From c9171ebe9a8c89c36e60326168a36e3c51247ca8 Mon Sep 17 00:00:00 2001
From: penguindark <57967770+penguindark@users.noreply.github.com>
Date: Mon, 21 Dec 2020 05:36:14 +0100
Subject: [PATCH] regex: fixes and optimizations (#7431)

---
 vlib/regex/README.md    | 1 +
 vlib/regex/regex.v      | 9 +++++++--
 vlib/regex/regex_test.v | 9 +++++++++
 3 files changed, 17 insertions(+), 2 deletions(-)

diff --git a/vlib/regex/README.md b/vlib/regex/README.md
index eba818ffa3..3645f7589a 100644
--- a/vlib/regex/README.md
+++ b/vlib/regex/README.md
@@ -475,6 +475,7 @@ pub fn new() RE
 #### **Custom initialization**
 For some particular need it is possible initialize a fully customized regex:
 ```v ignore
+pattern = r"ab(.*)(ac)"
 // init custom regex
 mut re := regex.RE{}
 re.prog = []Token    {len: pattern.len + 1} // max program length, can not be longer then the pattern
diff --git a/vlib/regex/regex.v b/vlib/regex/regex.v
index 210319dde2..d0ca5742f5 100644
--- a/vlib/regex/regex.v
+++ b/vlib/regex/regex.v
@@ -1825,9 +1825,9 @@ pub fn (mut re RE) match_base(in_txt byteptr, in_txt_len int ) (int,int) {
 			}
 
 			// manage here dot char
-			//println("Here we are, with stop: state buffer: [${state_list.len}]")
 
 			if state_list.len > 0 {
+				//println("Here we are, with stop: state buffer: [${state_list.len}]")
 				state = state_list.pop()
 
 				state.match_flag = true
@@ -1898,6 +1898,10 @@ pub fn (mut re RE) match_base(in_txt byteptr, in_txt_len int ) (int,int) {
 
 						re.groups[g_index+1] = state.i
 
+						if g_index > 0 && re.groups[g_index] <= re.groups[g_index-1] {
+							re.groups[g_index] = re.groups[g_index-1] 
+						}
+
 						if re.groups[g_index+1] >= in_txt_len {
 							//println("clamp group!")
 							re.groups[g_index+1] = in_txt_len-1
@@ -1946,6 +1950,7 @@ pub fn (mut re RE) match_base(in_txt byteptr, in_txt_len int ) (int,int) {
 				// if we are done with max go on dot char are dedicated case!!
 				if	re.prog[state.pc].rep >= re.prog[state.pc].rep_max 
 				{
+					state_list.pop()
 					m_state = .ist_next
 					continue
 				}
@@ -1986,7 +1991,7 @@ pub fn (mut re RE) match_base(in_txt byteptr, in_txt_len int ) (int,int) {
 					}
 
 				}
-
+				
 				// check if we must continue or pass to the next IST
 				if next_check_flag == true {
 					//println("save the state!!")
diff --git a/vlib/regex/regex_test.v b/vlib/regex/regex_test.v
index 21a967db05..b421866f96 100644
--- a/vlib/regex/regex_test.v
+++ b/vlib/regex/regex_test.v
@@ -138,6 +138,9 @@ match_test_suite = [
     TestItem{"*1*1(", r"(?:^|[*()])(\d+)([*])(\d+)(?:$|[*()])",0,5},
     TestItem{" 1*1(", r"(?:^|[*()])(\d+)([*])(\d+)(?:$|[*()])",-1,0},
     TestItem{"1*1 ", r"(?:^|[*()])(\d+)([*])(\d+)(?:$|[*()])",-1,0},
+
+    // particular groups
+    TestItem{"ababababac", r"ab(.*)(ac)",0,10},
 ]
 )
 
@@ -229,6 +232,12 @@ cgroups_test_suite = [
 		[0, 3, 4, 5, 5, 7],
 		map[string]int{}
 	},
+	TestItemCGroup{
+		"ababababac",
+		r"ab(.*)(ac)",0,10,
+		[2, 8, 8, 10],
+		map[string]int{}
+	},
 ]
 )