mirror of https://github.com/CrimsonTome/tldr.git
32 lines
966 B
Markdown
32 lines
966 B
Markdown
# zgrep
|
||
|
||
> Grep text patterns from files within compressed file (equivalent to grep -Z).
|
||
|
||
- Grep a pattern in a compressed file (case-sensitive):
|
||
|
||
`zgrep {{pattern}} {{path/to/compressed/file}}`
|
||
|
||
- Grep a pattern in a compressed file (case-insensitive):
|
||
|
||
`zgrep -i {{pattern}} {{path/to/compressed/file}}`
|
||
|
||
- Output count of lines containing matched pattern in a compressed file:
|
||
|
||
`zgrep -c {{pattern}} {{path/to/compressed/file}}`
|
||
|
||
- Display the lines which don’t have the pattern present (Invert the search function):
|
||
|
||
`zgrep -v {{pattern}} {{path/to/compressed/file}}`
|
||
|
||
- Grep a compressed file for multiple patterns:
|
||
|
||
`zgrep -e "{{pattern_1}}" -e "{{pattern_2}}" {{path/to/compressed/file}}`
|
||
|
||
- Use extended regular expressions (supporting `?`, `+`, `{}`, `()` and `|`):
|
||
|
||
`zgrep -E {{regular_expression}} {{path/to/file}}`
|
||
|
||
- Print 3 lines of [C]ontext around, [B]efore, or [A]fter each match:
|
||
|
||
`zgrep -{{C|B|A}} {{3}} {{pattern}} {{path/to/compressed/file}}`
|