Appended is a patch to trim trailing zeros after the decimal point, so that `10.000000' becomes `10', and `1.500000' becomes `1.5'.
(A couple of people have complained about our trailing zeros.)
Given that we're so near to release, I didn't want to commit without discussing here first, i.e. whether to wait until after 0.42.
This affects both SVG and eps/ps output, and possibly other formats.
Are there any situations where e.g. `4' would have a different meaning from `4.0' or `4.0000000' ? E.g. in C `4.0 / 3' means something different from `4 / 3'; are there any similar cases in postscript or something ?
Testing done so far:
Inkscape compiles, and passes its test suite (make check).
HEAD-inkscape --export-plain-svg=a0.svg share/examples/art-nouveau-P3.svg trim-inkscape --export-plain-svg=at.svg share/examples/art-nouveau-P3.svg diff -du a[0t].svg | less (manual checking for bad differences)
# Compare eps output from the trimmed and HEAD versions of inkscape, # in an attempt to pick up differences in meaning of the two SVG # files. HEAD-inkscape --export-eps=a0-0.eps a0.svg HEAD-inkscape --export-eps=at-0.eps at.svg diff a[0t]-0.eps # produces no output
trim-inkscape --export-eps=a0-t.eps a0.svg trim-inkscape --export-eps=at-t.eps at.svg diff a[0t]-t.eps # produces no output
diff -du a0-[0t].eps | less (manual checking for bad differences; though I'm not very familiar with postscript)
# Generalize the gives-same-eps test. for i in share/examples/*.{svg,svgz}; do HEAD-inkscape --export-plain-svg=$HOME/tmp/a0.svg $i if trim-inkscape --export-plain-svg=$HOME/tmp/at.svg $i; then :; else echo "$i to plain failed" fi for f in 0 t; do HEAD-inkscape --export-eps=$HOME/tmp/a$f-0.eps ~/tmp/a$f.svg done diff -u ~/tmp/a[0t]-0.eps|head|grep -q . && echo "$i has diffs" done
At the time of writing, I haven't actually run inkscape in its gui mode: I've done the change remotely. Of course I'll do that too before committing.
Another thing I'll do before committing (but haven't yet done) is to create a new unit test.
What are people's comments about whether to include this in 0.42 or not? What further testing would be needed if it were to be in 0.42 ?
pjrm.