#!/bin/sh
##
## Nautilus
## SCRIPT: 10_oneSVGfile_EXTRACTpath-etcXYdata_egrep-fold-awk-sed-txteditor.sh
##
## PURPOSE: For a user-selected SVG (Scalable Vector Graphics) file,
##          the file contents are passed through 'egrep' and 'fold' and 'awk'
##          and 'sed' utilities to extract and put the numeric
##          x,y coordinate data into a separate text file.
##
##          The x,y coordinate data is between '<path ' and
##          '/>' markup indicators --- and in some special SVG 'element'
##          statements such as '<rect ', '<line ', '<polyline ',
##          '<circle ', and '<text '.
##
##          The '<path ' and '/>' indicators may span multiple lines.
##          We may assume the others are on a single line.
##
##          Several types of character replacement are performed in
##          a 'path' statement's line of x,y data --- including:
##            - each occurrence of the 'm' or 'M' or 'c' or 'C'
##              and other such one-character commands is replaced
##              that character preceded by a line-feed.
##          This helps break up long 'path' statements into much
##          shorter lines of x,y data pairs.
##
##          (NOTE: We do not attempt to replace spline data with
##           data giving an accurate rendering the splines. We
##           simply pass along the control-point data --- hoping
##           that that data is near enough the actual spline curve.
##           However, that is often not the case.)
##
##          The output of 'egrep' and a pipeline of 'fold', 'awk',
##          and 'sed' commands is directed into a separate text file ---
##          in the same directory as the input file.
##
##         (An alternative is to put the output file into the /tmp
##          directory. This might be advisable if the output files
##          are typically huge. But most SVG files are not massive.)
##
##          NOTE:
##          The SVG coordinate data between '<path ' and
##          '/>' markup indicators can often 
##          go on for thousands of characters in a single line.
##          We basicly want to 'fold' these extra-long lines
##          into lines with relatively few pairs of x,y numeric values.
##          This is much more manageable to manually edit for
##          final touchup into 2 columns of x,y numbers.
##
## HOW TO USE: In Nautilus, select a SVG (text) file in a directory.
##             (The selected file should NOT be a directory.)
##             Right click and, from the 'Scripts >' submenus,
##             choose to run this script (name above).
##
## Created: 2017sep17 Based on the FE Nautilus 'MAPtools' script
##                    '00_oneKMLfile_REFORMAT-FoldLines_sed-txteditor.sh'
##                    and the FE Nautilus 'HTMLtools' script
##       '02_anyfile_HTMLfile-REMOVE-SCRIPTs-STYLEs_sed-awk-txteditor.sh'
## Changed: 2017

## FOR TESTING: (show statements as they execute)
# set -x

############################################
## Set a couple of input filename variables.
############################################

FILENAME="$1"

BASENAME=`basename $FILENAME`

############################################
## Set a couple of script name variables.
############################################

SCRIPTDIRNAME=`dirname $0`
SCRIPTBASENAME=`basename $0`

##############################################
## Set maximum length for lines in input to 'awk'.
## Set maximum length for lines in output file.
##############################################

MAXCHARSperLINEin=3071
MAXCHARSperLINEout=80

############################################
## Show an informative message with 'zenity'. 
############################################

Q1="'"
LT="&lt;"

MSG2USER="\
This utility expects the file you chose:
   $BASENAME
to be an SVG (Scalable Vector Graphics) file. This script processes
the file contents with the ${Q1}egrep${Q1} and ${Q1}fold${Q1} and ${Q1}awk${Q1} and
${Q1}sed${Q1} utilities --- to reformat the data into a separate text file.

The x,y coordinate data between ${Q1}${LT}path ${Q1} and ${Q1}/>${Q1} markup indicators
--- and in some special SVG ${Q1}element${Q1} statements such as ${Q1}${LT}rect ${Q1},
${Q1}${LT}line ${Q1}, ${Q1}${LT}polyline ${Q1}, ${Q1}${LT}circle ${Q1}, and ${Q1}${LT}text ${Q1} --- are retained.

Some extensive manual editing will typically be required to create
a file of x,y data in two columns -- for defining line-segments
--- especially if RELATIVE (lower-case) SVG commands are used.

The output file may include some ${Q1}left-over${Q1} character strings.
The main intent of this utility is to make a text file that is
easily edited to contain only a pair of x,y coordinate numbers
per line --- along with a comment line or two at the top that
may incorporate some description data, like the input filename
--- and comment lines to separate ${Q1}loops${Q1} of line segments.

A major purpose of this utility is to ${Q1}fold${Q1} extra-long
lines of coordinate data into relatively short lines containing
x,y coordinate numbers.

The user-selected file is processed and the output is put
in a file with a string like ${Q1}_EXTRACTED${Q1} appended to
the midname of the user-selected file.

The output file is put in the same directory with the
original selected file.

This script is:
   $SCRIPTBASENAME
in directory
   $SCRIPTDIRNAME"

## FOR TESTING:
#  echo "$MSG2USER"


zenity --warning --width=600 --no-wrap \
   --title "SVG file processor - EXTRACT x,y DATA lines" \
   --text "$MSG2USER" 


###############################################
## Exit if the selected file is a directory.
###############################################

if test -d "$FILENAME"
then
   exit
fi


####################################################
## Get the file extension and check that it is not
## blank. Skip the filename if it has no extension.
##  (Assumes one '.' in filename, at the extension.)
####################################################

FILEEXT=`echo "$FILENAME" | cut -d\. -f2`

if test "$FILEEXT" = "" 
then 
   exit
fi

####################################################
## Exit if the file extension is not 'svg'.
## COMMENTED, for now.
####################################################

# if test "$FILEEXT" != "svg"
# then
#    exit
# fi

####################################################
## Get the 'midname' of the file, the part before
## the period and the extension.
####################################################

MIDNAME=`echo "$FILENAME" | cut -d\. -f1`


###################################
## Make the output filename.
###################################

OUTNAME="${MIDNAME}_EXTRACTEDsvgXYdata.txt"

rm -f "$OUTNAME"

###########################################
## Put the input filename in a comment line
## at the top of the output file.
###########################################

echo "# $FILENAME" > "$OUTNAME"


##############################################################
## Use 'fold' and 'awk' and 'sed' to make the new output file.
##############################################################
## Could try running the command in a window,
## to see err msgs, if any.
## Could use zenity to offer this as an option.
##
## xterm -fg white -bg black -hold -e \
#######################################################
## The 'egrep' and the awk-and-sed pipe below does the following:
##  0) Use 'egrep' to concatenate the following 'element' lines
##     onto the output file: 
##          '<rect ... />'
##          '<line ... />'
##          '<polyline ... />'
##          '<circle ... />'
##          '<text ... />'
##     We assume each is on one line.
##  1) Use 'fold' to fold lines longer than $MAXCHARSperLINEin.
##        NOTE: The following awk will fail with a
##        'Input record too long' error that stops awk dead
##        when a line is longer than about 3071 chars.
##        The 'fold' prevents that.
##  2) Use an 'awk' program to print 
##     <path ...... /> sections even if they span more than
##      one line.
#   3) Use 'sed' to remove path parameters like
##     id="..." and style="..."
##     which would be broken up by the following sed.
##
##     NOTE: The regular expression   "[^"]*"   removes
##     the double-quotes, and the stuff between them, up to
##     the SECOND double-quote --- not to a LAST double-quote
##     in the line. Ref: page 54 of O'Reilly 'Sed and Awk' book.
##  4) Use 'sed' to replace each occurrence of the
##     'm' or 'M' or 'c' or 'C' or 'v' or 'V' or
##     'l' or 'L' etc. 'path' characters by a line-feed
##      followed by the character.
##     Also replace '-' by ' -'.
## Other not-so-critical changes on the "path" stuff:
##  5) Use 'sed' to replace '<path ' with '#P'.
##  6) Use 'sed' to replace trailing '/>' with null.
##  7) Use 'sed' to replace each occurrence of one or more space
##     characters with exactly one space character.
##  8) Use 'sed' to replace each occurrence of a comma character
##     by a space character.
##  9) Remove blank lines.
## 10) Remove double-quote characters.
## 5-10 are not necessarily in this order.
#######################################################
## In a '<path' element, the following single characters
## have the indicated meaning:
##   m means moveto-relative
##   M means Moveto-absolute
##   l means lineto-relative 
##   L means Lineto-absolute 
##   v means vertical-lineto-relative 
##   V means Vertical-lineto-absolute 
##   h means horizontal-lineto-relative 
##   H means Horizontal-lineto-absolute 
##   c means cubic-bezier-relative
##   C means Cubic-bezier-absolute
##   s means s-cubic-bezier-relative
##   S means S-Cubic-bezier-absolute
##   q means quadratic-bezier-relative
##   Q means Quadratic-bezier-absolute
##   t means t-quadratic-bezier-relative
##   T means T-Quadratic-bezier-absolute
##   a means arc-elliptic-relative
##   A means Arc-elliptic-absolute
##   z means z-closepath
#######################################################

## Concatenate the rect,line,polyline,circle,text lines on the output.
## (We assume these 'elements' are defined on a single line.)
## (We also assume we can extract these lines in no particular order,
##  from among 'path' lines. If this is a problem, we may have to
##  include this extraction in the following 'awk' program.)

egrep '<rect |<line |<polyline |<circle |<text ' "$FILENAME" >>  "$OUTNAME"


## FOR TESTING:
#   set -x

fold -b -w $MAXCHARSperLINEin "$FILENAME" | \
   awk \
'BEGIN  {
GetNEXTpathLine0or1 = 0
}
## END OF BEGIN
## START OF BODY
{
# HOLDline = tolower($0)
HOLDline = $0

## FOR TESTING:
# if  ( NR < 40 ) { print "HOLDline :" $HOLDline }

## If this is a continued "<path" line
## (as indicated by the GET flag equal to 1),
## check if it contains the end-indicator, "/>".
## If so, set the GET flag to 0.
## In any case, print this "path" line and  go to
## next line and start the "body" processing again.

if ( GetNEXTpathLine0or1 == 1 ) {
   if ( index(HOLDline,"/>") != 0  )  { 
      GetNEXTpathLine0or1 = 0
   }
   print
   next
} else {

   ## Check if this is a "<path" line. If it IS,
   ## print it ... AND
   ## if the line contains the end-indicator, "/>",
   ## set the GET flag to 0. If it does not contain
   ## the end-indicator, set the GET flag to 1. 
   ## In either case,  we are done with this
   ## "<path" line. Go to next line and start the
   ## "body" processing again.

   if ( index(HOLDline,"<path") != 0  )  {
      print
      if ( index(HOLDline,"/>") != 0  )  {
         GetNEXTpathLine0or1 = 0
      } else {
         GetNEXTpathLine0or1 = 1
      }
      next
   }
}


## No print of other lines if we reach this point ---
## because if we reach this point, the line is not
## part of a "path" line.

}' | \
   sed -e 's|id="[^"]*"||g' | \
   sed -e 's|style="[^"]*"||g' | \
   sed -e 's|transform="[^"]*"|\n#X|g' | \
   sed -e 's|sodipodi:[^"]*="[^"]*"|\n#Y|g' | \
   sed -e 's|<path|\n#P |g' | \
   sed -e 's|fill=|\nF |g' | \
   sed -e 's|d=||g' | \
   sed -e 's|-| -|g' | \
   sed -e 's|-| -|g' | \
   sed -e 's|m|\n#\nm |g' | \
   sed -e 's|M|\n#\nM |g' | \
   sed -e 's|l|\nl |g' | \
   sed -e 's|L|\nL |g' | \
   sed -e 's|v|\nv |g' | \
   sed -e 's|V|\nV |g' | \
   sed -e 's|h|\nh |g' | \
   sed -e 's|H|\nH |g' | \
   sed -e '/F/n;s|c|\nc |g' | \
   sed -e 's|C|\nC |g' | \
   sed -e 's|s|\ns |g' | \
   sed -e 's|S|\nS |g' | \
   sed -e 's|q|\nq |g' | \
   sed -e 's|Q|\nQ |g' | \
   sed -e 's|t|\nt |g' | \
   sed -e 's|T|\nT |g' | \
   sed -e '/F/n;s|a|\n#a |g' | \
   sed -e 's|A|\n#A |g' | \
   sed -e 's|z|\n#z |g' | \
   sed -e 's|Z|\n#Z |g' | \
   sed -e 's|/>||g' | \
   sed -e 's|"||g' | \
   sed -e 's|  *| |g' | \
   sed -e '/^ *$/d' | \
   sed -e 's|,| |g' | \
   fold -b -w $MAXCHARSperLINEout >> "$OUTNAME"


## FOR TESTING:
#   set -


###################################
## Show the output file.
###################################

## . $HOME/.gnome2/nautilus-scripts/.set_VIEWERvars.shi

. $HOME/.freedomenv/feNautilusScripts/set_DIR_NautilusScripts.shi
. $DIR_NautilusScripts/.set_VIEWERvars.shi

# $TXTVIEWER "$OUTNAME" &

$TXTEDITOR "$OUTNAME" &