Marmelator/pdf.js.mirror
1
0
Fork 0
mirror of https://github.com/mozilla/pdf.js.git synced 2026-06-27 02:25:47 +02:00
Code Issues Packages Projects Releases Wiki Activity

Escape punctuation signs in search query

Browse Source 
It fixes #20516.

The chars `*{}()[]\` are punctuation signs in regex, so they need to be escaped
when searching for them literally.
This commit is contained in:
calixteman 2025-12-22 17:41:50 +01:00
parent f7f963ef97
commit 923a778d26
No known key found for this signature in database
GPG Key ID: 0C5442631EE0691F
4 changed files with 27 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
test/pdfs/.gitignore vendored
View File

@ -765,3 +765,4 @@
!two_paragraphs.pdf !two_paragraphs.pdf
!paragraph_and_link.pdf !paragraph_and_link.pdf
!issue20225.pdf !issue20225.pdf
!issue20516.pdf

BIN
test/pdfs/issue20516.pdf Normal file
View File

Binary file not shown.

20
test/unit/pdf_find_controller_spec.js
View File

@ -1158,6 +1158,26 @@ describe("pdf_find_controller", function () {
}); });
}); });
it("performs a search with a group of punctuation signs to escape", async () => {
const { eventBus, pdfFindController } =
await initPdfFindController("issue20516.pdf");
await testSearch({
eventBus,
pdfFindController,
state: {
query: `("client")`,
},
matchesPerPage: [1],
selectedMatch: {
pageIndex: 0,
matchIndex: 0,
},
pageMatches: [[6]],
pageMatchesLength: [[10]],
});
});
describe("custom matcher", () => { describe("custom matcher", () => {
it("calls to the matcher with the right arguments", async () => { it("calls to the matcher with the right arguments", async () => {
const QUERY = "Foo bar"; const QUERY = "Foo bar";

9
web/pdf_find_controller.js
View File

@ -77,8 +77,8 @@ const DIACRITICS_EXCEPTION = new Set([
let DIACRITICS_EXCEPTION_STR; // Lazily initialized, see below. let DIACRITICS_EXCEPTION_STR; // Lazily initialized, see below.
const DIACRITICS_REG_EXP = /\p{M}+/gu; const DIACRITICS_REG_EXP = /\p{M}+/gu;
const SPECIAL_CHARS_REG_EXP = const SPECIAL_PUNCTUATION_CHARACTERS = /[.?*{}()[\]\\]/g;
/([*+^${}()|[\]\\])|(\p{P}+)|(\s+)|(\p{M})|(\p{L})/gu; const SPECIAL_CHARS_REG_EXP = /([+^$|])|(\p{P}+)|(\s+)|(\p{M})|(\p{L})/gu;
const NOT_DIACRITIC_FROM_END_REG_EXP = /([^\p{M}])\p{M}*$/u; const NOT_DIACRITIC_FROM_END_REG_EXP = /([^\p{M}])\p{M}*$/u;
const NOT_DIACRITIC_FROM_START_REG_EXP = /^\p{M}*([^\p{M}])/u; const NOT_DIACRITIC_FROM_START_REG_EXP = /^\p{M}*([^\p{M}])/u;
@ -739,7 +739,10 @@ class PDFFindController {
} }
if (p2) { if (p2) {
// Allow whitespaces around group of punctuation signs. // Allow whitespaces around group of punctuation signs.
return addExtraWhitespaces(p2, p2.replaceAll(/[.?]/g, "\\$&")); return addExtraWhitespaces(
p2,
p2.replaceAll(SPECIAL_PUNCTUATION_CHARACTERS, "\\$&")
);
} }
if (p3) { if (p3) {
// Replace spaces by \s+ to be sure to match any spaces. // Replace spaces by \s+ to be sure to match any spaces.