Source   Edit  

data wrangling

bookmark

TLDR

• for parsing configuration files, see packaging
• constructing regular expressions are expensive, save to a var if you can reuse it
• re
  • follows perl 5 (see pcre spec link)
  • is an impure module require C's PCRE to be available at runtime
• pegs meant to replace re
• scanf can be extended with arbitrary procs for data wrangling
• parseutils
  • provides many declarative wrappers utilizing while loops
  • prefer this over manually looping through haystacks looking for needles
  • sometimes faster than using re module

links

• intros
  • pcre specification
  • EBNF wikipedia
• examples
  • nimgrep source code
• high impact
  • peg matching
  • regex pcre wrapper
  • string scans
  • fusion matching
  • parse utils
• nitche
  • perl compatible regex

TODOs

• re
  • study the expression: no clue what this means

re

• everything works as expected, only things i found interesting are listed
• wrapper around pcre pkg
• supports up to 20 or 40 capturing subpatterns, not sure which is correct
• start param can change where scan starts, but output is always relative to ^input
• findAll and split can be iterated
• =~ is particularly useful

re metacharacters

• the usual suspects
• ddd octal code ddd or backreference
• x{hh} character with hex code hh

re exceptions

• RegexError re syntax invalid

re types

• Regex
• RegexFlag enum
  • reIgnoreCase
  • reMultiLine ^ $ match new lines
  • reDotAll . matches anything
  • reExtended ignore whitespace and # and comments
  • reStudy study the expression?

re consts

• MaxReBufSize high(cint)
• MaxSubPatterns 20

re procs

• find
• match