3f / regxwild Goto Github PK

• @L ... enabled ...
• @L ... enabled ... @L ... disabled ...

for example: @L12*57@L8+9

12
57819

v1 - Same char to escape

• 1024*2 -> 10247412
• 1024**2 -> 1024*2
• 1024***2 -> 1024*0xD2
• one|two -> one or two
• one||two -> one|two
• one|||two -> one| or two
• one||||two -> one||two
• one|||||two -> one|| or two
• one|*||two -> one or |two
• one|*||||two -> one or ||two

... for each available meta-symbol

upd: no, ESS version provides #, that means 1 any symbol, thus we cannot use this because of: track-num:*###7400 and similar.

v2 - markers via flags

We'll add @ + special flag ([a-z0-9] only) + @@ to escape this flag

Then we can use unified combinations of this, for example:

• @ + d for disable all meta-symbols before new @(?) -> @d ... @
  • 1024*2 -> 10247412
  • 1024@d*@2 -> 1024*2 or @d1024*2@ -> 1024*2
  • @d1024*@*2 -> 1024*0xD2
  • @@d1024*@*2 -> @@d1024*@0xD2 -> note: @* ok (@ + [a-z0-9])
  • one|two -> one or two
  • @done|two@ -> one|two
  • @done|@|two -> one| or two
  • @done||two@ -> one||two
  • @done||@|two -> one|| or two
  • one|@d|two@ -> one or |two
  • one|@d||two@ -> one or ||two

well, this is more powerful way, but probably more hard for view o_o and:

• v2: one|@d||two@ <- ?@done|@@d||two@@@
• v1: one|@d||two@ <- one||@@d||||two@@

v3 - markers via single tilde '~'

v2 can also be as simple ~ ... ~ + to escape ~~:

• one|@d||two@ <- ~one|@d||two@~
• one|~d||two~ <- ~one|~~d||two~~~

To use additional syntax, we can add special markers & flags

The entry point like:

• Single: @x ...
• Pair: @x ... @

Where x the any supported combination, like single symbol [0-9a-z] or complex expressions like {1,7}.

This related for #1 (v2 - markers via flags), but the main idea to extend syntax for support additional features like the following below.

Sub-Issues

• #4 - Capture data for all quantifiers
• #5 - Control of newline symbols
• #1 - Disable Meta-Symbols

Special symbol-logic like for current '>'

MS_ANYSP = _T('>'), // as [^/]* //TODO: >\>/ i.e. '>' + {symbol}

• @B> ...

etc.

Add any variant for using of whitespace & non-whitespace character/s

v1 - single meta-symbol + std. quantifier

• % - whitespace + quantifier *?+#
• & - non-whitespace + quantifier *?+#

for example:

• 1%+5 -> 125, 12345, ...
• 1%##5 -> 1235, 1775, ...
• 1%?5 -> 15, 125, 145, ...
• 1%*5 -> 123467895, 15, ...

configurable % + & like:

search(const TCHAR* data, const TCHAR* filter, const TCHAR* whsp, const TCHAR* nonwhsp)

_T("qwerty012345...")
_T("	 ")

or simply:

search(const TCHAR* data, const TCHAR* filter, const TCHAR* whsp)

_T("	 ") - all except this is a non-whitespace character

but the first variant is more flexible for using special set of characters for both cases of % + &

Your project is great.

The most lightweight way (+Speed) implement flag to enable capturing data by default for all quantifiers *?+#

• @c ... enabled ...
• @c ... enabled ... @c ... disabled ...

example:

• @c12*5 a?d 34+7
  • 12005 and 3417
    • -> $1 - 00
    • -> $2 - n
    • -> $3 - 1

Alternative

Speed - is most important, because the regex engine much more powerful anyway.

v2:

• @c ... @
  • 12@c*@7 - > 12457 accessor: $n

v3 (overhead):

• @c flag - switch to enable () for capturing
  • @c12(*)7 - > 12457 accessor: $1 - 45
  • escape \(\) or (())