) : Array(Endpoint)
Source
line_to_endpoint(content : String) : Endpoint
Source
line_to_param(content : String) : Param
Source
tech

Instance-side view of the same declaration. The per-file rescues live on this base class, which has no way to name the analyzer that is running inside them, so a skipped file could not be attributed to a tech. Deriving it from analyzer_for keeps the name written exactly once.

Source