implemented the ^~ pattern match construct: the rest of the pattern will be an Advanced Regular ExpressionHEADmaster-fmsbw-cimaster

* Allows searching by regular expressions.
  We will never support all ARE constructs in TECO patterns, so this is useful to have available.
* Can only be typed upcaret.
  This leaves ^E~q available as an escape-regexp string building construct.
* Once we replace the pattern2regexp converter with a custom terex lexer,
  we might want to restrict ^~ to the beginning of the pattern.
  Currently, however it can be anywhere, so you can mix TECO patterns with regular expressions.

3 files changed, 35 insertions, 9 deletions

diff --git a/doc/sciteco.7.template b/doc/sciteco.7.template
index 87e8574..41f3abf 100644
--- a/doc/sciteco.7.template
+++ b/doc/sciteco.7.template
@@ -2050,10 +2050,17 @@ the space and horizontal tab characters.
 Matches one in a list of patterns.
 Any pattern match construct may be used.
 The pattern alternatives must be separated by commas.
+.TP
+.SCITECO_TOPIC ^~
+.BI ^~
+The rest of the pattern is an Advanced Regular Expression (ARE)
+as in Tcl - see
+.BR re_syntax (3).
+This can only be typed with an upcaret.
 .LP
 All non-pattern-match-characters match themselves.
-Note however that currently, all pattern matching is performed
-.BR case-insensitive .
+Pattern matching is performed case-insensitive by default,
+but this can be controlled with the \fB^X\fP command.
 .
 .
 .SH FILE NAMES AND DIRECTORIES
diff --git a/src/search.c b/src/search.c
index 0df483f..ce4a338 100644
--- a/src/search.c
+++ b/src/search.c
@@ -425,14 +425,29 @@ teco_pattern2regexp(teco_string_t *pattern, teco_machine_qregspec_t *qreg_machin
 	g_auto(teco_string_t) re = {NULL, 0};
 
 	do {
-		/*
-		 * Previous character was caret.
-		 * Make sure it is handled like a control character.
-		 * This is necessary even though we have string building activated,
-		 * to support constructs like ^Q^Q (typed with carets) in order to
-		 * quote pattern matching characters.
-		 */
 		if (state == TECO_SEARCH_STATE_CTL) {
+			if (*pattern->data == '~') {
+				/* rest of pattern is a regular expression */
+				teco_string_append(&re, pattern->data+1, pattern->len-1);
+				/*
+				 * FIXME: In terex, it actually could contain null bytes.
+				 */
+				if (teco_string_contains(re, '\0')) {
+					g_set_error_literal(error, TECO_ERROR, TECO_ERROR_FAILED,
+					                    "Regular expression must not contain null-byte"
+					                    " - use \\0 instead");
+					return NULL;
+				}
+				return g_steal_pointer(&re.data) ? : g_strdup("");
+			}
+
+			/*
+			 * Previous character was caret.
+			 * Make sure it is handled like a control character.
+			 * This is necessary even though we have string building activated,
+			 * to support constructs like ^Q^Q (typed with carets) in order to
+			 * quote pattern matching characters.
+			 */
 			*pattern->data = TECO_CTL_KEY(g_ascii_toupper(*pattern->data));
 			state = TECO_SEARCH_STATE_START;
 		}
diff --git a/tests/testsuite.at b/tests/testsuite.at
index 0f7e32b..95c1336 100644
--- a/tests/testsuite.at
+++ b/tests/testsuite.at
@@ -293,6 +293,10 @@ TE_CHECK([[@I/XYZ/J @EB/foo/ @I/XZY/J @:FN/Z/0/"F(0/0)' Q*-2"N(0/0)'
            @:FN///"F(0/0)' Q*-1"N(0/0)']], 0, ignore, ignore)
 AT_CLEANUP
 
+AT_SETUP([Searches by regexp])
+TE_CHECK([[@I/ABC/J :@S/^~[A-Z]+/"F(0/0)']], 0, ignore, ignore)
+AT_CLEANUP
+
 AT_SETUP([Search and insertion ranges])
 # When deleting characters, the result of ^S/^Y must not change.
 TE_CHECK([[@I/XXYYZZ/^SC ."N(0/0)' C @S/YY/ HK ^YU1U0 Q0-2"N(0/0)' Q1-4"N(0/0)']], 0, ignore, ignore)