Extract text with awk or sed

Question

I would like to extract and remove some texts from the below

[root@test]# du -k./[a-zA-Z0-9] --max-depth=1 | sort -hr

Before

    7789696 ./b/bklee
    946792  ./a
    796588  ./b/bluecyn
    477860  ./b/bborikun
    473652  ./b/bluechiper
    220780  ./a/ara316
    144244  ./a/aceload
    131088  ./b/belivart
    79108   ./a/athlon85
    78644   ./b/beschur512
    66264   ./b/bogdanov
    52460   ./A

After

    796588  bluecyn
    477860  bborikun
    473652  bluechiper
    220780  ara316
    144244  aceload
    131088  belivart
    79108   athlon85
    78644   beschur512
    66264   bogdanov

what I want is to remove the repetitive pattern, "./a/" and lines that only prints out like "/a" I am trying to figure it out but since I am a beginner at AWK and SED, I need some help.

Thanks!

Answer 1

du -k ./[a-zA-Z0-9] --max-depth=1 | sort -hr | sed -e 's,\./[a-z]/,,; /\.\/[Aa-Zz]/d'

7789696 bklee
796588  bluecyn
477860  bborikun
473652  bluechiper
220780  ara316
144244  aceload
131088  belivart
79108   athlon85
78644   beschur512
66264   bogdanov

Answer 2

Like this, using GNU sed :

sed -E '/\s\.\/\w$/d; s!\./\w/?!!' file
7789696 bklee
796588 bluecyn
477860 bborikun
473652 bluechiper
220780 ara316
144244 aceload
131088 belivart
79108 athlon85
78644 beschur512
66264 bogdanov