summaryrefslogtreecommitdiffstats
path: root/perl_checker.src/parser_helper.ml
Commit message (Collapse)AuthorAgeFilesLines
* various detection of join() misusePascal Rigaux2004-04-281-0/+9
|
* get_pos_from_expr() (previously named get_pos_from_tree()) is useful, even ↵Pascal Rigaux2004-04-281-0/+34
| | | | in parser_helper
* fix typoPascal Rigaux2004-04-281-1/+1
|
* - deprecate "open F, ..." prefering "open(my $F, ...)"Pascal Rigaux2004-04-281-1/+14
| | | | | - fix an internal_error
* - do check_simple_pattern() directly in from_PATTERN()Pascal Rigaux2004-04-281-2/+4
| | | | | - don't die, log instead for some regexp expressions
* fix warning (would also need fixing detecting of boolean context vs scalarPascal Rigaux2004-03-311-1/+1
| | | | | context)
* in "$a ? $a : xxx", "xxx" can need short circuitPascal Rigaux2004-03-311-1/+1
| | | | | (even if i don't think any value which is_a_scalar can need_short_circuit)
* perl_checker: entries in generated pot file are sorted by filesPascal Rigaux2004-01-091-12/+14
|
* fix checking dropped value in some casesPascal Rigaux2004-01-051-1/+2
|
* use hash_ref() to check priority of the argumentsPascal Rigaux2004-01-051-1/+3
|
* disallowPascal Rigaux2003-12-161-12/+32
| | | | | | | - $a = (1, 2) - my $a = (1, 2) - my ($a, $b) = (1, 2, 3)
* check the number of parameters given to N()Pascal Rigaux2003-12-161-10/+19
|
* - drop internally operator "qw" by rewriting qw(...) into a list of stringsPascal Rigaux2003-12-161-2/+2
| | | | | | - as a side effect "foo" . qw(bar boo) is now detected - allow pop(@l) (with explicit parentheses)
* - pop behaves just like shiftPascal Rigaux2003-12-161-2/+8
| | | | | - check that the parameter to pop & shift is an array, and not more parameters are given
* fix checking value is dropped in tuplePascal Rigaux2003-12-161-1/+2
|
* - add the tuple(...) contextPascal Rigaux2003-12-161-10/+32
| | | | | | - check that the value xxx in "... or xxx" is dropped - handle "... or ..., ..."
* check what is on the left side of "xxx =~ s///"Pascal Rigaux2003-12-151-2/+8
|
* - check the context on left side of =~Pascal Rigaux2003-12-151-0/+4
| | | | | - !xx has the new context bool
* - specify the return context for N() and N_()Pascal Rigaux2003-12-151-1/+2
| | | | | - more precise position for warning "value is dropped"
* allow to specify the return context of functions instead of always saying ↵Pascal Rigaux2003-12-151-9/+10
| | | | M_unknown
* update "faking a normal lexbuf" code for new ocaml's parsing.mlPascal Rigaux2003-12-151-3/+3
|
* - correctly handle "shift" with no parameter at toplevelPascal Rigaux2003-12-151-2/+1
| | | | | - remove (hopefully) unused code rewriting shift -> shift(@_)
* in the generated .pot, try to use relative filenames (so that the .pot doesn'tPascal Rigaux2003-11-101-1/+1
| | | | | change at each build)
* new warning: you can replace "member($xxx, keys %yyy)" with "exists $yyy{$xxx}"Pascal Rigaux2003-11-101-0/+7
|
* also disallow map {...} %xxxPascal Rigaux2003-10-011-2/+9
|
* warn: foreach (%xxx) { ... }Pascal Rigaux2003-10-011-0/+2
|
* new warning: N_("xxx") . "yyy" is dumb since the string "xxx" will never get ↵Pascal Rigaux2003-09-301-0/+3
| | | | translated
* - allow $_o_XXX parameter name which is both unused and optional (same for ↵Pascal Rigaux2003-08-111-2/+5
| | | | | | | | | | | | $_b_XXX) - shift is a ONE_SCALAR_PARA so that $box->pack_start(shift @l, 0, 0, 4) is parsed correctly - in arrange_global_vars_declared(), don't keep anything in global_vars_declared, better create shadow packages to contain them - much better merging of multiple files defining functions in the same package. This fixes the bad behaviour when using the cache (esp. do_pkgs, but it was even worse with things in ugtk2.pm)
* - disallow return(...), prefering return ...Pascal Rigaux2003-05-271-13/+25
| | | | | - P_call_no_paren has a lower priority than P_comma
* better warning for: print $a . 'foo'Pascal Rigaux2003-05-261-1/+5
|
* add a special case to handle "arch => 1" without going through word_alone()Pascal Rigaux2003-05-261-4/+4
|
* - warn things like: if ($a = 1) { ... }Pascal Rigaux2003-05-261-0/+6
| | | | | - undef is always false
* check for "0 or ..."Pascal Rigaux2003-05-261-0/+5
|
* suggest using the functional map instead of the imperative foreach whenPascal Rigaux2003-05-231-0/+12
| | | | | possible
* fix detection of grep in scalar contextPascal Rigaux2003-05-231-3/+4
|
* add warning: you can replace "map { if_(..., $_) }" with "grep { ... }"Pascal Rigaux2003-05-231-0/+4
|
* - suggest any instead of grep in scalar contextPascal Rigaux2003-05-231-0/+6
| | | | | - suggest foreach instead of map in empty context
* correctly warn about the missing space in "my ($a,$b, $c);"Pascal Rigaux2003-05-161-1/+2
|
* map and grep do not return a scalarPascal Rigaux2003-04-301-0/+2
|
* allow "*foo = sub {}" with no prototypePascal Rigaux2003-04-301-6/+6
|
* ensure return values are usedPascal Rigaux2003-04-291-15/+70
|
* - handle empty prototypesPascal Rigaux2003-04-241-7/+11
| | | | | | | - check number of parameters in method calls (ignoring empty overloaded methods) - allow non-empty prototypes for methods not using @_
* basic "number of arguments" checkingPascal Rigaux2003-04-171-1/+2
|
* forbid "unless (...) { ... } else { ... }" and "unless (...) { ... } elsif ↵Pascal Rigaux2003-04-151-0/+4
| | | | (...) { ... }
* add basic "type" checking (using a very liberal lattice)Pascal Rigaux2003-04-151-20/+144
|
* use new types to have stricter type checking (prior to adding maybe_context)Pascal Rigaux2003-04-131-164/+174
|
* warn things like: ($foo) ||= ...Pascal Rigaux2003-04-111-0/+4
|
* enhance non_scalar case for some operators using is_not_a_scalarPascal Rigaux2003-04-101-10/+20
|
* correctly (in Perl way) handle priority for some special unary functions ↵Pascal Rigaux2003-04-101-7/+2
| | | | (length, exists, ref)
* warn xxx == "ia64", xxx eq 2Pascal Rigaux2003-04-101-0/+18
|