Binaries: Windows, Linux
Compiling: red -r parse.red
>parse.exe --help
Parse tool 29-Mar-2021 - Process input using Parse commands -
Syntax: parse [options] <input> <rule>
Options:
<input> File to parse
<rule> Rule to match against, written in Red PARSE
dialect
-l, --lines Parse input line-by-line (default: as a
single string)
-e, --enum Display line numbers together with the text
(implies /lines)
-c, --collect Collect matches and print to the console
-w, --write Write the contents back (incompatible with
--collect)
-v, --verbose Verbose output
--help Display full help text and exit
-h Display synopsis and exit
--version Display program version and exit
Parse tool works in 2 modes: LINE mode and FILE mode
1. In FILE mode, it matches full file text against the RULE
and returns 0 if RULE fully covers the file, or 1 if not.
(useful to check if file follows a certain structure)
If COLLECT option is provided, collected tokens (if any)
are also printed to the console.
(useful to gather info from the file)
If WRITE option is provided, and RULE changes the input,
file contents is also written back to the file.
(useful to modify the file)
2. In LINE mode, it splits text into lines, then matches RULE against every line.
If COLLECT option is provided, it prints result collected from ALL lines.
(useful to gather info from a file that is line-oriented)
If WRITE option is provided, and RULE changes at least one line,
file contents is written back to the file and no output is made.
(useful to modify a file that is line-oriented)
Otherwise, it prints each line that matches the RULE.
(useful to filter the lines)
Examples:
Displaying all lines containing 2 consecutive vowels:
parse -e FILE "(cs: charset {AEIOUaeiou}) to 2 cs"
List what datatypes a file contains:
for %%i in (integer! float! tuple! string! file!) do (
parse FILE "to %%i"
if not errorlevel 1 echo Contains %%i
)
Collect all mixed-case words:
parse -c FILE "any [thru [any { } copy w word!] (w: to string! w) opt [if (not any [w == uppercase copy w w == lowercase copy w]) keep (transcode/one w)]]"
Extract columns 8-15 from the text
parse -c -l FILE "0 8 skip keep copy _ 0 8 skip"
Extract all line comments from the script:
parse -c -l parse.red "to {;} keep to end"
See test*.bat files. Example output:
>test1.bat
Displaying all lines containing 2 consecutive vowels:
1 # Parse tool: command line interface to Parse
5 Parse tool 24-Mar-2021 - Process input using Parse commands -
7 Syntax: parse [options] <input> <rule>
9 Options:
11 <rule> Rule to match against, written in Red PARSE
12 dialect
13 -l, --lines Parse input line-by-line (default: as a
16 (implies /lines)
18 -v, --verbose Verbose output
19 --version Display program version and exit
23 Parse tool works in 2 modes: LINE mode and FILE mode
25 1. In FILE mode, it matches full file text against the RULE
28 If COLLECT option is provided, collected tokens (if any)
31 2. In LINE mode, it splits text into lines, then matches RULE against every line.
33 If COLLECT option is provided, it prints result collected from ALL lines.
34 If not, it prints each line that matches the RULE.
38 Displaying all lines containing 2 consecutive vowels:
39 parse -e FILE "(cs: charset {AEIOUaeiou}) to 2 cs"
41 List what datatypes a file contains:
42 for %%i in (integer! float! tuple! string! file!) do (
44 if not errorlevel 1 echo Contains %%i
48 parse -c FILE "any [thru [any { } copy w word!] (w: to string! w) opt [if (not any [w == uppercase copy w w == lowercase copy w]) keep (transcode/one w)]]"
51 parse -c -l FILE "0 8 skip keep copy _ 0 8 skip"
54 parse -c -l parse.red "to {;} keep to end"
57 See `test*.bat` files. Example output:
>test2.bat
What datatypes a file contains?
Contains integer!
Contains float!
Contains string!
Contains file!
>test3.bat
Displaying all mixed-case words:
[
Parse
Parse
Parse
-Mar-2021
Process
Parse
File
Rule
Red
Parse
Display
Collect
Verbose
Display
Display
Parse
In
If
In
If
If
Displaying
AEIOUaeiou
List
Contains
Collect
Extract
Extract
See
Example
]
>test4.bat |more
Extracting columns 8-15 from the text
[
"tool: co"
""
""
"xe --hel"
"ol 24-Ma"
""
"parse [o"
""
""
" "
" "
" "
"lines "
" "
"enum "
" "
"collect "
"verbose "
"version "
"help "
""
""
"ol works"
""
"LE mode,"
"eturns 0"
""
(...)
]
>test5.bat
>pfind README.md /
(implies /lines)
parse -c FILE "any [thru [any { } copy w word!] (w: to string! w) opt [if (not any [w == uppercase copy w w == lowercase copy w]) keep (transcode/one w)]]"
>test6.bat
Extracting all line comments from the script
[
{;@@ ideally should not be needed, but binary mode allows parsing by datatype, which is cool}
{;@@ TODO: "-" for stdin? not possible until ports though}
{; /case "Use case-sensitive comparison"^-;@@ BUG: always applies - #4862}
{;-- expose line-number to rule (e.g. `keep (line-number)`)}
";-- use uppercased name"
{;^} keep to end"}
]