Why do backslashes before braces make my regex match both and

Question

In notepad++ I tried to make a regex to match <tr> tags. At first, I thought that sharp braces had to be escaped, so I tried \<tr\>. However, this matched not just the beginning tags as I would have expected, but all of the <tr>s (both <tr> and </tr>). Why is this?

These are slashes, not backslashes... Also, which language? Some languages treat the first character as a separator in regexes, and as such, the "pattern" part is "``" which would be the part to replace the matched pattern... — ppeterka, Sep 06 '13 at 13:39
@ppeterka66 it was a typo, thanks. I'm not sure what regex processing Notepad++ uses. I'm looking into it. — 1252748, Sep 06 '13 at 13:40

score 7 · Accepted Answer · edited Aug 28 '18 at 01:39

7

\< and \> mean "word boundary" in certain regex implementations, including Notepad++. From the Notepad++ documentation:

\< This matches the start of a word using Scintilla's definitions of words.

\> This matches the end of a word using Scintilla's definition of words.

A word boundary is a zero-width match between a non-word character and a word character. Scintilla's definition of "word character" is:

A word is defined to be a character string beginning and/or ending with the characters A-Z a-z 0-9 and _. Scintilla extends this definition by user setting. The word must also be preceded and/or followed by any character outside those mentioned.

Thus, your regex \<tr\> actually matches the word boundary between < (or /) and t, followed by tr, followed by the word boundary between r and >.

edited Aug 28 '18 at 01:39

Pang

9,564
146
81
122

answered Sep 06 '13 at 13:46

nneonneo

171,345
36
312
383

1

You sure `\>` has the same definition about word boundary? – nhahtdh Sep 06 '13 at 13:51
1

To be precise, `\<` means non-word to word transition, and `\>` means word to non-word transition. – nneonneo Sep 06 '13 at 13:51
I mean the set of characters that is considered a word. – nhahtdh Sep 06 '13 at 13:52
@nhahtdh: Yes, I just amended the question. (Note that the "user setting" bit isn't really documented, but I guess it probably means locale-specific word characters can be added by the controlling application). – nneonneo Sep 06 '13 at 13:54
Nice. This is more a complete answer. – nhahtdh Sep 06 '13 at 13:55
+1, but your last paragraph is a bit difficult to read, at first it looked wrong. – Bernhard Barker Sep 06 '13 at 13:59

score 3 · Answer 2 · answered Sep 06 '13 at 13:41

3

escaping needs backslashes, e.g. \{
you don't need (shouldn't) escape < or >
if you escaped them, it means word boundary, not only <tr>, </tr> but also ,tr, will be matched

answered Sep 06 '13 at 13:41

Kent

189,393
32
233
301

What's a word boundary. Thank you. – 1252748 Sep 06 '13 at 13:43
@thomas http://stackoverflow.com/questions/1324676/what-is-a-word-boundary-in-regexes – Kent Sep 06 '13 at 13:44

Why do backslashes before braces make my regex match both and

2 Answers2