blob: e103fdebd233bc4125aeec270f36cb2c84c90c86 [file] [log] [blame]
// The analysis package defines a uniform interface for static checkers
// of Go source code. By implementing a common interface, checkers from
// a variety of sources can be easily selected, incorporated, and reused
// in a wide range of programs including command-line tools, text
// editors and IDEs, build systems, test frameworks, code review tools,
// and batch pipelines for large code bases. For the design, see
// https://docs.google.com/document/d/1-azPLXaLgTCKeKDNg0HVMq2ovMlD-e7n1ZHzZVzOlJk
//
// Each analysis is invoked once per Go package, and is provided the
// abstract syntax trees (ASTs) and type information for that package.
//
// The principal data types of this package are structs, not interfaces,
// to permit later addition of optional fields as the API evolves.
package analysis
import (
"flag"
"fmt"
"go/ast"
"go/token"
"go/types"
"reflect"
)
// An Analysis describes an analysis function and its options.
type Analysis struct {
// The Name of the analysis must be a valid Go identifier
// as it may appear in command-line flags, URLs, and so on.
Name string
// Doc is the documentation for the analysis.
Doc string
// Flags defines any flags accepted by the analysis.
// The manner in which these flags are exposed to the user
// depends on the driver which runs the analysis.
Flags flag.FlagSet
// Run applies the analysis to a package.
// It returns an error if the analysis failed.
Run func(*Unit) error
// RunDespiteErrors allows the driver to invoke
// the Run method of this analysis even on a
// package that contains parse or type errors.
RunDespiteErrors bool
// Requires is a set of analyses that must run successfully
// before this one on a given package. This analysis may inspect
// the outputs produced by each analysis in Requires.
// The graph over analyses implied by Requires edges must be acyclic.
//
// Requires establishes a "horizontal" dependency between
// analysis units (different analyses, same package).
Requires []*Analysis
// OutputType is the type of the optional Output value
// computed by this analysis and stored in Unit.Output.
// (The Output is provided as an Input to
// each analysis that Requires this one.)
OutputType reflect.Type
// LemmaTypes is the set of types of lemmas produced and
// consumed by this analysis. An analysis that uses lemmas
// may assume that its import dependencies have been
// similarly analyzed before it runs. Lemmas are pointers.
//
// LemmaTypes establishes a "vertical" dependency between
// analysis units (same analysis, different packages).
LemmaTypes []reflect.Type
}
func (a *Analysis) String() string { return a.Name }
// A Unit provides information to the Run function that
// applies a specific analysis to a single Go package.
//
// It forms the interface between the analysis logic and the driver
// program, and has both input and an output components.
type Unit struct {
// -- inputs --
Analysis *Analysis // the identity of the current analysis
// syntax and type information
Fset *token.FileSet // file position information
Syntax []*ast.File // the abstract syntax tree of each file
Pkg *types.Package // type information about the package
Info *types.Info // type information about the syntax trees
// Inputs provides the inputs to this analysis unit, which are
// the corresponding outputs of its prerequisite analysis.
// The map keys are the elements of Analysis.Required,
// and the type of each corresponding value is the required
// analysis's OutputType.
Inputs map[*Analysis]interface{}
// ObjectLemma retrieves a lemma associated with obj.
// Given a value ptr of type *T, where *T satisfies Lemma,
// ObjectLemma copies the value to *ptr.
//
// ObjectLemma may panic if applied to a lemma type that
// the analysis did not declare among its LemmaTypes,
// or if called after analysis of the unit is complete.
//
// ObjectLemma is not concurrency-safe.
ObjectLemma func(obj types.Object, lemma Lemma) bool
// PackageLemma retrives a lemma associated with package pkg,
// which must be this package or one if its dependencies.
// See comments for ObjectLemma.
PackageLemma func(pkg *types.Package, lemma Lemma) bool
// -- outputs --
// Findings is a list of findings about specific locations
// in the analyzed source code, such as potential mistakes.
// It is populated by the Run function.
Findings []*Finding
// SetObjectLemma associates a lemma of type *T with the obj,
// replacing any previous lemma of that type.
//
// SetObjectLemma panics if the lemma's type is not among
// Analysis.LemmaTypes, or if obj does not belong to the package
// being analyzed, or if it is called after analysis of the unit
// is complete.
//
// SetObjectLemma is not concurrency-safe.
SetObjectLemma func(obj types.Object, lemma Lemma)
// SetPackageLemma associates a lemma with the current package.
// See comments for SetObjectLemma.
SetPackageLemma func(lemma Lemma)
// Output is an immutable result computed by this analysis unit
// and set by the Run function.
// It will be made available as an input to any analysis that
// depends directly on this one; see Analysis.Requires.
// Its type must match Analysis.OutputType.
//
// Outputs are available as Inputs to later analyses of the
// same package. To pass analysis results between packages (and
// thus potentially between address spaces), use Lemmas, which
// are serializable.
Output interface{}
/* Further fields may be added in future. */
// For example, suggested or applied refactorings.
}
// Findingf is a helper function that creates a new Finding using the
// specified position and formatted error message, appends it to
// unit.Findings, and returns it.
func (unit *Unit) Findingf(pos token.Pos, format string, args ...interface{}) *Finding {
msg := fmt.Sprintf(format, args...)
f := &Finding{Pos: pos, Message: msg}
unit.Findings = append(unit.Findings, f)
return f
}
func (unit *Unit) String() string {
return fmt.Sprintf("%s@%s", unit.Analysis.Name, unit.Pkg.Path())
}
// A Lemma is an intermediate fact produced during analysis.
//
// Each lemma is associated with a named declaration (a types.Object).
// A single object may have multiple associated lemmas, but only one of
// any particular lemma type.
//
// A Lemma represents a predicate such as "never returns", but does not
// represent the subject of the predicate such as "function F".
//
// Lemmas may be produced in one analysis unit and consumed by another
// analysis unit even if these are in different address spaces.
// If package P imports Q, all lemmas about objects of Q produced during
// analysis of that package will be available during later analysis of P.
// Lemmas are analogous to type export data in a build system:
// just as export data enables separate compilation of several units,
// lemmas enable "separate analysis".
//
// Each unit of analysis starts with the set of lemmas produced by the
// same analysis applied to the packages directly imported by the
// current one. The analysis may add additional lemmas to the set, and
// they may be exported in turn. An analysis's Run function may retrieve
// lemmas by calling Unit.Lemma and set them using Unit.SetLemma.
//
// Each type of Lemma may be produced by at most one Analysis.
// Lemmas are logically private to their Analysis; to pass values
// between different analysis, use the Input/Output mechanism.
//
// A Lemma type must be a pointer. (Unit.GetLemma relies on it.)
// Lemmas are encoded and decoded using encoding/gob.
// A Lemma may implement the GobEncoder/GobDecoder interfaces
// to customize its encoding; Lemma encoding should not fail.
//
// A Lemma should not be modified once passed to SetLemma.
type Lemma interface {
IsLemma() // dummy method to avoid type errors
}
// A Finding is a message associated with a source location.
//
// An Analysis may return a variety of findings; the optional Category,
// which should be a constant, may be used to classify them.
// It is primarily intended to make it easy to look up documentation.
type Finding struct {
Pos token.Pos
Category string // optional
Message string
}