Dotaz: Bash: Jak dostanu ze souboru jen to, co chci?

[David Watzke]

Ahoj. To je zase problém. Potřebuju z HTMLka dostat jen to, co potřebuju. Ukázka HTML:

<table class="datatable" cellspacing="0" cellpadding="0"> <tbody> <tr> <th colspan="2">Browse Program</th><th style="text-align: right;" colspan="2"> </th> </tr> <tr> <td style="width: 150px;"><b>Title:</b></td> <td colspan="3">amarok</td> </tr> <tr> <td style="width: 150px;" valign='top'><b>Description:</b></td> <td colspan="3">amaroK - the audio player for KDE.</td> </tr> <tr> <td style="width: 150px;" valign='top'><b>Website:</b></td> <td colspan="3"> <a href="http://amarok.kde.org/">http://amarok.kde.org/</a>   </td> </tr> <tr> <td style="width: 150px;" valign="top"><b>Ebuilds:</b></td> <td colspan="3"> <table width="100%" cellpadding="0"> <tr> <td><a href="/ebuild?e=media-sound%2Famarok%2Famarok-1.3.7-r1.ebuild&browse=true"><img src="/images/magnify.png" alt="View Ebuild" height="16px" width="16px" /></a>   <a href="/ebuild?e=media-sound%2Famarok%2Famarok-1.3.7-r1.ebuild"><img src="/img/download.png" alt="Download Ebuild" height="16px" width="16px" /></a>   <b>1.3.7-r1 <span class='hardmask'></span></b></td> <td>~amd64 ~ppc ~ppc64 ~sparc ~x86 </td> </tr><tr> <td><a href="/ebuild?e=media-sound%2Famarok%2Famarok-1.3.7.ebuild&browse=true"><img src="/images/magnify.png" alt="View Ebuild" height="16px" width="16px" /></a>   <a href="/ebuild?e=media-sound%2Famarok%2Famarok-1.3.7.ebuild"><img src="/img/download.png" alt="Download Ebuild" height="16px" width="16px" /></a>   <b>1.3.7 <span class='hardmask'></span></b></td> <td>~amd64 ~ppc ~ppc64 ~sparc ~x86 </td> </tr><tr> <td><a href="/ebuild?e=media-sound%2Famarok%2Famarok-1.3.6.ebuild&browse=true"><img src="/images/magnify.png" alt="View Ebuild" height="16px" width="16px" /></a>   <a href="/ebuild?e=media-sound%2Famarok%2Famarok-1.3.6.ebuild"><img src="/img/download.png" alt="Download Ebuild" height="16px" width="16px" /></a>   <b>1.3.6 <span class='hardmask'></span></b></td> <td>~amd64 ppc ~sparc x86 </td> </tr><tr> <td><a href="/ebuild?e=media-sound%2Famarok%2Famarok-1.2.4.ebuild&browse=true"><img src="/images/magnify.png" alt="View Ebuild" height="16px" width="16px" /></a>   <a href="/ebuild?e=media-sound%2Famarok%2Famarok-1.2.4.ebuild"><img src="/img/download.png" alt="Download Ebuild" height="16px" width="16px" /></a>   <b>1.2.4 <span class='hardmask'></span></b></td> <td>amd64 ppc -sparc x86

a chtěl bych výstup:

/ebuild?e=media-sound%2Famarok%2Famarok-1.3.7-r1.ebuild
/ebuild?e=media-sound%2Famarok%2Famarok-1.3.7.ebuild
/ebuild?e=media-sound%2Famarok%2Famarok-1.3.6.ebuild
/ebuild?e=media-sound%2Famarok%2Famarok-1.2.4.ebuild

Nevíte jak na to?

Odpovědi

[NbS]

Co takhle nejaky ten regulerni vyraz v perlu ?

if ($text =~ /co hledam v textu/) { print promenna do ktere se nacte vysledek\n; }

[David Watzke]

V Perlu? To ne, potřebuju to v Bashi.

[Fuky]

sed 's/href="\(.+\)"/\1/g'

[David Watzke]

Divný, s tímhle příkazem se výstup nezmění ani o chlup. :/

[Fuky]

Omlouvám se napsal jsem to špatně.

sed -n 's/.*href="\([^"]\+\).*/\1/gp' <soutor

[David Watzke]

Teď je to fajn, díky moc!

[Michal Kubeček]

Na takovéto úlohy je obecně asi nejlepší použít XSLT, např. xsltproc.