Struct Detector 

pub struct Detector<'a> { /* private fields */ }

Searches a database for functions similar to checked ones.

Examples

let original = br#"class Search {
    static int find(int[] values, int key) {
        int low = 0, high = values.length - 1;
        while (low <= high) {
            int mid = (low + high) >>> 1;
            if (values[mid] < key) low = mid + 1;
            else if (values[mid] > key) high = mid - 1;
            else return mid;
        }
        return -(low + 1);
    }
}"#;
let package = PackageInfo {
    ecosystem: "maven".into(),
    name: "org.example:search".into(),
    version: "1.0".into(),
    ..PackageInfo::default()
};
let stored = extract_bytes("Search.java".as_ref(), Language::Java, original, &config).unwrap();
db.index(&package, &[stored]).unwrap();

// The same code with other names.
let copy = String::from_utf8_lossy(original)
    .replace("Search", "Finder").replace("values", "xs").replace("key", "target");
let checked = extract_bytes("Finder.java".as_ref(), Language::Java, copy.as_bytes(), &config).unwrap();
let detector = Detector::new(&db, Options { min_tokens: 20, ..Options::default() }).unwrap();
let reports = detector.detect(&[checked], |_, _| true).unwrap();
assert_eq!(reports[0].name, "find");
assert_eq!(reports[0].candidates[0].kind, MatchKind::Exact(pelicode_core::Level::N1));
assert_eq!(reports[0].candidates[0].occurrences[0].package, "org.example:search");

Implementations

impl<'a> Detector<'a>

pub fn new(db: &'a Database, options: Options) -> Result<Detector<'a>>

db must have been opened for features computed like the checked ones (see Database::config_from_meta).

Errors

Error when the database cannot be read.

Examples

use pelicode_core::Config;
use pelicode_db::Database;
use pelicode_detect::{Detector, Options};

let dir = tempfile::tempdir().unwrap();
let db = Database::open(dir.path().join("pelicode.db"), &Config::default()).unwrap();
let detector = Detector::new(&db, Options::default()).unwrap();

pub fn excluded_as_dependencies(&self) -> usize

Number of candidates left out so far because they were found only in Options::dependencies.

Examples

use pelicode_core::{Config, Language, extract_bytes};
use pelicode_db::{Database, PackageInfo};
use pelicode_detect::{Detector, Options, dependencies::Dependencies};

let dir = tempfile::tempdir().unwrap();
let config = Config { min_tokens: 0, ..Config::default() };
let mut db = Database::open(dir.path().join("pelicode.db"), &config).unwrap();
let source = b"def total(xs):\n    s = 0\n    for x in xs:\n        s += x * 2\n    return s\n";
let file = extract_bytes("calc.py".as_ref(), Language::Python, source, &config).unwrap();
let package = PackageInfo {
    ecosystem: "pypi".into(), name: "calc".into(), version: "1.0".into(), ..PackageInfo::default()
};
db.index(&package, &[file.clone()]).unwrap();

// The checked project depends on calc.
let mut dependencies = Dependencies::default();
dependencies.add_file("requirements.txt", "calc>=1\n");
let options = Options { min_tokens: 0, dependencies, ..Options::default() };
let detector = Detector::new(&db, options).unwrap();
assert!(detector.detect(&[file], |_, _| true).unwrap().is_empty());
assert_eq!(detector.excluded_as_dependencies(), 1);

pub fn detect( &self, files: &[FileFeatures], select: impl Fn(&FileFeatures, &FunctionFeatures) -> bool, ) -> Result<Vec<Report>>

Checks the functions of files accepted by select; only functions with at least one candidate are reported.

Errors

Error when the database cannot be read.

Examples

Only the functions accepted by select are checked, here those starting after line 1:

let original = br#"class Search {
    static int find(int[] values, int key) {
        int low = 0, high = values.length - 1;
        while (low <= high) {
            int mid = (low + high) >>> 1;
            if (values[mid] < key) low = mid + 1;
            else if (values[mid] > key) high = mid - 1;
            else return mid;
        }
        return -(low + 1);
    }
}"#;
let package = PackageInfo {
    ecosystem: "maven".into(),
    name: "org.example:search".into(),
    version: "1.0".into(),
    ..PackageInfo::default()
};
let stored = extract_bytes("Search.java".as_ref(), Language::Java, original, &config).unwrap();
db.index(&package, &[stored]).unwrap();

// The same code with other names.
let copy = String::from_utf8_lossy(original)
    .replace("Search", "Finder").replace("values", "xs").replace("key", "target");
let checked = extract_bytes("Finder.java".as_ref(), Language::Java, copy.as_bytes(), &config).unwrap();
let detector = Detector::new(&db, Options { min_tokens: 20, ..Options::default() }).unwrap();
let reports = detector.detect(&[checked], |_, f| f.start_line > 1).unwrap();
assert_eq!(reports.len(), 1);

pub fn candidates(&self, function: &FunctionFeatures) -> Result<Vec<Candidate>>

Candidates for one function, best first.

Unlike Detector::detect, the function is checked whatever its size.

Errors

Error when the database cannot be read.

Examples

let original = br#"class Search {
    static int find(int[] values, int key) {
        int low = 0, high = values.length - 1;
        while (low <= high) {
            int mid = (low + high) >>> 1;
            if (values[mid] < key) low = mid + 1;
            else if (values[mid] > key) high = mid - 1;
            else return mid;
        }
        return -(low + 1);
    }
}"#;
let package = PackageInfo {
    ecosystem: "maven".into(),
    name: "org.example:search".into(),
    version: "1.0".into(),
    ..PackageInfo::default()
};
let stored = extract_bytes("Search.java".as_ref(), Language::Java, original, &config).unwrap();
db.index(&package, &[stored]).unwrap();

// The same code with other names.
let copy = String::from_utf8_lossy(original)
    .replace("Search", "Finder").replace("values", "xs").replace("key", "target");
let checked = extract_bytes("Finder.java".as_ref(), Language::Java, copy.as_bytes(), &config).unwrap();
let detector = Detector::new(&db, Options::default()).unwrap();
let candidates = detector.candidates(&checked.functions[0]).unwrap();
assert_eq!(candidates[0].similarity, 1.0);
assert_eq!(candidates[0].function.name, "find");