#!/usr/bin/python -utt
# vim:et:sw=4
'''Unzip an ODF document and list or kill/substitute text styles.

Usage: kstyles INPUT OUTPUT [-l] [-dsty1]... [sty1=sty2]...
    INPUT
        name of file to read

    OUTPUT
        name of file to write new (modified) file

    -l
        list styles

    -dsty1
        Show text spans having property sty1

    sty1=sty2
        Text elements which are sty1 are assigned sty2

$Id: kstyles.py,v 0.6 2014/05/10 22:19:13 collin Exp collin $
'''

import codecs
import os
import sys
import xml.dom.minidom
import zipfile

CONTENT = 'content.xml'
STYLES = 'styles.xml'
STYLE = 'style:style'
SFAMILY = 'style:family'
SNAME = 'style:name'
TSTYLENAME = 'text:style-name'
TSPAN = 'text:span'

def main(args):
    '''Unpack ODF/XML (zipfile).  Discover text styles.
    Find # of text elements which have each style; if "-l", display.
    If "-dXX", display text:spans with style XX
    For each "sty1=sty2" provided:
        change any text elements of sty1 to sty2'''
    try:
        infile_name = args[0]
        outfile_name = args[1]
        ops = args[2:]
    except:
        usage()
    if not os.path.exists(infile_name):
        print "Couldn't find input file %s" % infile_name
        usage()
    INFILE = zipfile.ZipFile(infile_name, 'r')

    if os.path.exists(outfile_name):
        if os.path.realpath(infile_name) == os.path.realpath(outfile_name):
            # Don't ask...
            print "I can't update a file in place: %s" % outfile_name
            usage()
        print "Output file exists: %s" % outfile_name
        print "Overwrite? Hit <ENTER>, or ^C to abort now. ", 
        junk = sys.stdin.readline()

    # Sanity-check input file before doing anything else.
    if CONTENT not in INFILE.namelist():
        print "Couldn't find %s in %s's zip archive" % (CONTENT, infile_name)
        print 'Is it an ODF file?'
        sys.exit(1)
    # Read and parse content.
    adom = dict()
    text_styles = list()
    for afile in [X for X in CONTENT, STYLES if X in INFILE.namelist()]:
        adom[afile] = xml.dom.minidom.parseString(INFILE.read(afile))
        cstyles = adom[afile].getElementsByTagName(STYLE)
        text_styles.extend([X for X in cstyles
                                if X.getAttribute(SFAMILY) == 'text'])
    text_style_names = [X.getAttribute(SNAME) for X in text_styles]
    cdom = adom[CONTENT]

    # Track how many spans of each style.  Names should be unique
    style_counts = dict.fromkeys(text_style_names, 0)
    n, k = len(text_style_names), len(style_counts)
    if n != k:
        print('NOTE: %d text_style_names, %d unique' % (n, k))

    for aspan in [X for X in cdom.getElementsByTagName(TSPAN)
                    if X.hasAttribute(TSTYLENAME)]:
        astyle = aspan.getAttribute(TSTYLENAME)
        if astyle not in style_counts:
            print "? unknown style: %s" % astyle
        style_counts[astyle] = style_counts.get(astyle, 0) + 1

    if '-l' in ops:
        # print text_styles
        for aname, astyle in zip(text_style_names, text_styles):
            if style_counts[aname]:
                print '%s: %d spans' % (aname, style_counts[aname])
                print '\t%s' % astyle.toxml()
        for astyle in [X for X in style_counts if X not in text_style_names]:
            print '??? %s: %d spans' % (astyle, style_counts[astyle])
        while '-l' in ops:
            ops.remove('-l')

    # Before the following fun stuff, make stdout be utf8
    utf8_enc = codecs.getencoder('utf8')

    for op in ops:
        if op.startswith('-d'):
            astyle = op[2:]
            if astyle not in style_counts:
                print "*** Couldn't find style %s" % astyle
                continue
            print "=== Text style %s:" % astyle
            for aspan in [X for X in cdom.getElementsByTagName(TSPAN)
                            if X.getAttribute(TSTYLENAME) == astyle]:
                mydata = ' '.join([X.data for X in aspan.childNodes
                                        if X.nodeType == X.TEXT_NODE])
                print utf8_enc(mydata)[0]
            continue
        styles = op.split('=')
        if len(styles) > 2:
            print >> sys.stderr, "Can't parse: '%s'" % op
            usage()
        if len(styles) < 2:
            print >> sys.stderr, 'Not yet implemented: %s' % op
            continue
        print 'Changing style "%s" to "%s".' % (styles[0], styles[1])
        for aspan in [X for X in cdom.getElementsByTagName(TSPAN)
                        if X.getAttribute(TSTYLENAME) == styles[0]]:
            aspan.setAttribute(TSTYLENAME, styles[1])

    if os.path.exists(outfile_name):
        os.unlink(outfile_name)
    OUTFILE = zipfile.ZipFile(outfile_name, 'w', zipfile.ZIP_DEFLATED)
    for oldinfo in INFILE.infolist():
        fname = oldinfo.filename
        fsize = oldinfo.file_size
        #print 'archive member "%s", %d bytes' % (fname, fsize)
        if fsize > 0:
            if fname == CONTENT:
                OUTFILE.writestr(fname, utf8_enc(cdom.toxml())[0])
            else:
                OUTFILE.writestr(fname, INFILE.read(fname))
        else:
            OUTFILE.writestr(fname, '')
    OUTFILE.close()


def usage():
    print >> sys.stderr, __doc__
    sys.exit(1)

if __name__ == '__main__':
    main(sys.argv[1:])