Automated (Macro?) upper case to lower case

GIS_USER · Post by **GIS_USER** » Wed Jul 07, 2004 3:32 pm

Is there a way to quickly automate HTML tags across several documents from upper case to lower case? I need this done for all tags across multiple files.

Here is a small portion of what I need done...

<BODY LINK="#3366CC" BGCOLOR="WHITE" MARGINWIDTH=0 MARGINHEIGHT=0>
<TABLE WIDTH="100%" CELLPADDING=0 CELLSPACING=0 BORDER=0> to

<body link="#3366cc" bgcolor="white" marginwidth=0 marginheight=0>
<table width="100%" cellpadding=0 cellspacing=0 border=0>

Thanks!

ben_josephs · Post by **ben_josephs** » Wed Jul 07, 2004 4:30 pm

Here's one way, using regular expression search and replace:

Find what: <[^>]+>
Replace with: \L&

using

Configure | Preferences | Editor | Use POSIX regular expression syntax

Post by **MudGuard** » Wed Jul 07, 2004 5:22 pm

ben, your expression replaces too much - imagine this:

Code: Select all

<IMG SRC="BLA.PNG" ALT="Picture of George Washington">

would end up as

Code: Select all

<img src="bla.png" alt="picture of george washington">

Not only the alt text would be changed, but also the URL - which could lead to broken links.

Another thing - attributes like alt or title might contain the > character.

HTML can't be parsed with regexes alone - this is the reason why I do not offer a better solution...

What can be done is changing the element name:

Search for

Code: Select all

</?[A-Za-z]+

Replace by

Code: Select all

\L&

Automatically finding/changing the attribute names is much more difficult - as attributes may contain quotes, blanks, and = and >
e.g.

Code: Select all

title="title='> > >'"

or

Code: Select all

title='title=">">">"'

are both valid HTML attributes but give lots of trouble when trying to use regexes on them...

Post by **MudGuard** » Wed Jul 07, 2004 5:27 pm

Forgot something:

There are tools that have an HTML parser inside and can therefore do the job properly, e.g.
HTML Tidy

ben_josephs · Post by **ben_josephs** » Wed Jul 07, 2004 8:27 pm

MudGuard wrote:ben, your expression replaces too much

You're quite right, of course. But it's OK for the example given. Apologies if it misled.

HTML can't be parsed with regexes alone

Indeed. Anything with arbitrarily nested structures can't be.

GIS_USER · Post by **GIS_USER** » Thu Jul 08, 2004 12:51 pm

Thanks for the info everyone. MudGuard was correct in assuming the following scenario:

<IMG SRC="BLA.PNG" ALT="Picture of George Washington">

<img src="bla.png" alt="picture of george washington">

Of course, idealy the end result would be:

<img src="bla.png" alt="Picture of George Washington">

UNIX constraints also require file names (ie. bla.png) to be lower case, however any non-code text would remain the same. We are translating our HTML to XHTML

Post by **MudGuard** » Thu Jul 08, 2004 3:37 pm

UNIX constraints also require file names (ie. bla.png) to be lower case.

No. UNIX has case sensitive file names. Same goes for all derivates of UNIX that I know (including HP-UX, AIX, SINIX, Linux)

You can, e.g. have the following files in one folder:

bla
blA
bLa
bLA
Bla
BlA
BLa
BLA

without any problem.