marcus/scour
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Remove XML pretty printing due to erroneous injection of whitespace on text nodes

Browse source
This commit is contained in:
JSCHILL1 2009-04-28 13:18:29 -05:00
parent f37fef89df
commit 02602edde0
2 changed files with 30 additions and 16 deletions
Download patch file Download diff file Expand all files Collapse all files

18
release-notes.html
View file

@ -9,6 +9,24 @@
<p>Copyright 2009, Jeff Schiller</p> <p>Copyright 2009, Jeff Schiller</p>
<section id="0.11">
<header>
<h2><a href="#0.11">Version 0.11</a></h2>
</header>
<ul>
<li>convert gradient stop offsets from percentages to float</li>
<li>convert gradient stop offsets to integers if possible (0 or 1)</li>
<li>fix bug in line-to-hz conversion</li>
<li>handle non-ASCII characters (Unicode)</li>
<li>remove empty line or curve segments from path</li>
<li>added option to prevent style-to-xml conversion</li>
<li>handle compressed svg (svgz) on the input and output</li>
<li>added total time taken to the report</li>
<li>Removed XML pretty printing because of <a href="http://ronrothman.com/public/leftbraned/xml-dom-minidom-toprettyxml-and-silly-whitespace/">this problem</a>.</li>
</ul>
</section>
<section id="0.10"> <section id="0.10">
<header> <header>
<h2><a href="#0.10">Version 0.10</a></h2> <h2><a href="#0.10">Version 0.10</a></h2>

28
scour.py
View file

@ -34,8 +34,6 @@
# * Convert RGB colours from RGB(r,g,b) to #RRGGBB format # * Convert RGB colours from RGB(r,g,b) to #RRGGBB format
# * Convert RGB colours from #RRGGBB to #RGB if possible # * Convert RGB colours from #RRGGBB to #RGB if possible
# * Clean up paths # * Clean up paths
# * Detect vertical/horizontal lines and replace.
# * Eliminate empty path segments
# * Eliminate last segment in a polygon # * Eliminate last segment in a polygon
# * Collapse straight curves. # * Collapse straight curves.
# * Process Transformations # * Process Transformations
@ -46,14 +44,7 @@
# * Put id attributes first in the serialization (or make the d attribute last) # * Put id attributes first in the serialization (or make the d attribute last)
# Next Up: # Next Up:
# + convert gradient stop offsets from percentages to float # - text elements are shoved over to the right
# + convert gradient stop offsets to integers if possible (0 or 1)
# + fix bug in line-to-hz conversion
# + handle non-ASCII characters (Unicode)
# + remove empty line or curve segments from path
# + added option to prevent style-to-xml conversion
# + handle compressed svg (svgz) on the input and output
# - display how long it took to scour the file in the report
# - prevent elements from being stripped if they are referenced in a <style> element # - prevent elements from being stripped if they are referenced in a <style> element
# (for instance, filter, marker, pattern) - need a crude CSS parser # (for instance, filter, marker, pattern) - need a crude CSS parser
# - Remove unnecessary units of precision on attributes (use decimal: # - Remove unnecessary units of precision on attributes (use decimal:
@ -669,7 +660,6 @@ def cleanPath(element) :
# however, this parser object has some ugliness in it (lists of tuples of tuples of # however, this parser object has some ugliness in it (lists of tuples of tuples of
# numbers and booleans). we just need a list of (cmd,[numbers]): # numbers and booleans). we just need a list of (cmd,[numbers]):
# TODO: remove empty path segments
path = [] path = []
for (cmd,dataset) in pathObj: for (cmd,dataset) in pathObj:
if cmd in ['M','m','L','l','T','t']: if cmd in ['M','m','L','l','T','t']:
@ -688,10 +678,7 @@ def cleanPath(element) :
# one or more numbers # one or more numbers
nums = [] nums = []
for n in dataset: for n in dataset:
if n != 0: nums.append(Decimal(str(n)))
nums.append(Decimal(str(n)))
else:
numPathSegmentsRemoved += 1
if nums: if nums:
path.append( (cmd, nums) ) path.append( (cmd, nums) )
@ -1072,7 +1059,10 @@ def scourString(in_string, options=[]):
properlySizeDoc(doc.documentElement) properlySizeDoc(doc.documentElement)
# output the document as a pretty string with a single space for indent # output the document as a pretty string with a single space for indent
out_string = doc.documentElement.toprettyxml(' ') # NOTE: removed pretty printing because of this problem:
# http://ronrothman.com/public/leftbraned/xml-dom-minidom-toprettyxml-and-silly-whitespace/
# out_string = doc.documentElement.toprettyxml(' ')
out_string = doc.documentElement.toxml()
# now strip out empty lines # now strip out empty lines
lines = [] lines = []
@ -1165,6 +1155,9 @@ def parseCLA():
if __name__ == '__main__': if __name__ == '__main__':
startTimes = os.times()
# print times[0], times[1]
(input, output, options, inputfilename, outputfilename) = parseCLA() (input, output, options, inputfilename, outputfilename) = parseCLA()
# if we are not sending to stdout, then print out app information # if we are not sending to stdout, then print out app information
@ -1182,10 +1175,13 @@ if __name__ == '__main__':
input.close() input.close()
output.close() output.close()
endTimes = os.times()
# output some statistics if we are not using stdout # output some statistics if we are not using stdout
if bOutputReport : if bOutputReport :
if inputfilename != '': if inputfilename != '':
print ' File:', inputfilename print ' File:', inputfilename
print ' Time taken:', str(endTimes[0]-startTimes[0]) + 's'
print ' Number of unreferenced id attributes removed:', numIDsRemoved print ' Number of unreferenced id attributes removed:', numIDsRemoved
print ' Number of elements removed:', numElemsRemoved print ' Number of elements removed:', numElemsRemoved
print ' Number of attributes removed:', numAttrsRemoved print ' Number of attributes removed:', numAttrsRemoved