summaryrefslogtreecommitdiffstats
path: root/txr.1
diff options
context:
space:
mode:
Diffstat (limited to 'txr.1')
-rw-r--r--txr.110
1 files changed, 5 insertions, 5 deletions
diff --git a/txr.1 b/txr.1
index a1427118..b4b6e625 100644
--- a/txr.1
+++ b/txr.1
@@ -2959,8 +2959,8 @@ subexpression [*][^/] can match "**", since a * is not a /. If the next
character in the input is /, we missed a comment close. To fix the problem we
revise to this:
- ([^*]|[*][^*/])*
-
+ ([^*]|[*][^*/])*
+
(The interior of a C language comment is a any
mixture of zero or more non-asterisks, or digraphs consisting of an asterisk
followed by something other than a slash or another asterisk). Oops, now we
@@ -2970,7 +2970,7 @@ not simply match asterisk-non-asterisk digraphs, but rather sequences of one or
more asterisks followed by a non-asterisk:
([^*]|[*]*[^*/])*
-
+
This is still not right, because, for instance, it fails to match the interior
of a comment which is terminated by asterisks, including the simple test cases
where the comment interior is nothing but asterisks. We have no provision in
@@ -2980,11 +2980,11 @@ is to add on a subexpression which optionally matches a run of zero or more
interior asterisks before the comment close:
([^*]|[*]*[^*/])*[*]*
-
+
Thus our the semi-final regular expression is
[/][*]([^*]|[*]*[^*/])*[*]*[*][/]
-
+
(A C comment is an interior string enclosed in /* */, where this interior part
consists of a mixture of non-asterisk characters, as well as runs of asterisk
characters which are terminated by a character other than a slash, except for