public class | source

Inspector

Extends:

lib/inspector.js~EventEmitter → Inspector

Constructor Summary

Public Constructor
public	constructor(filePaths: string[], opts: object) Creates a new Inspector, which extends EventEmitter.

Method Summary

Public Methods
public	run() Runs the inspector on the given file paths, as provided in the constructor.

Private Methods
private	_analyze() Traverses the keys at which the various nodes are stored.
private	_expand(nodeArrays: Node[][]) Expands each instance of a match to the largest common sequence of nodes with the same type, and optionally identifiers.
private	_getMapKey(nodes: Node[]): string Generates a key based on the type of each of the passed nodes, returned as a base64-encoded sha1 hash.
private	_group(groups: Node[][][], fn: function): Node[][][] Accepts a multi-dimensional array of nodes and groups them based on the supplied function, which is expected to return a string.
private	_groupByMatchingIdentifiers(groups: Node[][][]): Node[][][] Iterates over the multi-dimensional array of nodes, and returns a new array grouping them based on matching identifiers.
private	_groupByMatchingLiterals(groups: Node[][][]): Node[][][] Iterates over the multi-dimensional array of nodes, and returns a new array grouping them based on matching literals.
private	_insert(nodes: Node[]) Generates a key based on the combined types of each of the supplied nodes.
private	_omitOverlappingInstances(nodeArrays: Node[][]) Removes overlapping instances from a group of node arrays.
private	_prune(nodeArrays: Node[][]) Removes the nodes from consideration in any additional matches.
private	_removeNode(node: Node) Removes all occurrences of a given node.
private	_walk(node: Node, fn: function) Walks a given node's AST, building up arrays of nodes that meet the inspector's threshold.
private	_walkSiblings(node: Node, fn: function) Walks sibling nodes under a parent, grouping their DFS traversals, and invoking the callback for those that wouldn't otherwise meet the threshold. Helpful for nodes like BlockStatements that hold a sequence. Note that this will generate overlapping instances, and so _omitOverlappingInstances helps cleanup the results.

Public Constructors

public constructor(filePaths: string[], opts: object) source

Creates a new Inspector, which extends EventEmitter. filePaths is expected to be an array of string paths. Also accepts an options object with any combination of the following: threshold, identifiers literals, and minInstances. Threshold indicates the minimum number of nodes to analyze. Identifiers indicates whether or not the nodes in a match should also have matching identifiers, and literals whether or not literal values should match. minInstances specifies the min number of instances for a match. An instance of Inspector emits the following events: start, match and end.

Params:

Name	Type	Attribute	Description
filePaths	string[]		The files on which to run the inspector
opts	object	optional	Options to set for the inspector

Public Methods

public run() source

Runs the inspector on the given file paths, as provided in the constructor. Emits a start event, followed by a series of match events for any detected similarities, and an end event on completion.

Private Methods

private _analyze() source

Traverses the keys at which the various nodes are stored. A key containing an array of more than a single entry indicates a potential match. The nodes are then grouped if identifier matching is enabled. A match results in the relevant nodes being removed from any future results. This pruning ensures that we only include the greatest common parent in a set of matches.

private _expand(nodeArrays: Node[][]) source

Expands each instance of a match to the largest common sequence of nodes with the same type, and optionally identifiers. Each array of nodes is modified in place.

Params:

Name	Type	Attribute	Description
nodeArrays	Node[][]

private _getMapKey(nodes: Node[]): string source

Generates a key based on the type of each of the passed nodes, returned as a base64-encoded sha1 hash.

Params:

Name	Type	Attribute	Description
nodes	Node[]		The nodes for which to generate the key

Return:

string

private _group(groups: Node[][][], fn: function): Node[][][] source

Accepts a multi-dimensional array of nodes and groups them based on the supplied function, which is expected to return a string.

Params:

Name	Type	Attribute	Description
groups	Node[][][]		The groups of nodes to further group
fn	function		Synchronous function for generating group ids

Return:

Node[][][]

private _groupByMatchingIdentifiers(groups: Node[][][]): Node[][][] source

Iterates over the multi-dimensional array of nodes, and returns a new array grouping them based on matching identifiers.

Params:

Name	Type	Attribute	Description
groups	Node[][][]

Return:

Node[][][]

private _groupByMatchingLiterals(groups: Node[][][]): Node[][][] source

Iterates over the multi-dimensional array of nodes, and returns a new array grouping them based on matching literals.

Params:

Name	Type	Attribute	Description
groups	Node[][][]

Return:

Node[][][]

private _insert(nodes: Node[]) source

Generates a key based on the combined types of each of the supplied nodes. Pushes the array to another array at the generated key in _map. Nodes are updated to keep a reference to all their occurrences in _map.

Params:

Name	Type	Attribute	Description
nodes	Node[]

private _omitOverlappingInstances(nodeArrays: Node[][]) source

Removes overlapping instances from a group of node arrays. That is, if one instance has nodes abcd, and another has bcde, then bcde will be removed from the array.

Params:

Name	Type	Attribute	Description
nodeArrays	Node[][]

private _prune(nodeArrays: Node[][]) source

Removes the nodes from consideration in any additional matches.

Params:

Name	Type	Attribute	Description
nodeArrays	Node[][]

private _removeNode(node: Node) source

Removes all occurrences of a given node.

Params:

Name	Type	Attribute	Description
node	Node		The node to remove

private _walk(node: Node, fn: function) source

Walks a given node's AST, building up arrays of nodes that meet the inspector's threshold. When found, the callback is invoked and passed the array of nodes.

Params:

Name	Type	Attribute	Description
node	Node		The node to traverse
fn	function		The callback to invoke

private _walkSiblings(node: Node, fn: function) source

Walks sibling nodes under a parent, grouping their DFS traversals, and invoking the callback for those that wouldn't otherwise meet the threshold. Helpful for nodes like BlockStatements that hold a sequence. Note that this will generate overlapping instances, and so _omitOverlappingInstances helps cleanup the results.

Params:

Name	Type	Attribute	Description
node	Node		The node to traverse
fn	function		The callback to invoke