werc/bin/cgilib.rc

# Useful CGI stuff

fn dprint { echo $* >[1=2] }
fn dprintv { { for(v in $*) { echo -n $v^'#'^$#$v^'=' $$v '; '  }; echo } >[1=2] }
fn echo {if(! ~ $1 -n || ! ~ $2 '') /bin/echo $*}
fn escape_html { sed 's/&/\&amp;/g; s/</\&lt;/g; s/>/\&gt;/g' $* }

fn http_redirect {
    if(~ $1 http://* https://*)
        t=$1
    if not if(~ $1 /*)
        t=$"base_url^$1
    if not
        t=$"base_url^$"req_path^$1
    exec /bin/echo 'Status: '^$2^'
Location: '^$t^'

'
    exit
}
fn perm_redirect { http_redirect $1 '301 Moved Permanantly' }
fn post_redirect { http_redirect $1 '303 See Other' }


# Note: should check if content type is application/x-www-form-urlencoded?
# Should compare with http://www.shelldorado.com/scripts/cmds/urlgetopt.txt
fn load_post_args {
    if(~ $REQUEST_METHOD POST && ~ $#post_args 0) {
        ifs='&
'       for(pair in `{cat}) {
            ifs='=' { pair=`{echo -n $pair} }
            n='post_arg_'^`{echo $pair(1)|nurldecode|tr -cd 'a-zA-Z0-9_'}
            post_args=( $post_args $n )
            ifs=() { $n=`{echo -n $pair(2)|nurldecode|tr -d '
'} }
        }
        pair=()
    }
    if not
        status='No POST or post args already loaded'
}
# Status is () if at least one arg is found. DEPRECATED: access vars directly.
fn get_post_args {
    load_post_args
    _status='No post arg matches'
    for(n in $*) {
        v=post_arg_$n
        if(! ~ $#$v 0) {
            $n=$$v
            _status=()
        }
    }
    status=$_status
}

# This seems slightly improve performance, but might depend on httpd buffering behavior.
fn awk_buffer {
    awk '{
        buf = buf $0"\n"
        if(length(buf) > 1400) {
            printf "%s", buf
            buf = ""
        }
    }
    END { printf "%s", buf }'
}

fn nurldecode { urlencode -d || url_decode}	# GROSS

fn url_decode {
awk '
BEGIN {
    hextab ["0"] = 0; hextab ["8"] = 8;
    hextab ["1"] = 1; hextab ["9"] = 9;
    hextab ["2"] = 2; hextab ["A"] = hextab ["a"] = 10
    hextab ["3"] = 3; hextab ["B"] = hextab ["b"] = 11;
    hextab ["4"] = 4; hextab ["C"] = hextab ["c"] = 12;
    hextab ["5"] = 5; hextab ["D"] = hextab ["d"] = 13;
    hextab ["6"] = 6; hextab ["E"] = hextab ["e"] = 14;
    hextab ["7"] = 7; hextab ["F"] = hextab ["f"] = 15;
}
{
    decoded = ""
    i = 1
    len = length ($0)
    while ( i <= len ) {
        c = substr ($0, i, 1)
        if ( c == "%" ) {
            if ( i+2 <= len ) {
                c1 = substr ($0, i+1, 1)
                c2 = substr ($0, i+2, 1)
                if ( hextab [c1] == "" || hextab [c2] == "" ) {
                    print "WARNING: invalid hex encoding: %" c1 c2 | "cat >&2"
                } else {
                    code = 0 + hextab [c1] * 16 + hextab [c2] + 0
                    c = sprintf ("%c", code)
                    i = i + 2
                }
            } else {
                print "WARNING: invalid % encoding: " substr ($0, i, len - i)
            }
        } else if ( c == "+" ) {
            c = " "
        }
        decoded = decoded c
        ++i
    }
    printf "%s", decoded
}
'
}

fn nurlencode { urlencode || url_encode }	# GROSS

fn url_encode {
    awk '
    BEGIN {
    # We assume an awk implementation that is just plain dumb.
    # We will convert an character to its ASCII value with the
    # table ord[], and produce two-digit hexadecimal output
    # without the printf("%02X") feature.

    EOL = "%0A"     # "end of line" string (encoded)
    split ("1 2 3 4 5 6 7 8 9 A B C D E F", hextab, " ")
    hextab [0] = 0
    for ( i=1; i<=255; ++i ) ord [ sprintf ("%c", i) "" ] = i + 0
    if ("'^$"EncodeEOL^'" == "yes") EncodeEOL = 1; else EncodeEOL = 0
    }
    {
    encoded = ""
    for ( i=1; i<=length ($0); ++i ) {
        c = substr ($0, i, 1)
        if ( c ~ /[a-zA-Z0-9.-]/ ) {
        encoded = encoded c     # safe character
        } else if ( c == " " ) {
        encoded = encoded "+"   # special handling
        } else {
        # unsafe character, encode it as a two-digit hex-number
        lo = ord [c] % 16
        hi = int (ord [c] / 16);
        encoded = encoded "%" hextab [hi] hextab [lo]
        }
    }
    if ( EncodeEOL ) {
        printf ("%s", encoded EOL)
    } else {
        print encoded
    }
    }
    END {
        #if ( EncodeEOL ) print ""
    }
' $* 
}

# Cookies
fn set_cookie {
    # TODO: should check input values more carefully
    name=$1
    val=$2
    extraHttpHeaders=( $extraHttpHeaders 'Set-cookie: '^$"name^'='^$"val^'; path=/;' )
}
fn get_cookie {
    ifs=';' { co=`{echo $HTTP_COOKIE} }

    # XXX: we might be adding a trailing new line?
    # The ' ?' is needed to deal with '; ' inter-cookie delimiter
    { for(c in $co) echo $c } | sed -n 's/^ ?'$1'=//p' 
}


fn static_file {
    echo -n 'Content-Type: '
    select_mime $1
    echo
    exec cat $1
}

fn select_mime {
    m='text/plain'
    if(~ $1 *.css)
        m='text/css'
    if not if(~ $1 *.ico)
        m='image/x-icon'
    if not if(~ $1 *.png)
        m='image/png'
    if not if(~ $1 *.jpg *.jpeg)
        m='image/jpeg'
    if not if(~ $1 *.gif)
        m='image/gif'
    if not if(~ $1 *.pdf)
        m='application/pdf'
    echo $m
}

##############################################
# Generic rc programming helpers

# Manage nested lists
fn ll_add {
    _l=$1^_^$#$1
    $_l=$*(2-)
    $1=( $$1 $_l )
}
# Add to the head: dangerous if you shrink list by hand!
fn ll_addh {
    _l=$1^_^$#$1
    $_l=$*(2-)
    $1=( $_l $$1 )
}


NEW_LINE='
'

# crop_text [max_lenght [ellipsis]] 
# TODO: Option to crop only at word-delimiters.
fn crop_text {
    m=512
    e='...'
    if(! ~ $#1 0)
        m=$1
    if(! ~ $#2 0)
        e=$2

    awk -v 'max='^$"m -v 'ellipsis='$e '
    {
        nc += 1 + length;
        if(nc > max) {
            print substr($0, 1, nc - max) " " ellipsis
            exit
        }
        print
    }' 
}