hyung-hwan/hawk
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

enhanced the tokenizer to recognize the octal notation as much as possible in a regular expression
All checks were successful
continuous-integration/drone/push Build is passing
Details

Browse Source
This commit is contained in:
hyung-hwan
2025-10-04 02:36:41 +09:00
parent 2bc122f23a
commit 1692e9ac2b
3 changed files with 23 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
README.md
View File

@ -15,7 +15,7 @@ The library is stable, portable, and designed for projects that need a scripting
- [Pragmas](#pragmas) - [Pragmas](#pragmas)
- [@pragma entry](#pragma-entry) - [@pragma entry](#pragma-entry)
- [@pragma implicit](#pragma-implicit) - [@pragma implicit](#pragma-implicit)
- [@pragma sriprecspc](#pragma-sriprecspc) - [@pragma striprecspc](#pragma-striprecspc)
- [@include and @include\_once](#include-and-include_once) - [@include and @include\_once](#include-and-include_once)
- [Comments](#comments) - [Comments](#comments)
- [Reserved Words](#reserved-words) - [Reserved Words](#reserved-words)
@ -364,7 +364,7 @@ This feature can be beneficial for catching potential variable misspellings or u
If you don't want to enforce variable declarations, you can simply omit the `@pragma implicit off` directive or specify `@pragma implicit on`, and Hawk will behave like traditional awk, allowing implicit variable declarations. If you don't want to enforce variable declarations, you can simply omit the `@pragma implicit off` directive or specify `@pragma implicit on`, and Hawk will behave like traditional awk, allowing implicit variable declarations.
### @pragma sriprecspc ### @pragma striprecspc
The `@pragma striprecspc` directive in Hawk controls how the interpreter handles leading and trailing blank fields in input records when using a regular expression as the field separator (FS). The `@pragma striprecspc` directive in Hawk controls how the interpreter handles leading and trailing blank fields in input records when using a regular expression as the field separator (FS).

22
lib/parse.c
View File

@ -6559,7 +6559,17 @@ static int get_string (
} }
else else
{ {
ADD_TOKEN_UINT32(hawk, tok, c_acc); if (digit_count == 1 && end_char == HAWK_T('/'))
{
/* inside a regular expression, it's likely a backreference */
hawk_ooch_t oc = c_acc + HAWK_T('0');
ADD_TOKEN_CHAR(hawk, tok, esc_char);
ADD_TOKEN_CHAR(hawk, tok, oc);
}
else
{
ADD_TOKEN_UINT32(hawk, tok, c_acc);
}
escaped = 0; escaped = 0;
} }
} }
@ -6600,7 +6610,6 @@ static int get_string (
} }
else else
{ {
if (digit_count == 0) if (digit_count == 0)
{ {
hawk_ooch_t ec; hawk_ooch_t ec;
@ -6619,7 +6628,7 @@ static int get_string (
else ADD_TOKEN_UINT32(hawk, tok, c_acc); else ADD_TOKEN_UINT32(hawk, tok, c_acc);
escaped = 0; escaped = 0;
/* carray on to handle the current character */ /* carry on to handle the current character */
} }
} }
else if (escaped == 99) else if (escaped == 99)
@ -6671,10 +6680,11 @@ static int get_string (
else if (c == HAWK_T('b')) c = HAWK_T('\b'); else if (c == HAWK_T('b')) c = HAWK_T('\b');
else if (c == HAWK_T('v')) c = HAWK_T('\v'); else if (c == HAWK_T('v')) c = HAWK_T('\v');
else if (c == HAWK_T('a')) c = HAWK_T('\a'); else if (c == HAWK_T('a')) c = HAWK_T('\a');
else if (c >= HAWK_T('0') && c <= HAWK_T('7') && end_char != HAWK_T('/')) else if (c >= HAWK_T('0') && c <= HAWK_T('7'))
{ {
/* i don't support the octal notation for a regular expression. /* treat it as an octal notation first and
* it conflicts with the backreference notation between \1 and \7 inclusive. */ * check if it's a backreference between \1 and \7 inclusive
* in the `if (escaped == 3)` block. */
escaped = 3; escaped = 3;
digit_count = 1; digit_count = 1;
c_acc = c - HAWK_T('0'); c_acc = c - HAWK_T('0');

9
t/h-002.hawk
View File

@ -557,10 +557,11 @@ function main()
{ {
## back reference in a regular expression ## back reference in a regular expression - use the character class
tap_ensure (("a2b" ~ /(a)\12b/), 0, @SCRIPTNAME, @SCRIPTLINE); ## notation to avoid escaping - \1[2]
tap_ensure (("aa2b" ~ /(a)\12b/), 1, @SCRIPTNAME, @SCRIPTLINE); tap_ensure (("a2b" ~ /(a)\1[2]b/), 0, @SCRIPTNAME, @SCRIPTLINE);
tap_ensure (("aaa2b" ~ /(a)\12b/), 1, @SCRIPTNAME, @SCRIPTLINE); tap_ensure (("aa2b" ~ /(a)\1[2]b/), 1, @SCRIPTNAME, @SCRIPTLINE);
tap_ensure (("aaa2b" ~ /(a)\1[2]b/), 1, @SCRIPTNAME, @SCRIPTLINE);
} }