From a69793a9898086cc8c209d05a7c7dcc5356fd245 Mon Sep 17 00:00:00 2001
From: koalaman <vidar@vidarholen.net>
Date: Sat, 1 Mar 2014 16:24:53 -0800
Subject: [PATCH] Created SC2063 (markdown)

---
 SC2063.md | 26 ++++++++++++++++++++++++++
 1 file changed, 26 insertions(+)
 create mode 100644 SC2063.md

diff --git a/SC2063.md b/SC2063.md
new file mode 100644
index 0000000..9f9dd7d
--- /dev/null
+++ b/SC2063.md
@@ -0,0 +1,26 @@
+# Grep uses regex, but this looks like a glob.
+
+### Problematic code:
+
+    grep '*foo*'
+
+### Correct code:
+
+    grep 'foo'   # or more explicitly,  grep '.*foo.*'
+
+### Rationale:
+
+In globs, `*` matches any number of any character.
+
+In regex, `*` matches any number of the preceding character. 
+
+`grep` uses regex, not globs, so this means that `grep '*foo'` is nonsensical because there's no preceding character for `*`. 
+
+If the intention was to match "any number of characters followed by foo", use `'.*foo'`. Also note that since grep matches substrings, this will match "fishfood". Use anchors to prevent this, e.g. `foo$`.
+
+This also means that `f*` will match "hello", because `f*` matches 0 (or more) "f"s and there are indeed 0 "f" characters in "hello". Again, use `grep 'f'` to find strings containing "f", or `grep '^f'` to find strings starting with "f".
+
+
+### Contraindications
+
+If you're aware of the differences between globs and regex, you can ignore this message.
\ No newline at end of file