hare: regex: allow ^ at start of every whole-expression alternation v3 APPLIED

Max Schillinger: 1
 regex: allow ^ at start of every whole-expression alternation

 2 files changed, 21 insertions(+), 3 deletions(-)

[PATCH hare v3] regex: allow ^ at start of every whole-expression alternation Export this patch

Signed-off-by: Max Schillinger <max@mxsr.de>
---
Implements Vlad's suggestion on v2.

 regex/+test.ha |  7 +++++++
 regex/regex.ha | 17 ++++++++++++++---
 2 files changed, 21 insertions(+), 3 deletions(-)

diff --git a/regex/+test.ha b/regex/+test.ha
index f5db579b..279eeea7 100644
--- a/regex/+test.ha
+++ b/regex/+test.ha
@@ -585,6 +585,9 @@ fn run_rawreplace_case(
 		(`ab$|cd`, "ab", matchres::MATCH, 0, 2),
 		(`ab$|cd`, "abc", matchres::NOMATCH, 0, 0),
 		(`ab|cd$`, "cde", matchres::NOMATCH, 0, 0),
+		(`ab|^cd`, "bcd", matchres::NOMATCH, 0, 0),
+		(`ab|^cd`, "cde", matchres::MATCH, 0, 2),
+		(`ab\|^cd`, "cde", matchres::ERROR, 0, 0),
 		// multiple alternation
 		(`a|b|c|d|e`, "e", matchres::MATCH, 0, -1),
 		(`a|b|c|d|e`, "xe", matchres::MATCH, 1, -1),
@@ -593,6 +596,10 @@ fn run_rawreplace_case(
 		(`a|b$|c$|d$|e`, "ax", matchres::MATCH, 0, 1),
 		(`a|b$|c$|d$|e`, "cx", matchres::NOMATCH, 0, 0),
 		(`a|b$|c$|d$|e`, "ex", matchres::MATCH, 0, 1),
+		(`a|^b|^c|^d|e`, "cd", matchres::MATCH, 0, 1),
+		(`a|^b|^c|^d|e`, "xa", matchres::MATCH, 1, 2),
+		(`a|^b|^c|^d|e`, "xc", matchres::NOMATCH, 0, 0),
+		(`a|^b|^c|^d|e`, "xe", matchres::MATCH, 1, 2),
 		// TODO: nested capture groups
 		(`((a))`, "abc", matchres::ERROR, 0, -1),
 		// (`((a))`, "abc", matchres::MATCH, 0, -1),
diff --git a/regex/regex.ha b/regex/regex.ha
index aabc5cbc..540c11fb 100644
--- a/regex/regex.ha
+++ b/regex/regex.ha
@@ -227,6 +227,7 @@ export fn compile(expr: str) (regex | error) = {
 	let skip_charclass_rest = false;
 	let bracket_idx = -1;
 	let is_charset_positive = true;
+	let was_prev_rune_pipe = false;
 	let n_reps = 0z;
 	let n_groupstarts = 0;
 
@@ -267,8 +268,11 @@ export fn compile(expr: str) (regex | error) = {
 				r_idx += 1;
 			};
 		case '^' =>
-			if (r_idx != 0) {
-				return `Anchor '^' not at start`: error;
+			if (n_groupstarts > 0) {
+				return `Anchor '^' in capture groups is unsupported`: error;
+			};
+			if (!(r_idx == 0 || was_prev_rune_pipe)) {
+				return `Anchor '^' not at start of whole pattern or alternation`: error;
 			};
 		case '$' =>
 			if (n_groupstarts > 0) {
@@ -319,7 +323,13 @@ export fn compile(expr: str) (regex | error) = {
 			append(jump_idxs, len(insts) - 1);
 			// add skip if it's a whole-expression alternation
 			if (origin == 0) {
-				append(insts, inst_skip);
+				const peek1 = strings::next(&iter);
+				if (peek1 is rune) {
+					if (peek1 as rune != '^') {
+						append(insts, inst_skip);
+					};
+					strings::prev(&iter);
+				};
 			};
 		case '{' =>
 			let origin = len(insts) - 1;
@@ -408,6 +418,7 @@ export fn compile(expr: str) (regex | error) = {
 		case =>
 			append(insts, r: inst_lit);
 		};
+		was_prev_rune_pipe = (r == '|');
 		r_idx += 1;
 	};
 
-- 
2.45.2

builds.sr.ht <builds@sr.ht>

7 months ago

hare/patches: FAILED in 1m20s

[regex: allow ^ at start of every whole-expression alternation][0] v3 from [Max Schillinger][1]

[0]: https://lists.sr.ht/~sircmpwn/hare-dev/patches/54013
[1]: mailto:max@mxsr.de

✗ #1280954 FAILED hare/patches/freebsd.yml https://builds.sr.ht/~sircmpwn/job/1280954
✗ #1280955 FAILED hare/patches/netbsd.yml  https://builds.sr.ht/~sircmpwn/job/1280955
✗ #1280953 FAILED hare/patches/alpine.yml  https://builds.sr.ht/~sircmpwn/job/1280953
✗ #1280956 FAILED hare/patches/openbsd.yml https://builds.sr.ht/~sircmpwn/job/1280956

Vlad-Stefan Harbuz <vlad@vladh.net>

7 months ago

Looks great, thank you! :) Applied both patches.

to git@git.sr.ht:~sircmpwn/hare
  2e40a947..dc65431e  master -> master