- Rewrite to ESM.
- Fix
http://
incorrectly returned as a link by matchStart.
- Add
matchAtStart
method to match full URLs at the start of the string. - Fixed paired symbols (
()
,{}
,""
, etc.) after punctuation. ---
option now affects parsing of emails (e.g.user@example.com---
)
- Fixed #98. Don't count
;
at the end of link (when followed with space).
- Proper fix for #54. Allow multiple
!
in links (but not at the end).
- Reverted #54 fix (allowed multiple
!
in links), and added collision sample.
- Allow unlimited
.
inside link params, #81. This should not be breaking, but bumped version for sure. - Allow
..&
in params, #87. - Allow multiple
!
in links, #54. - Deps bump.
- Rewrite build scripts.
- Improved quoted email detect (disable
"
at email start), #72. - Fix some google links (allow more consecutive
.
), #66.
- Allow
--
(and more dashes) in domain names, #63.
- Process
|
(asian vertical pipe 0xFF5C) as valid text separator.
- Allow dashes in local domains, #43.
- Restrict user:pass@... content - prohibit "()[]" chars in auth, #41.
---
no longer terminates link. Use option{ '---': true }
to return old behaviour..onCompile()
hook to modify base regexp constants.- Allow
foo'-bar
in path
- Consider
<
&>
as invalid in links. - Support links in lt/gt braces:
<user@domain.com>
,<http://example.com>
.
- Allow digits in local domains, #36.
- Restrict user/pass (prohibit [@/] chars) to avoid wrong domain fetch.
- More restrictions for protocol-transparent links. Don't allow single-level (local) domains, except '//localhost', #19.
- Security fix: due problem in
Any
class regexp from oldunicode-7.0.0
package (used inuc-micro
), hang happend with astral char patterns like😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡😡 .com
if fuzzy options used. New installs will use fixeduc-micro
automatically. Old installs need to be updated. #36. - Unicode rules updated to 8.+ version.
- Fix detect email after opening parenthesis:
(my@email.com)
, #32.
- Allow dash at the end of url, thanks to @Mumakil.
- Allow ".." in link paths.
- Added options to control fuzzy links recognition (
fuzzyLink: true
,fuzzyEmail: true
,fuzzyIP: false
). - Disabled IP-links without schema prefix by default.
- More strict default 2-characters tlds handle in fuzzy links, to avoid
false positives for
node.js
,io.js
and so on.
- Version bump to 1.0.0 for semver.
- Removed
Cf
class from whitespace & punctuation sets (#10). - API change. Exported regex names renamed to reflect changes. Update your
custom rules if needed:
src_ZPCcCf
->src_ZPCc
src_ZCcCf
->src_ZCc
- Fixed special chars handling (line breaks).
- Fixed demo permalink encode/decode.
- Allow
..
and...
inside of link paths (#9). Useful for github links with commit ranges. - Added
.pretest()
method for speed optimizations. - Autogenerate demo sample from fixtures.
- Maintenance release. Deps update.
- Fixed blockquoted links (some symbols exclusions), thanks to @MayhemYDG.
- Fixed demo permalinks, thanks to @MayhemYDG.
- Moved unicode data to external package.
- Demo permalink improvements.
- Docs update.
- First release.