CalvinBackup/anonymous_github
1
0
Fork 0
mirror of https://github.com/tdurieux/anonymous_github.git synced 2026-05-15 14:38:03 +02:00
Code Issues Packages Projects Releases Wiki Activity

feat: per-term anonymization output via term=>replacement

Browse Source 
A term entered as "Anonymous=>ABC" now scrubs "Anonymous" to "ABC"
instead of "XXXX-N". Lets users keep anonymized identifiers valid in
source code (no hyphen) and align tokens between paper text and repo.
Indexing for default-mask terms is unchanged: a list of
"Alpha=>AAA", "Beta" still produces XXXX-2 for Beta.

Fixes #285.
This commit is contained in:
tdurieux
2026-05-04 09:31:31 +02:00
parent db2ac5307d
commit 4bc83db416
4 changed files with 92 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files
test/anonymize-utils.test.js
+42 -9
View File
@@ -5,6 +5,7 @@ require("ts-node/register/transpile-only");
const {
withWordBoundaries,
termVariants,
parseTermSpec,
} = require("../src/core/term-matching");
/**
@@ -97,11 +98,16 @@ class ContentAnonimizer {
replaceTerms(content) {
const terms = this.opt.terms || [];
for (let i = 0; i < terms.length; i++) {
let term = terms[i];
if (term.trim() == "") {
const spec = terms[i];
if (spec.trim() == "") {
continue;
}
const mask = ANONYMIZATION_MASK + "-" + (i + 1);
const parsed = parseTermSpec(spec);
let term = parsed.term;
const mask =
parsed.replacement !== null
? parsed.replacement
: ANONYMIZATION_MASK + "-" + (i + 1);
try {
new RegExp(term, "gi");
} catch {
@@ -140,19 +146,22 @@ class ContentAnonimizer {
function anonymizePath(path, terms) {
for (let i = 0; i < terms.length; i++) {
let term = terms[i];
if (term.trim() == "") {
const spec = terms[i];
if (spec.trim() == "") {
continue;
}
const parsed = parseTermSpec(spec);
let term = parsed.term;
const mask =
parsed.replacement !== null
? parsed.replacement
: ANONYMIZATION_MASK + "-" + (i + 1);
try {
new RegExp(term, "gi");
} catch {
term = term.replace(/[-[\]{}()*+?.,\\^$|#]/g, "\\$&");
}
path = path.replace(
new RegExp(term, "gi"),
ANONYMIZATION_MASK + "-" + (i + 1)
);
path = path.replace(new RegExp(term, "gi"), mask);
}
return path;
}
@@ -236,6 +245,30 @@ describe("ContentAnonimizer", function () {
expect(result).to.include("XXXX-1");
});
// #285 — `term=>replacement` uses the user-supplied replacement
// instead of XXXX-N, so anonymized identifiers can stay valid in code.
it("uses a custom replacement when the term is 'term=>replacement'", function () {
const a = new ContentAnonimizer({ terms: ["Anonymous=>ABC"] });
const result = a.anonymize("class Anonymous extends Base {}");
expect(result).to.equal("class ABC extends Base {}");
});
it("supports custom and default-mask terms together with stable indices", function () {
const a = new ContentAnonimizer({
terms: ["Alpha=>AAA", "Beta"],
});
const result = a.anonymize("Alpha and Beta");
// Beta uses XXXX-2 (its 1-based index in the list), even though
// Alpha had a custom replacement.
expect(result).to.equal("AAA and XXXX-2");
});
it("falls back to the default mask when the entry has no replacement", function () {
const a = new ContentAnonimizer({ terms: ["Foo=>"] });
const result = a.anonymize("Foo bar");
expect(result).to.equal(" bar");
});
// #280 — accented terms should match both the accented and unaccented
// variants so "Davó" scrubs "Davo" (and vice versa).
it("matches accented and unaccented variants of the same term", function () {