js> ''+function z(){x={1:1}} function z() { x = {, 1}; } this came from google reader, it's a big regression. i believe from the last week

http://bonsai.mozilla.org/cvsquery.cgi?treeid=default&module=all&branch=JS_1_7_ALPHA_BRANCH&branchtype=match&dir=&file=&filetype=match&who=&whotype=match&sortby=Date&hours=2&date=explicit&mindate=2006-06-15+12%3A00&maxdate=2006-06-16+12%3A00&cvsroot=%2Fcvsroot i believe this is what killed it. note that i'm fairly rusty wrt chasing regressions along branches.

specifically: mozilla/js/src/jsemit.c 3.154.2.18 JS_1_7_ALPHA_BRANCH 94/33 did it, which isn't surprising based on a read of what it's talking about, i don't expect to figure out more.

I forgot to change the source note used by the decompiler! This is a zero-risk fix and trivially correct. /be

This should be fixed for beta 1. /be

Checking in regress-344120.js; /cvsroot/mozilla/js/tests/js1_5/Function/regress-344120.js,v <-- regress-344120.js initial revision: 1.1

Checked into trunk (and JS_1_7_ALPHA_BRANCH, for posterity -- last time I will hit that branch). /be

Fixed on 1.8 branch too. /be

verified fixed 1.8.1, trunk 20060723 win/linux/mac(ppc|tel) and verified no regression on 1.8.0.5

This should have been decompilation test that failed with recent changes. I've updated it and moved it. Removing js1_5/Function/regress-344120.js; /cvsroot/mozilla/js/tests/js1_5/Function/regress-344120.js,v <-- regress-344120.js new revision: delete; previous revision: 1.2 done RCS file: /cvsroot/mozilla/js/tests/js1_5/decompilation/regress-344120.js,v done Checking in js1_5/decompilation/regress-344120.js; /cvsroot/mozilla/js/tests/js1_5/decompilation/regress-344120.js,v <-- regress-344120.js initial revision: 1.1 done